en de
16th Proceeding of the ADµP Developer Network
Please select a language version in the header line, one of the following internet links or headings from the table of contents!
(«) First Proceeding (<) Previous Proceeding (>) Following Proceeding (») Last one Proceeding (^) Author (=) Editorial Staff ($) Order of the detailed text (&) List of proceedings (§) List of company documents
Manner of proceedings:
Patent application DPMA 10 2007 006 084 (abridgment)
Topic:
Procedures for the timely determine of the characteristics, harmonics and non-harmonics of rapidly changing signals with additional output of derivative pattern, control signals, event stamp the post-processing and a weighting of results
Author:
Christian E. Jacob
top     (0) Summary
The invention concerns a process to determine the characteristics timely, harmonics and non-harmonics rapidly changing signals of additional output derived pattern, control signals, post processing event stamp and a weighting of results, whose operational steps at multiple levels of procedure the in Fig. 18 demonstrates, expire. The process is described using events associated with the respective levels of procedure. Each procedure level occupies procedure States this specific. The signal to analyze is processed side by side at the procedure level one, two, four and five. A reference signal and the relation to the reference signal for the first and second level of the procedure must be provided as additional input signals. These two additional input signals are needed to control the process procedure. The process is also affected by an input size of the third procedure level, which in turn affects the first procedure level. Of a reset signal that the first and the third procedure level monitoring, the process can be started or stopped. Step-the sampling over the signal to be analyzed, required by the procedure level one, two, four, and five is variable. The timeline, as to added sampling step-over, is required for the first and second level of the procedure, and return transformations from the frequency in the time range. This phase of harmonics is either internal or external referencing this timeline. The procedure level 2 provides surface elements under the signal to be analyzed and cached sample values for the first procedure level. Procedure-level top is four and five provided the lower envelope of the signal to be analyzed. The sixth level of the procedure are obtained from arithmetical average, and period time predicted values influence over two additional input sizes. There which are above also Weight and the frequency of the signal to be analyzed. At the seventh and may further process levels be determined phases of basic vibration and other harmonics are needed for the Rücktransformierten. Two sizes as a time reference from the first procedure level are also provided for the reverse transformation. Formed directly from the in according to the invention physical processes from the referred to pass values, harmonics and non-harmonics and spent time discrete to evaluate in a neural post-processing is not dispute, of in according to the invention physical referred pattern. Progresses preferably in fast programmable logic devices.
top     (1) Prior art
For the pictorial representation of in according to the invention physical procedure in Fig. 1 to 14 special, defined here will be used graph networks that allow procedure.

Graph network element to pictorial describe of a procedure step

The graph networks used here contact from graph power elements appropriately Fig. 15A consisting of an old state, a time-dependent  event and a new State together. Multiple input and output States about each event, as in [L1. 01] described, are not allowed. This limitation excluded that a procedure simultaneously or alike can occupy two States after an event. An event is always a process step that runs when the event related zeitbedingte transition condition is met.

Transition condition to the stage of proceedings must be time-dependent in any case. She may also depend on any other system size. Contains a graph power element with a then systemic transition condition Fig. 15B. It consists of the foregoing State, systemic event, and the downstream State.

Graph power to the pictorial description of a procedure

Graph network elements are linked to a graph power by the initial state of a graph power element alike is the input state of subsequent graph network element. The approach to hand of Fig. Discusses 15 c. The common sampling procedure with two graph network elements are connected to a ring is described there.

The first State of the procedure applies a value for the sampling step-over control of procedure procedure from outside in an action. The procedure enters its second State if the sampling step-over elapsed. To perform the transition, the graph power defines a first event with a time-dependent  transition condition. In the second State of the procedure the current system time passed the current value of the continuous signal and the space for the sampling time in a second action space of the time discrete value in a first action.

To give the possibility the sampling step-over again from outside to take, the procedure in a closed graph power from the second switches back to its first. For this change, you must insert a second event. Lt. definition [L1. 02] a closed graph network must contain at least an event with an outstanding transition condition. I.e. a closed graph network transition conditions may not (time-dependent) at the same time and (by design) equally satisfied. This is by default of a delay in the second State in a third action here and their evaluation in the second event reached with a time-dependent  transition condition.

The presented example of a known technique includes two graph network elements in a closed structure: the first is input and the second is the initial state of first graph network element that is associated with the first event. Second as input and the first as baseline condition is associated with the second event, however. So a procedure can be represented figurative with almost infinite sample events with only two graph network elements in a closed graph power.

In Fig. 15D represent between the continuous signal and the associated time discrete sample values. The in according to the invention physical procedure shown others and all results were found with a graph Network Simulator. The syntax used draws on [L1. 03] at. Is the in [F1. 01] similar used State graphs. In [F1. 02] State graphs are design and in [L1. 04] used to verify technical comments. To be [F1. 03] used in particular to the design of LSI circuits.

Dominant vibration as reference signal

With in Fig. Sampling procedure described 15 c samples in accordance with the sampling theorem can [L1. 05] - here known as regular scanning - but executed multiple rates procedure. Reg. sampled subsequently referred values, the sampling frequency determines the upper vibration maximum frequency. It applies to [L1. 05]

f max sample/2 = f . GL. 1.01

To characterize technical entity of signal transmission and processing their bandwidth is preferably

Β = {q}{min; f max} GL. 1.02

any subharmonics which are fundamental and most harmonics of the signal, specify. The basic vibration usually dominant vibration in this vibration mixture. Is derived directly from the source or filtered out from the signal to be analyzed using means are State of the art and as a reference signal

c.z = c.y (ny = 1) glasses 1.03.

provided for the in according to the invention procedure. In the fonts [F1. 04] until [F1. 09] such reference signal is used.

Segmenting signals

Useful signals have random [L1. 06 sec 2.1.3] and deterministic [L1. 07] shares. One of those securities disturbs usually, whose effect is suppressed. The other portion is desired, it is separated with technical AIDS, processed and evaluated. A predominantly stochastic usable signal such as a sound or a bang requires statistical out value procedure. In one such signal a process event to recognize is by [F1. 10] until [F1. 12] technically expensively. After [L1. 08] also proposes a segmentation of stochastic signal. In [F1. 13] divides the voice box and processed sequentially. Segmentation is defined but also the distribution of signal using tape passports in groups and its continuous analysis [F1. 14].

Technical area

in according to the invention physical method, you are rapidly changing signals analyzes characteristics, harmonics and non-harmonic determined and printed thereof derived pattern control signals, event stamp and a weight of results for the post-processing.

Unlike deterministic signals such as the sound, the sound, the music, the language can to stochastic or fully describe other modulated signals by their time or their spectral function. To signals with no strict determinacy [L1. 09], but with a physical reason (the source is a natural or technical vibration capable structure), to undergo an analysis through in according to the invention dominant vibration, derived also from signal source is fed next to the time function of the signal to be analyzed as reference signal. The half life of period or a multiple of half period duration of the reference signal is set as time discrete distance for the analysis. I.e. this time discrete distance is determined by a reference signal may rapidly changing. Often characteristics are accordingly (common share, RMS, RMS of basic vibration etc.) of the signal to analyze detects and spent time discrete periodically. The dominant vibration of the analysed signal is used as reference signal in according to the invention physical process. In addition to the reference fashion described in the in according to the invention physical procedure still voice-over, tandem and Y_O_U fashion are defined. The latter modes used to analyze signal itself as a reference signal.

 


top     (2) Underlying problem
 

A subspecialty of signal analysis is the evaluation of simulation data. In the post process, is (offline) or directly during the simulation (online). The development of a variety of event-oriented analysis, calculation and control models was the development of in according to the invention physical process [L2. 01] advance. Fig. 16 provides a such simulation example to clarify the problem. There, the voltage of a capacitor in a subdued measuring is parsed.

The in according to the invention procedure provides timeline directly related to the analysed signal etc.

- the upper and lower envelope, envelopes, distance

- Figures such as the RMS,.

- Real and imaginary part of harmonics,.

- Scanning and event stamp,.

- along usable synchronization pulses

- weight analysis results.

- time continuous retransformed of basic vibration,.

- more time-continuous retransformed of harmonics,

- the represented time continuous sum of harmonics, calculated from the analyze signal less of mean and Rücktransformierten of basic vibration and

- the time continuous shown total of pro rata harmonics which further from the analyze signal minus the mean Rücktransformierten of basic vibration and minus one Rücktransformierten are calculated.

To the necessary general application for an electronic circuit where the in according to the invention procedure runs to achieve a variable sampling step-over is allowed for the signal to analyze. The necessary accuracy of signal analysis is still achieved by controlled by the in according to the invention physical procedure itself faces formed between two sample steps are, which either left the so-called or be assigned the so-called right of the first procedure level but.


top     (3) Description of the invention
 

in according to the invention shows duly figurative using seven graph networks. To formulate the claims, each graph network is assigned a procedure level. On each process levels can in according to the invention procedures take a particular State and complete the respective state its actions. Several actions are assigned, a State they are processed in figurative this order. Because this order as timing is to understand, it is defined here order as logical actions within a Verfahrensschrittes? or a logical Reihenfolge? briefly. The transition of States describe the progress of the procedure. Changing the State at each process level is immediate. Perform the actions associated with the respective States at the procedure level as well in a logical order. The actions

- of the third procedure level are in this order above,.

below are the actions each

-the second

-the first

-the fourth

-the fifth

- of the sixth and

-seventh

Procedure-level processed in a logical order.

Operational steps

Events in the graph to describe the progress of the procedure. Appropriate events include:

- the raising of external reset signal;

- break of external reset signal and

- reaching an externally specified sampling step

- no change of the reference signal or but

a change of the reference signal after a further sampling. Other events occur after

-a output of analysis it results if

·No,.

·a small change or

·on the other hand, a large change

the reference to the reference signal is caused.

Also enter events,

- if the value of proceedings progress meter too small or

- if this value

is sufficient.

Procedure-level 1

The in according to the invention procedure occupies one sixteen different States procedure-level: in a first State three output synchronization pulses are reset. The absence of external reset signal is monitored as first event. This first event occurs, changes the in according to the invention procedure at the first level of procedure in the second State.

Beginning of the second State in according to the invention procedure applies the current values of the signal to be analyzed, the reference signal and the reference to the reference signal and moves exactly after an externally specified sampling in the third State. There the current values of the signal to be analyzed, the reference signal and the reference to the reference signal is again read with beginning this State:

Reference signal and relative to the reference signal are identical in fourth State the first procedure level initially at changing waiting and changed again in the second State.

In the third State are calculated in a logical order of actions the half sampling step-over and various surface elements under the signal to be analyzed.

The reference signal is greater than the relation to the reference signal in the third State otherwise changes the in according to the invention procedure on this so-called left of the first procedure level to the fifth State to the so-called right of the first procedure level in the sixth State. The logical order of actions of States both is identical. Symmetrically shows in the graph network both referred to. However, they are distinguished by their events and the associated memory locations used in the actions:

In the fifth State

- held the Abszisse value to determine of the duration of the period of reference signal.

- in the third or the sixteenth state calculated surface elements to initialize Summation stores used and

-totaled to each externally specified sampling step surface elements calculated on the second level of the procedure.

In the sixth State are also

- held the Abszisse value to determine of the duration of the period of reference signal.

- in a third State or the thirteenth State calculated surface elements to initialize other Summation storage used and

- to each externally specified sampling step again accumulated surface elements calculated on the second level of the procedure.

Is less than the reference becomes the reference signal in the fifth State the reference signal in according to the invention procedure on the left side of the first level of procedures in the seventh State.

On the right side of the first procedure level in according to the invention procedure from the sixth to the eighth State changes if the reference signal is greater than the reference to the reference signal.

In the seventh State which is

- output value of left event meter wide count,.

- determines the relative value of Abszisse of Schnittpunktes between reference signal and the reference to the reference signal

- calculates the absolute value of Abszisse this Schnittpunktes

- this intersection on the signal to analyze projeziert

- held the Abszisse value to determine of the duration of the period of reference signal.

- already in the third State or the sixteenth state calculated surface elements again calculated and

- adds the Summation memory used in the fifth State by above surface elements.

In the eighth State which is

- output value of the right event counter further count,.

- determines the relative value of Abszisse of Schnittpunktes between reference signal and the reference to the reference signal

- calculates the absolute value of Abszisse this Schnittpunktes

- this intersection on the signal to analyze projeziert

- held the Abszisse value to determine of the duration of the period of reference signal.

- already in the third State or the thirteenth State calculated surface elements again calculated and

- adds the Summation storage by above surface elements used in the sixth State.

Before from the until date accumulated left and right the output parameters are calculated associated surface elements in according to the invention physical procedure procedure progress will be monitored:

The procedure cumulative count is less than ?Zwei? in according to the invention procedures at the first procedure level from the seventh through the tenth switches back to its sixth or eighth on the twelfth back in the fifth State.

Against the procedure cumulative count is greater than or equal to ?Zwei? in according to the invention procedure at the first level of proceedings of the seventh State ninth in the changes or from the 8th to the 11th State.

Be the first procedure level in the ninth and eleventh States of in according to the invention physical process

- Synchronisierimpule for process layer creates four, five, six and seven and spent to further leverage outside of in according to the invention physical procedure;

the time discrete

- recorded to date used period duration of the reference signal,.

- determined new period duration of the reference signal

time discrete

- recorded to date used mean value of the signal to be analyzed.

- determined new average of the signal to be analyzed and issued as characteristic,.

the / the time discrete / n

- Gleichrichten mean

- RMS,

- RMS of removable share

- Crest factor related to the positive peak

- Crest factor related to the negative peak

- Form factor,

- Vibration levels

- Ripple,

- Bulb factor

- Realteil of basic vibration

- Pattern of Realteils of basic vibration

- Imaginary part of the basic vibration

- Pattern of basic vibration imaginary part

- RMS of basic vibration

- Basic vibration levels

- Upper vibration ripple

- THD

- Realteil of the ny-th upper vibration,.

- Pattern of Realteils of the ny-th upper vibration,.

- Imaginary part of the ny-th upper vibration,.

- Pattern of imaginary part of the ny-th upper vibration,.

- RMS of ny-th upper vibration,.

promptly determined and each issued as characteristic of the signal to be analyzed.

This output is time discrete according to preset outside half period time interval or a multiple of half period duration of the reference signal.

Switches between the 9th or 11th State in according to the invention procedure in the thirteenth, Fourteenth , fifteenth or sixteenth state to

-to verify, mapping reference signal and relate to the reference signal

- If necessary, the procedure progress counter reset,

-to identify, the relative value of Abszisse of Schnittpunktes between reference signal and the reference to the reference signal

- to compute the absolute value of Abszisse this Schnittpunktes

- to projezieren this intersection on the signal to analyze and

- recalculate calculated already in third State surface elements.

The referred review of mapping reference signal and relative to the reference signal is particularly emergency slightly in addition enable the voice-over, tandem and Y_O_U mode without changing the process of in according to the invention physical process (structure of graph network) to the reference fashion.

Immediately the in according to the invention procedure from left to right of the first procedure level from the thirteenth returns to its sixth .

Also immediately in according to the invention procedure from right to left of the first procedure level from the sixteenth returns to its fifth .

Until more output of referred characteristics is possible, the in according to the invention procedures at the first procedure level occupies some which described again States:

Regardless of the described here sequence of States can in according to the invention procedure return of external reset signal, fifth, sixth, fourteenth, and fifteenth in the start or start state return.

Procedure-level 2

The second level of the procedure, the in according to the invention procedure occupies only two States, which are numbered to formulate claims with seventeen and eighteen:

In the seventeenth state become by first scanning

- the reference signal, the relation to the reference signal and the signal to analyze multiple cached,.

- of the count of samples: set to ?Null? and

more cells for calculating

- of the square

- of cosinus-weighted part

- of the sinus-weighted part

- cosinus-weighted part corrected with the atomic number and

- of sinus-weighted part corrected with the atomic number

Initializes the signal to be analyzed.

In the eighteenth State are to all other samples

- Sample values taken from the previous sampling of reference signal, the Datum to the reference signal and the signal to be analyzed to calculate surface elements.

- Values of the current sampling of reference signal, the Datum to the reference signal and the signal to be analyzed cached,.

- of the count of samples: increments to each sampling also

for calculating (trapezoid) surface elements

- square,.

- cosinus-weighted part,.

- sine-weighted part,.

- corrected the with the atomic number cosinus-weighted part and

- corrected the with the atomic number sine-weighted part

the signal to be analyzed, and

it as (trapezoid) surface elements between previous and current sampling

- surface the simple (arithmetic)

- fill the square (geometric)

- cosinus-weighted surface;

- sine-weighted surface;

- with the atomic number corrected cosinus-weighted surface and

- with the atomic number corrected sinus-weighted surface

deployed as repeatedly used sizes.

Procedure-level 3

The third level of the procedure, the in according to the invention procedure occupies as only two States which are numbered to formulate claims with nineteen twenty:

The in according to the invention procedure occupies the nineteenth state the third level of the procedure if the external reset signal exists or respawn. Then be

- of memory provide the procedure fortschrittes and

- of counter of procedure fortschrittes

initialized with ?Eins? and

This procedure level event stamp on the back to ?Null? set,

- left.

- the first procedure level right and

- of the sixth procedure level

increments and are output.

The twentieth State of in according to the invention physical process is taken up the left and right at the first procedure level event stamp were incremented once. Then pass an externally prescribed value memory provide the procedure fortschrittes and thus the output discrete time distance set.

Procedure-level 4 and 5

The in according to the invention procedure takes on the fourth conditions twenty-one, twenty-two and 23 and the fifth level of the procedure, the States a twenty-four, Fünfundzwanzig, and twenty-six.

The fourth level of the procedure, during the positive half-vibration of the reference signal determines the maximum value of the signal to be analyzed and issued as upper envelope.

Regardless, determines the minimum value of the signal to analyze the fifth level of the procedure during the negative half-vibration of the reference signal and issued as lower envelope.

At two levels of procedure determines each distance of envelopes and issued.

Procedure-level 6

The sixth level of procedures can occupy a total of five States. To formulate the claims, the States were numbered from siebenundzwanzig to einunddreißig. You are required to determine the weight of the signal analysis and output:

The common share changes to the signal to be analysed very quickly analysis is complicated by. Results still usable, in according to the invention physical procedures suggests a predictive value for the DC component. This prediction was correct, reviewing the in according to the invention physical procedure time discrete periodically and as weight of common share issued.

For integrated Fourier analysis, the in according to the invention physical procedure suggests a predictive value for basic vibration period duration. This prediction was correct, reviewing the in according to the invention physical procedure also periodically time discrete and issued as weight the duration of the period.

From the weights of common share and period duration in according to the invention physical procedures in time discrete internally synchronized intervals determined the total weight and issued.

The predictive value for the DC component can be affected by outside of in according to the invention physical process. Similarly, the predictive value for the period life of outside the in according to the invention physical process basic vibration can be corrected. The above mentioned weights can about a prediction matrix [F3. 01] outside of in according to the invention physical process can improve.


 

Procedure-level 7

The seventh level of procedures can occupy a total of six States. To formulate the claims, the States were numbered from Zweiunddreißig to Siebenunddreißig. Are required to by Fourier analysis phase of the vibration of the ny. h1-th ordinal in wheel to determine and deg and issue.

Further levels of procedure

Further levels of procedure are not described here. You are possible and complement the in according to the invention procedure with a structure accordingly Fig. 13 determine basic vibration and other harmonics phases.

Calculation and output of continuous Rücktransformierter

Become independent from the States of process levels each sampling step

- a continuous average derived from the time discrete mean,.

- it as a continuous exchange of the difference between the signal to be analyzed and the continuous average determines

- from RMS and phase of basic vibration continuous retransformed of basic vibration taking a second vorgebbaren from outside the in according to the invention physical process time basis determined

- thereof as upper vibration of the difference between continuous exchange rate and the continuous return transformierten basic vibration determine

- from RMS and phase of the ny-th upper vibration continuous retransformed of the ny-th upper vibration taking into account the second vorgebbaren from outside the in according to the invention physical process time basis determined,.

- it as a vibration of the difference between continuous exchange rate and the continuous Rücktransformierten of the ny-th upper vibration determine and

- again as a change of the difference between continuous vibration unit and the continuous Rücktransformierten of the ny-th upper vibration calculate

and issued. Continuous expenditure referred and others used to segment the signal to be analyzed in subsequent instances of in according to the invention physical process


top     (4) Industrial applicability
 

The in according to the invention process is suitable for Off-and analysis of deterministic signals that are subject to rapid Intensitäts-and/or frequency variations. The operational steps and the associated actions run preferably in programmable logic modules (PLM) or other mikroelektronischen building blocks as a pure hardware solution. Hardware solutions with multiple instances of in according to the invention physical procedure are suitable especially for multi-channel building for modal or structural. Hardware solutions are also useful when multiple instances of in according to the invention physical process successively segmented a signal with high information density as a voice.

In other cases, runs the in according to the invention procedure on hardware with digital signal processor (DSP), microcontroller (µC), hybrid module (HyperStone), or default processor (i386) as a pure software solution (program). Software solutions are preferably for what exist only in a few numbers or regularly receive an update.

Output channels for harmonic and non-harmonic analysis

The in according to the invention procedure supports no algorithmic processing. Therefore, the algorithm of the fast Fourier transform (FFT) is not implemented. Trozdem undergoes the analyze signal using a harmonic or non harmonischen analysis of in according to the invention physical process. Transformation channels each intended for harmonics and/or Nichtharmonischen. An ordinal is specially assigned each transformation channel. The atomic number is ganzzahlig, a harmonious runs otherwise non equation analysis.

Evaluation of simulation data

The in according to the invention procedure provides accurate results with a variable sampling step-over of the signal to be analyzed. The sampling step-over can be changed after each sample. It is especially possible, simulation data (SPICE, Simplorer, MatLab, SABER,...), which a variable step-computing over who evaluate directly.

Orientation in space parsing waves

Wave analysis, instead of sampling time a grid of the area covered by a positioning system and the current Abszisse value derived from it. Also, a point in space at appropriate spatial intervals are moving can be identified instead of the system time. Abscissa in the wave propagation direction should be on the basis of where referred to.

Handling biological signals

The sweep of the signal to analyze eliminates running surface elements discovery in biological systems - as in the cochlea-. This way of a signal analysis is useful if the results as biological signals are used.

Tool for design and execution

The used fachsprachliche representation using Graph networks in Fig. 1 to 14 was base for technical design of in according to the invention physical procedure and their verification. Also becomes the implementation of the design in a hardware description language (VHDL, VHDL-AMS, MAST,...) and/or programming language (C, C++, Java,...) used.


 

Applikationsbreite for electronic circuit

The development of an electronic circuit is justified if necessary Applikationsbreite exists. The in according to the invention procedure to the

- qualitative and/or quantitative characterization to the

- Including amplitude, frequency and phase, identify parameters to the

- Analysis of harmonics and Nichtharmonischen, to the

- Creating patterns and their classification and the

- Filter conditioning and/or encrypt

 

used the signal to analyze. More applications

- to assess and be a flusen of electrical parameters

- to identify, in particular the time-dependent behavior, a signal transmission path

- ordinal and signature analysis

- to diagnose machines, vehicles and units and

-Automatic speech recognition and-Übersetzung

are possible.

Programming system at the AD interface

Control flow, process steps and some associated actions run in the future in a specific microprocessor. The other part of the actions can be adapted to the respective Applikationsaufgabe by this novel microprocessor programmer. To be comfortable programming environment at the interface between the analog and digital worlds.


top     (5) Beneficial effect of invention
 

In [F5. 01] is the online processing of process data in a neural network (predictive adaptive resonance theory - ARTMAP) described. The determination of the AR parameters [F5. 01Seite 13 line 33] occurs after [L5. 01]. Process data be transformed into pattern [F5. 01Sheet 3 Fig. 4 and page 10 line 39: dam acquisition modules 410], which are then processed by a neural network. Other means and a procedure [F5. 02 page 2 line 1] a signal is decomposed so-called basic elements in there. The decomposition is a conversion unit with characteristic extraction device [F5. 02 page 7 lines 22] made. In two fonts database consisting of acquired and standard elements is exposed not for neural processing obvious reasons. The technical execution of acquisition of elements referred is not described. An assessment of the fonts in this respect is not type.

[F5. 03] co-founded a generally usable universal procedure [F5. 03 page 2 line 26] for establishing the basic vibration for rapidly changing signals such as Voice signals. The technical effort there earn filter [F5. 03 page 13 Fig. 1] is significant. Reference signal and relate to the reference signal must be provided for the in according to the invention procedure. It requires only two filters.

Is a variable sampling step-over [F5. 04] achieved by specified sampling events the signal directly from a Tachometers pulses to analyze [F5. 04 page 7 line 10]. Enables synchronous averaging [F5. 04 page 5 line 57], a fast Fourier transform (FFT) [F5. 04 page 7 line 18] and the localization of needles by using a corresponding detector in the analyte signal angle on rotating parts [F5. 04 page 6 line 33]. The program remains limited to error detection machines. This narrowing does in according to the invention procedure. It is broadly applicable.

After [F5. 05 page 3 line 17-25] meets a window interval with a window function sufficiently precisely determine the period duration of the signal to be analyzed. Just described technical effort to synchronize the sample steps on the duration of the period is not emergency slightly. This solution is suitable only for pass values of signals with almost constant fundamental frequency. The in according to the invention procedure allows contrast to a large Variationsbreite of the fundamental frequency of the signal to be analyzed in more beneficial ways.

In [F5. 06] be temporal frequency shifts of the formants [F5. 06 page 01 rows 41-65] a voice signal, primarily to speech recognition [F5. 06 page 03 lines 32-35] be used, separated. Identify three or four formants in the spectrum one runs each over short periods (8-10 ms) Fast Fourier Transforms (FFT) [F5. 06 page 07 lines 33 - page 08 lines 12]. With the expansion of transformation channels in the in according to the invention physical procedure described creates a similar spectrum. Time discrete history basic vibration and the formants, and its patterns are in Fig. 17 for real and imaginary part separately represented. A basic vibration and formants separately after amount and phase is also possible. This procedure of in according to the invention physical each updates the tracks after the half period of the dominant vibration. Compared to the application in [F5. 06] time discrete sequence arranged frequency spectrums resolution is improved significantly.

In contrast to [F5. 06] segmented formants in other instances of the in according to the invention physical process. This time discrete resolution of sequentially ordered frequency spectrums of basic vibration can be refined independently much.

Resolution of consonants in the voice is described as problematic [F5. 06 page 02 lines 43-48]. Will solve the problem of the in according to the invention procedure by standardised and scaled with the atomic number pattern for neural processing to the distance of the envelopes. The scale with the atomic number is new. She can be performed only with the in according to the invention physical procedure.

In [F5. 07 page 2 line 56 to page 3 line 18] outward and return transformation is harmonious and non-harmonious shares (42 and 83 1/3 Hz in one 16 2/3 Hz-power) described an AC current. Thus determined shares of AC current to zero will be regulated through this procedure and this device. The in according to the invention procedure allows also the outward and return transformation of harmonious and non-harmonious shares of the signal to be analyzed. Return transformed share is affected but not (by a Governor). He is directly output through in according to the invention or the subtracts the signal to analyze process the rest of the signal to be analyzed in a further instance of in according to the invention physical process.

Event stamp are incremented and printed to index the determined time discrete characteristic values, harmonics and non-harmonics for any further processing, representing in patterns or other graphical output. Referred to counter support back trekking from neuronal evaluation back up in the pattern. The numbering of events in the fonts [F1. 10], [F1. 11] and [F1. 12] serves other objectives. It serves not indexing the determined time discrete characteristic values, harmonics and non-harmonics there.

The in [F1. 04] until [F1. 09] described signal converter or analysis need a reference signal as well the in according to the invention procedure. It is derived directly from the source or filtered out from the signal to be analyzed using means are State of the art. What is new is that a further input signal in according to the invention physical procedure, namely the reference to the reference signal is needed. The reference signal and the relation to the reference signal be solely used to control the process. In terms of reference signal the in according to the invention procedure differs in principle from fonts [F1. 04] until [F1. 09].

The output of the analysis results weight is also new. The weight is the successful enforcement of in according to the invention physical process time discrete periodically and avoids the usual for measurement and analysis tasks accuracy viewing user. Applications range basic of signal analysis as the user to successfully run of the in according to the invention physical process in the most diverse in [L5. 02] and [L5. 03] conceptually processed.


top     (6) Execution sample
 

Succeed drain a reference signal from the source or filter out the signal that provides in according to the invention procedure characteristics harmonic patterns derived tax impulses, event stamp, and a weight of results, as in Fig. 17A to p shows.

Fig. the execution example 17A above are two low-pass filter (Tschebischew 3) Order; limit frequencies = √2 * or 1/√2 * frequency of base vibration) for filtering out of reference signal Fig. 17A below, relate to the reference signal Fig. 17 b above uses.

Fig. as analysierendes signal 17A Middle used a voice. There different consonants (at the front and back) and vowels (in the Center and in the second third) significantly by their intensity.

In Fig. 17 b Middle is by in according to the invention physical processes from reference signal Fig. through the abscissa time discrete history of frequency, 17A below, relate to the reference signal Fig. 17 b above determined, issued.

Fig. 17 b below contains the synchronization pulses internally used to synchronize of different procedure levels and external control of processing procedures.

Fig. 17 c above contains time discrete representations of Gleichrichten average and Middle average and below the continuous representation of mean.

Fig. 17D above contains the event stamp used to index all time discrete expenditure outside the in according to the invention physical process. Fig. 17D Centre contains the represented time discrete RMS and Fig. 17D below shown as time discrete RMS of removable component of the signal to be analyzed.

Contain more time discrete representations

Fig. 17E top to form factor in the Middle to the vibration levels and down to the ripple of the signal to be analyzed,

Fig. 17F top to Realteil in the Center to the imaginary part and below to the RMS current of basic vibration of the signal to be analyzed, and

Fig. 17 g above the stage in wheel and in the Middle phase in deg of basic vibration of the signal to be analyzed.

On the other hand contains Fig. 17 g below continuous representation of return transformierten basic vibration of the signal to be analyzed.

Below contain time discrete representations

in Fig. 17 h above the basic vibration content in the Center of the basic vibration ripple and down in relation to the basic vibration of the signal to be analyzed, the THD

in Fig. 17I above time of period, the Middle weight weight of common share and below the weight of analysis of signal to the respective time step

in Fig. 17J above Realteil of the ny. h1-th upper vibration in the Middle the imaginary part of the ny. h1-th upper vibration and below the RMS ny. h1-th upper vibration in the first transformation channel and

Fig. 17 k above the phase ny. h1-th upper vibration in wheel and in the Middle phase of ny. h1 deg-th upper vibration in .

On the other hand contains Fig. 17 k below the continuous representation of Rücktransformierten ny. h1-th upper vibration from the first transformation channel.

The subsequent continuous representations are derived from above continuous representations to each of the signal to analyze sampling step:

Contains to Fig. 17 l above the history of the signal to analyze minus the mean difference, in the Middle of analyze signal less of mean and minus the difference back transformierten basic vibration and below the difference of analyze signal less of mean and less Rücktransformierten of the ny. h1-th upper vibration in the first transformation channel. To non equation number for the first transformation channel with in the example was ny. h1 4,22 = (first formant after base vibration in the voice) externally specified.

More time discrete representations concerning

in Fig. 17 m above upper envelope in the Center of lower envelope and below the distance of the signal to be analyzed, envelopes

in Fig. 17N above the crest factor hergeleitet from the upper envelope in the Middle of the crest factor inferred from the lower envelope and down inferred distance envelopes of the signal to be analyzed, the bulb factor

in Fig. 17o above a pattern derived from the Realteil in the middle of a further patterns derived from the imaginary part of the signal to be analyzed, basic vibration

in Fig. 17p above a third model derived from the Realteil ny. h1-th upper vibration, in the middle of a last pattern derived from the imaginary part of the ny. h1-th upper vibration of ny. h1-th upper vibration in the first transformation channel and down values of a second abscissa used as an input for the same and more instances of in according to the invention physical process and here about the first abscissa plotted is.

Finally is in Fig. 17o below continuous history of analyze signal less of continuous mean minus the difference back transformierten basic vibration and less Rücktransformierten NY. h1 represented -th upper vibration in the first transformation channel. This original size contains segments of signal to be analyzed into downstream instances of in according to the invention physical procedure for analysis.

The execution example matches Edition/version 4.40 in according to the invention physical process in an application as a Simplorer model [L1. 03] in reference mode. A graph power structure of Edition/version has already been 2.10 under [F6. 01] declared. At the time was used to develop of the so-called TANDEM procedure for period duration and wavelength measurement.


top     (7) Reference character
 

Modes of operation-input sizes and repatriation of output quantities

RMReference fashion--> inputs:c.z, c.y and c. z_0--> no repatriation

SMSynchronous fashion--> inputs:c.z c = c.y and . z_0--> no repatriation

TMTandem mode--> receipt:c.z = c.y--> repatriationc. z_0 = d. z_0

UMY_O_U fashion--> receipt:c.z = c.y--> repatriationc. z_0 = c. z_0

 

Procedure layers and processing order of actions

VE1Fig. 1 procedure level 2 to 8 processing order 3

VE2Fig. 2 procedure level 1 Processing order 2

VE3Fig. 3 procedure level 9 Processing order 1

VE4Fig. 4 procedure level 10 Processing order 4

VE5Procedure-level fig. 5 11 Processing order 4

VE6Fig. 6 procedure level 12 Processing order 4

VE7Fig. 7 procedure level 13 Processing order 5

 

Procedure States

1

.

Level

 

 

1

.

State

 

Fig. 2:--> Start state : s3cRS1

2

.

State

 

Fig. 2:--> First sampling: s3cFS1

3

.

State

 

Fig. 2:--> Second sampling: s3cFS3

4

.

State

 

Fig. 2:--> Waiting for changing reference signal: s3cNO2

5

.

State

 

Fig. 3:--> Actions during the left half-vibration of the reference signal above the reference signal: s3cHL2

6

.

State

 

Fig. 4:--> Actions while the right half-vibration of the reference signal below the reference signal: s3cHR2

7

.

State

 

Fig. 3:--> End the left half-vibration of the reference signal:s3cEL2

8

.

State

 

Fig. 4:--> End the right half-vibration of the reference signal:s3cER2

9

.

State

 

Fig. 5:--> Output characteristics after the left half-vibration of the reference signals3cFL2

10

.

State

 

Fig. 4:--> Increase of procedure forsch spread meter on the left side: s3cPL2

11

.

State

 

Fig. 6:--> Output of characteristics after the right half-vibration of the reference signals3cFR2

12

.

State

 

Fig. 3:--> Increase of procedure forsch spread meter on the right: s3cPR2

13

.

State

 

Fig. 7:--> Change to the right:s3cCL2

14

.

State

 

Fig. 7:--> Setting of proceedings progress meter when jump down:s3cJL3

15

.

State

 

Fig. 8:--> Setting of proceedings progress meter when leap upward:s3cJH3

16

.

State

 

Fig. 8: Left--> change to the page:s3cCR2

2

.

Level

 

 

17

.

State

 

Fig. 1:--> Provide the old values in the first sampling step: s2cCA1

18

.

State

 

Fig. 1:--> Provide the old values: s2cCA3

3

.

Level

 

 

19

.

State

 

Fig. 9:--> Reset of proceedings progress meter: s1cSP1

20

.

State

 

Fig. 9:--> Set of memory provide the procedure fortschrittes: s1cSP3

4

.

Level

 

 

21

.

State

 

Fig. 10:--> Accept maximum: s4cVL1

22

.

State

 

Fig. 10:--> Left waiting for maximum value: s4cVL3

23

.

State

 

Fig. 10:--> Issue top envelope: s4cVL6

5

.

Level

 

 

24

.

State

 

Fig. 11:--> Minimum right accept: s4cVR1

25

.

State

 

Fig. 11:--> Right waiting minimum value: s4cVR3

26

.

State

 

Fig. 11:--> Emit lower envelope: s4cVR6

6

.

Level

 

 

27

.

State

 

Fig. 12:--> Output weight: s4cWP7

28

.

State

 

Fig. 12:--> Boundary conditions to determine of the weight: s4cWP0

29

.

State

 

Fig. 12:--> Period about prediction: e4cWP1

30

.

State

 

Fig. 12:--> Period under prediction: e4cWP2

31

.

State

 

Fig. 12:--> Wrong period: e4cWP3

7

.

Level

 

 

32

.

State

 

Fig. 13:--> Phase of transformation channel ny. h1 spend: s5cQ19

33

.

State

 

Fig. 13:--> Constraints to determine the phase: s5cQ10

34

.

State

 

Fig. 13:--> Positive sinus valuated share of the spectral line: s5cQ11

35

.

State

 

Fig. 13:--> Positive cosinus valuated share of the spectral line: s5cQ12

36

.

State

 

Fig. 13:--> Negative sine valuated share of the spectral line: s5cQ13

37

.

State

 

Fig. 13:--> Negative cosinus valuated share of the spectral line: s5cQ14

Procedures (-events)

1

.

Level

 

 

1

.

Event

 

Fig. 2:--> Eliminating reset signal: e3cRS2

2

.

Event

 

Fig. 2:--> Second sampling: e3cFS2

3

.

Event

 

Fig. 2:--> Unchanged reference signal: e3cNO1

4

.

Event

 

Fig. 3:--> The left half-vibration of the reference signal above the reference signal processing: e3cHL1

5

.

Event

 

Fig. 4:--> The right half-vibration of the reference signal below the reference signal processing: e3cHR1

6

.

Event

 

Fig. 2:--> Changing reference signal: e3cNO3

7

.

Event

 

Fig. 2:--> External reset : e3cNO4

8

.

Event

 

Fig. 3:--> At the end of the left half-vibration of the reference signal: e3cEL1

9

.

Event

 

Fig. 2:--> External reset: e3cHL3

10

.

Event

 

Fig. 4:--> End the right half-vibration of the reference signal: e3cER1

11

.

Event

 

Fig. 2:--> External reset: e3cHR3

12

.

Event

 

Fig. 5:--> Output of characteristics after the left half-vibration of the reference signal: e3cFL1

13

.

Event

 

Fig. 4:--> Increase of proceedings progress meter on the left side: e3cPL1

14

.

Event

 

Fig. 6:--> Output of characteristics after the right half-vibration of the reference signal: e3cFR1

15

.

Event

 

Fig. 3:--> Increase of proceedings progress meter on the right: e3cPR1

16

.

Event

 

Fig. 7:--> Small change to the reference to the reference signal c. z_0 linksseitig: e3cCL1

17

.

Event

 

Fig. 7:--> Linksseitiger jump of the reference to the reference signal c. z_0 down: e3cJL1

18

.

Event

 

Fig. 8:--> Linksseitiger jump of the reference to the reference signal c. z_0 upwards: e3cJH2

19

.

Event

 

Fig. 8:--> Small change to the reference to the reference signal c. z_0 rechtsseitig: e3cCR1

20

.

Event

 

Fig. 8:--> Rechtsseitiger jump of the reference to the reference signal c. z_0 upwards: e3cJH1

21

.

Event

 

Fig. 7:--> Rechtsseitiger jump of the reference to the reference signal c. z_0 down: e3cJL2

22

.

Event

 

Fig. 7:--> Preparing to move to the right:e3cJL5

23

.

Event

 

Fig. 7:--> External reset: e3cJL4

24

.

Event

 

Fig. 7: Left--> preparing to go to the page:e3cJH5

25

.

Event

 

Fig. 8:--> External reset: e3cJH4

2

.

Level

 

 

26

.

Event

 

Fig. 1:--> Provide the old values: e2cCA2

3

.

Level

 

 

27

.

Event

 

Fig. 9:--> Set of memory provide the procedure fortschrittes: e1cSP2

28

.

Event

 

Fig. 9:--> Reset of proceedings progress meter: e1cSP1

4

.

Level

 

 

29

.

Event

 

Fig. 10:--> Left await the next sample stage: e4cVL2

30

.

Event

 

Fig. 10:--> Maximum left accept: e4cVL4

31

.

Event

 

Fig. 10:--> Issue top envelope: e4cVL5

32

.

Event

 

Fig. 10:--> Upper envelope again update: e4cVL7

5

.

Level

 

 

33

.

Event

 

Fig. 11:--> Right await the next sample stage: e4cVR2

34

.

Event

 

Fig. 11:--> Minimum right accept: e4cVR4

35

.

Event

 

Fig. 11:--> Emit lower envelope: e4cVR5

36

.

Event

 

Fig. 11:--> Lower envelope again update: e4cVR7

6

.

Level

 

 

37

.

Event

 

Fig. 12:-->Synchronize: e4cWP7

38

.

Event

 

Fig. 12:--> Period about prediction: e4cWP1

39

.

Event

 

Fig. 12:--> Period under prediction: e4cWP2

40

.

Event

 

Fig. 12:--> Wrong period: e4cWP3

41

.

Event

 

Fig. 12:--> Synchronize output after survival prediction: e4cWP4

42

.

Event

 

Fig. 12:--> Output after synchronizing, prediction: e4cWP5

43

.

Event

 

Fig. 12:--> Synchronize by incorrect period: e4cWP6

7

.

Level

 

 

44

.

Event

 

Fig. 13:--> Phase calculation synchronize: e5cQ10

45

.

Event

 

Fig. 13:--> Positive sinus valuated share of the spectral line: e5cQ11

46

.

Event

 

Fig. 13:--> Positive cosinus valuated share of the spectral line: e5cQ12

47

.

Event

 

Fig. 13:--> Negative sine valuated share of the spectral line: e5cQ13

48

.

Event

 

Fig. 13:--> Negative cosinus valuated share of the spectral line: e5cQ14

49

.

Event

 

Fig. 13:--> Synchronize output of phase after positive sinus bewertetem share of the spectral line: e5cQ15

50

.

Event

 

Fig. 13:--> Synchronize output of phase after positive cosinus bewertetem share of the spectral line: e5cQ16

51

.

Event

 

Fig. 13:--> Synchronize output of phase after negative sine bewertetem share of the spectral line: e5cQ17

52

.

Event

 

Fig. 13:--> Synchronize output of phase after negative cosinus bewertetem share of the spectral line: e5cQ18

Input sizes of procedure

1

.

Input size

 

Fig. 2: --> e3cRS2 External reset signal: b.reset

2

.

Input size

 

Fig. 2: s3cFS1--> analysierendes signal: c.y

3

.

Input size

 

Fig. 2: --> s3cFS3 External sampling step-over: c.Dx

4

.

Input size

 

Fig. 2: --> s3cFS3 Of current Abszisse value: c.x

5

.

Input size

 

Fig. 12: --> s4cWP7 Atomic number: ny. h1

6

.

Input size

 

Fig. 2: --> e3cNO1Reference signal: c.z

7

.

Input size

 

Fig. 2: --> e3cNO1 Reference to the reference signal: c. z_0

8

.

Input size

 

Fig. 9: s1cSP3--> number of half vibrations:k.character

9

.

Input size

 

Fig. 12: s4cWP7--> correction of prediction of arithmetic mean:kPdV. y_0

10

.

Input size

 

Fig. 12: s4cWP7--> correction of forecast period duration:kPdV.T

11

.

Input size

 

Fig. 14:--> ContinuedOutputs Secondary abscissa: c.Bx

12

.

Input size

 

Fig. 5: --> s3cFL2 Synchronization pulse length: k.synch

Memory locations of the procedure

1

.

Memory cell

 

Fig. 2: s3cFS1--> age discrete average: od. y_0

2

.

Memory cell

 

Fig. 2: s3cFS1--> predictive value of discrete mean:PD. y_0

3

.

Memory cell

 

Fig. 1: s2cCA1--> old value of the signal to analyze:OC.y

4

.

Memory cell

 

Fig. 2: --> s3cFS3 Of split sampling step: cPoS.Dx

5

.

Memory cell

 

Fig. 2: --> s3cFS3 Abszisse value of split sampling step: c.Cx

6

.

Memory cell

 

Fig. 2: --> s3cFS3 Trapezoidal surface on the split sampling step up to the analyte signal: aPoS. y_0

7

.

Memory cell

 

Fig. 2: --> s3cFS3 Trapezoidal surface on the split sampling step up to the square of the signal to analyze: aPoS. y_r

8

.

Memory cell

 

Fig. 2: --> s3cFS3 Trapezoidal surface on the split sampling step up cosinus valued part of the signal to analyze: aPoS. y_1a

9

.

Memory cell

 

Fig. 2: --> s3cFS3 Trapezoidal surface on the split sampling step up sinus valued part of the signal to analyze: aPoS. y_1b

10

.

Memory cell

 

Fig. 2: --> s3cFS3 Trapezoidal surface on the split sampling step up to the with ny. h1 corrected cosinus valued part of the signal to analyze: aPoS. y_h1a

11

.

Memory cell

 

Fig. 2: --> s3cFS3 Trapezoidal surface on the split sampling step up to the with ny. h1 corrected sinus valued part of the signal to analyze: aPoS. y_h1b

12

.

Memory cell

 

Fig. 3: s3cHL2--> beginning left timing:cL.Bx

13

.

Memory cell

 

Fig. 4: s3cHR2--> beginning right timing:cR.Bx

14

.

Memory cell

 

Fig. 3: s3cHL2--> total linear surface elements on the left:sLa. y_0

15

.

Memory cell

 

Fig. 4: s3cHR2--> total linear surface elements on the right:sRa. y_0

16

.

Memory cell

 

Fig. 3: s3cHL2--> sum of the absolute linear surface elements on the left:sLa. y_0R

17

.

Memory cell

 

Fig. 4: s3cHR2--> sum of the absolute linear surface elements on the right:sRa. y_0R

18

.

Memory cell

 

Fig. 3: s3cHL2--> sum of square surface elements on the left:sLa. y_r

19

.

Memory cell

 

Fig. 4: s3cHR2--> sum of square surface elements on the right:sRa. y_r

20

.

Memory cell

 

Fig. 3: s3cHL2--> sum cosinus-weighted surface elements on the left:sLa. y_1a

21

.

Memory cell

 

Fig. 4: s3cHR2--> sum cosinus-weighted surface elements on the right:sRa. y_1a

22

.

Memory cell

 

Fig. 3: s3cHL2--> sum sine-weighted surface elements on the left:sLa. y_1b

23

.

Memory cell

 

Fig. 4: s3cHR2--> sum sine-weighted surface elements on the right:sRa. y_1b

24

.

Memory cell

 

Fig. 3: s3cHL2 sum--> the withny. h1 corrected cosinus-weighted surface elements on the left: sLa. y_h1a

25

.

Memory cell

 

Fig. 4: s3cHR2 sum--> the withny. h1 corrected cosinus-weighted surface elements on the right: sRa. y_h1a

26

.

Memory cell

 

Fig. 3: s3cHL2 sum--> the withny. h1 corrected sinus-weighted surface elements on the left: sLa. y_h1b

27

.

Memory cell

 

Fig. 2: --> s3cFS3 Trapezoidal surface on the split sampling step up to the with ny. h1 corrected sinus valued part of the signal to analyze: aPoS. y_h1b

28

.

Memory cell

 

Fig. 1: --> s2cCA3 Trapezoidal surface under the signal to analyze: a. y_0

29

.

Memory cell

 

Fig. 1: --> s2cCA3 Trapezoidal surface to the square of the signal to analyze: a. y_r

30

.

Memory cell

 

Fig. 1: --> s2cCA3 Trapezoidal area until cosinus valued part of the signal to analyze: a. y_1a

31

.

Memory cell

 

Fig. 1: --> s2cCA3 Trapezoidal area until the sinus valued part of the signal to analyze: a. y_1b

32

.

Memory cell

 

Fig. 1: --> s2cCA3 Trapezoidal surface up to the with NY. h1 corrected cosinus valued part of the signal to analyze: a. y_h1a

33

.

Memory cell

 

Fig. 1: --> s2cCA3 Trapezoidal surface up to the with NY. h1 corrected sinus valued part of the signal to analyze: a. y_h1b

34

.

Memory cell

 

Fig. 3: --> s3cEL2 End left timing: cL.Ex

35

.

Memory cell

 

Fig. 3: --> s3cEL2 Of discrete left era dL.T

36

.

Memory cell

 

Fig. 4: --> s3cER2 End right timing: cR.Ex

37

.

Memory cell

 

Fig. 4: --> s3cER2 Of discrete left era dR.T

38

.

Memory cell

 

Fig. 3: s3cEL2--> function value of the signal to be analyzed at the intersection of the reference signals:cPoS.y

39

.

Memory cell

 

Fig. 5: --> e3cFL1 Counter of procedure fortschrittes: i.prog

40

.

Memory cell

 

Fig. 5: s3cFL2--> old discrete period duration:Od.T

41

.

Memory cell

 

Fig. 5: s3cFL2--> discrete period duration: d.T

42

.

Memory cell

 

Fig. 1: s2cCA1--> old value of the reference signal:OC.z

43

.

Memory cell

 

Fig. 9: --> s1cSP1 Memory provide the procedure fortschrittes: m.prog

44

.

Memory cell

 

Fig. 1: s2cCA1--> cache of the reference signal:OOC.z

45

.

Memory cell

 

Fig. 1: s2cCA1--> cache of the reference to the reference signal:OOC. z_0

46

.

Memory cell

 

Fig. 1: s2cCA1--> old value of the reference to the reference signal:OC. z_0

47

.

Memory cell

 

Fig. 1: s2cCA1--> second cache of signal to analyze:oooc.y

48

.

Memory cell

 

Fig. 1: s2cCA1--> first cache of signal to analyze:OOC.y

49

.

Memory cell

 

Fig. 1: s2cCA1--> count of samples: i.CA

50

.

Memory cell

 

Fig. 1: s2cCA1--> old square of the signal to analyze:oc. y_r

51

.

Memory cell

 

Fig. 1: s2cCA1--> age cosinus valued part of the signal to analyze:OC. y_1a

52

.

Memory cell

 

Fig. 1: s2cCA1--> age sinus valued part of the signal to analyze:OC. y_1b

53

.

Memory cell

 

Fig. 1: s2cCA1--> age withny. h1 korrigierter cosinus valued part of the signal to analyze: OC. y_h1a

54

.

Memory cell

 

Fig. 1: s2cCA1--> age withny. h1 korrigierter sinus valued part of the signal to analyze: OC. y_h1b

55

.

Memory cell

 

Fig. 1: s2cCA3--> square of the signal to analyze:c. y_r

56

.

Memory cell

 

Fig. 1: s2cCA3--> cosine-valued part of the signal to analyze:c. y_1a

57

.

Memory cell

 

Fig. 1: s2cCA3--> sine-valued part of the signal to analyze:c. y_1b

58

.

Memory cell

 

Fig. 1: s2cCA3--> withny. h1 korrigierter cosinus valued part of the signal to analyze: (c). y_h1a

59

.

Memory cell

 

Fig. 1: s2cCA3--> withny. h1 korrigierter sinus valued part of the signal to analyze: (c). y_h1b

60

.

Memory cell

 

Fig. 10: --> s4cVL1 Maximum: c. y_h

61

.

Memory cell

 

Fig. 11: --> s4cVR1 Minimum: c. y_l

62

.

Memory cell

 

Fig. 12: s4cWP7--> increase of arithmetic mean:GD. y_0

63

.

Memory cell

 

Fig. 12: s4cWP7--> predictive value of arithmetic mean:PD.T

64

.

Memory cell

 

Fig. 12: s4cWP7--> predictive value of frequency:PD.f

65

.

Memory cell

 

Fig. 12: s4cWP7--> predictive value of Kreisfrequenz:PD.O

66

.

Memory cell

 

Fig. 12: s4cWP7--> predictive value of Kreisfrequenz for the transformation channelny. h1: pd. O_h1

Output quantities of procedure

1

.

Original size

 

Fig. 2: --> s3cRS1 Of inverse synchronization Pulse: bj.synch

2

.

Original size

 

Fig. 2: --> s3cRS1 Linker inverse synchronization Pulse: bLj.synch

3

.

Original size

 

Fig. 2: --> s3cRS1 Rechter inverse synchronization Pulse: bRj.synch

4

.

Original size

 

Fig. 2: --> s3cFS3 Of discrete average: d. y_0

5

.

Original size

 

Fig. 3: s3cEL2--> event stamp of the left:eL.stamp

6

.

Original size

 

Fig. 4: s3cER2--> event stamp of the right:eR.stamp

7

.

Original size

 

Fig. 5: s3cFL2 cosine-valued part--> the inny. h1 given spectral line: d. y_h1a

8

.

Original size

 

Fig. 5: s3cFL2 sine-valued part--> the inny. h1 given spectral line: d. y_h1b

9

.

Original size

 

Fig. 5: s3cFL2 pattern of cosinus-weighted part--> the inny. h1 given spectral line: d. y_h1ap

10

.

Original size

 

Fig. 10: s4cVL6--> distance of envelopes: m. y_nm

11

.

Original size

 

Fig. 5: s3cFL2 pattern of sine-weighted part--> the inny. h1 given spectral line: d. y_h1bp

12

.

Original size

 

Fig. 3: s1cSP1--> event stamp:e.stamp

13

.

Original size

 

Fig. 10: --> s4cVL6 Upper envelope : m. y_m

14

.

Original size

 

Fig. 11: --> s4cVR6 Lower envelope: m. y_n

15

.

Original size

 

Fig. 12: s4cWP7--> weight of determined characteristic values:w.y

16

.

Original size

 

Fig. 12: s4cWP7--> part weight of determined characteristic values related to the period duration:w.T

17

.

Original size

 

Fig. 12: s4cWP7--> part weight of determined characteristic values related to the arithmetic mean:w. y_0

18

.

Original size

 

Fig. 12: --> s4cWP1 Discrete frequency: d.f

19

.

Original size

 

Fig. 13: s5cQ19--> degree of phase of transformation channelny. h1: d. y_h1d

20

.

Original size

 

Fig. 13: s5cQ11--> phase of transformation channelny. h1: d. y_h1p

21

.

Original size

 

Fig. 14: ContinuedOutputs--> continuous average:c. y_0

22

.

Original size

 

Fig. 14: ContinuedOutputs--> continuous change share:c. y_y0

23

.

Original size

 

Fig. 14: ContinuedOutputs--> continuous retransformed of basic vibration:c. y_1

24

.

Original size

 

Fig. 5: s3cFL2--> RMS of basic vibration:d. y_1c

25

.

Original size

 

Fig. 14: ContinuedOutputs--> continuous upper vibration unit:c. y_y01

26

.

Original size

 

Fig. 14: ContinuedOutputs--> continuous retransformed of upper vibration from the transformation channel NY. h1 : c. y_h1

27

.

Original size

 

Fig. 5: s3cFL2--> RMS of upper vibration from the transformation channelny. h1: d. y_h1c

28

.

Original size

 

Fig. 14: ContinuedOutputs--> continuous vibration unit:c. y_y0h1

29

.

Original size

 

Fig. 14: ContinuedOutputs--> continuous partial signal:c. y_y01h1

 


top     (8) References
 

Use graph networks

[F1. 01] Hiroaki SAKOE: System for recognizing a word sequence by dynamic programming and by the use of a state transition slide gram; US 4 326 101 of the 06.08.1980

[F1. 02] TAKAI et al.: apparatus and method of supporting functional design of logic circuit and apparatus and method of verifying functional design of logic circuit; US 5 751 592 of the 27.12.1995

[F1. 03] FUKAZAWA et al.: LSI design support system; US 5 371 683 of the 03.10.1991

Use reference signal

[F1. 04] Hans WERBA: analog/digital converter; US 4 584 565 of the 21.03.1985

[F1. 05] Avery WANG: Formant-and frequency verriegelte basic frequency sequence circuit and system to separate sounds; EP 0 803 116 of the 11.01.1996

[F1. 06] Walter David FLYNN: digital to analogue converter with reference signal; EP 0 897 617 of the 07.05.1996

[F1. 07] REBEL Reimund: process to match a band pass and circuit order to perform the procedure; DE 199 04 604 of the 05.02.1999

[F1. 08] KAWAHARA et al.: method and apparatus for fundamentally frequency extraction or detection in speech; US 7 085 721 of the 05.07.2000

[F1. 09] Wolfgang KERNCHEN: measuring system with a reference signal between a signal generator and a signal Analyzer; EP 1 608 089 of the 26.06.2001

Identify process events

[F1.10]ERYUREK et al.: device in a progress system for detecting events; US 6 017 143 of the 28.03.1996

[F1. 11] Tommaso MAJO: event counter; EP 0 990 993 of the 28.07.1998

 

[F1.12]Limin SONG: System and methodology for vibration analysis and condition monitoring; US 7 133 801 of the 12.01.2005

Signalbreaker

[F1. 13] RAJASEKARAN et al.: method of encoding speech signal involving the extraction of speech formant candidates in real time; US 4 922 539 of the 26.01.1989

 

[F1. 14] JIN et al.: interactive adaptive filter and interactive adaptive filtering method thereof; US 7 133 886 from the 06.06.2003

The language encoding prediction matrix

[F3. 01] LAFLAMME, Claude et al.: quantization split prediction matrix with spectral parameters to effective language encoding; EP 819 303 of the 02.04.1996

Signal analysis

[F5. 01] WANG et al.: machine fault diagnostic system at method; US 5 566 092 of the 30.12.1993

[F5. 02] E. VLIEMBERGEN: equipment and procedures for syntactic signal analysis; DE 693 31 044 of the 19.02.1993

[F5. 03] Hldeki KAWAHARA: method and apparatus for signal analysis; EP 0 853 309 of the 14.01.1997

[F5. 04] ROBINSON et al.: machine fault detection using vibration signal peak detector;US 5 895 857 of the 17.04.1997

[F5. 05] F. MARKS: digital signal processing methods for the determination of electrical parameters in AC networks; en 197 26 988 of the 25.06.1997

[F5. 06] John N HOLMES: speech processing system using formant analysis ; US 6 292 775 of the 13.10.1997

[F5. 07] Andreas EISELE: method for the control of a non-harmonic network current portion of a converter, and device for carrying out the method; EP 1 017 157 of the 13.12.1999

Period length measurement

[F6. 01] JACOB, Christian: procedure OS 195 period duration and wavelength measurement of dominant oscillations and waves, and Spektralanalyse oscillations and waves; 20 836 of the 31.05.1995


top     (8) Bibliography
 

[L1. 01] Christian E. JACOB: the interdisciplinary simulation of continuous and discontinuous technical systems; Special Edition SAMS; Sep. 1994

[L1. 02] Wolfgang REISIG: Petri nets - an introduction (Paperback), Springer-Verlag Berlin, 1986

[L1. 03] anonymous: SIMPLORER user's Guide; company SIMEC GmbH & CoKG, output 01.12.1999 simulation system

[L1. 04] Hans EVEKING: verification of digital systems ? an introduction in the correct digital systems design; in guides and monographs of computer science, B.G.Teubner Stuttgart 1991

[L1. 05] Claude Elwood SHANNON: communication in the presence of noise; proc. IRE, vol. 37, No. 1 (Jan.) 1949; emphasis in proc. IEEE, vol. 86, No. 2 (Feb) 1998 S. 447 457

[L1. 06] Daniel ch. of GRÜNIGEN: digital signal processing; Carl Hanser 2001

[L1. 07] Bernard PICINBONO: principles of signal and systems: deterministic signal; Artech House 1988

[L1. 08] Bulgari Kim boundary: slice method for the analysis of stochastic signals by means of mathematical statistics, dissertation TU Chemnitz Zwickau 1994

[L1. 09] Mike RETERS: Psychoakustische signal enhancement and noise reduction in motor vehicles; dissertation University Kaiserslautern 2002

[L2. 01] Christian E. JACOB: own simulation with event-oriented analysis, calculation and control models; progress in simulation technology band 6; 8.Symposium ASIM in Berlin 1993

[L5. 01] Lawrence S. MARPLE jr: digital spectral analysis with applications; Prentice-Hall Inc. 1987

[L5. 02] Manfred UNÜBERWACHTES, Eckhard HUBER: tele school II physics ? vibration and wave; TR publishing Union München 1987

[L5. 03] Christian BIEN MÜLLER: assisted Fourier analysis and synthesis in the physics internship; housework to degree 1993