Flame detection system and method for generating a learning model

Abstract

To make it possible to ascertain a degree of deterioration of a measuring unit in a flame detection system.SOLUTION: A flame detection system according to the present invention comprises: a measuring unit that measures a light intensity with a sensor; a flame detection unit that detects a flame; and a deterioration degree determination unit that determines the degree of deterioration of the measuring unit. A stable term specifying unit is provided to specify a term in which an environmental factor is stable, and a degree of deterioration is determined based on night vision noise which is an intensity of light measured during that term. For instance, a deterioration of an amplifier is determined based on an average value of the night vision noise, and a deterioration of a sensor is determined based on the index of its fluctuation. In addition, the stable time period may also be identified based on the detection results of other sensors such as an optical camera, a microphone, an object detection sensor, and a vibration sensor.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a technique for maintaining a flame detection system.

Background Art

[0002] There are cases where a disaster prevention system may malfunction due to deterioration. Various techniques have been proposed to prevent such inconveniences. For example, in Patent Document 1, when a fire is detected, the current value flowing through a signal line connecting a terminal device that outputs a fire signal to a disaster prevention receiving board and the disaster prevention receiving board is monitored, and when the current value changes significantly, it is determined as a sign of a failure caused by insulation deterioration of the signal line, etc., and a disaster prevention system equipped with a function of outputting a sign warning is proposed.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] There is a flame detection system that detects flames with a sensor that is sensitive to the light emitted by the flames. One of the causes of the malfunction of such a flame detection system is the deterioration of the components of the measurement unit that measures the light intensity with the sensor. For example, when the sensor of the measurement unit provided in the flame detection system deteriorates, the amplitude value of the signal output from the sensor may show irregular fluctuations regardless of the presence or absence of light. Also, when the amplifier that amplifies the signal generated by the sensor deteriorates, the amplitude value of the signal from the sensor amplified by the amplifier may constantly show a high value regardless of the presence or absence of light.

[0005] As the above-mentioned components deteriorate, the accuracy of the flame detection system's flame detection judgment decreases. Therefore, if the degree of deterioration of the measurement unit can be known, the accuracy of the flame detection judgment required of the flame detection system can be maintained by replacing the parts, etc.

[0006] In view of these circumstances, the present invention aims to make it possible to determine the degree of deterioration of the measuring unit of a flame detection system. [Means for solving the problem]

[0007] To solve the above problems, the present invention provides, in a first embodiment, a flame detection system comprising: a measuring unit that measures the intensity of light using a sensor; a flame detection unit that detects a flame based on the intensity of light measured by the measuring unit; a stable period identification unit that identifies a period during which environmental factors affecting the measurement results of the measuring unit are stable; and a storage unit that stores data indicating the intensity of light measured by the measuring unit during the period identified by the stable period identification unit as data indicating noise.

[0008] According to the first embodiment of the flame detection system, the degree of deterioration of the measuring unit of the flame detection system can be determined by the noise indicated by the stored data.

[0009] In the flame detection system according to the first embodiment, a second embodiment may be adopted in which the stability period identification unit identifies a period in which environmental factors are stable based on the measurement results of the measurement unit.

[0010] According to the flame detection system of the second embodiment, there is no need to install a separate sensor to identify the period during which environmental factors are stable.

[0011] In the flame detection system according to the first embodiment, when the sensor is a first sensor, a third embodiment may be adopted in which a second sensor different from the first sensor that measures the physical quantity of the surroundings is provided, and the stability period identification unit identifies a period in which the environmental factors are stable based on the measurement results of the second sensor.

[0012] According to the third embodiment of the flame detection system, by including a second sensor, the period during which ambient light is stably irradiated onto the first sensor can be excluded from the period during which environmental factors are stable. Therefore, the reliability of the stored noise data is higher compared to the case without the second sensor.

[0013] In the third embodiment of the flame detection system, a fourth embodiment may be adopted in which the second sensor is a sensor provided in a device of the same type as the measuring unit, located within a predetermined distance from the measuring unit.

[0014] For example, in the case of an existing flame detection system having multiple identical measuring units arranged at intervals along the direction of travel of a vehicle, the flame detection system according to the fourth embodiment can be realized without making any physical changes to the existing flame detection system by identifying the period during which the environmental factors of a certain measuring unit are stable using sensors of other measuring units located near that measuring unit.

[0015] A fifth embodiment of the flame detection system according to any of the first to fourth embodiments may include a thermometer for measuring temperature and a correction unit for correcting the light intensity measured by the measurement unit during the period specified by the stabilization period specification unit based on the temperature measured by the thermometer, wherein the storage unit stores data indicating the intensity after correction by the correction unit as data indicating noise.

[0016] According to the fifth embodiment of the flame detection system, even if the light intensity measured by the measuring unit is affected by temperature and results in errors, the reliability of the data indicating noise is maintained.

[0017] Furthermore, in a sixth embodiment of the present invention, a flame detection system according to any of the first to fifth embodiments may be provided with a degradation determination unit that determines the degree of degradation of the measurement unit based on the change in noise over time indicated by the data stored in the storage unit.

[0018] According to the flame detection system according to the sixth aspect, the degree of deterioration of the measurement unit of the flame detection system is determined.

Advantages of the Invention

[0019] According to the present invention, the degree of deterioration of the measurement unit of the flame detection system can be known.

Brief Description of the Drawings

[0020] [Figure 1] A diagram showing the overall configuration of a flame detection system according to an embodiment. [Figure 2] A diagram schematically showing the hardware configuration of a flame detector according to an embodiment. [Figure 3] A diagram schematically showing the functional configuration of a flame detector according to an embodiment. [Figure 4] A diagram illustrating the configuration of a measurement value log table according to an embodiment. [Figure 5] A diagram illustrating the configuration of a temperature correction constant table according to an embodiment. [Figure 6] A graph for explaining a method by which a stabilization period specifying unit according to an embodiment specifies a stabilization period. [Figure 7] A diagram schematically showing the hardware configuration of a disaster prevention receiving board according to an embodiment. [Figure 8] A diagram schematically showing the functional configuration of a disaster prevention receiving board according to an embodiment. [Figure 9] A diagram illustrating the configuration of a dark vision noise table according to an embodiment. [Figure 10] A graph for explaining a method by which a deterioration degree determination device according to an embodiment determines a deterioration degree and estimates a deteriorated component. [Figure 11] A graph for explaining a method by which a deterioration degree determination device according to an embodiment determines a deterioration degree and estimates a deteriorated component. [Figure 12] A diagram illustrating the configuration of a deterioration diagnosis result table according to an embodiment. [Figure 13] A diagram illustrating a deterioration diagnosis screen displayed by a terminal device according to an embodiment. [Figure 14]A schematic diagram showing the hardware configuration of a flame detector according to one modified example. [Figure 15] A schematic diagram showing the functional configuration of a flame detector according to one modified example.

[0021] [Embodiment] The following describes a flame detection system 1 according to one embodiment of the present invention. Figure 1 shows the overall configuration of the flame detection system 1. The flame detection system 1 is a system for detecting flames that occur in a tunnel TN.

[0022] The flame detection system 1 comprises n flame detectors, namely flame detectors 11(1), 11(2), 11(3), ..., 11(n), installed at approximately equal intervals along the direction of vehicle travel inside the tunnel TN. Hereinafter, these n flame detectors will be collectively referred to as flame detector 11.

[0023] Each flame detector 11 effectively integrates two flame detectors. Specifically, each flame detector 11 integrates a flame detector that monitors a predetermined area to the right of the flame detector 11 (hereinafter referred to as the "right-side flame detector") and a flame detector that monitors a predetermined area to the left (hereinafter referred to as the "left-side flame detector").

[0024] The tunnel TN is divided into monitoring areas A(1), A(2), A(3), ..., A(n-1). Hereafter, these (n-1) monitoring areas will be collectively referred to as monitoring area A. Each of monitoring area A is monitored by two adjacent flame detectors 11. For example, monitoring area A(1) is monitored by the left flame detector of flame detector 11(1) and the right flame detector of flame detector 11(2). Therefore, even if one of the two adjacent flame detectors 11 fails, monitoring of monitoring area A will not be interrupted unless the other also fails at the same time.

[0025] The flame detection system 1 includes flame detectors 11, a disaster prevention receiving panel 12 that is connected to each of the flame detectors 11, a server device 13 that is connected to the disaster prevention receiving panel 12, and a terminal device 14 that can communicate with the server device 13.

[0026] Figure 2 is a schematic diagram showing the hardware configuration of the flame detector 11. The flame detector 11 comprises a computer 101, four sensors connected to the computer 101, namely sensor 111R, sensor 112R, sensor 111L, and sensor 112L, four amplifiers corresponding to each of the four sensors, namely amplifiers 113R, amplifier 114R, amplifier 113L, and amplifier 114L, and a thermometer 115 connected to the computer 101.

[0027] Sensors 111R and 112R are optical sensors for monitoring the monitoring area A on the right side as viewed from the flame detector 11. Sensors 111L and 112L are optical sensors for monitoring the monitoring area A on the left side as viewed from the flame detector 11.

[0028] Sensors 111R and 111L are long-wavelength optical sensors that respond with high sensitivity to the longer wavelength range emitted by a flame (heat source). For example, optical sensors using pyroelectric elements are employed as sensors 111R and 111L. Hereafter, sensors 111R and 111L will be collectively referred to as sensor 111.

[0029] Sensors 112R and 112L are short-wavelength optical sensors that respond with high sensitivity to the short-wavelength wavelength range emitted by a flame (heat source). For example, optical sensors using photodiodes are employed as sensors 112R and 112L. Hereafter, sensors 112R and 112L will be collectively referred to as sensor 112.

[0030] Amplifier 113R is connected between computer 101 and sensor 111R and amplifies the signal generated by sensor 111R in response to light. Amplifier 114R is connected between computer 101 and sensor 112R and amplifies the signal generated by sensor 112R in response to light. Amplifier 113L is connected between computer 101 and sensor 111L and amplifies the signal generated by sensor 111L in response to light. Amplifier 114L is connected between computer 101 and sensor 112L and amplifies the signal generated by sensor 112R in response to light.

[0031] Sensor 111R, amplifier 113R, the wiring and contacts connecting them, and the wiring and contacts connecting amplifier 113R and computer 101 constitute one measuring unit. Also, sensor 112R, amplifier 114R, the wiring and contacts connecting them, and the wiring and contacts connecting amplifier 114R and computer 101 constitute one measuring unit. Also, sensor 111L, amplifier 113L, the wiring and contacts connecting them, and the wiring and contacts connecting amplifier 113L and computer 101 constitute one measuring unit. Also, sensor 112L, amplifier 114L, the wiring and contacts connecting them, and the wiring and contacts connecting amplifier 114L and computer 101 constitute one measuring unit.

[0032] In the following description, when referring to the signal output from sensor 111 or sensor 112, unless otherwise specified, it means the signal output from sensor 111 or sensor 112 and amplified by amplifier 113 or amplifier 114.

[0033] The thermometer 115 measures the temperature inside the casing of the flame detector 11 and outputs temperature data indicating the measured temperature to the computer 101.

[0034] The computer 101 includes a processor 1011 that processes data according to a program, a memory 1012 that stores various data including the program, an input / output interface 1013 that receives signal input from four sensors via four amplifiers and also receives temperature data input from a thermometer 115, and a communication interface 1014 that communicates data with the disaster prevention receiving panel 12.

[0035] In addition to the components shown in Figure 2, the flame detector 11 also includes components such as an A / D converter that converts the analog signals output by the four sensors into digital signals. However, these components are unrelated to the features of the present invention and are therefore omitted from Figure 2, as well as from the following description.

[0036] Figure 3 is a schematic diagram showing the functional configuration of the flame detector 11. That is, the processor 1011 of the computer 101 executes processing according to the program of this embodiment, thereby realizing the flame detector 11 equipped with the determination device indicated by reference numeral 116 in Figure 3. The functional configuration of the determination device 116 will be described below.

[0037] The determination device 116 includes a determination device 116R that determines flame detection using signals output from sensors 111R and 112R, and a determination device 116L that determines flame detection using signals output from sensors 111L and 112L. Since the configurations of determination devices 116R and 116L are the same, the configuration of determination device 116R will be described below as an example, and the description of the configuration of determination device 116L will be omitted.

[0038] The determination device 116R includes a storage unit 1161R, an acquisition unit 1162R, a correction unit 1163R, a stabilization period identification unit 1164R, a flame detection unit 1165R, a transmission / reception unit 1166R, and a timing unit 1167R.

[0039] The storage unit 1161R is implemented by memory 1012, which operates under the control of processor 1011. The storage unit 1161R stores various types of data. The data stored by the storage unit 1161R includes a measurement log table, a temperature correction constant table, and steady-state correction constant data.

[0040] Figure 4 is an example diagram illustrating the configuration of the measurement log table. A measurement log table is prepared according to each of the sensors 111R and 112R. The measurement log table has columns for "Time", "Temperature", "Intensity (Measured)", "Intensity (Temperature Corrected)", and "Intensity (Steady-State Corrected)". The "Time" column stores the time measured by the timing unit 1167R when a signal is output from the sensor. The "Temperature" column stores the temperature measured by the thermometer 115 when a signal is output from the sensor. The "Intensity (Measured)" column stores the amplitude value of the signal output from the sensor. The "Intensity (Temperature Corrected)" column stores the value obtained by the correction unit 1163R correcting the amplitude value stored in the "Intensity (Measured)" column using a temperature correction constant described later. The "Intensity (Steady-State Corrected)" column stores the value obtained by the correction unit 1163R correcting the amplitude value stored in the "Intensity (Temperature Corrected)" column using a steady-state correction constant described later.

[0041] Figure 5 illustrates the configuration of the temperature compensation constant table. A temperature compensation constant table is prepared for each of the sensors 111R and 112R. The temperature compensation constant table has columns for "Temperature" and "Temperature Compensation Constant". The "Temperature" column stores various temperatures. The "Temperature Compensation Constant" column stores the temperature compensation constant corresponding to the temperature stored in the "Temperature" column. The amplitude values ​​of the signals generated by sensors 111R and 112R are biased by the effect of temperature. The temperature compensation constant is a value used to cancel the bias due to the effect of temperature by subtracting it from the amplitude value of the signal generated by sensor 111R or sensor 112R, and is a value that has been measured in advance through experiments using sensor 111R or sensor 112R.

[0042] The steady-state correction constant data is data that indicates the steady-state correction constant. The amplitude values ​​of the signals generated by sensors 111R and 112R are subject to biases unrelated to temperature, in addition to the bias due to the temperature effect described above. The steady-state correction constant is a value that cancels out the bias unrelated to temperature by subtracting it from the amplitude value of the signal generated by sensor 111R or sensor 112R. In this embodiment, the average value of the amplitude values ​​shown in the night vision noise data described later is used as the steady-state correction constant.

[0043] The explanation of the functional configuration of the flame detector 11 (Figure 3) continues. The acquisition unit 1162R is implemented by an input / output interface 1013 that operates under the control of the processor 1011. The acquisition unit 1162R continuously acquires the signals output from sensor 111R and the signals output from sensor 112R. The amplitude values ​​of these signals acquired by the acquisition unit 1162R are stored in a measurement log table along with the time and temperature at that time.

[0044] The correction unit 1163R is implemented by the processor 1011. When new data is stored in the "Time," "Temperature," and "Intensity (Measured)" columns of the measurement log table, the correction unit 1163R refers to the temperature correction constant table to identify the temperature correction constant corresponding to the temperature stored in the "Temperature" column, and stores the value obtained by subtracting the temperature correction constant from the amplitude value stored in the "Intensity (Measured)" column in the "Intensity (After Temperature Correction)" column. Subsequently, the correction unit 1163R stores the value obtained by subtracting the steady-state correction constant from the value stored in the "Intensity (After Temperature Correction)" column in the "Intensity (After Steady-State Correction)" column.

[0045] The stability period identification unit 1164R is implemented by the processor 1011. For each of the sensors 111R and 112R, the stability period identification unit 1164R identifies the period during which environmental factors affecting the measurement results of those sensors are stable (hereinafter referred to as the "stability period"), based on the data stored in the measurement value log tables corresponding to those sensors.

[0046] Figure 6 is a graph illustrating how the stabilization period identification unit 1164R identifies the stabilization period. The graph in Figure 6 shows the change in light intensity over time, indicated by the amplitude value stored in the "Intensity (Measured)" column of the measurement log table corresponding to the sensor 111, for example. Since the graph in Figure 6 covers a short period of a few minutes, the value stored in the "Intensity (Temperature Corrected)" column or the "Intensity (Steady-State Corrected)" column may be used instead of the amplitude value stored in the "Intensity (Measured)" column.

[0047] The stability period identification unit 1164R identifies a period as a stability period based on the data stored in the measurement log table, during which the intensity fluctuation remains within a predetermined threshold for a predetermined time (e.g., 1 minute) or longer. In the example graph shown in Figure 6, the stability period identification unit 1164R identifies period Q1 as a stability period.

[0048] Let's continue explaining the functional configuration of the flame detector 11 (Figure 3). The flame detection unit 1165R is implemented by the processor 1011. The flame detection unit 1165R refers to the measurement log table and, if it determines that the amplitude value of the signal output from sensor 111R (value in the "Intensity (after steady-state correction)" column) and the amplitude value of the signal output from sensor 112R (value in the "Intensity (after steady-state correction)" column) satisfy predetermined conditions, it stores data indicating that a flame has been detected in the storage unit 1161R. Conversely, if the flame detection unit 1165R determines that the conditions are not met, it stores data indicating that no flame has been detected in the storage unit 1161R.

[0049] The following are examples of conditions used by the flame detection unit 1165R to determine whether a flame has been detected. (Condition 1) The amplitude value of the signal output from sensor 111R is greater than or equal to threshold T1. (Condition 2) The amplitude value of the signal output from sensor 112R is greater than or equal to threshold T2. (Condition 3) The ratio of the amplitude value of the signal output from sensor 111R to the amplitude value of the signal output from sensor 112R is greater than or equal to threshold T3 and less than or equal to threshold T4 (however, T3 <T4)である。

[0050] The flame detection unit 1165R determines that a flame has occurred if all of the above conditions 1 to 3 have been met a predetermined number of times or more within a predetermined period of time (for example, 10 seconds) in the past.

[0051] The transmitting / receiving unit 1166R is implemented by a communication interface 1014 that operates under the control of the processor 1011. The transmitting / receiving unit 1166R continuously outputs a flame detection signal to the disaster prevention receiving panel 12 while data indicating that a flame has been detected is stored in the storage unit 1161R. The transmitting / receiving unit 1166R also outputs data from the measurement value log table that falls within the stable period identified by the stable period identification unit 1164R as noise data (hereinafter referred to as "night vision noise data") to the disaster prevention receiving panel 12.

[0052] Furthermore, the transmitting / receiving unit 1166R receives steady-state correction constant data transmitted from the disaster prevention receiving panel 12. The steady-state correction constant data received by the transmitting / receiving unit 1166R overwrites the steady-state correction constant data already stored in the storage unit 1161R.

[0053] The timing unit 1167R is implemented by the processor 1011. The timing unit 1167R continuously measures the elapsed time from a reference time, determines the current time, and generates a time signal indicating the determined current time.

[0054] The fire prevention receiver panel 12 (Figure 1), which constitutes the flame detection system 1, is installed inside the tunnel TN and, upon receiving a flame detection signal from the flame detector 11, alerts people in the vicinity through displays and sounds, and also sends a notification to the server device 13 that a flame has been detected. In addition to the functions of such a general fire prevention receiver panel, the fire prevention receiver panel 12 also has a function to determine the degree of deterioration of the flame detector 11 based on night vision noise data output from the flame detector 11.

[0055] Figure 7 is a schematic diagram showing the hardware configuration of the disaster prevention receiving panel 12. The disaster prevention receiving panel 12 includes a computer 102, a display 121 connected to the computer 102, and an operation unit 122.

[0056] The computer 102 includes a processor 1021 that processes data according to a program, a memory 1022 that stores various data including the program, an input / output interface 1023 that inputs and outputs signals to and from the display 121 and the operation unit 122, and a communication interface 1024 that communicates data with n flame detectors 11 and the server device 13.

[0057] Figure 8 is a schematic diagram showing the functional configuration of the disaster prevention receiving panel 12. That is, the processor 1021 of the computer 102 executes processing according to the program of this embodiment, thereby realizing the disaster prevention receiving panel 12 equipped with the deterioration degree determination device shown by reference numeral 123 in Figure 8. The functional configuration of the deterioration degree determination device 123 will be described below.

[0058] The deterioration degree determination device 123 includes a storage unit 1231, an acquisition unit 1232, a deterioration degree determination unit 1233, a transmission unit 1234, a display control unit 1235, and an operation reception unit 1236.

[0059] The storage unit 1231 is implemented by memory 1022, which operates under the control of processor 1021, and stores various types of data. For example, night vision noise data transmitted from flame detector 11 and acquired by acquisition unit 1232 is stored in the storage unit 1231. Figure 9 is an example diagram illustrating the configuration of a table (hereinafter referred to as the "night vision noise table") that the storage unit 1231 stores for storing night vision noise data. The storage unit 1231 stores a night vision noise table corresponding to each of the four sensors for each of the multiple flame detectors 11.

[0060] The night vision noise table has the following columns: "Date and Time," "Night Vision Noise Data," "Average Value," "Standard Deviation," and "Data Anomaly." The "Date and Time" column stores, for example, the date and time of the first day of the night vision period indicated by the night vision noise data. The "Night Vision Noise Data" column stores the night vision noise data. Note that each night vision noise data entry contains multiple records, as it is extracted from the measurement log table (Figure 4). The "Average Value" column stores the average value of the values ​​stored in the "Intensity (Temperature Corrected)" column of the night vision noise data. The "Standard Deviation" column stores the standard deviation of the values ​​stored in the "Intensity (Temperature Corrected)" column of the night vision noise data. The "Data Anomaly" column stores data indicating whether the night vision noise data stored in the "Night Vision Noise Data" column is anomaly data.

[0061] Let's return to the explanation of the functional configuration of the disaster prevention receiver panel 12 (Figure 8). The acquisition unit 1232 is implemented by a communication interface 1024 that operates under the control of the processor 1021. The acquisition unit 1232 acquires flame detection signals and night vision noise data from each of the n flame detectors 11. The night vision noise data acquired by the acquisition unit 1232 is stored in a night vision noise table.

[0062] The degradation level determination unit 1233 is implemented by the processor 1021. When new night vision noise data is stored in any of the night vision noise tables, the degradation level determination unit 1233 calculates the average value and standard deviation of the values ​​stored in the "Intensity (after temperature correction)" column of that night vision noise data and stores them in the corresponding column of the night vision noise table. In addition, the degradation level determination unit 1233 determines that any night vision noise data stored in the night vision noise table whose average value or standard deviation deviates by a predetermined threshold or more from the average value or standard deviation of other night vision noise data before or after it is abnormal, and stores data indicating the determination result in the "Abnormal Data" column.

[0063] Furthermore, the degradation determination unit 1233 performs the following determinations and estimations for each of the four sensors of each of the multiple flame detectors 11, based on the data stored in the night vision noise table corresponding to those sensors (excluding the data of records in which the "Data Abnormality" column contains data indicating that the data is abnormal).

[0064] (1) If the average intensity of the most recent night vision noise data is equal to or greater than the predetermined threshold A1, it is determined that the component needs to be replaced. The amplifier is also estimated to be the component that should be replaced. (2) If the standard deviation of the intensity shown in the most recent night vision noise data is greater than or equal to a predetermined threshold B1, it is determined that the part needs to be replaced. The sensor is also estimated to be the part that should be replaced. (3) Based on the time-dependent change in the average intensity of night vision noise data over a predetermined period in the past, if the day on which the estimated average intensity reaches a predetermined threshold A1 is within a predetermined number of days from the present, it is determined that the component will need to be replaced soon. The amplifier is also estimated to be a component that should be replaced. (4) Based on the time-series change in the standard deviation of the intensity shown in the night vision noise data over a predetermined period in the past, if the day on which the standard deviation of the intensity reaches a predetermined threshold B1 is within a predetermined number of days from the present, it is determined that the part will need to be replaced soon. The sensor is also estimated to be a part that should be replaced.

[0065] Figures 10 and 11 are graphs illustrating the method by which the degradation level determination device 123 performs the above determination and estimation. The graphs in Figures 10(a) and 11(a) are curves that approximate points corresponding to the values ​​in the "Date and Time" and "Average Value" columns of the night vision noise table. The solid lines of the curves show the trend of the average value up to the present, while the dashed-dotted lines show the trend of the average value predicted for the future.

[0066] The graphs in Figures 10(b) and 11(b) are curves that approximate points corresponding to the values ​​in the "Date and Time" and "Standard Deviation" columns of the night vision noise table. The solid lines on the curves show the trend of the standard deviation up to the present, while the dashed-dotted lines show the projected trend of the standard deviation in the future.

[0067] The graph in Figure 10(a) shows that at the current date and time d1, the average intensity of the night vision noise data does not reach threshold A1. Similarly, the graph in Figure 10(b) shows that at the current date and time d1, the standard deviation of the intensity of the night vision noise data does not reach threshold B1. Therefore, based on this data, the degradation determination unit 1233 determines that the sensor or the amplifier connected to that sensor does not need to be replaced.

[0068] Furthermore, the graph in Figure 10(a) indicates that, within a predetermined number of days D1 from the current date and time d1, the average value of the intensity shown by the night vision noise data is estimated to reach threshold A1 at date and time d2. Furthermore, the graph in Figure 10(b) indicates that, within a predetermined number of days D1 from the current date and time d1, the standard deviation of the intensity shown by the night vision noise data is estimated not to reach threshold B1. Therefore, based on this data, the degradation determination unit 1233 determines that the sensor or the amplifier connected to the sensor will soon need to be replaced, and in that case, it estimates that the part to be replaced is the amplifier.

[0069] The graph in Figure 11(a) shows that at the current date and time d1, the average intensity of the night vision noise data has not reached threshold A1. The graph in Figure 11(b) shows that at the current date and time d1, the standard deviation of the intensity of the night vision noise data has reached threshold B1. Therefore, based on this data, the degradation determination unit 1233 determines that the sensor or the amplifier connected to the sensor needs to be replaced, and estimates that the part to be replaced is the sensor.

[0070] The degradation degree determination unit 1233 stores data indicating the results of the determination and estimation performed as described above in the storage unit 1231. Figure 12 is a diagram illustrating the configuration of a table (hereinafter referred to as the "degradation diagnosis result table") for storing data indicating the results of the determination and estimation performed by the degradation degree determination unit 1233. The degradation diagnosis result table has a "device ID" column, a "right long wavelength" column, a "right short wavelength" column, a "left long wavelength" column, and a "left short wavelength" column. The "device ID" column stores the identification information of the flame detector 11. The "right long wavelength" column, the "right short wavelength" column, the "left long wavelength" column, and the "left short wavelength" column are columns corresponding to sensors 111R, 112R, 111L, and 112L, respectively, and for those sensors, data is stored according to one of the following formats (a) to (c).

[0071] (a) "Normal": Indicates that no part replacement is currently required, and that no part replacement will be required within the specified number of days D1 from now. (b) "Needs replacement (##)": Indicates that a part needs to be replaced, and it is estimated that the part indicated by "##" (either the sensor or the amplifier, or both) needs to be replaced. (c) "Replacement needed soon (##) (Month XX, Day XX)": This indicates that it is estimated that a part will need to be replaced around Month XX, Day XX, and that the part indicated by "##" (either the sensor or the amplifier, or both) will need to be replaced.

[0072] Let's return to the explanation of the functional configuration of the disaster prevention receiver panel 12 (Figure 8). The transmitter 1234 is realized by a communication interface 1024 that operates under the control of the processor 1021. The transmitter 1234 transmits the night vision noise table and the degradation diagnosis result table to the server device 13. In addition, the transmitter 1234 periodically transmits data indicating the value in the "average value" column of the most recent record in the night vision noise table (excluding records in which data indicating that the data is abnormal is stored in the "data abnormality" column) to the flame detector 11 as steady-state correction constant data.

[0073] The display control unit 1235 is implemented by the processor 1021. The display control unit 1235 controls the display 121 to display various images. For example, when it receives a flame detection signal from any of the flame detectors 11, the display control unit 1235 generates image data representing the text "Flame Detection Area ##" and displays the image represented by that image data on the display 121. Here, "Area ##" is the identification information of the monitoring area A corresponding to the right flame detector or left flame detector of the flame detector 11 that detected the flame.

[0074] The operation reception unit 1236 is implemented by an input / output interface 1023 that operates under the control of the processor 1021. The operation reception unit 1236 receives operations performed by users, such as maintenance personnel, on the operation unit 122. Operations performed by users on the disaster prevention receiving panel 12 using the operation unit 122 include, for example, operations to instruct the fire extinguishing system controlled by the disaster prevention receiving panel 12 to start operating when a flame is detected.

[0075] The above is a description of the disaster prevention receiving panel 12. The server device 13 (see Figure 1) is a general-purpose server device equipped with web server functionality, so its hardware configuration and functional configuration will be omitted.

[0076] Terminal device 14 (see Figure 1) is a terminal device used by maintenance personnel. Since terminal device 14 is a general terminal device equipped with a web browser function, a description of its hardware configuration and functional configuration will be omitted.

[0077] The server device 13 receives the night vision noise table and the degradation diagnosis result table from the disaster prevention receiver panel 12 and stores the received data. In addition, the server device 13 generates display instruction data (e.g., HTML data) that instructs the terminal device 14 to display a screen like the one shown in Figure 13 (hereinafter referred to as the "degradation diagnosis screen"), and sends it to the terminal device 14. The terminal device 14 displays the degradation diagnosis screen according to the display instruction data sent from the server device 13.

[0078] The deterioration diagnosis screen includes a table displaying information about flame detectors 11 that currently require parts replacement, and another table displaying information about flame detectors 11 that are estimated to require parts replacement in the near future. Maintenance personnel can easily identify flame detectors 11 that require maintenance by looking at the deterioration diagnosis screen.

[0079] Furthermore, when a maintenance worker touches or clicks on any row in the table on the deterioration diagnosis screen, the terminal device 14 sends a request to the server device 13 that includes the identification information of the sensor corresponding to that row. In response to the request from the terminal device 14, the server device 13 generates display instruction data that instructs the display of a screen (hereinafter referred to as the "graph display screen") that includes a graph related to the sensor identified by the identification information included in the request, and sends it to the terminal device 14. The graph included in the graph display screen is a graph like the one shown in Figure 10 or Figure 11. The terminal device 14 displays the graph display screen according to the display instruction data sent from the server device 13. By looking at the graph display screen, the maintenance worker can check the deterioration status of the flame detector 11, which currently or in the near future will require replacement of parts.

[0080] [Differentiation] The embodiments described above are specific examples of the present invention and can be modified in various ways within the scope of the technical idea of ​​the present invention. Examples of such modifications are shown below. Two or more of the following modifications may be combined as appropriate.

[0081] (1) The stabilization period identification unit 1164 of the flame detector 11 in the flame detection system 1 described above identifies the stabilization period based on the measurement results of sensor 111 or sensor 112. The method by which the stabilization period identification unit 1164 identifies the stabilization period is not limited to this. For example, the flame detector 11 may be equipped with a sensor that measures a physical quantity in the surroundings (a sensor different from sensors 111 and 112), and the stabilization period identification unit 1164 may identify the stabilization period based on the measurement results of that sensor.

[0082] Figure 14 is a schematic diagram showing the hardware configuration of the flame detector 21 according to this modified example. In addition to the components of the flame detector 11 (see Figure 2) according to the above embodiment, the flame detector 21 includes a sensor 117 (an example of a second sensor) connected to the computer 101. The sensor 117 is an optical camera that continuously captures images of the area around the flame detector 11 and outputs image data representing the generated images.

[0083] Figure 15 is a schematic diagram showing the functional configuration of the flame detector 21. Specifically, the processor 1011 of the computer 101 of the flame detector 21 executes processing according to the program related to this modified example, thereby realizing the flame detector 21 equipped with a determination device indicated by reference numeral 116 in Figure 15. The acquisition unit 1162 of the determination device 116 of the flame detector 21 acquires image data output from the sensor 117.

[0084] The stability period identification unit 1164 of the determination device 116 of the flame detector 21 compares images represented by image data continuously acquired by the acquisition unit 1162 with images adjacent to each other in time series, and determines that there is no change between those images if, for example, the number of changing pixels is less than a predetermined threshold. Subsequently, the stability period identification unit 1164 identifies the period during which the state of no change between images continued for a predetermined time (for example, 1 minute) or longer as the stability period.

[0085] The flame detector 21 also extracts data from the measurement log table for the stable period as night vision noise data and outputs it to the disaster prevention receiver panel 12.

[0086] The type of sensor 117 in the flame detector 21 is not limited to an optical camera. For example, a microphone that picks up ambient sounds may be used as the sensor 117. In this case, the stabilization period identification unit 1164 identifies the period during which the amplitude value of the sound measured by the sensor 117 remains below a predetermined threshold for a predetermined time or longer as the stabilization period.

[0087] Alternatively, an object detection sensor that detects moving objects in the surroundings may be used as the sensor 117. In this case, the stabilization period determination unit 1164 determines the period during which the sensor 117 does not detect any objects for a predetermined time or longer as the stabilization period.

[0088] Alternatively, a vibration sensor that detects vibrations may be used as the sensor 117. In this case, the stabilization period identification unit 1164 identifies the period during which the amplitude value of the vibration measured by the sensor 117 remains below a predetermined threshold for a predetermined time or longer as the stabilization period.

[0089] Furthermore, the sensor 117 does not have to be part of the flame detector 11. For example, the sensor 117 may be a separate device located within a predetermined distance from the flame detector 11 and connected to the flame detector 11 wirelessly or by wire.

[0090] Furthermore, with respect to a certain flame detector 11, the sensor 111 or sensor 112 of another flame detector 11 located within a predetermined distance from that flame detector 11 may be used as sensor 117. For example, flame detector 11(2) may acquire data indicating the amplitude value of the signal output from the sensor 111 or sensor 112 of flame detector 11(1) or flame detector 11(3) installed next to flame detector 11(2), and the stabilization period identification unit 1164 of flame detector 11(2) may identify the stabilization period based on that data.

[0091] (2) The flame detector 11 provided in the flame detection system 1 described above is a dual-wavelength flame detector, but the number of wavelength bands used by the flame detector provided in the flame detection system 1 to detect flames may be three or more.

[0092] (3) Some of the processing that the fire prevention receiving panel 12 is supposed to perform in the fire detection system 1 described above may be performed by the fire detector 11, the server device 13, or the terminal device 14. For example, the fire detector 11 may be equipped with a deterioration degree determination unit 1233. Also, some of the processing that the fire detector 11 is supposed to perform in the fire detection system 1 described above may be performed by the fire prevention receiving panel 12, the server device 13, or the terminal device 14.

[0093] (4) Although the flame detector 11 of the flame detection system 1 described above is intended to monitor the space inside the tunnel TN, the area that the flame detection system 1 monitors is not limited to the inside of the tunnel. For example, the flame detector 11 may monitor the space inside a factory where there is a risk of fire occurring.

[0094] (5) The flame detector 11 provided in the flame detection system 1 described above includes a right-side flame detector and a left-side flame detector, but the flame detector 11 may also be a monocular flame detector that monitors only one area.

[0095] (6) In the flame detection system 1 described above, the flame detector 11 transmits night vision noise data to the disaster prevention receiving panel 12 without waiting for a request from the disaster prevention receiving panel 12. However, the flame detector 11 may transmit night vision noise data to the disaster prevention receiving panel 12 in response to a request from the disaster prevention receiving panel 12.

[0096] (7) In the flame detection system 1 described above, the method for determining the degree of deterioration of the flame detector 11 explained using the graphs in Figures 10 and 11 is just one example, and various other methods may be adopted. For example, instead of the standard deviation, variance, coefficient of variation, or the difference between the maximum and minimum amplitude values ​​may be used as an indicator of the variability of amplitude values ​​included in the night vision noise data. In addition, the degree of deterioration and the estimation of parts to be replaced may be performed based on the shape of the graphs as illustrated in Figures 10 and 11. For example, a learning model may be generated by machine learning using a large amount of training data, with the types of parts replaced in past maintenance work and the degree of deterioration of those parts as the target variables, and the night vision noise data immediately before replacement corresponding to those parts as the explanatory variables. Night vision noise data output from the flame detector 11 in operation may be input into the learning model, and the learning model may output the types of parts to be replaced and the degree of deterioration of those parts.

[0097] (8) In the flame detection system 1 described above, the degree of deterioration determined by the deterioration determination unit 1233 is one of three stages: (a) parts need to be replaced now, (b) parts will need to be replaced soon, or (c) normal. The deterioration determination unit 1233 may also determine the degree of deterioration using a different expression format. For example, the ratio of the average value of the amplitude values ​​included in the night vision noise data to threshold A1, or the ratio of the standard deviation of the amplitude values ​​included in the night vision noise data to threshold B1, may be used as the degree of deterioration of the flame detector 11.

[0098] (9) In the flame detection system 1 described above, the components of the flame detector 11 that are likely to deteriorate are the sensor and the amplifier, but the components that are likely to deteriorate are not limited to these. For example, deterioration of wiring, contacts, etc. may also be suspected. [Explanation of Symbols]

[0099] 1...Flame detection system, 11...Flame detector, 12...Disaster prevention receiving panel, 13...Server device, 14...Terminal device, 21...Flame detector, 101...Computer, 102...Computer, 111...Sensor, 112...Sensor, 113...Amplifier, 114...Amplifier, 115...Thermometer, 116...Judgment device, 117...Sensor, 121...Display, 122...Operation unit, 123...Degradation degree determination device, 1011...Processor, 1012...Memory, 1013...Input / Output interface -, 1014...Communication interface, 1021...Processor, 1022...Memory, 1023...Input / Output interface, 1024...Communication interface, 1161...Storage unit, 1162...Acquisition unit, 1163...Correction unit, 1164...Stability period specification unit, 1165...Flame detection unit, 1166...Transmit / receive unit, 1167...Timekeeping unit, 1231...Storage unit, 1232...Acquisition unit, 1233...Degradation degree determination unit, 1234...Transmit unit, 1235...Display control unit, 1236...Operation reception unit.

Claims

1. A measuring unit comprising a sensor and an amplifier for amplifying the signal generated by the sensor, wherein the sensor measures the intensity of light, A flame detection unit that detects a flame based on the light intensity measured by the aforementioned measuring unit, A degradation determination unit that determines the degree of degradation of the amplifier based on the average value of the light intensity measured by the measurement unit during a past period, and determines the degree of degradation of the sensor based on an index of the variation in light intensity measured by the measurement unit during a past period. A flame detection system equipped with [the necessary components].

2. A method for generating a learning model for a flame detection system comprising a sensor, an amplifier for amplifying a signal generated by the sensor, a measuring unit for measuring the intensity of light, and a flame detection unit for detecting a flame based on the intensity of light measured by the measuring unit, the method being used to generate a learning model for a flame detection system, A step to generate a learning model by machine learning using training data in which the type of sensor or amplifier that has been replaced in the past and the degree of degradation of said part are the target variables, and the light intensity measured by the measurement unit during the period immediately preceding the replacement of said part are the explanatory variables. A method for generating a learning model that includes the following features.

3. A measuring unit that includes a sensor and an amplifier for amplifying the signal generated by the sensor, for measuring the intensity of light, A flame detection unit that detects a flame based on the light intensity measured by the aforementioned measuring unit, A learning model generated by the generation method described in claim 2 is input to the light intensity measured by the measurement unit during a past period, and a degradation determination unit determines the degree of degradation of the measurement unit based on the results output from the learning model. A flame detection system equipped with [the necessary components].

Images