Electrical fire monitoring and early warning method and system

Abstract

The invention relates to the technical field of fire early warning, in particular to an electrical fire monitoring and early warning method and system, and the method comprises the steps: collecting the current electrical parameter data of a monitored object, and obtaining the electrical parameter data before multiple historical fire early warning and in multiple normal states; analyzing the first necessity of each historical fire early warning according to the characteristics of abnormal data in the data collected before each historical fire early warning and the correlation between the abnormal data and the fire early warning; analyzing the second necessity of each historical fire early warning according to the difference between any two kinds of data in each historical normal state and the correlation between the abnormal condition and the fire early warning; and combining the first necessity and the second necessity to calculate the disaster early warning necessity of each moment in the current and previous preset time periods so as to judge the current disaster early warning necessity. According to the invention, the historical correlation change among various electrical parameters is analyzed, the fire risk is evaluated more comprehensively, and the accuracy of early warning is improved.

Description

Technical Field

[0001] The present application relates to the field of fire warning technology, and in particular to an electrical fire monitoring and warning method and system. Background Art

[0002] Electrical fire monitoring and early warning uses a series of technical means and equipment to monitor the operating status of electrical lines and equipment in real time, so as to promptly detect potential electrical fire hazards and issue early warnings before a fire occurs, so as to take appropriate measures to prevent the occurrence of fire accidents.

[0003] The existing electrical fire monitoring algorithms are mainly based on simple threshold judgments of a single parameter in real time, and have problems such as low sensitivity, high false alarm rate, and inability to provide early warning. Summary of the invention

[0004] In order to solve the above technical problems, the present application provides an electrical fire monitoring and early warning method and system, and the technical solutions adopted are as follows: In a first aspect, an embodiment of the present application provides an electrical fire monitoring and early warning method, the method comprising the following steps: Collect the current electrical parameter data of the monitored object, and obtain the electrical parameter data before multiple fire warnings and multiple times in normal state; For each item of electrical parameter data, the correlation between the data collected before each fire warning in history and the fire warning is obtained according to the proportion of abnormal data in the data collected before each fire warning in history and the abnormal score; and the mean value of the correlation between the data collected before all historical warnings and the fire warning is analyzed, and the weighted sum of the correlation of the corresponding item of electrical parameter data before each historical fire warning is performed to obtain the first necessity of each historical fire warning; Based on the difference between the ratio of any two data collected at the same time under each normal state in history and the ratio of the mean of any two data under the normal state in history, the difference between any two data under normal state is obtained; the number of abnormal ratios and their abnormal scores in the ratio of any two data collected at the same time before each fire warning in history are used to determine the correlation between the abnormal conditions of any two data before each fire warning in history and the disaster warning; and the average value of the correlation between the abnormal conditions of any two data before all fire warnings in history and the disaster warning is used as the correlation between the abnormal conditions of any two data and the fire warning; The differences between any two data under normal conditions, the correlation between abnormal conditions and fire warnings, and the correlation between abnormal conditions and disaster warnings between any two data before each fire warning in history are forward integrated to obtain the second necessity of each fire warning in history; The necessity of disaster warning at each moment in the current and previous preset time periods is calculated in combination with the first and second necessities to determine the necessity of the current disaster warning.

[0005] Preferably, the electrical parameter data includes temperature, voltage, current and carbon dioxide concentration.

[0006] Preferably, the abnormal data is obtained by performing a LOF abnormality detection algorithm on various electrical parameter data before all historical fire warnings and under all normal conditions.

[0007] Preferably, the method for obtaining the correlation between the data collected before each fire warning in history and the fire warning is: the correlation between the temperature data collected before the jth warning in history and the fire warning is recorded as , ;in represents the correlation between the temperature data collected before the jth warning in history and the fire warning, norm represents the normalization function, represents the proportion of abnormal data in the temperature data collected before the jth fire warning in history, represents the number of abnormal data in the temperature data collected before the jth fire warning in history, represents the anomaly score of the i-th anomaly data.

[0008] Preferably, before using the mean value of the correlation between the data collected before all historical warnings and the fire warning to weight the correlation of the corresponding electrical parameter data before each historical fire warning, the mean value of the correlation between each electrical parameter data collected before all historical warnings and the fire warning is normalized.

[0009] Preferably, the difference between any two data under normal conditions is determined by the cumulative sum of all the differences calculated under all historical sub-normal conditions.

[0010] Preferably, the method for obtaining the correlation between the abnormal conditions of any two data before each fire warning in history and the disaster warning is: record the correlation between the abnormal conditions of the temperature data and the current data before the a-th fire warning in history and the disaster warning as ; Where m represents the number of historical fire warnings obtained, norm represents the normalization function, represents the number of abnormal ratios of the temperature to current data before the a-th fire warning in history, g represents the number of moments of each historical collection, Indicates the abnormal score corresponding to the ratio of temperature to current data when the abnormal ratio occurs for the bth time.

[0011] Preferably, the necessity of the disaster warning at each moment in the current and previous preset time periods is determined by the sum of the first necessity and the second necessity at each moment in the current and previous preset time periods.

[0012] Preferably, the method for determining the necessity of the current disaster warning is: Threshold segmentation is performed on all disaster warning necessities obtained in the current and previous preset time periods, and a fire warning is issued when the current disaster warning necessity is greater than the segmentation threshold.

[0013] In a second aspect, an embodiment of the present application further provides an electrical fire monitoring and early warning system, the system comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above-mentioned electrical fire monitoring and early warning methods when executing the computer program.

[0014] This application has at least the following beneficial effects: The advantage of the present application over the prior art is that, before the electrical parameters reach a dangerous threshold, the present application can comprehensively analyze the historical data and real-time data of the electrical parameters, and combine data anomaly analysis and data correlation anomaly analysis of multiple electrical parameters. It not only considers the anomaly of a single electrical parameter, but also analyzes the historical correlation changes between multiple electrical parameters, so as to more comprehensively assess the fire risk, discover fire hazards in advance, reduce the false alarm rate, and improve the accuracy of early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0016] Figure 1 A flowchart of an electrical fire monitoring and early warning method provided in one embodiment of the present application; Figure 2 A flow chart of a method for obtaining the necessity of current disaster warning provided in one embodiment of the present application. DETAILED DESCRIPTION

[0017] Example 1 An electrical fire monitoring and early warning method provided by an embodiment of the present application is specifically referred to in Figure 1 , the method comprises the following steps: Step 1: Collect the current electrical parameter data of the monitored object, and obtain the electrical parameter data before multiple fire warnings and multiple times under normal conditions.

[0018] Electrical parameter data related to the fire of the monitored object, including temperature, voltage, current and carbon dioxide concentration, is installed near the monitored object. The temperature W, voltage D, current L and carbon dioxide CO2 concentration C in the monitored object environment are obtained through temperature sensors, voltage sensors, current sensors and carbon dioxide concentration sensors respectively.

[0019] In this embodiment, the collection frequency of all data is set to be evenly collected 10 times per minute, and the collection time is set to the current time and within 1 hour before it, which can be set by the implementer.

[0020] Furthermore, the electrical parameter data before multiple historical fire warnings and multiple normal states are obtained, and the early warning analysis of the current fire monitoring data is completed by analyzing the data change characteristics in all historical data.

[0021] In this embodiment, the electrical parameter data of n hours before each fire warning in history are obtained, and the characteristic analysis of these parameter data is performed to obtain the closeness between the fire precursor and these data. The number of historical fire warnings obtained is recorded as m, and the number of data obtained in the normal state is recorded as Q. In this embodiment, Q=200, then m+Q groups of data are obtained, and each group of data constitutes a sequence, and the length of the sequence is , each element in the sequence is a column vector of all the electrical parameter data at each acquisition moment, and this application performs electrical data analysis based on these data. The length of the temperature data sequence under normal conditions is also In this embodiment, n=1, that is, only one hour of data is collected.

[0022] Step 2: Analyze the necessity of each fire warning in history based on the characteristics of abnormal data in the data collected before each fire warning and the correlation with the fire warning.

[0023] In this application, for each electrical parameter data, the correlation between the data collected before each fire warning in history and the fire warning is obtained based on the proportion of abnormal data in the data collected before each fire warning in history and the abnormal score.

[0024] Specifically, this embodiment takes temperature data as an example, and first obtains m groups of temperature data collected before the historical fire warning. The LOF anomaly detection algorithm is used to calculate the anomaly score for all the temperature data collected historically, and the data with an anomaly score greater than 1 is recorded as abnormal data. Among them, the LOF anomaly detection algorithm is a well-known technology and will not be described in detail. In other embodiments, other suitable anomaly detection algorithms can also be used to calculate the anomaly score to obtain abnormal data.

[0025] In this embodiment, the correlation between the temperature data collected before the jth warning in history and the fire warning can be obtained according to the proportion of abnormal data in the temperature data collected before the jth warning in history and the abnormal score: in represents the correlation between the temperature data collected before the jth warning in history and the fire warning, norm represents the normalization function, It represents the proportion of abnormal data in the temperature data collected before the jth fire warning in history, which is determined by the ratio of the number of abnormal data obtained by anomaly detection to the number of all temperature data collected before the fire warning; represents the number of abnormal data in the temperature data collected before the jth fire warning in history, Represents the abnormal score of the ith abnormal data. That is, before the fire warning is required, the more abnormal temperature data there are and the greater the abnormality of the temperature data, the stronger the correlation between abnormal temperature change and fire warning. That is, the greater the degree of abnormal temperature change, the greater the necessity of fire warning.

[0026] Similarly, the correlation between the voltage, current, and carbon dioxide concentration data collected before each historical warning and the fire warning can be obtained and recorded as , and .

[0027] Furthermore, in the present application, the mean correlation between the data collected before all historical warnings and the fire warning is analyzed, and the weighted sum of the correlation of the corresponding electrical parameter data before each historical fire warning is performed to obtain the first necessity of each historical fire warning, which is used to characterize the necessity of each historical fire warning.

[0028] In this embodiment, the average value of the correlation between the temperature data collected before all the warnings and the fire warning is calculated. , which is used to characterize the correlation between various electrical parameters corresponding to historical collection and fire warning.

[0029] Similarly, the correlation mean values ​​of the voltage, current, and carbon dioxide concentration data collected before all historical warnings and the fire warnings can be obtained and recorded as , and .

[0030] Before using the mean value of the correlation between the data collected before all historical warnings and the fire warning to weight the correlation of the corresponding electrical parameter data before each historical fire warning, the mean value of the correlation between each electrical parameter data collected before all historical warnings and the fire warning is normalized. Specifically, a probability normalization operation can be used, which is set by the implementer.

[0031] Step 3: Analyze the second necessity of each fire warning in history based on the difference between any two data under each normal state in history and the correlation between abnormal conditions and fire warnings.

[0032] At the same time, there is a certain correlation between the electrical parameter data when the electrical system is running smoothly. When there are signs of fire, the correlation between these data may be destroyed and changed. Therefore, the correlation between the electrical parameter data can be monitored and the changes in the correlation between these data before the fire warning can be analyzed to obtain the relationship between the abnormal changes in the correlation between these data and the signs of fire.

[0033] In this application, the difference between any two data under normal conditions is obtained based on the difference between the ratio of any two data collected at the same time under each historical normal state and the ratio of the mean values ​​of any two data under the historical normal state.

[0034] Preferably, the difference between any two data under normal conditions is determined by the cumulative sum of all the differences calculated under all historical sub-normal conditions.

[0035] This embodiment uses temperature and current as examples to obtain the differences between these two electrical parameter data under historical normal conditions, analyzes the temperature and current data obtained at the same acquisition time, calculates the ratio of temperature to current data at the same acquisition time, and performs a difference analysis with the ratio of all temperature averages to all current averages under each historical normal state at the acquisition time, and obtains the difference between temperature and current under normal conditions, that is, the degree of uncorrelation between the changes in the two electrical parameter data under normal operation. The corresponding difference between temperature and current data under normal conditions is The calculation method is as follows: Where Q represents the number of data acquired under normal conditions, g represents the number of moments of each historical acquisition, , They represent the mean values ​​of all temperatures and all currents in the uth normal state in history, that is, when the historical data is analyzed, the smaller the change of the two ratios is, the smaller the change of the required value is. The smaller it is, the more stable the ratio of the two data is under normal circumstances, that is, the stronger the correlation between the two data is, and further, the greater the reliability of disaster warning based on the change in the correlation. It should be noted that when the data is under normal conditions, there are other reasons that cause the data ratio to be abnormal. When calculated according to the above method, the calculated parameter correlation is low, and the reference for disaster warning based on such parameters is reduced according to its correlation. Therefore, for some parameters that are easily disturbed by other factors in normal times, their parameter reference is low.

[0036] Furthermore, in the present application, the number of abnormal ratios and their abnormal scores in the ratios of any two data collected at the same time before each historical fire warning are used to determine the correlation between the abnormal conditions of any two data before each historical fire warning and the disaster warning; and the average value of the correlation between the abnormal conditions of any two data before all historical fire warnings and the disaster warning is used as the correlation between the abnormal conditions of any two data and the fire warning.

[0037] In this embodiment, for the abnormal ratios of the temperature and current data collected at the same time before all historical fire warnings, the LOF anomaly detection algorithm is used in the same way as above to obtain the abnormal ratios and their anomaly scores, and the number of abnormal ratios is recorded as R.

[0038] Furthermore, in this embodiment, the calculation method for the correlation between abnormal conditions of temperature data and current data and fire warning is as follows: in It indicates the correlation between abnormal conditions of temperature data and current data and fire warning. norm indicates the normalization function. represents the correlation between abnormal conditions of temperature data and current data before the a-th fire warning in history and disaster warning, m represents the number of historical fire warnings obtained, represents the number of abnormal ratios of the temperature to current data before the a-th fire warning in history, and g represents the number of moments of each historical collection. It represents the proportion of abnormal ratios of temperature to current data before the a-th fire warning in history, Indicates the abnormal score corresponding to the ratio of temperature to current data when the abnormal ratio occurs for the bth time.

[0039] It should be understood that, that is, when the ratio of the abnormality in the ratio of temperature and current data obtained by analyzing the temperature and current data before the a-th fire warning in history is greater, and the abnormality score of the corresponding abnormal ratio is also greater, it means that the abnormality of the abnormal ratio is more likely to be a sign of a disaster, and the effect of predicting fire using the abnormality of the abnormal ratio is better, that is, the correlation between the abnormal change of the abnormal ratio and the fire warning is stronger.

[0040] Similarly, the correlation between the abnormal conditions of any two electrical parameter data and the fire warning is obtained, and recorded as , , , and .

[0041] The second necessity of disaster warning can be obtained based on the above information. This application forwardly integrates the differences between any two data under normal conditions, the correlation between abnormal conditions and fire warnings, and the correlation between abnormal conditions and disaster warnings between any two data before each fire warning in history to obtain the second necessity of each fire warning in history.

[0042] It can be understood that forward fusion is a fusion method such as addition and multiplication between data. The specific forward fusion method is determined by the implementer according to the actual situation, and this application does not impose any special restrictions.

[0043] In this embodiment, taking temperature and current data as an example, the difference between temperature and current data under normal conditions, the correlation between abnormal conditions of temperature and current data and fire warning, and the correlation between abnormal conditions of temperature and current data and disaster warning before the hth fire warning in history are forward fused, and the average of the forward fusion results of all two arbitrary data is used as the second necessity of the hth fire warning in history. The specific calculation method is as follows: First, the correlation weight factor is calculated by calculating the correlation between the two data and the product of the abnormal change in the correlation between the two data and the correlation between the disaster warning, that is, ; represents the correlation weight factor between temperature and current data, norm represents the normalization function, Indicates the difference between temperature and current data under normal conditions, Indicates the correlation between abnormal conditions of temperature data and current data and fire warning. That is, the stronger the correlation between the two data, and the stronger the correlation between the change in the correlation of the two data and the disaster warning, the The larger the value is, the more the abnormal correlation between the two types of data can reflect whether a disaster needs early warning. After normalizing it, it is used as the correlation weight factor of the pairwise correlation anomaly reflecting the disaster early warning characteristics, and the specific pairwise data correlation anomaly weight can be obtained.

[0044] The corresponding calculation method of the second necessity of each fire warning in history based on the correlation between temperature and current data is as follows: For example: ; represents the correlation weight factor between temperature and current data, Indicates the correlation between abnormal conditions of temperature data and current data before the h-th fire warning in history and disaster warning. That is, when the correlation weight factor between the required temperature and current data is larger, and the abnormal change of the correlation between temperature and current data in the h-th fire warning in history is larger, it means that the necessity of disaster warning based on the analysis of temperature and current data is greater.

[0045] The same method is used to obtain the second necessity of temperature and voltage, temperature and carbon dioxide, voltage and current, voltage and carbon dioxide, and current and carbon dioxide in the hth fire warning in history, respectively, and recorded as , , , , , , sum up the above six results to obtain the second necessity of the h-th fire warning in history.

[0046] Step 4: Combine the first and second necessities to calculate the necessity of disaster warning at each moment in the current and previous preset time periods to determine the necessity of the current disaster warning.

[0047] According to the data attributes analyzed above and the correlation analysis between the data, the necessity of each disaster warning in history can be obtained BY: that is, the first necessity of the disaster warning and the second necessity of the disaster warning are added together to obtain it.

[0048] The necessity of disaster warning can be obtained for all collection moments in the current and previous preset time periods according to the above method, wherein the values ​​in the time period that are less than the length g are supplemented according to the average value in the time period. In this embodiment, the length of the current and previous preset time period obtained is taken as 50, and the implementer can set it according to the actual situation.

[0049] Threshold segmentation is performed on all disaster warning necessities obtained in the current and previous preset time periods, and a fire warning is performed when the current disaster warning necessity is greater than the segmentation threshold. In this embodiment, the Otsu threshold is used for threshold segmentation, and the Otsu threshold is a well-known technology and will not be described in detail.

[0050] In this embodiment, by analyzing the abnormality of changes in various data based on the data before the warning, disaster signs can be discovered in advance before the detection values ​​of various monitoring data reach the threshold, providing more processing time to avoid the occurrence of disasters.

[0051] In this embodiment, the method flow chart of obtaining the necessity of current disaster warning is as shown in the attached figure. Figure 2 shown.

[0052] Based on the same inventive concept as the above method, another embodiment of the present application also provides an electrical fire monitoring and early warning system, including a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, the steps of any one of the above-mentioned electrical fire monitoring and early warning methods are implemented.

[0053] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention herein. The present application is intended to cover any variations, uses or adaptations of the present application, which follow the general principles of the present application and include common knowledge or customary technical means in the art that are not invented by the present application.

[0054] It should be understood that the present application is not limited to the exact construction that has been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof.

Claims

1. An electrical fire monitoring and early warning method, characterized in that: The method comprises the following steps: Collect the current electrical parameter data of the monitored object, and obtain the electrical parameter data before multiple fire warnings and multiple times in normal state; For each item of electrical parameter data, the correlation between the data collected before each fire warning in history and the fire warning is obtained according to the proportion of abnormal data in the data collected before each fire warning in history and the abnormal score; and the mean value of the correlation between the data collected before all historical warnings and the fire warning is analyzed, and the weighted sum of the correlation of the corresponding item of electrical parameter data before each historical fire warning is performed to obtain the first necessity of each historical fire warning; Based on the difference between the ratio of any two data collected at the same time under each normal state in history and the ratio of the mean of any two data under the normal state in history, the difference between any two data under normal state is obtained; the number of abnormal ratios and their abnormal scores in the ratio of any two data collected at the same time before each fire warning in history are used to determine the correlation between the abnormal conditions of any two data before each fire warning in history and the disaster warning; and the average value of the correlation between the abnormal conditions of any two data before all fire warnings in history and the disaster warning is used as the correlation between the abnormal conditions of any two data and the fire warning; The differences between any two data under normal conditions, the correlation between abnormal conditions and fire warnings, and the correlation between abnormal conditions and disaster warnings between any two data before each fire warning in history are forward integrated to obtain the second necessity of each fire warning in history; The necessity of disaster warning at each moment in the current and previous preset time periods is calculated in combination with the first and second necessities to determine the necessity of the current disaster warning.

2. The electrical fire monitoring and early warning method according to claim 1, characterized in that: The electrical parameter data include temperature, voltage, current and carbon dioxide concentration.

3. The electrical fire monitoring and early warning method according to claim 1, characterized in that: The abnormal data is obtained by performing a LOF abnormality detection algorithm on the electrical parameter data before all historical fire warnings and under all normal conditions.

4. The electrical fire monitoring and early warning method according to claim 1, characterized in that: The method for obtaining the correlation between the data collected before each fire warning in history and the fire warning is as follows: the correlation between the temperature data collected before the jth warning in history and the fire warning is recorded as , ;in represents the correlation between the temperature data collected before the jth warning in history and the fire warning, norm represents the normalization function, represents the proportion of abnormal data in the temperature data collected before the jth fire warning in history, represents the number of abnormal data in the temperature data collected before the jth fire warning in history, represents the anomaly score of the i-th anomaly data.

5. The electrical fire monitoring and early warning method according to claim 1, characterized in that: Before using the mean value of the correlation between the data collected before all historical warnings and the fire warning to weight the correlation of the corresponding electrical parameter data before each historical fire warning, the mean value of the correlation between each electrical parameter data collected before all historical warnings and the fire warning is normalized.

6. The electrical fire monitoring and early warning method according to claim 1, characterized in that: The difference between any two data under normal conditions is determined by the cumulative sum of all the differences calculated under all historical subnormal conditions.

7. The electrical fire monitoring and early warning method according to claim 1, characterized in that: The method for obtaining the correlation between the abnormal conditions of any two data before each fire warning in history and the disaster warning is as follows: the correlation between the abnormal conditions of the temperature data and the current data before the a-th fire warning in history and the disaster warning is recorded as ; Where m represents the number of historical fire warnings obtained, norm represents the normalization function, represents the number of abnormal ratios of the temperature to current data before the a-th fire warning in history, g represents the number of moments of each historical collection, Indicates the abnormal score corresponding to the ratio of temperature to current data when the abnormal ratio occurs for the bth time.

8. The electrical fire monitoring and early warning method according to claim 1, characterized in that: The necessity of the disaster warning at each moment in the current and previous preset time periods is determined by the sum of the first necessity and the second necessity at each moment in the current and previous preset time periods.

9. The electrical fire monitoring and early warning method according to claim 1, characterized in that: The method for judging the necessity of current disaster warning is: Threshold segmentation is performed on all disaster warning necessities obtained in the current and previous preset time periods, and a fire warning is issued when the current disaster warning necessity is greater than the segmentation threshold.

10. An electrical fire monitoring and early warning system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, an electrical fire monitoring and early warning method as described in any one of claims 1 to 9 is implemented.

Images