EP0865646B1 - Method for analyzing the signals of a danger alarm system and danger alarm system for implementing said method - Google Patents

Abstract

The method involves a rapid wavelet transformation and fuzzy logic. The original signal is passed through a cascade of high/low pass filter pairs during wavelet transformation. The rapid wavelet transformation is combined with a fuzzy logic evaluation, whereby a corresp. function is generated from the resultant of the high pass filter for each filter stage. The corresp. function is used for the further analysis of the frequency signal using fuzzy logic rules.

Description

The present invention relates to a method for analyzing the signal of a hazard detector using frequency analysis and fuzzy logic evaluation, as well as a hazard detector to carry out this procedure. The hazard detector can, for example a flame detector, noise detector, fire detector, passive infrared detector or the like.

The output signals from hazard detectors are often due to their typical frequency spectra characterized. By analyzing these frequency spectra, the origin of the Signals are determined, and above all, real alarm signals from interference signals differentiate and avoid false alarms. Especially with flame detectors the typical low frequency flickering of a flame is analyzed to determine the Radiation from real flames from that of a source of interference, such as reflected Sunlight, or a flickering light source.

The output signals from hazard detectors are analyzed, for example, using Fourier analysis, Fast Fourier analysis, zero crossing method or turning point method analyzed. The latter is used in GB-A 2 277 989 for flame detectors described, the time spans between radiation maxima measured and on their Regularities and irregularities checked and irregular ones Radiation maxima can be interpreted as a flame and regularly as a disturbance.

Fuzzy logic is well known. With regard to the present invention, it should be emphasized that that signal values according to so-called fuzzy sets, or unsharp quantities be assigned to a membership function, the value of the membership function, or the degree of belonging to a fuzzy set, between zero and is one. It is important that the membership function can be normalized, i.e. the sum of all values of the membership function is equal to one, which results in the fuzzy logic evaluation allows a clear interpretation of the signal.

In a flame detector described in EP-A 0 718 814, the frequency of the detected radiation and analyzed between regular and irregular Differentiated signals in certain frequency ranges. The evaluation of the different Signals in the given frequency ranges are based on several fuzzy logic rules. This procedure enables a more precise distinction between real flame signals and other interference signals and thus false alarm security allows. The frequency spectrum is generated here, for example, by fast Fourier transform, which is what is required for the transformation Time, the necessary processor and the processor costs is expensive. For the Determination of a detected signal can take up to three seconds. For certain applications, however, a shorter evaluation time and response time is up to desired for alarming, using methods such as the zero crossing or turning point method or wavelet analysis accelerate the decision-making process, but are less accurate.

The object of the invention is a method for frequency analysis of a signal create a hazard detector that is combined with a fuzzy logic evaluation, and with a smaller number compared to prior art analysis methods is carried out by arithmetic steps, so that a result of same or higher accuracy is achieved. Furthermore, the method is intended with a simpler processor and therefore less expensive to carry out.

According to the invention, the object is achieved in that the original signal in the fast wavelet transform a multi-stage filter cascade of high / low pass filter pairs is carried out, and that at each filter stage of the wavelet transformation from the results of the high-pass filter a membership function is generated for further analysis of the frequency signal is used according to fuzzy logic rules.

The wavelet transform is a transformation or mapping of a signal from the Time domain in the frequency domain (see, for example, "The Fast Wavelet Transform" by Mac A. Cody in Dr. Dobb's Journal, April 1992); so it is fundamental similar to the Fourier transform and Fast Fourier transform. It makes a difference differ from these by the basic function of the transformation, according to which the Signal is developed. In a Fourier transform, there is a sine and cosine function used, which is localized sharply in the frequency domain and indefinitely in the time domain is. With a wavelet transformation, a so-called wavelet or wave packet is used used. There are different types of this, such as a Gaussian, Spline or Haar wavelet, each with two parameters in the time domain can be shifted and stretched or compressed in the frequency range.

A wavelet transform can therefore be used in both the time and frequency domain localized signals are transformed. A fast wavelet transformation is carried out by the Mallat pyramid algorithm, which is used repeatedly a low-pass and high-pass filter, through which the low-frequency be separated from the high-frequency signal components. In each case, the Output signal of the low-pass filter in turn fed to a pair of low / high-pass filters. A series of approximations of the original signal results, each of which has a coarser resolution than the previous one. The number of operations for The transformation required is proportional to the length of the original Signals, while in the Fourier transform this number is disproportionate to the signal length. The fast wavelet transformation can also be carried out inversely by dividing the original signal from the approximated values and coefficients is restored for reconstruction. The decomposition algorithm and reconstruction of the signal and a table of the coefficients of the decomposition and Reconstruction using the example of a spline wavelet in "An Introduction to Wavelets" by Charles K. Chui (Academic Press, San Diego, 1992).

When used in a hazard detector, the results of the fuzzy evaluation allow a decision as to whether there is an alarm or an interference signal. The number of for The wavelet analysis required arithmetic steps compared to Fourier analyzes significantly reduced. This is the necessary computer time to identify the Signals are shortened and the processor costs are reduced.

According to the invention, the original digitized signal is initially replaced by a fast wavelet transformation analyzed. For this, the signal is based on the algorithm from Mallat through several stages of a cascade of high and low pass filter pairs guided. The results of the high-pass filter then become one at each filter stage Affiliation function that generates the sum of the calculated values from the High pass filter contains and by the sum of the squares of the original signal values is divided. The sum of membership functions here at each filter level generated is equal to or almost equal to one. These normalized membership functions are then in this form for a continuation of the frequency analysis used with fuzzy logic.

A frequency analysis of this type gives the following advantages: The high-pass filter of the wavelet transform first provide information about the high-frequency signals. This is particularly advantageous in the flame report, since with the information about the higher Frequencies speeds up the identification of the type of signal and its accuracy can be increased. For example, a high-frequency signal of over 15 Hz discovered, this is interpreted as an interference signal. The subsequent message, interference signal or alarm signal, occurs earlier and is more certain to be correct. Wavelets are in often very simple in shape, such as a hair wavelet, and allow one Analysis with a few calculation steps, what the computing time and the decision time additionally shortened. The shortening of the decision time is not, however, with one Losses connected in the accuracy of the signal identification. Are fewer rows code is required, an inexpensive processor can also be used.

A first preferred embodiment of the method according to the invention is characterized in that that the wavelet used for the fast wavelet transformation orthonormal or semi-orthonormal wavelet or a wavelet packet basis is, and that the membership functions generated each by the wavelet coefficients weighted sum of the squared values of the high pass filter and the Sum of the squared values of the original signal included and in normalized Form used for further analysis of the frequency signal according to fuzzy logic rules become.

In a second preferred embodiment, this is for fast wavelet transformation Wavelet used an orthonormal or semi-orthonormal wavelet or contain a wavelet packet base and the membership functions generated the sum of the squared output values of the high-pass filter and the sum of the squared values of the original signal from the hazard detector and are given in normalized form for the evaluation of the frequency signal according to fuzzy logic rules used.

The hazard detector according to the invention for carrying out the method mentioned contains a sensor for a hazard parameter, evaluation electronics with means for processing the output signal of the sensor and a microprocessor with a Fuzzy controller. This hazard detector is characterized in that the microprocessor has a software program according to which the fuzzy controller is part of a Fuzzy wavelet controller is, and that that processed by the evaluation electronics and the signal supplied to the fuzzy controller is wavelet-transformed.

Fig. 1

ein Blockschema eines Verfahrens mit einer schnellen Wavelet-Analyse durch mehrere Filterstufen und Weiteranalyse durch Fuzzy-Logik,

Fig. 2

Darstellungen von Zugehörigkeitsfunktionen am Beispiel einer Frequenzanalyse mittels einer schnellen Haar-Wavelet-Transformation,

Fig. 3

ein Blockschema eines Gefahrenmelders zur Durchführung des Verfahrens von Fig. 1; und

Fig. 4

ein Blockschema für die Implementierung des Verfahrens von Fig. 1 in einem Gefahrenmelder.

The invention is explained in more detail below on the basis of an exemplary embodiment illustrated in the drawings; it shows:

Fig. 1

1 shows a block diagram of a method with a fast wavelet analysis using several filter stages and further analysis using fuzzy logic,

Fig. 2

Representation of membership functions using the example of a frequency analysis using a fast Haar-Wavelet transformation,

Fig. 3

a block diagram of a hazard detector for performing the method of Fig. 1; and

Fig. 4

a block diagram for the implementation of the method of FIG. 1 in a hazard detector.

und die Ausgangswerte eines Tiefpassfilters durch

ausgedrückt.1, the output signal x _{0, k is} first used to carry out a fast wavelet transformation 1 by means of any wavelet of the type known from the prior art. An orthonormal or semi-orthonormal wavelet or a wavelet packet base is preferably used. In the figure, the signal values are denoted by x _{i, k} and y _{i, k} , where x is the original signal values and the values from the low-pass filters (LP) and y are the values from the high-pass filters (HP). The index i denotes the level of the filter cascade in increasing numbers, the original signal being at level zero. The index k denotes an individual value of a signal. An original signal x _{0, k} at zero level is assumed, which is transformed by several filterings. The output signal of the first high-pass filter gives the values y _{1, k} and the output signal of the first low-pass filter, which also forms the input signal for the second filter stage, gives the values x _{1, k} . The output signal of the second high-pass filter gives the values y _{2, k} , that of the second low-pass filter x _{2, k} is fed to a third filter pair, etc. It should be noted here that the number of values which result from the filter stages in each stage is different. More specifically, the number of values decreases by a factor of two at each level. At stage i + 1, for example, the output values of a high-pass filter are checked

and the output values of a low-pass filter

expressed.

.The coefficients a and b for the transformation are generally known and can be calculated using the Chui book mentioned. For example, for a Haar wavelet, a ₀ = a ₁ = 1/2, b ₀ = 1/2 and b ₁ = -1 / 2. Index 1 takes integer values for which the coefficients are not equal to zero. The original signal is reconstructed in stages by creating the values of each filter stage from the values of the previous stage, namely

,

The coefficients p and q for the wavelet reconstruction are in the already mentioned To find book.

und

wobei N die Anzahl der Filterstufen ist. Die letztere Funktion µ_N+1 wird also durch die Ausgangswerte des letzten Tiefpassfilters gebildet. Diese Zugehörigkeitsfunktionen sind normalisiert, indem

.The membership functions µ _{i are then} generated from the output values of the high-pass filter of the respective filter stage and the associated coefficients q for the wavelet reconstruction. It is

and

where N is the number of filter stages. The latter function µ _{N + 1} is thus formed by the output values of the last low-pass filter. These membership functions are normalized by

,

und

An often good approximation of these membership functions is given by the following equation:

and

.With this approach, the function is almost normalized by

,

und

In a special embodiment of the method, the digitized raw values x ₀ , k are subjected to a quick hair analysis. Membership functions µ _{i are} formed from the values y _{i, k of} each filter stage i, namely:

and

ist. In this case, these membership functions are normalized by

is.

In FIG. 2, membership functions μ, which have been generated from the results of a fast Haar wavelet transformation, are shown as a function of the frequency. Of the various curves, µ _{N + 1} illustrate the degree of affiliation of very low frequencies, µ _N that of low frequencies, and µ ₁ and µ ₂ the degree of affiliation of high and medium frequencies. It can be clearly seen here that the sum of the curve values is one for each selected frequency.

In all of the executions of the method, these membership functions become one Fuzzy logic controller 2 (FIG. 1) for evaluation according to fuzzy logic rules, whereupon a decision is made as to whether an alarm signal is triggered or that Signal is evaluated as a disturbance.

This method is suitable for differentiation when used in flame detectors between interference signals, such as periodic signals of over 15 Hz, and real flame signals, such as narrow-band, low-frequency signals or broadband signals in the low frequency range. Because of the fast Identification of high frequency signals are the interference signals of this frequency and whose resonance frequencies are eliminated from the signal, which the frequency analysis of the signal accelerated. By accelerating frequency analysis through the wavelet transformation can be the time required for a decision on the type of signal and the message to be submitted has been reduced from three seconds to one second, for example become. The method described is also passive for noise detectors Infrared detector, for the spectral analysis of the signals of individual pixels in image processing as well as for various sensors such as gas and vibration sensors.

Figure 3 shows a diagram of a hazard detector 3 for performing the described Process. According to the illustration, the hazard detector 3 has a sensor 4 for detection a hazard parameter, evaluation electronics 5, a microprocessor 6 and the fuzzy controller 2. The hazard parameter can be, for example, the intensity the radiation emitted by a flame, the acoustic signal of a noise, the infrared radiation emitted by a warm body or the output signal a CCD camera.

The output signal of the sensor 4 is fed to the evaluation electronics 5, which has suitable means for processing the signal, such as amplifiers, and passes from the evaluation electronics 5 into the microprocessor 6. The fuzzy controller 2 (FIG. 1) is integrated here as software in the microprocessor 6. In particular, the Fuzzy controller Part of a fuzzy wavelet controller based on fuzzy logic theory linked to the wavelet theory. The microprocessor 6 contains, for example Software program of the type shown in Figure 4, which the input signal of a Undergoes wavelet transformation. The resulting transformed signal is then fed to the fuzzy controller 2. Should that result from the fuzzy controller 2 If the signal is evaluated as an alarm, this will be an alarm delivery device 7 or an alarm center.

FIG. 4 shows a block diagram for the implementation of the method according to the invention in the microprocessor of a hazard detector, this microprocessor having a fuzzy wavelet controller 8. After evaluation by the evaluation electronics 5 (FIG. 3), the output signal of the sensor 4 is fed to the fuzzy wavelet controller 8, in which the signal is first passed through a cascade of filters 9. The membership functions μ _{i are} formed from the results 10 of each filter 9 according to equation 1. These functions are then fed to the fuzzy controller 2 for fuzzy analysis, which may send a signal to the alarm output device 7.

Claims (5)

1. A method of analyzing the signal of a hazard detector (3) by frequency analysis and fuzzy logic analysis, wherein a fast wavelet transformation (1) is performed as frequency analysis, characterised in that the original signal (x_0,k) in the fast wavelet transformation is conducted through a multi-stage filter cascade of pairs of high-pass/low-pass filters (HP, LP), and that in each filter stage of the wavelet transformation an association function (µ_i) is in each case produced from the results of the high-pass filter (HP), which association function (µ_i) is used for the further analysis of the frequency signal in accordance with fuzzy logic rules.
2. A method according to Claim 1, characterised in that the wavelet used for the fast wavelet transformation (1) is an orthonormal or semi-orthonormal wavelet or a wavelet packet base, and that the produced association functions (µ_i) in each case contain the sum, weighted by the wavelet coefficients, of the squared values of the high-pass filter (HP) and the sum of the squared values of the original signal (x_0,k) of the hazard detector (3) and are used in normalised form for the further analysis of the frequency signal in accordance with fuzzy logic rules.
3. A method according to Claim 1, characterised in that the wavelet used for the fast wavelet transformation (1) is an orthonormal or semi-orthonormal wavelet or a wavelet packet base, and that the produced association functions(µ_i) in each case contain the sum of the squared output values of the high-pass filter (HP) and the sum of the squared values of the original signal (x_0,k) of the hazard detector (3) and are used in normalised form for the analysis of the frequency signal in accordance with fuzzy logic rules.
4. A method according to one of Claims 1 to 3, characterised in that the output signals are those of a flame detector and the frequency analysis and the analysis of the output signals of the flame detector has a duration of 100 ms to 10 s.
5. A hazard detector (3) for the implementation of the method according to one of Claims 1 to 3 with a sensor (4) for a hazard characteristic variable, an analysis electronics unit (5) with means for processing the output signal of the sensor (4) and a microprocessor (6) with a fuzzy controller (2), characterised in that the microprocessor (6) has a software program in accordance with which the fuzzy controller (2) is part of a fuzzy wavelet controller (8), and that the signal processed by the analysis electronics unit (5) and fed to the fuzzy controller (2) is wavelet-transformed.

Images