EP0865646A1 - Method for analyzing the signals of a danger alarm system and danger alarm system for implementing said method - Google Patents

Method for analyzing the signals of a danger alarm system and danger alarm system for implementing said method

Info

Publication number
EP0865646A1
EP0865646A1 EP97939930A EP97939930A EP0865646A1 EP 0865646 A1 EP0865646 A1 EP 0865646A1 EP 97939930 A EP97939930 A EP 97939930A EP 97939930 A EP97939930 A EP 97939930A EP 0865646 A1 EP0865646 A1 EP 0865646A1
Authority
EP
European Patent Office
Prior art keywords
wavelet
signal
evaluation
fuzzy
pass filter
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
EP97939930A
Other languages
German (de)
French (fr)
Other versions
EP0865646B1 (en
Inventor
Marc Pierre Thuillard
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Siemens AG
Original Assignee
Cerberus AG
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Cerberus AG filed Critical Cerberus AG
Priority to EP97939930A priority Critical patent/EP0865646B1/en
Publication of EP0865646A1 publication Critical patent/EP0865646A1/en
Application granted granted Critical
Publication of EP0865646B1 publication Critical patent/EP0865646B1/en
Anticipated expiration legal-status Critical
Expired - Lifetime legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B17/00—Fire alarms; Alarms responsive to explosion
    • G08B17/02—Mechanical actuation of the alarm, e.g. by the breaking of a wire
    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B29/00—Checking or monitoring of signalling or alarm systems; Prevention or correction of operating errors, e.g. preventing unauthorised operation
    • G08B29/18—Prevention or correction of operating errors
    • G08B29/185—Signal analysis techniques for reducing or preventing false alarms or for enhancing the reliability of the system
    • G08B29/186—Fuzzy logic; neural networks

Definitions

  • the present invention relates to a method for analyzing the signal emes a hazard detector using frequency analysis and fuzzy logic evaluation, and to a hazard detector for carrying out this method.
  • the hazard detector can be, for example, a flame detector, noise detector, fire detector, passive infrared detector or the like
  • the output signals from hazard detectors are often characterized by frequency spectra typical of them. By analyzing these frequency spectra, the origin of the signals can be determined and, above all, real alarm signals can be distinguished from fault signals and false alarms can be avoided in this way in order to be able to distinguish the radiation from real flames from that of a source of disturbance, such as reflected sunlight or a flickering light source
  • the output signals from hazard detectors are analyzed, for example, by means of Fou ⁇ er analysis, Fast-Fou ⁇ er analysis, Zero-Crossmg method or Turmng-Pomt method.
  • Fou ⁇ er analysis for example, by means of Fou ⁇ er analysis, Fast-Fou ⁇ er analysis, Zero-Crossmg method or Turmng-Pomt method.
  • the latter is described in GB-A 2 277 989 m using flame detectors, the time periods between Radiation maxi mia measured and checked for their regularities and irregularities and irregularly occurring radiation maxima are interpreted as a flame and regular ones as a disturbance
  • the Fn7.7y-T .og1k is generally known.
  • signal values so-called fuzzy sets, or fuzzy sets, are allocated in accordance with an association function, the value of the association function or the degree of association An unsharp amount, between zero and ems. It is important that the accessibility function can be normalized, ie the sum of all values of the membership function is equal to one, whereby the fuzzy logic evaluation allows a clear interpretation of the signal.
  • the frequency of the detected radiation is analyzed and a distinction is made between regular and irregular signals in certain frequency ranges.
  • the various signals in the given frequency ranges are evaluated according to several fuzzy logic rules. This procedure enables a more precise distinction between real flame signals and other interference signals and thus false alarm security.
  • the frequency spectrum is generated here, for example, by fast Fourier transformation, which is complex in terms of the time required for the transformation, the necessary processor and the processor costs. It sometimes takes up to three seconds to determine a detected signal. For certain applications, however, a shorter evaluation time and response time before the alarm is desired, although methods such as the zero crossing or turning point method or wavelet analysis speed up the decision-making process, but are less accurate.
  • the invention has for its object to provide a method for frequency analysis of a signal from a hazard detector, which is combined with a fuzzy logic evaluation, and is carried out in comparison with analysis methods of the prior art with a smaller number of computing steps, so that in a result of the same or higher accuracy is achieved in a shorter time. Furthermore, the method should be able to be carried out with a simpler processor and therefore more cost-effectively.
  • the object is achieved in that a rapid wavelet transformation is carried out as frequency analysis and the original signal is passed through a multi-stage filter cascade of high / low-pass filter pairs, and in that for each filter stage of the wavelet transformation a membership function is generated from the results of the high-pass filter is used for further analysis of the frequency signal according to fuzzy logic rules.
  • the wavelet transformation is a transformation or mapping of a signal from the time domain to the frequency domain (see, for example, "The Fast Wavelet-Transfo ⁇ n" by Mac A Cody in Dr Dobb's Journal, Ap ⁇ l 1992), so it is basically the Fou ⁇ er transformation and fast -Fou ⁇ er-Transformation similar It differs from these but by the basic function of the transformation, according to which the signal is developed.
  • Wavelet transformation is used in a so-called wavelet or wave packet.
  • wavelet or wave packet There are various types, such as a Gaussian, Sphne or Haar wavelet, which can be shifted in the time domain and stretched or compressed in the frequency domain using two parameters
  • localized signals can be transformed in the time as well as in the frequency domain by means of a wavelet transformation.
  • a fast wavelet transformation is carried out by the pyramid algorithm according to Mallat, which is based on repeated use of a low-pass and high-pass filter, by means of which the median frequencies of The high-frequency signal components are separated.
  • the output signal of the low-pass filter is in turn fed to a pair of high-pass / high-pass filters. This results in a number of approximations of the original signal, each of which has a higher resolution than the previous one. The number of operations required for the transformation.
  • the results of the fuzzy evaluation permit a decision as to whether an alarm or a fault signal is present.
  • the number of for the wavelet analysis of the required computing steps is significantly reduced compared to Fou ⁇ er analyzes. This shortens the computing time required to identify the signal and reduces the processor costs
  • the original digitized signal is first analyzed by a fast wavelet transformation.
  • the signal according to Mallat's algorithm is passed through several stages of a cascade of high-pass and low-pass filter pairs.
  • the results of the high-pass filters then result in an access function for each filter stage which contains the sum of the calculated values from the high-pass filter and is divided by the sum of the squares of the original signal values.
  • the sum of the access functions that are generated here for each filter level is equal to or almost equal to one.
  • a frequency analysis of this type yields the following advantages.
  • the high-pass filters of the wavelet transformation first provide information about the high-frequency signals. This is particularly advantageous in flame reporting, since the information about the higher frequencies accelerates the identification of the type of signal and its accuracy can be increased For example, if a high-frequency signal of over 15 Hz is detected, this is interpreted as a disturbance signal.
  • the subsequent message, disturbance signal or alarm signal occurs earlier and is, with greater certainty, often very simple in form, such as a Haar-Wavelet, and enable an analysis with wemgen arithmetic, which additionally shortens the computing time and the decision time. However, the shortening of the decision time is not associated with a loss of accuracy in the signal identification. If more lines of code are required, an inexpensive processor can also be used be set
  • a first preferred embodiment of the method according to the invention is characterized in that the wavelet used for the fast wavelet transformation is an orthonormal or semi-orthonormal wavelet or also a wavelet packet basis, and that the membership functions generated in each case are those generated by the wavelet Contain coefficient weighted sum of the squared values of the high-pass filter and the sum of the squared value of the original signal and used in normalized form for the further analysis of the frequency signal according to fuzzy logic rules
  • the wavelet used for the fast wavelet transformation is an orthonormal or semi-orthonormal wavelet or a wavelet packet basis
  • the generated functions each contain the sum of the squared output values of the high-pass filter and the sum of the squared values of original signal of the hazard detector and are used in a normalized form for the evaluation of the frequency signal according to fuzzy logic rules
  • the hazard detector according to the invention for carrying out the method mentioned contains a sensor for a hazard parameter, an evaluation electronics with means for processing the output signal of the sensor and a microprocessor with an 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, and that the signal processed by the evaluation electronics and supplied to the fuzzy controller is wavelet-transformed
  • FIG. 1 shows a block diagram of a method with a rapid wavelet analysis by means of several filter stages and further analysis by fuzzy logic
  • FIG. 2 shows representations of associated functions using the example of a frequency analysis using a fast Haar-Wavelet transformation
  • FIG. 3 shows a block diagram of a hazard detector for performing the method of
  • the output signal XQ k is first used to perform a fast wavelet transformation 1 using any wavelet of the type known from the prior art.
  • An orthonormal or semi-orthonormal wavelet or a wavelet packet base is preferably used.
  • the signal values are denoted by XJ ⁇ and VJ ⁇ , 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. It is from an original signal XQ au k f of the zero level assumed, which is transformed by a plurality of filtering.
  • the output signal of the first high-pass filter gives the values yi k and the output signal of the first low-pass filter, which also forms the input signal for the second filter stage, the values x ⁇ ⁇
  • the output signal of the second high-pass filter gives the values y2 k > that of the second low-pass filter X2 k wu "d added to a third pair of filters, etc. It should be noted here that the number of values resulting from the filter stages is different for each stage. More specifically, for each stage the number of values decreases by a factor of two.
  • the coefficients p and q for the wavelet reconstruction can be found in the book already mentioned.
  • the access functions ⁇ , 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
  • the digitized raw values x f jk are subjected to a quick hair analysis.
  • the values y ⁇ of each filter stage I are used to form access functions ⁇ t , namely
  • these access functions are fed to a 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 the signal is evaluated as a fault
  • this method is suitable for distinguishing between interference signals, such as pe ⁇ odic signals of over 15 Hz, and real flame signals, such as narrow-band signals of medium frequency or broadband signals of medium frequency range.Through the rapid identification of high-frequency signals, the Disturbance signals of this frequency and their resonance frequencies are eliminated from the signal, which speeds up the frequency analysis of the signal. By accelerating the frequency analysis by means of the wavelet transformation, the time required for a decision on the type of signal and the message to be issued can be from three seconds to one, for example Second reduced
  • the method described is also suitable for noise detectors, passive infrared detectors, for the spectral analysis of the signals of individual pixels in the image processing as well as for various sensors such as gas and vibration sensors
  • FIG. 3 shows a diagram of a hazard detector 3 for carrying out the method described.
  • the hazard detector 3 has a sensor 4 for the detection of hazard indicators, an evaluation electronics 5, a microprocessor 6 and the fuzzy controller 2.
  • the hazard indicators can, for example, determine the intensity of emer Flame emitted radiation, the acoustic signal emes noise, the Ixi infrared radiation emitted by a warm body or the output signal from a CCD camera sem
  • the output signal of the sensor 4 is fed to the evaluation electronics 5, which has suitable means for processing the signal, such as an amplifier, and passes from the evaluation electronics 5 into the microprocessor 6.
  • the fuzzy controller 2 (FIG. 1) is shown here as Software integrated in the microprocessor 6.
  • the fuzzy controller is part of a fuzzy wavelet controller that links the fuzzy logic theory with the wavelet theory.
  • the microprocessor 6 contains, for example, a software program of the type shown in FIG. 4, which subjects the input signal to a wavelet transformation. The resulting, transformed signal is then fed to the fuzzy controller 2. If the signal resulting from the fuzzy controller 2 is evaluated as an alarm, it is fed to 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.
  • 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 ⁇ j 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 optionally sends a signal to the alarm output device 7.

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Fuzzy Systems (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • Automation & Control Theory (AREA)
  • Evolutionary Computation (AREA)
  • Computer Security & Cryptography (AREA)
  • Mathematical Physics (AREA)
  • Emergency Management (AREA)
  • Business, Economics & Management (AREA)
  • Alarm Systems (AREA)
  • Fire-Detection Mechanisms (AREA)
  • Lighting Device Outwards From Vehicle And Optical Signal (AREA)
  • Geophysics And Detection Of Objects (AREA)
  • Photometry And Measurement Of Optical Pulse Characteristics (AREA)
  • Feedback Control In General (AREA)
  • Radar Systems Or Details Thereof (AREA)

Abstract

In a method for frequency analysis of a danger alarm system signal, a wavelet transformation (1) is combined with a fuzzy logic evaluation. The original signal (x0,k) of a multistage filter cascade of high-/low- pass filter pairs (HP, LP) is transmitted in the transformation carried out by means of an orthonormal or semi-orthonormal wavelet. A membership function (νi) is worked out in each filter stage based on the results of the high-pass filter, the wavelet coefficients and the values of the original signal (x0,k). These functions are standardized and used in this form for further evaluation (2) according to fuzzy logic rules. The disclosed method is specially suitable for evaluating the output signals of a danger alarm system, such as a fire alarm system, a noise warning system or the like. The wavelet transformation(1) and the fuzzy logic evaluation (2) are conducted by means of a small amount of lines of a processor code. The evaluation can be performed with an economical processor and can be carried out much faster and with an identical or increased precision.

Description

Verfahren zur Analyse des Signals eines Gefahrenmelders und Gefahrenmelder zur Durchführung des VerfahrensProcess for analyzing the signal of a hazard detector and hazard detector for carrying out the method
Die vorliegende Erfindung betrifft em Verfahren zur Analyse des Signals emes Gefahrenmelders mittels Frequenzanalyse und Fuzzy-Logik-Auswertung, sowie einen Gefahrenmelder zur Durchfuhrung dieses Verfahrens Der Gefahrenmelder kann beispielsweise ein Flammenmelder, Gerauschmelder, Brandmelder, passiver Infrarotmelder oder dergleichen semThe present invention relates to a method for analyzing the signal emes a hazard detector using frequency analysis and fuzzy logic evaluation, and to a hazard detector for carrying out this method. The hazard detector can be, for example, a flame detector, noise detector, fire detector, passive infrared detector or the like
Die Ausgangssignale von Gefahrenmeldern smd häufig durch für sie typische Frequenzspektren gekennzeichnet Durch Analyse dieser Frequenzspektren kann die Herkunft der Signale bestimmt werden, und es können vor allem echte Alarmsignale von Storsignalen unterschieden und dadurch Fehlalarme vermieden werden Insbesondere bei Flammenmeldern wird das typische niederfrequente Flackern einer Flamme analysiert, um die Strahlung von echten Flammen von derjenigen emer Storquelle, wie zum Beispiel reflektiertem Sonnenlicht, oder emer flackernden Lichtquelle, unterscheiden zu könnenThe output signals from hazard detectors are often characterized by frequency spectra typical of them. By analyzing these frequency spectra, the origin of the signals can be determined and, above all, real alarm signals can be distinguished from fault signals and false alarms can be avoided in this way in order to be able to distinguish the radiation from real flames from that of a source of disturbance, such as reflected sunlight or a flickering light source
Die Ausgangssignale von Gefahrenmeldern werden beispielsweise mittels Fouπer- Analyse, Fast-Fouπer-Analyse, Zero-Crossmg-Methode oder Turmng-Pomt-Methode analysiert Die letztere ist m der GB-A 2 277 989 m Anwendung an Flammenmeldern beschrieben, wobei die Zeitspannen zwischen Strahlungsmaxnria gemessen und auf ihre Regelmassigkeiten und Unregelmassigkeiten überprüft und unregelmassig auftretende Strahlungsmaxima als Flamme und regelmassige als Störung interpretiert werdenThe output signals from hazard detectors are analyzed, for example, by means of Fouπer analysis, Fast-Fouπer analysis, Zero-Crossmg method or Turmng-Pomt method. The latter is described in GB-A 2 277 989 m using flame detectors, the time periods between Radiation maxi mia measured and checked for their regularities and irregularities and irregularly occurring radiation maxima are interpreted as a flame and regular ones as a disturbance
Die Fn7.7y-T .og1k ist allgemein bekannt In bezug auf die vorliegende Erfindung ist hervorzuheben, dass Signalwerte sogenannten Fuzzy sets, oder unscharfen Mengen, gemass emer Zugehoπgkeitsfunktion zugeteilt werden, wobei der Wert der Zugehoπgkeitsfunk- tion, oder der Grad der Zugehörigkeit zu emer unscharfen Menge, zwischen null und ems betragt Wichtig dabei ist, dass die Zugehongkeitsfunktion normalisierbar smd, d h die Summe aller Werte der Zugehörigkeitsfunktion gleich eins ist, wodurch die Fuzzy- Logik-Auswertung eine eindeutige Interpretation des Signals erlaubt.The Fn7.7y-T .og1k is generally known. With regard to the present invention, it should be emphasized that signal values, so-called fuzzy sets, or fuzzy sets, are allocated in accordance with an association function, the value of the association function or the degree of association An unsharp amount, between zero and ems. It is important that the accessibility function can be normalized, ie the sum of all values of the membership function is equal to one, whereby the fuzzy logic evaluation allows a clear interpretation of the signal.
Bei einem in der EP-A 0 718 814 beschriebenen Flammenmelder wird die Frequenz der detektierten Strahlung analysiert und dabei zwischen regelmässigen und unregelmässi- gen Signalen in bestimmten Frequenzbereichen unterschieden. Die Auswertung der verschiedenen Signale in den gegebenen Frequenzbereichen erfolgt nach mehreren Fuzzy- Logik-Regeln. Durch dieses Verfahren ist eine genauere Unterscheidung zwischen echten Flammensignalen und anderen Störsignalen und somit die Fehlalarmsicherheit ermöglicht. Die Erzeugung des Frequenzspektrums erfolgt hier zum Beispiel durch schnelle Fourier-Transformation, was bezüglich der für die Transformation erforderlichen Zeit, des notwendigen Prozessors und der Prozessorkosten aufwendig ist. Für die Bestimmung eines detektierten Signals sind zum Teil bis zu drei Sekunden erforderlich. Für bestimmte Anwendungen ist jedoch eine kürzere Auswertezeit und Reaktionszeit bis zur Alarmgebung erwünscht, wobei Verfahren wie die Zero-Crossing- oder Turning- Point-Methode oder Wavelet-Analyse zwar den Entscheidungsprozess beschleunigen, aber weniger genau sind.In the case of a flame detector described in EP-A 0 718 814, the frequency of the detected radiation is analyzed and a distinction is made between regular and irregular signals in certain frequency ranges. The various signals in the given frequency ranges are evaluated according to several fuzzy logic rules. This procedure enables a more precise distinction between real flame signals and other interference signals and thus false alarm security. The frequency spectrum is generated here, for example, by fast Fourier transformation, which is complex in terms of the time required for the transformation, the necessary processor and the processor costs. It sometimes takes up to three seconds to determine a detected signal. For certain applications, however, a shorter evaluation time and response time before the alarm is desired, although methods such as the zero crossing or turning point method or wavelet analysis speed up the decision-making process, but are less accurate.
Der Erfindung ist die Aufgabe gestellt, ein Verfahren zur Frequenzanalyse eines Signals eines Gefahrenmelders zu schaffen, das mit einer Fuzzy-Logik-Auswertung vereinigt ist, und im Vergleich zu Analyseverfahren des Standes der Technik mit einer kleineren Anzahl von Rechenschritten durchgeführt wird, so dass in kürzerer Zeit ein Resultat von gleicher oder höherer Genauigkeit erzielt wird. Ferner soll das Verfahren mit einem einfacheren Prozessor und dadurch kostengünstiger durchführbar sein.The invention has for its object to provide a method for frequency analysis of a signal from a hazard detector, which is combined with a fuzzy logic evaluation, and is carried out in comparison with analysis methods of the prior art with a smaller number of computing steps, so that in a result of the same or higher accuracy is achieved in a shorter time. Furthermore, the method should be able to be carried out with a simpler processor and therefore more cost-effectively.
Die Aufgabe wird erfmdungsgemäss dadurch gelöst, dass als Frequenzanalyse eine schnelle Wavelet-Transformation durchgeführt und das ursprüngliche Signal dabei durch eine mehrstufige Filterkaskade von Hoch/Tiefpassfilterpaaren geführt wird, und dass bei jeder Filterstufe der Wavelet-Transformation aus den Resultaten des Hochpassfilters jeweils eine Zugehörigkeitsfunktion erzeugt wird, die zur Weiteranalyse des Frequenzsignals nach Fuzzy-Logik-Regeln verwendet wird. Die Wavelet-Transformation ist eine Transformation oder Abbildung eines Signals vom Zeitbereich in den Frequenzbereich (siehe dazu beispielsweise "The Fast Wavelet- Transfoπn" von Mac A Cody m Dr Dobb's Journal, Apπl 1992), sie ist also grundsätzlich der Fouπer-Transformation und Fast-Fouπer-Transformation ähnlich Sie unterscheidet sich von diesen aber durch die Basisfunktion der Transformation, nach der das Signal entwickelt wird Bei einer Fouπer-Transformation wird eine Sinus- und Cosmus- Funktion verwendet, die im Frequenzbereich scharf lokalisiert und im Zeitbereich unbestimmt ist Bei einer Wavelet-Transformation wird em sogenanntes Wavelet oder Wellenpaket verwendet Hiervon gibt es verschiedene Typen wie zum Beispiel ein Gauss-, Sphne- oder Haar-Wavelet, die jeweils durch zwei Parameter beliebig im Zeitbereich verschoben und im Frequenzbereich gedehnt oder komprimiert werden könnenAccording to the invention, the object is achieved in that a rapid wavelet transformation is carried out as frequency analysis and the original signal is passed through a multi-stage filter cascade of high / low-pass filter pairs, and in that for each filter stage of the wavelet transformation a membership function is generated from the results of the high-pass filter is used for further analysis of the frequency signal according to fuzzy logic rules. The wavelet transformation is a transformation or mapping of a signal from the time domain to the frequency domain (see, for example, "The Fast Wavelet-Transfoπn" by Mac A Cody in Dr Dobb's Journal, Apπl 1992), so it is basically the Fouπer transformation and fast -Fouπer-Transformation similar It differs from these but by the basic function of the transformation, according to which the signal is developed. With a Fouπer transformation a sine and cosmos function is used, which is localized sharply in the frequency domain and indefinitely in the time domain Wavelet transformation is used in a so-called wavelet or wave packet.There are various types, such as a Gaussian, Sphne or Haar wavelet, which can be shifted in the time domain and stretched or compressed in the frequency domain using two parameters
Es können also durch eme Wavelet-Transformation sowohl im Zeit- als auch im Frequenzbereich lokalisierte Signale transformiert werden Eine schnelle Wavelet-Transformation erfolgt durch den Pyramiden- Algorithmus nach Mallat, der auf wiederholter Anwendung emes Tiefpass- und Hochpassfilters beruht, durch welche die mederfrequenten von den hochfrequenten Signalkomponenten getrennt werden Dabei wird jeweils das Ausgangs Signal des Tiefpassfilters wiederum emem Tιef-/Hochpassfilterpaar zugeführt Es resultiert eine Reihe von Approximationen des ursprünglichen Signals, wovon jede eme gröbere Auflösung besitzt als die vorhergehende Die Anzahl Operationen, die für die Transformation erforderlich smd, ist jeweils proportional zur Lange des ursprünglichen Signals, wahrend bei der Fouπer-Transformation diese Anzahl uberproportional zur Signallange ist Die schnelle Wavelet-Transformation kann auch mvers durchgeführt werden, indem das ursprüngliche Signal aus den approximierten Werten und Koeffizienten für die Rekonstruktion wiederhergestellt wird Der Algoπthmus für die Zerlegimg und Rekonstruktion des Signals und eme Tabelle der Koeffizienten der Zerlegung und Rekonstruktion smd am Beispiel für em Splme Wavelet m "An Introduction to Wave- lets" von Charles K Chui (Academic Press, San Diego, 1992) angegebenThus, localized signals can be transformed in the time as well as in the frequency domain by means of a wavelet transformation. A fast wavelet transformation is carried out by the pyramid algorithm according to Mallat, which is based on repeated use of a low-pass and high-pass filter, by means of which the median frequencies of The high-frequency signal components are separated. The output signal of the low-pass filter is in turn fed to a pair of high-pass / high-pass filters. This results in a number of approximations of the original signal, each of which has a higher resolution than the previous one. The number of operations required for the transformation. is proportional to the length of the original signal, while in the Fouπer transformation this number is disproportionate to the signal length d Coefficients for the reconstruction is restored The algorithm for the decomposition and reconstruction of the signal and a table of the coefficients for the decomposition and reconstruction smd using the example of an Splme Wavelet m "An Introduction to Wavelets" by Charles K Chui (Academic Press, San Diego, 1992)
Bei Anwendung m emem Gefahrenmelder erlauben die Resultate der Fuzzy-Auswertung emen Entscheid darüber, ob em Alarm- oder em Storsignal vorliegt Die Anzahl der für die Wavelet-Analyse erforderlicher Rechenschritte ist im Vergleich zu Fouπer-Analysen bedeutend reduziert Dadurch ist die notwendige Rechnerzeit zur Identifizierung des Signals verkürzt, und es verπngern sich die Kosten für den ProzessorWhen using a hazard detector, the results of the fuzzy evaluation permit a decision as to whether an alarm or a fault signal is present. The number of for the wavelet analysis of the required computing steps is significantly reduced compared to Fouπer analyzes. This shortens the computing time required to identify the signal and reduces the processor costs
Gemass der Erfindung wird das ursprungliche digitalisierte Signal zunächst durch eine schnelle Wavelet-Transformation analysiert Hierfür wird das Signal nach dem Algoπth- mus von Mallat durch mehrere Stufen einer Kaskade von Hoch- und Tiefpassfilterpaaren geführt Aus den Resultaten der Hochpassfilter wird sodann bei jeder Filterstufe eine Zugehoπgkeitsfünktion erzeugt, welche die Summe der gerechneten Werte aus dem Hochpassfilter enthalt und durch die Summe der Quadrate der ursprunglichen Signalwerte dividiert ist Die Summe der Zugehoπgkeitsfunktionen, die hier bei jeder Filterstufe erzeugt werden, ist gleich oder nahezu gleich eins Diese normalisierten Zugeho- πgkeitsfünktionen werden sodann m dieser Form für eme Weiterführung der Frequenzanalyse mit Fuzzy-Logik verwendetAccording to the invention, the original digitized signal is first analyzed by a fast wavelet transformation. For this purpose, the signal according to Mallat's algorithm is passed through several stages of a cascade of high-pass and low-pass filter pairs. The results of the high-pass filters then result in an access function for each filter stage which contains the sum of the calculated values from the high-pass filter and is divided by the sum of the squares of the original signal values. The sum of the access functions that are generated here for each filter level is equal to or almost equal to one. These normalized access functions are then m This form is used for the continuation of frequency analysis with fuzzy logic
Eme Frequenzanalyse dieser Art ergibt folgende Vorteile Die Hochpassfilter der Wavelet-Transformation ergeben zuerst Informationen über die hochfrequenten Signale Dies ist msbesondere m der Flammenmeldung vorteilhaft, da mit der Information über die höheren Frequenzen die Identifizierung der Art des Signals beschleunigt und ihre Genauigkeit erhöht werden kann Wird zum Beispiel em hochfrequentes Signal von über 15 Hz entdeckt, wird dieses als Storsignal gedeutet Die darauffolgende Meldung, Storsignal oder Alarmsignal, erfolgt früher und ist mit grosserer Sicherheit πchtig Wavelets smd in ihrer Form oft sehr einfach, wie zum Beispiel em Haar-Wavelet, und ermöglichen eme Analyse mit wemgen Rechenschπtten, was die Rechenzeit und die Entscheidungszeit zusätzlich verkürzt Die Verkürzung der Entscheidungszeit ist jedoch nicht mit emer Embusse in der Genauigkeit der Signahdentifizierung verbunden Smd wemger Zeilen von Code erforderlich, kann auch em kostengünstiger Prozessor eingesetzt werdenA frequency analysis of this type yields the following advantages. The high-pass filters of the wavelet transformation first provide information about the high-frequency signals. This is particularly advantageous in flame reporting, since the information about the higher frequencies accelerates the identification of the type of signal and its accuracy can be increased For example, if a high-frequency signal of over 15 Hz is detected, this is interpreted as a disturbance signal.The subsequent message, disturbance signal or alarm signal, occurs earlier and is, with greater certainty, often very simple in form, such as a Haar-Wavelet, and enable an analysis with wemgen arithmetic, which additionally shortens the computing time and the decision time. However, the shortening of the decision time is not associated with a loss of accuracy in the signal identification. If more lines of code are required, an inexpensive processor can also be used be set
Eme erste bevorzugte Ausführung des erfindungsgemassen Verfahrens ist dadurch gekennzeichnet, dass das für die schnelle Wavelet-Transformation verwendete Wavelet em orthonormales oder semi-orthonormales Wavelet oder auch eme Wavelet-Paket-Basis ist, und dass die erzeugten Zugehoπgkeitsfunktionen jeweils die durch die Wavelet- Koeffizienten gewichtete Summe der quadπerten Werte des Hochpassfilters und die Summe der quadπerten Wert des ursprünglichen Signals enthalten und in normalisierter Form für die Weiteranalyse des Frequenzsignals nach Fuzzy-Logik-Regeln verwendet werdenA first preferred embodiment of the method according to the invention is characterized in that the wavelet used for the fast wavelet transformation is an orthonormal or semi-orthonormal wavelet or also a wavelet packet basis, and that the membership functions generated in each case are those generated by the wavelet Contain coefficient weighted sum of the squared values of the high-pass filter and the sum of the squared value of the original signal and used in normalized form for the further analysis of the frequency signal according to fuzzy logic rules
Bei einer zweiten bevorzugten Ausführungsform ist das für die schnelle Wavelet-Transformation verwendete Wavelet ein orthonormales oder semi-orthonormales Wavelet oder eine Wavelet-Paket-Basis und die erzeugten Zgehoπgkeitsf nktionen enthalten jeweils die Summe der quadπerten Ausgangswerte des Hochpassfilters und die Summe der quadπerten Werte des ursprünglichen Signals des Gefahrenmelders und werden in normalisierter Form für die Auswertung des Frequenzsignals nach Fuzzy-Logik-Regeln verwendetIn a second preferred embodiment, the wavelet used for the fast wavelet transformation is an orthonormal or semi-orthonormal wavelet or a wavelet packet basis, and the generated functions each contain the sum of the squared output values of the high-pass filter and the sum of the squared values of original signal of the hazard detector and are used in a normalized form for the evaluation of the frequency signal according to fuzzy logic rules
Der erfindungsgernasse Gefahrenmelder zur Durchführung des genannten Verfahrens enthalt emen Sensor für eme Gefahrenkenngrosse, eme Auswerteelektronik mit Mitteln zur Verarbeitung des Ausgangssignals des Sensors und emen Mikroprozessor mit emem Fuzzy-Controller Dieser Gefahrenmelder ist dadurch gekennzeichnet, dass der Mikroprozessor em Software-Programm aufweist, nach dem der Fuzzy-Controller Teil eme Fuzzy- Wavelet-Controllers ist, und dass das durch die Auswertelektronik verarbeitete und dem Fuzzy-Controller zugeführte Signal wavelet-transformiert istThe hazard detector according to the invention for carrying out the method mentioned contains a sensor for a hazard parameter, an evaluation electronics with means for processing the output signal of the sensor and a microprocessor with an 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, and that the signal processed by the evaluation electronics and supplied to the fuzzy controller is wavelet-transformed
Im folgenden wird die Erfindung anhand emes m den Zeichnungen dargestellten Aus- führungsbeispiels naher erläutert, es zeigtThe invention is explained in more detail below with reference to the exemplary embodiment shown in the drawings, it shows
Fig 1 ein Blockschema emeses Verfahrens mit emer schnellen Wavelet-Analyse durch mehrere Filterstufen und Weiteranalyse durch Fuzzy-Logik, Fig 2 Darstellungen von Zugehoπgkeitsfünktionen am Beispiel einer Frequenzanalyse mittels einer schnellen Haar-Wavelet-Transformation, Fig 3 em Blockschema emes Gefahrenmelders zur Diuchführung des Verfahrens von1 shows a block diagram of a method with a rapid wavelet analysis by means of several filter stages and further analysis by fuzzy logic, FIG. 2 shows representations of associated functions using the example of a frequency analysis using a fast Haar-Wavelet transformation, FIG. 3 shows a block diagram of a hazard detector for performing the method of
Fig l, und Fig 4 em Blockschema für die Implementierung des Verfahrens von Fig 1 m ememFig l, and Fig 4 em block diagram for the implementation of the method of Fig 1 m emem
Gefahrenmelder Gemass Fig. 1 wird mit dem Ausgangssignal XQ k zunächst eine schnelle Wavelet-Transformation 1 mittels eines beliebigen Wavelet der aus dem Stand der Technik bekannten Art durchgeführt. Vorzugsweise wird ein orthonormales oder semi-orthonormales Wavelet oder eine Wavelet-Paket-Basis verwendet. In der Figur sind die Signalwerte mit XJ ^ und VJ ^ bezeichnet, wobei x die ursprünglichen Signalwerte und die Werte aus den Tiefpassfiltem (LP) und y die Werte aus den Hochpassfiltern (HP) bedeuten. Der Index i bezeichnet in steigender Zahl die Stufe der Filterkaskade, wobei das ursprüngliche Signal auf Stufe Null ist. Der Index k bezeichnet einen individuellen Wert eines Signals. Es wird von einem ursprünglichen Signal XQ k auf der Stufe Null ausgegangen, das durch mehrere Filterungen transformiert wird. Das Ausgangssignal des ersten Hochpassfilters ergibt die Werte yi k und das Ausgangssignal des ersten Tiefpassfüters, das zugleich das Eingangssignal für die zweite Filterstufe bildet, die Werte x\ ^ Das Ausgangssignal des zweiten Hochpassfilters ergibt die Werte y2 k> das des zweiten Tiefpassfüters X2 k wu"d einem dritten Filterpaar zugeführt, usw. Es ist hier zu bemerken, dass die Anzahl Werte, die aus den Filterstufen hervorgehen, jeweils bei jeder Stufe verschieden ist. Genauer gesagt, bei jeder Stufe verkleinert sich die Anzahl der Werte um den Faktor zwei. Bei der Stufe i+1 werden beispielsweise die Ausgangswerte eines Hochpassfilters durch yM k = ∑ tl kxt l und die Ausgangswerte eines Tiefpassfüters durch xM k = ∑b,_;ix,,, ausgedrückt.Hazard detector 1, the output signal XQ k is first used to perform a fast wavelet transformation 1 using 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 XJ ^ and VJ ^, 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. It is from an original signal XQ au k f of the zero level assumed, which is transformed by a plurality of filtering. The output signal of the first high-pass filter gives the values yi k and the output signal of the first low-pass filter, which also forms the input signal for the second filter stage, the values x \ ^ The output signal of the second high-pass filter gives the values y2 k > that of the second low-pass filter X2 k wu "d added to a third pair of filters, etc. It should be noted here that the number of values resulting from the filter stages is different for each stage. More specifically, for each stage the number of values decreases by a factor of two. at stage i + 1, the output values are, for example, a high pass filter by M y k = Σ t lk x tl and the output values of a Tiefpassfüters by x k = M .sigma..sub.B, _; i x, expressed.
Die Koeffizienten a und b für die Transformation sind im allgemeinen bekannt und können mit Hilfe des genannten Buches von Chui berechnet werden. Beispielsweise sind für ein Haar-Wavelet aυ =aι=l/2, bo=l/2 und bι=-l/2. Der Index 1 nimmt jeweils ganzzahlige Werte an, für die die Koeffizienten ungleich null sind. Die Rekonstruktion des ursprünglichen Signals erfolgt stufenweise, indem die Werte jeder Filterstufe aus den Werten der vorherigen Stufe erstellt werden, nämlich xι k = ∑ (/ -:Λ+ι., + Qk-ιιy,+\.ι)- lThe coefficients a and b for the transformation are generally known and can be calculated using the Chui book mentioned. For example, for a hair wavelet there are a υ = aι = 1/2, bo = 1/2 and bι = -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 x ι k = ∑ (/ - : Λ + ι ., + Q k -ι ι y, + \ . Ι ) - l
Die Koeffizienten p und q für die Wavelet-Rekonstruktion sind in dem schon genannten Buch zu finden. Anschliessend werden aus den Ausgangswerten des Hochpassfilters der jeweiligen Filterstufe und den dazugehoπgen Koeffizienten q für die Wavelet-Rekonstruktion die Zugehoπgkeitsfunktionen μ,erzeugtThe coefficients p and q for the wavelet reconstruction can be found in the book already mentioned. The access functions μ, 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
Dabei ist μ = — ; — füπ=l, 2, , N (Gleichung 1)Where μ = -; - fiπ = 1, 2,, N (equation 1)
Σ ( „/)" Σ ( "/ )"
und μ. , = , wobei N die Anzahl der Filterstufen ist Die letztere Funktion μjsj+j wird also durch die Ausgangswerte des letzten Tiefpassfüters gebildet Diese Zugehoπgkeitsfunktionen smd normalisiert, mdem ∑ , = 1and μ. , =, where N is the number of filter stages the latter function μjsj + j is thus formed by the output values of the last low-pass filter. These access functions are normalized smd, mdem ∑, = 1
Eme oft gute Annäherung dieser Zugehoπgkeitsfünktionen ist durch folgende Gleichung gegebenAn often good approximation of these functions is given by the following equation
Bei dieser Annäherung ist die Funktion nahezu normalisiert, mdem ∑μ, ~ 1With this approach, the function is almost normalized, mdem ∑μ, ~ 1
Bei emer besonderen Ausführung des Verfahrens werden die digitalisierten Rohwerte xfj k emer schnellen Haar- Analyse unterworfen Aus den Werten y^ jeder Filterstufe I werden Zugehoπgkeitsfünktionen μt gebildet, nämlichIn the case of a special implementation of the method, the digitized raw values x f jk are subjected to a quick hair analysis. The values y ^ of each filter stage I are used to form access functions μ t , namely
μ, = für ι=N+l μ, = for ι = N + l
Diese Zugehoπgkeitsfünktionen smd m diesem Fall normalisiert, mdem ∑μ, = \ ist In Figur 2 smd Zugehoπgkeitsfünkttonen μ, die aus den Resultaten emer schnellen Haar- Wavelet-Transformation erzeugt worden sind, als Funktion der Frequenz gezeigt Von den verschiedenen Kurven ülustπeren μj\j+ 1 den Grad der Zugehoπgkeit von sehr tiefen Frequenzen, μjyj den von tiefen Frequenzen, und μ\ und \xr den Grad der Zugehoπgkeit von hohen beziehungsweise mittleren Frequenzen Es ist hier klar ersichtlich, dass bei jeder gewählten Frequenz die Summe der Kurvenwerte ems betragtThese features are normalized in this case, since ∑μ, = \ In FIG. 2, the five-point tonality μ which have been generated from the results of a fast Haar-Wavelet transformation is shown as a function of the frequency. The various curves show the degree of the association of very low frequencies, μjyj that of the low frequencies , and μ \ and \ xr the degree of affiliation of high and medium frequencies. It is clearly evident here that the sum of the curve values is ems for each frequency selected
Bei allen Ausführungen des Verfahrens werden diese Zugehoπgkeitsfünktionen einem Fuzzy-Logik-Controller 2 (Fig 1) für die Auswertung nach Fuzzy-Logik-Regeln zugeführt, worauf eme Entscheidung gefallt wird, ob ein Alarmsignal ausgelost oder das Signal als Störung bewertet wirdIn all embodiments of the method, these access functions are fed to a 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 the signal is evaluated as a fault
Bei der Anwendung m Flammenmeldern eignet sich dieses Verfahren zur Unterscheidung zwischen Storsignalen, wie zum Beispiel peπodischen Signalen von über 15 Hz, und echten Flammensignalen, wie zum Beispiel schmalbandigen Signalen mederer Frequenz oder breitbandigen Signalen m mederem Frequenzbereich Durch die schnelle Identifizierung von hochfrequenten Signalen werden die Storsignale dieser Frequenz und deren Resonanzfrequenzen vom Signal eliminiert, was die Frequenzanalyse des Signals beschleunigt Durch die Beschleunigung der Frequenzanalyse durch die Wavelet-Transformation kann die erforderliche Zeit für eme Entscheidung über die Art des Signals und die abzugebende Meldung von zum Beispiel bisher drei Sekunden auf eine Sekunde verringert werden Das beschriebene Verfahren ist weiter auch für Gerauschmelder, passive Infrarotmelder, für die Spektralanalyse der Signale einzelner Pixel m der Büdverar- beitung sowie für verschiedene Sensoren wie Gas- und Vibrationssensoren geeignetWhen using m flame detectors, this method is suitable for distinguishing between interference signals, such as peπodic signals of over 15 Hz, and real flame signals, such as narrow-band signals of medium frequency or broadband signals of medium frequency range.Through the rapid identification of high-frequency signals, the Disturbance signals of this frequency and their resonance frequencies are eliminated from the signal, which speeds up the frequency analysis of the signal. By accelerating the frequency analysis by means of the wavelet transformation, the time required for a decision on the type of signal and the message to be issued can be from three seconds to one, for example Second reduced The method described is also suitable for noise detectors, passive infrared detectors, for the spectral analysis of the signals of individual pixels in the image processing as well as for various sensors such as gas and vibration sensors
Figur 3 zeigt em Schema emes Gefahrenmelders 3 zur Durchführung des beschriebenen Verfahrens Darstellungsgemass weist der Gefahrenmelder 3 emen Sensor 4 zur Detektion emer Gefahrenkenngrosse, eme Auswerteelektronik 5, emen Mikroprozessor 6 und den Fuzzy-Controller 2 auf Die Gefahrenkenngrosse kann zum Beispiel die Intensität der von emer Flamme abgegebenen Strahlung, das akustische Signals emes Geräusches, die von einem warmen Korper abgegebenen Ixifrarotstrahlung oder das Ausgangssignal emer CCD-Kamera sem Das Ausgangssignal des Sensors 4 wird der Auswerteelektronik 5 zugeführt, welche geeignete Mittel zur Verarbeitung des Signals, wie zum Beispiel Verstärker, aufweist, und gelangt von der Auswerteelektronik 5 in den Mikroprozessor 6. Der Fuzzy-Controller 2 (Fig. 1) ist hier als Software im Mikroprozessor 6 integriert. Insbesondere ist der Fuzzy-Controller Teil eines Fuzzy-Wavelet-Controllers, der die Fuzzy-Logik-Theorie mit der Wavelet-Theorie verknüpft. Der Mikroprozessor 6 enthält beispielsweise ein Software-Programm der in Figur 4 gezeigten Art, welches das Eingangs-Signal einer Wavelet-Transformation unterzieht. Das resultierende, transformierte Signal wird sodann dem Fuzzy-Controller 2 zugeführt. Sollte das aus dem Fuzzy-Controller 2 resultierende Signal als Alarm gewertet werden, wird dieses einer Alarmabgabevorrichtung 7 oder einer Alarmzentrale zugeführt.FIG. 3 shows a diagram of a hazard detector 3 for carrying out the method described. According to the illustration, the hazard detector 3 has a sensor 4 for the detection of hazard indicators, an evaluation electronics 5, a microprocessor 6 and the fuzzy controller 2. The hazard indicators can, for example, determine the intensity of emer Flame emitted radiation, the acoustic signal emes noise, the Ixi infrared radiation emitted by a warm body or the output signal from a CCD camera sem The output signal of the sensor 4 is fed to the evaluation electronics 5, which has suitable means for processing the signal, such as an amplifier, and passes from the evaluation electronics 5 into the microprocessor 6. The fuzzy controller 2 (FIG. 1) is shown here as Software integrated in the microprocessor 6. In particular, the fuzzy controller is part of a fuzzy wavelet controller that links the fuzzy logic theory with the wavelet theory. The microprocessor 6 contains, for example, a software program of the type shown in FIG. 4, which subjects the input signal to a wavelet transformation. The resulting, transformed signal is then fed to the fuzzy controller 2. If the signal resulting from the fuzzy controller 2 is evaluated as an alarm, it is fed to an alarm delivery device 7 or an alarm center.
Figur 4 zeigt ein Blockschema für die Implementierung des erfindungsgemässen Verfahrens im Mikroprozessor eines Gefahrenmelders, wobei dieser Mikroprozessor einen Fuzzy-Wavelet Controller 8 aufweist. Das Ausgangssignal des Sensors 4 wird nach Auswertung durch die Auswerteelektronik 5 (Fig. 3) dem Fuzzy-Wavelet Controller 8 zugeführt, in dem zunächst das Signal durch eine Kaskade von Filtern 9 geführt wird. Aus den Resultaten 10 jedes Filters 9 werden nach Gleichung 1 die Zugehörigkeitsfünktionen μj gebildet. Diese Funktionen werden sodann dem Fuzzy-Controller 2 zur Fuzzy-Analy- se zugeführt, der gegebenenfalls ein Signal an die Alarmabgabevorrichtung 7 sendet. 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 μj 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 optionally sends a signal to the alarm output device 7.

Claims

Patentansprüche claims
1 Verfahren zur Analyse des Signals emes Gefahrenmelders (3) mittels Frequenzanalyse und Fuzzy-Logik-Auswertung, dadurch gekennzeichnet, dass als Frequenzanalyse eine schnelle Wavelet-Transformation ( 1) durchgeführt und das ursprungliche Signal (x0,k) dabei durch eine mehrstufige Filterkaskade von Hoch-/Tιefpassfüterpaaren (HP, LP) gefuhrt wird, und dass bei jeder Filterstufe der Wavelet-Transformation aus den Resultaten des Hochpassfilters (HP) jeweils eme Zugehoπgkeitsfunktion (μt) erzeugt wird, die zur Weiteranalyse des Frequenzsignals nach Fuzzy-Logik-Regeln verwendet1 Method for analyzing the signal of a hazard detector (3) by means of frequency analysis and fuzzy logic evaluation, characterized in that a fast wavelet transformation (1) is carried out as frequency analysis and the original signal ( x 0 , k) is carried out by a multi-stage filter cascade is carried out by high / low pass filter pairs (HP, LP), and that at each filter stage of the wavelet transformation, the results of the high pass filter (HP) are used to generate an associated function (μ t ), which is used for further analysis of the frequency signal according to fuzzy logic. Rules used
2 Verfahren nach Anspruch 1, dadurch gekennzeichnet, dass das für die schnelle Wavelet-Transformation (1) verwendete Wavelet em orthonormales oder semi-orthonormales Wavelet oder eme Wavelet-Paket-Basis ist, und dass die erzeugten Zugehoπgkeitsfünktionen (μj) jeweils die durch die Wavelet- Koeffizienten gewichtete Summe der quadπerten Werte des Hochpassfilters (HP) und die Summe der quadπerten Werte des ursprünglichen Signals (xυ k) des Gefahrenmelders (3) enthalten und m normalisierter Form für die Weiteranalyse des Frequenzsignals nach Fuzzy-Logik-Regeln verwendet werden2. The method according to claim 1, characterized in that the wavelet used for the fast wavelet transformation (1) is an orthonormal or semi-orthonormal wavelet or an wavelet packet basis, and that the generated functions (μ j ) each have the functions contain the wavelet coefficients weighted sum of the squared values of the high-pass filter (HP) and the sum of the squared values of the original signal (x υ k) of the hazard detector (3) and used in normalized form for the further analysis of the frequency signal according to fuzzy logic rules become
3 Verfahren nach Anspruch 1, dadurch gekennzeichnet, dass das für die schnelle Wavelet-Transformation (1) verwendete Wavelet em orthonormales oder semi orthonormales Wavelet oder eme Wavelet-Paket-Basis ist, und dass die erzeugten Zugehoπgkeitsfünktionen (μj) jeweils die Summe der quadπerten Ausgangswerte des Hochpassfilters (HP) und die Summe der quadπerten Werte des ursprunglichen Signals (xo5k) des Gefahrenmelders (3) enthalten und m normalisierter Form für die Auswertung des Frequenzsignals nach Fuzzy-Logik-Regeln verwendet werden3. The method according to claim 1, characterized in that the wavelet used for the fast wavelet transformation (1) is an orthonormal or semi-orthonormal wavelet or a wavelet packet basis, and that the generated functions (μ j ) are each the sum of the contain squared output values of the high pass filter (HP) and the sum of the squared values of the original signal (xo 5 k) of the hazard detector (3) and are used in normalized form for the evaluation of the frequency signal according to fuzzy logic rules
4 Verfahren nach emem der Ansprüche 1 bis 3, dadurch gekennzeichnet, dass die Ausgangssignale die emes Flammenmelders smd und die Frequenzanalyse und Auswertung der Ausgangssignale des Flammenmelders 100 ms bis 10 s dauert 4. The method according to claims 1 to 3, characterized in that the output signals emes the flame detector smd and the frequency analysis and evaluation of the output signals of the flame detector takes 100 ms to 10 s
5. Gefahrenmelder (3) zur Durchführung des Verfahrens nach emem der Ansprüche 1 bis 3 mit einem Sensor (4) für eine Gefahrenkenngrosse, emer Auswerteleektronik (5) mit Mitteln zur Verarbeitung des Ausgangssignals des Sensors (4) und einem Mikroprozessor (6) mit einem Fuzzy-Controller (2), dadurch gekennzeichnet, dass der Mikroprozessor (6) ein Software-Programm aufweist, nach dem der Fuzzy-Controller (2) Teil eines Fuzzy-Wavelet Controllers (8) ist, und dass das durch die Auswertelektronik (5) verarbeitete und dem Fuzzy-Controller (2) zugeführte Signal wavelet-transformiert ist. 5. Hazard detector (3) for carrying out the method according to emem of claims 1 to 3 with a sensor (4) for a hazard, emer evaluation electronics (5) with means for processing the output signal of the sensor (4) and a microprocessor (6) with a fuzzy controller (2), characterized in that the microprocessor (6) has a software program, according to which the fuzzy controller (2) is part of a fuzzy wavelet controller (8), and that the evaluation electronics ( 5) processed and supplied to the fuzzy controller (2) is wavelet-transformed.
EP97939930A 1996-10-04 1997-09-19 Method for analyzing the signals of a danger alarm system and danger alarm system for implementing said method Expired - Lifetime EP0865646B1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
EP97939930A EP0865646B1 (en) 1996-10-04 1997-09-19 Method for analyzing the signals of a danger alarm system and danger alarm system for implementing said method

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
EP96115952 1996-10-04
EP96115952A EP0834845A1 (en) 1996-10-04 1996-10-04 Method for frequency analysis of a signal
EP97939930A EP0865646B1 (en) 1996-10-04 1997-09-19 Method for analyzing the signals of a danger alarm system and danger alarm system for implementing said method
PCT/CH1997/000354 WO1998015931A1 (en) 1996-10-04 1997-09-19 Method for analyzing the signals of a danger alarm system and danger alarm system for implementing said method

Publications (2)

Publication Number Publication Date
EP0865646A1 true EP0865646A1 (en) 1998-09-23
EP0865646B1 EP0865646B1 (en) 2002-03-13

Family

ID=8223268

Family Applications (2)

Application Number Title Priority Date Filing Date
EP96115952A Withdrawn EP0834845A1 (en) 1996-10-04 1996-10-04 Method for frequency analysis of a signal
EP97939930A Expired - Lifetime EP0865646B1 (en) 1996-10-04 1997-09-19 Method for analyzing the signals of a danger alarm system and danger alarm system for implementing said method

Family Applications Before (1)

Application Number Title Priority Date Filing Date
EP96115952A Withdrawn EP0834845A1 (en) 1996-10-04 1996-10-04 Method for frequency analysis of a signal

Country Status (9)

Country Link
US (1) US6011464A (en)
EP (2) EP0834845A1 (en)
JP (1) JP2000503438A (en)
KR (1) KR19990071873A (en)
CN (1) CN1129879C (en)
AT (1) ATE214504T1 (en)
DE (1) DE59706608D1 (en)
PL (1) PL327070A1 (en)
WO (1) WO1998015931A1 (en)

Families Citing this family (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6742183B1 (en) 1998-05-15 2004-05-25 United Video Properties, Inc. Systems and methods for advertising television networks, channels, and programs
US6219373B1 (en) * 1998-06-15 2001-04-17 The United States Of America As Represented By The Secretary Of The Navy Wavelet-based interference filtering for spread-spectrum signal
US6879253B1 (en) * 2000-03-15 2005-04-12 Siemens Building Technologies Ag Method for the processing of a signal from an alarm and alarms with means for carrying out said method
US6184792B1 (en) * 2000-04-19 2001-02-06 George Privalov Early fire detection method and apparatus
BR0209543A (en) * 2001-05-11 2005-04-26 Detector Electronics Flame detection and fire detection method and apparatus
FR2841424A1 (en) * 2002-06-25 2003-12-26 Koninkl Philips Electronics Nv METHOD FOR DETECTING BLOCK ARTEFACTS
US7202794B2 (en) * 2004-07-20 2007-04-10 General Monitors, Inc. Flame detection system
MX2008001357A (en) * 2005-07-29 2008-04-16 V&M Deutschland Gmbh Method for error-free checking of tubes for surface faults.
WO2008088325A1 (en) * 2007-01-16 2008-07-24 Utc Fire & Security Corporation System and method for video based fire detection
US8094015B2 (en) * 2009-01-22 2012-01-10 International Business Machines Corporation Wavelet based hard disk analysis
US8941734B2 (en) * 2009-07-23 2015-01-27 International Electronic Machines Corp. Area monitoring for detection of leaks and/or flames
US8359616B2 (en) 2009-09-30 2013-01-22 United Video Properties, Inc. Systems and methods for automatically generating advertisements using a media guidance application
US8949901B2 (en) 2011-06-29 2015-02-03 Rovi Guides, Inc. Methods and systems for customizing viewing environment preferences in a viewing environment control application
CN103501205B (en) * 2013-10-11 2016-05-11 北京理工大学 Target Frequency Hopping Signal recognition methods based on fuzzy comprehensive evoluation

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4866420A (en) * 1988-04-26 1989-09-12 Systron Donner Corp. Method of detecting a fire of open uncontrolled flames
US5453733A (en) * 1992-07-20 1995-09-26 Digital Security Controls Ltd. Intrusion alarm with independent trouble evaluation
US6310963B1 (en) * 1994-09-30 2001-10-30 Sensormatic Electronics Corp Method and apparatus for detecting an EAS (electronic article surveillance) marker using wavelet transform signal processing
ATE203118T1 (en) * 1994-12-19 2001-07-15 Siemens Building Tech Ag METHOD AND ARRANGEMENT FOR DETECTING A FLAME
US5815198A (en) * 1996-05-31 1998-09-29 Vachtsevanos; George J. Method and apparatus for analyzing an image to detect and identify defects

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See references of WO9815931A1 *

Also Published As

Publication number Publication date
CN1205094A (en) 1999-01-13
EP0834845A1 (en) 1998-04-08
KR19990071873A (en) 1999-09-27
WO1998015931A1 (en) 1998-04-16
JP2000503438A (en) 2000-03-21
ATE214504T1 (en) 2002-03-15
US6011464A (en) 2000-01-04
CN1129879C (en) 2003-12-03
EP0865646B1 (en) 2002-03-13
DE59706608D1 (en) 2002-04-18
PL327070A1 (en) 1998-11-23

Similar Documents

Publication Publication Date Title
EP0865646B1 (en) Method for analyzing the signals of a danger alarm system and danger alarm system for implementing said method
DE19828409B4 (en) Circuit for detecting an accident noise
EP1134712B1 (en) Method for the processing of the signal in a danger detector, and detector with means for the implementation of such method
EP1439396B1 (en) Method for the analysis and monitoring of the partial discharge behavior of an electrical operating means
DE69920645T2 (en) Method and device for measuring light absorption spectra
EP0718814B1 (en) Method and device for flame detection
DE4207728A1 (en) Quality testing of test objects - externally stimulating object to emit sound waves which are analysed to classify objects using neural network taught using objects of known class
DE19535094B4 (en) A method of detecting misfire on one or more continuously-deploying cylinders
DE2748558B2 (en) Device for checking the authenticity of banknotes
DE4013660C2 (en)
EP1496483B1 (en) Method and apparatus for the detection of flames
DE3726585A1 (en) Method for locating noises from leaks, and device for carrying out the method
DE2924605C2 (en) Process for the optical differentiation of test objects
DE69920680T2 (en) Method and device for detecting plastic materials by means of optical measurements
DE102005028669A1 (en) Method and device for recognizing a coin entered in a coin validator using its embossed image
DE4002829A1 (en) METHOD FOR DETECTING METAL OBJECTS
EP0957471A2 (en) Measuring process for loudness quality assessment of audio signals
WO1992015171A1 (en) Process and device for filtering signals
EP1209458B1 (en) Procedure for determining the noise level of an internal combustion engine
DE19906118C2 (en) Method and device for comparing acoustic input signals fed into an input device with acoustic reference signals stored in a memory
DE69726458T2 (en) DEVICE AND METHOD FOR ANALYZING AND FILTERING NOISE
EP0654676B1 (en) Method for determining the fundamental frequencies of the frequency spectrum of an acoustic locating device
DE69729724T2 (en) ELECTRONIC GOODS MONITORING SYSTEM WITH CHAMBER FILTRATION THROUGH MULTIPHASE DECOMPOSITION AND NONLINEAR FILTERING OF SUB-SEQUENCES
EP0939308A2 (en) Apparatus for recognising or analysis of damage to machines
DE19626329C2 (en) Procedure for the analysis of sound events

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AT BE CH DE ES FI FR GB IT LI NL PT SE

17P Request for examination filed

Effective date: 19980929

RAP1 Party data changed (applicant data changed or rights of an application transferred)

Owner name: SIEMENS BUILDING TECHNOLOGIES AG

17Q First examination report despatched

Effective date: 20001020

RAP1 Party data changed (applicant data changed or rights of an application transferred)

Owner name: SIEMENS BUILDING TECHNOLOGIES AG

GRAG Despatch of communication of intention to grant

Free format text: ORIGINAL CODE: EPIDOS AGRA

GRAG Despatch of communication of intention to grant

Free format text: ORIGINAL CODE: EPIDOS AGRA

GRAG Despatch of communication of intention to grant

Free format text: ORIGINAL CODE: EPIDOS AGRA

GRAH Despatch of communication of intention to grant a patent

Free format text: ORIGINAL CODE: EPIDOS IGRA

GRAH Despatch of communication of intention to grant a patent

Free format text: ORIGINAL CODE: EPIDOS IGRA

REG Reference to a national code

Ref country code: GB

Ref legal event code: IF02

GRAA (expected) grant

Free format text: ORIGINAL CODE: 0009210

AK Designated contracting states

Kind code of ref document: B1

Designated state(s): AT BE CH DE ES FI FR GB IT LI NL PT SE

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: NL

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20020313

Ref country code: FI

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20020313

REF Corresponds to:

Ref document number: 214504

Country of ref document: AT

Date of ref document: 20020315

Kind code of ref document: T

REG Reference to a national code

Ref country code: CH

Ref legal event code: EP

GBT Gb: translation of ep patent filed (gb section 77(6)(a)/1977)

Effective date: 20020314

REF Corresponds to:

Ref document number: 59706608

Country of ref document: DE

Date of ref document: 20020418

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: SE

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20020613

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: PT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20020614

NLV1 Nl: lapsed or annulled due to failure to fulfill the requirements of art. 29p and 29m of the patents act
ET Fr: translation filed
PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: ES

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20020925

PLBE No opposition filed within time limit

Free format text: ORIGINAL CODE: 0009261

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: NO OPPOSITION FILED WITHIN TIME LIMIT

26N No opposition filed

Effective date: 20021216

REG Reference to a national code

Ref country code: CH

Ref legal event code: PFA

Owner name: SIEMENS BUILDING TECHNOLOGIES AG C-IPR

Free format text: SIEMENS BUILDING TECHNOLOGIES AG#BELLERIVESTRASSE 36#8034 ZUERICH (CH) -TRANSFER TO- SIEMENS BUILDING TECHNOLOGIES AG C-IPR#GUBELSTRASSE 22#6300 ZUG (CH)

REG Reference to a national code

Ref country code: FR

Ref legal event code: TP

Ref country code: FR

Ref legal event code: CD

REG Reference to a national code

Ref country code: CH

Ref legal event code: PUE

Owner name: SIEMENS AKTIENGESELLSCHAFT

Free format text: SIEMENS BUILDING TECHNOLOGIES AG C-IPR#GUBELSTRASSE 22#6300 ZUG (CH) -TRANSFER TO- SIEMENS AKTIENGESELLSCHAFT#WITTELSBACHERPLATZ 2#80333 MUENCHEN (DE)

Ref country code: CH

Ref legal event code: NV

Representative=s name: SIEMENS SCHWEIZ AG

REG Reference to a national code

Ref country code: GB

Ref legal event code: 732E

Free format text: REGISTERED BETWEEN 20090514 AND 20090520

BECA Be: change of holder's address

Owner name: SIEMENS A.G.WITTELSBACHERPLATZ 2, DE-80333 MUENCHE

Effective date: 20100423

BECH Be: change of holder

Owner name: SIEMENS A.G.

Effective date: 20100423

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: DE

Payment date: 20101119

Year of fee payment: 14

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: BE

Payment date: 20101019

Year of fee payment: 14

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: CH

Payment date: 20111213

Year of fee payment: 15

BERE Be: lapsed

Owner name: SIEMENS A.G.

Effective date: 20110930

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: BE

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20110930

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: GB

Payment date: 20120917

Year of fee payment: 16

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: FR

Payment date: 20121002

Year of fee payment: 16

Ref country code: IT

Payment date: 20120927

Year of fee payment: 16

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: AT

Payment date: 20120810

Year of fee payment: 16

REG Reference to a national code

Ref country code: CH

Ref legal event code: PL

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: CH

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20120930

Ref country code: DE

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20130403

Ref country code: LI

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20120930

REG Reference to a national code

Ref country code: DE

Ref legal event code: R119

Ref document number: 59706608

Country of ref document: DE

Effective date: 20130403

REG Reference to a national code

Ref country code: AT

Ref legal event code: MM01

Ref document number: 214504

Country of ref document: AT

Kind code of ref document: T

Effective date: 20130919

GBPC Gb: european patent ceased through non-payment of renewal fee

Effective date: 20130919

REG Reference to a national code

Ref country code: FR

Ref legal event code: ST

Effective date: 20140530

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: GB

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20130919

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: IT

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20130919

Ref country code: FR

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20130930

Ref country code: AT

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20130919