Method for detecting flame by using statistical analysis multiband infrared sensor

Abstract

The invention belongs to the technical field of fire detection and alarm, and particularly relates to a method for detecting flame by using a multiband infrared sensor through statistical analysis. Firstly, the multi-band infrared sensor collects signals through the multi-path filter circuit, the main control chip obtains the signals and calculates the average value, the overall standard deviation, the discrete coefficient and the like of sample data, then the fluctuation range is calculated and compared with a reference sensor, and then fire alarm or early warning is judged according to different conditions. And finally, a plurality of interference sources are used for simulating a daily environment to carry out an anti-interference test so as to verify the rejection capability of an algorithm on interference factors, ensure stable operation of the detector and reduce false alarms, and the method is suitable for a multi-band flame detector, and has a wide application prospect. And the multi-environment stability is improved.

Description

Technical Field

[0001] The invention belongs to the technical field of fire detection and alarm, and in particular relates to a method for detecting flames by using a statistical analysis multi-band infrared sensor. Background Art

[0002] There are many types of detectors for fire alarm, with smoke, temperature, image, carbon monoxide and spectral radiation as the main detection targets and analysis and judgment. From the perspective of practicality, stability and environmental compatibility, the use of optical analysis characteristics is a detection solution that has regular patterns and fast responses and is worthy of research and continuous optimization.

[0003] Flame detectors that combine multiple bands of infrared and ultraviolet optical sensors can capture and monitor infrared radiation generated by the combustion of ordinary carbon compounds in the atmosphere from an optical perspective according to hardware functions. Most sensor manufacturers and flame detector manufacturers know that infrared sensors can be used to detect infrared radiation in bands such as 3.8, 4.3, 5.0, and 10um. Reasonable signal preprocessing circuits, signal amplification circuits, and multi-order filtering circuits meet the signal acquisition design. However, due to the changing environment and the complexity of the fire scene, the judgment and processing of the signals captured by the sensor has always been a difficult problem in the actual production and development of products. Irregular and unscientific logical algorithms will greatly affect the performance of the designed products. If the interference signal is not extracted and the long-distance flame signal is not analyzed, small flames or long-distance flames will be confused and masked by interference signals due to weak signals, and it will be impossible to achieve both sensitive and stable operation.

[0004] Now, in order to meet the flame detector's ability to adapt to a variety of situations, it is necessary to constantly adjust the program by simulating known environments, adding test conditions and recording the corresponding output data, calculating average power, and performing mathematical analysis and comparisons such as signal size comparison and multiple judgment. Such algorithms have a single typical value, poor representativeness, and poor regular compatibility, resulting in more and more judgment conditions, and cannot cover the complex conditions of the large environment. Therefore, the logic algorithm of signal analysis plays a decisive role in the performance of the detector and even whether it can meet the standards. Summary of the invention

[0005] In view of the technical problems existing in the above-mentioned background technology, the present invention proposes a method for detecting flames by using a multi-band infrared sensor with statistical analysis.

[0006] In order to achieve the above-mentioned purpose, the technical scheme adopted by the present invention is as follows: S1, firstly, the multi-band infrared sensor collects signals through a multi-channel filtering circuit to ensure that the front-stage power supply is stable and has an anti-interference circuit, so that the main control chip collects stable signals, and the hardware circuit design has a reference voltage, which is 1 / 2 of the sensor power supply voltage; S2, secondly, the main control chip continuously obtains signals in the signal acquisition circuit, takes a value every 50ms and counts it into the sample array, collects the external channel voltage value, takes 20 values ​​as a group of sample values, and collects the average value, overall standard deviation and dispersion coefficient of the data; S3, calculates the data fluctuation range of the main flame detector in the same time, and compares it with the reference sensor; S4, judges whether it is a fire alarm or early warning according to different conditions, and the specific judgment conditions are as follows: when the average value of the reference sensor voltage is greater than 2V and less than 3V, the overall standard deviation of the environmental background interference is less than 500, the overall standard deviation of the biological interference is less than 1000, and the number of effective pulses of the main sensor is three times the number of reference pulses, the number of similar pulse times is greater than or equal to 10 times, or the number of pulses reaches 30 When the voltage fluctuation range of the main flame detector is greater than that of the reference sensor and the overall standard deviation is greater than 500, the overall standard deviation of the main sensor is 3 times the overall standard deviation of the environmental background interference and greater than the overall standard deviation of the biological interference, or the overall standard deviation of the main sensor is greater than 1000, the fire alarm warning flag Fire is set to increase by one; when Fire is greater than 2, it is considered to be a valid fire alarm; when the overall standard deviation of the main sensor is greater than 999 and the voltage fluctuation range is greater than or equal to 1V, and the number of pulses of the main sensor is greater than 20, which is more than 3 times the number of pulses of the environmental background and biological interference, it is considered to be a valid fire alarm; when the overall standard deviation of the main sensor is greater than 6 times the overall standard deviation of the environmental background interference and greater than 3 times the overall standard deviation of the biological interference, the fire alarm warning flag Fire is set to increase by one, and when Fire is greater than 2 S5. Finally, anti-environmental interference tests are conducted using sunlight, high-temperature equipment, human waving and walking, incandescent lamps, fluorescent lamps, etc. to simulate daily interference environments, in order to verify the algorithm's ability to eliminate interference factors, ensure the detector's stable operation in complex environments, and reduce the occurrence of false alarms.

[0007] Preferably, the calculation expression of the dispersion coefficient of the overall data in step S2 is: , where X is the original data, is the population standard deviation, where The calculation formula is: , where n is the total amount of data.

[0008] Preferably, in the pulse time processing part in step S4, pulse times less than 50 ms, which are mostly caused by circuit interference, are discarded and not stored.

[0009] Preferably, in the time data processing in step S4, the time of capturing the incoming signal is compared with the last time to determine whether the pulse times are the same, and the similar frequency radiation of the stable flame is identified.

[0010] Preferably, in the step S4, during the main sensor data calculation and processing, the data initialization flag Intdada is set to be incremented by one after each judgment of the fire alarm signal, and the fire alarm warning flag Fire = 0, the data initialization flag Intdada = 0, the pulse number pulseXCnt of the main flame sensor = 0, the environmental background interference pulse number pulseYCnt = 0, the biological interference pulse number pulseZCnt = 0, and the main flame sensor similar pulse time number Same_TCnt = 0 after every 8 judgments; within 8 times, when the overall standard deviation of the main flame sensor is less than 100 and the sampling voltage fluctuation range is less than 100, it is determined that no fire has occurred and the fire alarm warning flag is reset to 0.

[0011] Preferably, the PCA module of the main control chip is set to work in capture mode, samples the jump of the external circuit level input, obtains the pulse time of the sensor, and transmits the command line parameters by VoltageDataRecord (Voltage, X) and CapturedTimeRecord (Time, X), and stores the sampled voltage and capture time in a two-dimensional structure array.

[0012] Compared with the prior art, the advantages and positive effects of the present invention are that the working environment of the detector is not affected by the distance and sensitivity within the effective range, and the stability of multiple environments is greatly improved. Based on many years of market research and customer needs, a large number of tests have been conducted to ensure stable operation in complex environments such as high-temperature equipment factories, frequent movement of workers, direct outdoor sunlight, small flames in small enclosed boxes, 5S quick reports, and near lighting lamps. A set of scientific sample statistical results and concise and clear logical algorithms have been summarized to meet the strict monitoring of fire detection performance. It is suitable for multi-band flame detectors, not limited to dual-band, three-band or even red-ultraviolet composite detectors. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0014] Figure 1 This is an example diagram of three-way stable data; Figure 2 It is a stable state waveform diagram; Figure 3 This is a stable state separation comparison waveform diagram; Figure 4 This is an example diagram of flame occurrence data; Figure 5 This is an example of the waveform when flame occurs Figure 1 ; Figure 6 This is an example of the waveform when flame occurs Figure 2 ; Figure 7 This is an example diagram of interference occurrence data; Figure 8 It is the main program flow chart of the present invention; Fig. 9 Flowchart for flame recognition algorithm analysis. DETAILED DESCRIPTION

[0015] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described below in conjunction with the accompanying drawings and embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0016] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways than those described herein. Therefore, the present invention is not limited to the specific embodiments of the following disclosure.

[0017] Embodiment, the multi-band infrared sensor collects signals through a multi-channel filtering circuit to ensure the stability of the front-stage power supply, and has a certain anti-interference circuit, so that the signals collected by the main control chip will not receive a large number of false signals in a stable environment without fire, interfering with the calculation results. The reference voltage is used as a consideration for determining whether the result is stable in this algorithm, so this algorithm needs a reference voltage based on the hardware circuit design. The reference voltage is required to be 1 / 2 of the sensor power supply voltage, so that the sensor output voltage is stable above and below the reference voltage, avoiding the one-sided data caused by the simple voltage increase. After the fluctuation occurs, the signal of the reference voltage floating up and down is output to the main control chip for analog-to-digital conversion, and the sensor voltage value and pulse width time are collected. The algorithm used includes mathematical related judgment and analysis of pulse frequency, pulse width time, average power comparison, etc., the data in the collection array is compared with the law, and multiple mathematical calculations are integrated for fuzzy logic programming. The basic characteristics of the sampled voltage fluctuation range and the discrete distribution degree of the voltage value in the data set are represented by the extreme difference in the same time period as the primary condition, and the condition judgment is divided according to the flame program type.

[0018] Now let's take the set main sensor data as an example to explain. The main control chip continuously obtains signals in the signal acquisition circuit, takes a value every 50ms and counts it into the sample array, collects the external channel voltage value, and uses 20 values ​​as a group of sample values ​​in the program. The calculation results are all integers, and the calculation expression of the overall data discrete coefficient is: , where X is the original data, is the population standard deviation, where The calculation formula is: , where n is the total amount of data.

[0019] When in stable state, the flame sensor output value is 2660,2625,2605,2590,2595,2565,2560,2555,2550,2555,2550,2545,2540,2535,2540,2575,2595,2590,2475,2580, such as Figure 1As shown. First, calculate the mean of the data set: (2660, 2625, 2605, 2590, 2595, 2565, 2560, 2555, 2550, 2555, 2550, 2545, 2540, 2535, 2540, 2575, 2595, 2590, 2475, 2580) / 20 = 2569. Then, calculate the degree of deviation of each value from the mean and get (91, 56, 36, 21, 26, -4, -9, -14, -19, -14, -19, -24, -29, -34, -29, 6, 26, 21, -94, 11). Square these deviations and then find the average to get the variance: (91²+56²+36²+21²+26²+(-4²)+(-9²)+(-14²)+(-19²)+(-14²)+(-19²)+(-24²)+(-29²)+(-34²)+(-29²)+6²+26²+21²+(-94²)+11²) / 20=1428. Take the square root of the variance to get the standard deviation: √1503=37. At this time, the mean of this sample is 2569, and the overall standard deviation is 37. Multiply the ratio of the standard deviation to the mean by 100% to get the dispersion coefficient: 37 / 2569*100%=1.4%. From the above calculation results, it can be seen that the calculation results of the variance, standard deviation and dispersion coefficient of this set of data are 1428, 37 and 1.4% respectively. The deviation here is relatively small, and the values ​​of the standard deviation and the coefficient of dispersion indicate that the data is relatively concentrated, with small overall fluctuations. From the data to verify the waveform image, when the two sets of sensors are in the monitoring state and there is no flame or obvious interference in the environment, the waveform is relatively smooth, such as Figure 2 , Figure 3 As shown, the serial port data fluctuates within 500mV above or below the reference voltage.

[0020] When a flame is detected, the flame sensor output value is 535,435,285,240,255,1005,1835,2485,2335,1850,1470,1175,955,1145,1905,2575,3085,3465,4430,4555, such as Figure 4 As shown. First, calculate the average value of the data set: the average number of samples is 1801, the overall variance is 1657754, the overall standard deviation is 1287, and the coefficient of dispersion is 71.4%. When the sensor detects a flame, the waveform changes dramatically, as shown in Figure 5 , Figure 6As shown in the figure, due to the flickering characteristics of the flame during the combustion process, the infrared light signal is continuously radiated to the surrounding space in the CO2 atmosphere, so that the acquisition circuit continuously captures the peak and trough transformation waveforms of different amplitudes. The discrete degree of the voltage value captured during the transformation process is relatively large, and the sample data fluctuates more than 1V above and below the reference voltage. This argument describes the calculation of the flame sensor's response to the characteristics of the flame light. From the data analysis and waveform comparison, it can be concluded that when the sample voltage value is significantly different from the average value, that is, the corresponding sensor sampling data fluctuates greatly near the reference voltage, showing that the distribution of the value size is relatively dispersed, the sum of the squares of the differences between each sampling voltage value and the average value is large, and the dispersion coefficient is large; when the sampling data distribution is relatively concentrated, it is around the size of the reference voltage value, and the sum of the squares of the differences between each sampling data and the average value is small. Therefore, the larger the dispersion coefficient, the greater the fluctuation amplitude of the sample value, indicating that the flame characteristics are detected; the smaller the dispersion coefficient, the smaller the fluctuation amplitude of the sample value, indicating that the sensor output is in a stable state. This algorithm of discrete coefficient is different from calculating the average power for judgment. The two sets of data with the same or similar average power may not have the same degree of numerical dispersion. Since the environment in which the detector is located is quite complex and changeable, modern industry is developed, there are many types of equipment, and the range of movement of personnel in the factory is large, causing a variety of unpredictable interferences and frequent instantaneous interferences. When using the average power method, the average power cannot reflect the fluctuation state of the value, and it is very likely that the false alarm value will be used as the alarm value, or the alarm value will be submerged in the false alarm value.

[0021] The above discussion is a method for normalizing the collected data set. The actual judgment of the flame still needs further judgment. This is the outstanding significance of the cooperation of multiple multi-band sensors. The hazards of fire should not be underestimated. Missed reports damage life and property, but frequent false alarms also greatly occupy the manpower cost. The following describes the key ideas and algorithms for eliminating false alarms. The main control chip converts the ADC channel in turn to collect the sensor voltage value Voltage=Get_ADC10bitResult(x)*k obtained by the external sampling circuit. The PCA module of the main control chip is set to work in capture mode, sampling the jump of the external circuit level input to obtain the pulse time of the sensor. The command line parameters are passed by VoltageDataRecord(Voltage,X) and CapturedTimeRecord(Time,X), and the sampled voltage and capture time are stored in a two-dimensional structure array. At this time, the pulse time less than 50ms is automatically discarded and not stored, because it is found through experiments that the pulse time less than 10ms is mostly obtained by the interference of the circuit itself. The LastTime variable is set in the time data processing part. The incoming time is captured and compared with the previous time each time. If the pulse time is not greater than 50ms or less than 50ms, the pulse time is considered to be the same. Same_TCnt is incremented by one each time it occurs. If Same_TCnt does not continue to occur and the current value is less than 5, the initial value will be restored to 0. It is considered that the previous signal was an interference signal, and there is no flame at this time. This algorithm has a good recognition effect on stable flames, such as alcohol lamps, lighters, candles, or small flames in the early stage. The pulseXCnt variable is set to count each valid pulse time to prepare for the final conditional judgment.

[0022] In the flame determination analysis function, a for loop is used to traverse the two-dimensional array, and each sensor data set is processed in detail and compared with each other. For the reference sensor, the maximum and minimum values ​​are retained. For the main sensor, the maximum and minimum values ​​are removed to obtain the discrete coefficient. The purpose is to eliminate interference signals and effectively reduce false alarms. The sample data set capacity is Ndata, which is generally set to 20. It can be adjusted appropriately according to the characteristics of sensors of different brands. This algorithm performs a flame characteristic determination after the data collected by each sensor reaches 20.

[0023] According to the test data, when the average voltage of the reference sensor is greater than 2V and less than 3V, the overall standard deviation of the environmental background interference is less than 500, and the overall standard deviation of the biological interference is less than 1000 ((meanY>2000&&meanY<3000)&&(meanZ>2000&&meanZ<3000)&&vartY<500&&vartZ<1000)), there are fewer interference sources in the detection environment and the reference sensor does not fluctuate significantly. Under this condition, if the number of effective pulses of the main sensor is three times the number of reference pulses (pulseXt>=3*pulseYCnt), the number of similar pulse times is greater than or equal to 10 times (Same_TCnt>=10), or the number of pulses reaches more than 30 times (pulseXCnt>30), it can be determined that the current is a valid fire alarm.

[0024] If the fluctuation range of the main flame detector is greater than that of the reference sensor and the overall standard deviation is greater than 500 (RangeX>RangeY)&&(RangeX>RangeY)&&(vartX>500) at the same time, when the overall standard deviation of the main sensor is 3 times the overall standard deviation of the environmental background interference and greater than the overall standard deviation of the biological interference, or the overall standard deviation of the main sensor is greater than 1000, (vartX>2*vartY)&&(vartX>vartZ)||(vartX>1000), in order to prevent misjudgment, the fire alarm warning flag Fire is set to increase by one, and the alarm will not be issued immediately. At this time, it is in the warning stage. When Fire is greater than 2, that is, if 20 sample values ​​are taken twice and both meet this condition, it is determined that the current is a valid fire alarm.

[0025] If the overall standard deviation of the main sensor is greater than 999 and the voltage fluctuation range is greater than or equal to 1V, the sensor has detected a large flame that is burning relatively violently. There is no need to determine whether the reference sensor is stable. However, a certain proportional coefficient must be satisfied between them. When the pulse number of the main sensor is greater than 20, which is more than 3 times the number of environmental background and biological interference pulses, it is determined that the current is a valid fire alarm.

[0026] When the overall standard deviation of the main sensor is greater than 6 times the overall standard deviation of the environmental background interference and greater than 3 times the overall standard deviation of the biological interference, the fire alarm warning flag Fire will be incremented by 1. When Fire is greater than 2, it is considered to be a valid fire alarm.

[0027] For a real flame, the combustion characteristics will always exist, which is different from the signal value that meets the flame characteristics appearing once due to interference coincidence, and the matching values ​​will not be continuous. Therefore, after adding the processing of removing the maximum and minimum values ​​in the main sensor data calculation process, the fire alarm signal analysis function is also added to set the data initialization flag Intdada to increment by one after each fire alarm signal is judged. After every 8 judgments, the fire alarm warning flag Fire=0, the data initialization flag Intdada=0, the pulse number of the main flame sensor pulseXCnt=0; the number of environmental background interference pulses pulseYCnt=0, the number of biological interference pulses pulseZCnt=0, and the number of similar pulse times of the main flame sensor Same_TCnt=0. Within 8 times, when the overall standard deviation of the main flame sensor is less than 100 and the sampling voltage fluctuation range is less than 100, regardless of whether interference exists, it can be determined that no fire has occurred. At this time, the fire alarm warning flag is reset to 0. The purpose is to prevent multiple erroneous signals from accumulating for a long time and failing to clearly distinguish the actual characteristics of the flame. Figure 7 As shown, the data is in an interference state. Although the overall standard deviation of the main sensor reaches 346 and the fluctuation range exceeds 1V, the reference sensor detects biological interference and the overall standard deviation value calculated by the collected data reaches 540, with a fluctuation range of more than 2V, which is much higher than the main sensor signal. However, the average power of the two is 2652 and 2639 respectively. It can be seen that the information obtained by average power has considerable limitations. Finally, the main program flow of the present invention is as follows Figure 8 As shown in Figure 2, the process of the flame recognition algorithm is as follows: Fig. 9 shown.

[0028] It is easy for flame detectors to alarm, but difficult to prevent false alarms. Flame detectors are relatively easy to detect flames and trigger alarms, but they may face greater challenges in preventing false alarms. Preventing false alarms of flame detectors requires comprehensive consideration of multiple factors, including equipment selection, layout optimization, mobile detection, etc. This experiment verifies the anti-interference ability of the flame judgment algorithm. Sunlight, high-temperature equipment, human waving and walking, incandescent lamps, and fluorescent lamps are used to interfere with the detectors respectively, simulating the interference encountered in daily use environments. As shown in the table, the algorithm can eliminate interference factors. Table 1 below is an interference test under a simulated environment.

[0029] Table 1: Simulated environmental interference test

[0030]

[0031] The above description is only a preferred embodiment of the present invention and does not limit the present invention in other forms. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A method for detecting flames using a multi-band infrared sensor using statistical analysis, characterized in that: The following steps are involved: S1. First, the multi-band infrared sensor collects signals through a multi-channel filtering circuit to ensure that the front-stage power supply is stable and has an anti-interference circuit, so that the main control chip collects stable signals. The hardware circuit design has a reference voltage, which is 1 / 2 of the sensor power supply voltage; S2. Secondly, the main control chip continuously acquires signals in the signal acquisition circuit, takes values ​​every 50ms and counts them into the sample array, collects the voltage value of the external channel, takes 20 values ​​as a group of sample values, and collects the average value, overall standard deviation and dispersion coefficient of the data; S3, calculate the data fluctuation range of the main flame detector in the same time, and compare it with the reference sensor; S4. Determine whether it is a fire alarm or a warning according to different conditions. The specific judgment conditions are as follows: When the average voltage of the reference sensor is greater than 2V and less than 3V, the overall standard deviation of the environmental background interference is less than 500, the overall standard deviation of the biological interference is less than 1000, and the number of effective pulses of the main sensor is three times the number of reference pulses, the number of similar pulse times is greater than or equal to 10 times, or the number of pulses reaches more than 30 times, it is considered to be a valid fire alarm; When the fluctuation range of the main flame detector is greater than that of the reference sensor and the overall standard deviation is greater than 500, the overall standard deviation of the main sensor is 3 times the overall standard deviation of the environmental background interference and greater than the overall standard deviation of the biological interference, or the overall standard deviation of the main sensor is greater than 1000, the fire alarm warning flag Fire is set to increase by one; when Fire is greater than 2, it is considered to be a valid fire alarm; When the overall standard deviation of the main sensor is greater than 999 and the voltage fluctuation range is greater than or equal to 1V, and the pulse number of the main sensor is greater than 20, which is more than 3 times the number of environmental background and biological interference pulses, it is considered to be a valid fire alarm; When the overall standard deviation of the main sensor is greater than 6 times the overall standard deviation of the environmental background interference and greater than 3 times the overall standard deviation of the biological interference, the fire alarm warning flag Fire will increase by 1. When Fire is greater than 2, it is considered to be a valid fire alarm. S5. Finally, we used sunlight, high-temperature equipment, human waving and walking, incandescent lamps, and fluorescent lamps to simulate daily interference environments to conduct anti-environmental interference tests to verify the algorithm's ability to eliminate interference factors, ensure the detector's stable operation in complex environments, and reduce the occurrence of false alarms.

2. The method for detecting flames using a multi-band infrared sensor using statistical analysis according to claim 1, characterized in that: The calculation expression of the dispersion coefficient of the overall data in step S2 is: , where X is the original data, is the population standard deviation, where The calculation formula is: , where n is the total amount of data.

3. The method for detecting flames using a multi-band infrared sensor using statistical analysis according to claim 1, characterized in that: In the pulse time processing part in step S4, the pulse time less than 50 ms due to the interference of the circuit itself is discarded and not stored.

4. The method for detecting flames using a multi-band infrared sensor with statistical analysis according to claim 1, characterized in that: In the time data processing in step S4, the incoming capture time is compared with the previous time to determine whether the pulse time is the same, and the similar frequency radiation of the stable flame is identified.

5. The method for detecting flames using a multi-band infrared sensor with statistical analysis according to claim 1, characterized in that: In the step S4, during the main sensor data calculation and processing, the data initialization flag Intdada is set to be incremented by one after each fire alarm signal is judged, and the fire alarm warning flag Fire = 0, the data initialization flag Intdada = 0, the pulse number pulseXCnt of the main flame sensor = 0, the environmental background interference pulse number pulseYCnt = 0, the biological interference pulse number pulseZCnt = 0, and the main flame sensor similar pulse time number Same_TCnt = 0 after each 8 judgments; within 8 times, when the overall standard deviation of the main flame sensor is less than 100 and the sampling voltage fluctuation range is less than 100, it is determined that no fire has occurred and the fire alarm warning flag is reset to 0.

6. The method for detecting flames using a multi-band infrared sensor with statistical analysis according to claim 1, characterized in that: The PCA module of the main control chip is set to work in capture mode, sampling the jump of the external circuit level input to obtain the pulse time of the sensor. The command lines are passed by VoltageDataRecord (Voltage, X) and CapturedTimeRecord (Time, X) to store the sampled voltage and capture time in a two-dimensional structure array.

Images