Coal field fire area monitoring method

Abstract

The invention relates to the technical field of coal field monitoring, in particular to a coal field fire area monitoring method, which comprises the following steps of: 1, setting a simulation monitoring model, and determining a three-dimensional coordinate position of monitoring equipment; 2, dividing the monitoring parameters into main analysis parameters and influence parameters, constructing a simulation operation model, and obtaining a parameter correlation model of each monitoring parameter; and step 3, acquiring monitoring parameters acquired in real time, calculating real-time change parameters, respectively acquiring influence parameters in each parameter correlation model, inputting the influence parameters into the corresponding parameter correlation models to obtain theoretical change values of the monitoring parameters, and determining fire area predicted values of the coal field fire area according to the theoretical change values and the real-time change parameters. Meanwhile, determining an abnormal risk signal according to the fire area predicted value; and step 4, based on the three-dimensional coordinate position of the monitoring device generating the abnormal risk signal, carrying out abnormity confirmation on the coal field fire area so as to improve the prediction and detection capability of the abnormal condition of the coal field fire area.

Description

Technical Field

[0001] The invention relates to the technical field of coalfield monitoring, and in particular to a method for monitoring a coalfield fire zone. Background Art

[0002] Coalfield fire zone refers to the area where spontaneous combustion of coal seams in the coal-bearing area forms a fire. The formation of coalfield fire zone is mainly due to spontaneous combustion of coal seams. This process is affected by both internal factors, such as the degree of coal metamorphism and physical and chemical properties, and external factors, including external conditions such as climate, topography and human activities.

[0003] Prior art CN103760619A discloses a method and device for monitoring coalfield fire areas, including: obtaining a hyperspectral thermal infrared image of a target area on the ground; obtaining thermal radiation information of objects in the target area through the hyperspectral thermal infrared image of the target area; obtaining an abnormal temperature area in the target area based on the thermal radiation information of the objects in the target area; and determining the location of the coalfield fire area in the hyperspectral thermal infrared image based on the abnormal temperature area in the target area.

[0004] When monitoring a coalfield fire area, infrared imaging or thermal radiation is often used to monitor the temperature in the coalfield fire area in real time. However, when using infrared imaging or thermal radiation for monitoring, if the coalfield is not completely burned, the fire response signal will have hysteresis, which will affect the timing of fire extinguishing. Summary of the invention

[0005] The purpose of the present invention is to solve the problems in the background technology and to propose a monitoring method for coalfield fire areas.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A method for monitoring a coalfield fire zone, the monitoring method specifically comprising the following steps:

[0008] Step 1: Collect regional information of the coalfield fire area, set up a simulation monitoring model based on the regional information, and then determine the three-dimensional coordinate position of each monitoring device in the simulation monitoring model, and use the monitoring equipment to monitor the coalfield fire area in real time;

[0009] Step 2: arbitrarily select a monitoring parameter as the main analysis parameter, and use the remaining monitoring parameters as influencing parameters, obtain the historical monitoring data of each fire, determine the pre-data and post-data according to the time of fire occurrence, and use the pre-data and post-data to respectively determine the data change value of the monitoring parameter, the data change value of the influencing parameter, and the data change value of the main analysis parameter to construct a simulation operation model; train the simulation operation model to obtain a parameter association model of the monitoring parameter;

[0010] Step 3: Obtain the monitoring parameters collected in real time, and combine them with the monitoring parameters before the Ty time period to calculate the real-time change parameters. Then, according to the parameter association model, take the influencing parameters in the real-time change parameters and input them into the corresponding parameter association model. Then, mark the operation output result as the theoretical change value, obtain the theoretical change values ​​of all monitoring parameters, combine the theoretical change values ​​with the corresponding real-time change parameters, determine the fire zone prediction value, and then determine the abnormal risk signal based on the fire zone prediction value.

[0011] Step 4: Detect abnormal risk signals and identify the monitoring equipment that generates abnormal risk signals, obtain the three-dimensional coordinate position of the corresponding monitoring equipment, and transmit the three-dimensional coordinate position and the abnormal risk signal to the terminal device of the relevant management personnel at the same time. The management personnel confirm the abnormality of the coalfield fire area based on the three-dimensional coordinate position.

[0012] As a further solution of the present invention, the simulation monitoring model is a three-dimensional stereo model and is provided with a three-dimensional spatial coordinate system. At the same time, based on the regional information of the coalfield fire zone, the monitoring equipment in the target fire zone and the position of the monitoring equipment are obtained. According to the position of the monitoring equipment in the target fire zone, the three-dimensional coordinate position of the monitoring equipment in the simulation monitoring model is determined. At the same time, the simulation monitoring model restores the actual coalfield fire zone according to a preset scale.

[0013] As a further solution of the present invention, a method for setting a parameter association model includes:

[0014] S1: According to the fires that have occurred in the historical data, the historical monitoring data of each fire is obtained, wherein the historical monitoring data includes the monitoring data before the fire occurs and the monitoring data when the fire occurs. At the same time, the monitoring data before the fire occurs is marked as the pre-data, and the monitoring data when the fire occurs is marked as the post-data;

[0015] S2: Subtract the pre-data from the post-data in each fire to obtain the data change value FGj, where j represents different monitoring parameters. When subtracting the pre-data from the post-data, only the monitoring data of the same data type are subtracted;

[0016] Extract the data value corresponding to the main analysis parameter in the data change value FGj, and mark the remaining monitoring data in the data change value FGj as influencing parameters. Then mark the data change values ​​in the same fire as a data group, obtain all the data groups in the historical data, and divide the data groups into training data and test data;

[0017] S3: First, a simulation operation model is constructed based on the training data, and then the simulation operation model is evaluated using the test data, and the final parameter association model is determined based on the evaluation data.

[0018] As a further solution of the present invention, when selecting the monitoring data before the fire occurs, the fire occurrence time is taken as a reference, and the fire occurrence time is subtracted from the Ty preset time period to obtain the leading time point. Thereafter, the corresponding monitoring data is obtained at the position of the leading time point and used as the monitoring data before the fire occurs, wherein the Ty time period is a preset time period and a fixed value.

[0019] As a further solution of the present invention, the process of constructing the simulation operation model is: selecting a neural network algorithm for operation, using the influencing parameters as the input layer of the neural network algorithm, the hidden layer as a number of neurons, and the main analysis parameters as the output layer; the process of determining the parameter association model is: inputting the training data into the neural network, and performing simulation operations to obtain the parameter association model GCj, wherein GCj represents the parameter association model with the monitoring parameter j as the main analysis parameter.

[0020] As a further solution of the present invention, a method for calculating the fire zone prediction value includes:

[0021] Acquire the monitoring parameters collected in real time, and identify the monitoring parameters before the Ty time period based on the Ty time period and the real time time as the time node, and then subtract the monitoring parameters before the Ty time period from the monitoring parameters collected in real time to obtain the real-time change parameter BHj;

[0022] Obtain the parameter association model GCj of the monitoring parameter j, then obtain the real-time change parameter corresponding to the influencing parameter in the parameter association model GCj, and input the real-time change parameter corresponding to the influencing parameter into the parameter association model GCj for calculation, and mark the value obtained after the calculation as the theoretical change value LBj;

[0023] Obtain the theoretical change value LBj of all monitoring parameters, and then use the formula The fire zone prediction value FIR is obtained, where J represents the total amount of monitoring parameters, a1 is a constant coefficient, and 0<a1<1.

[0024] As a further solution of the present invention, a method for determining an abnormal risk signal includes:

[0025] The fire zone prediction value FIR is compared with the risk coefficient. If the fire zone prediction value FIR is less than the risk coefficient, a normal monitoring signal is generated. When the monitoring system detects a normal monitoring signal, a normal monitoring instruction is transmitted to the monitoring equipment at this location. Conversely, if the fire zone prediction value FIR is greater than or equal to the risk coefficient, an abnormal risk signal is generated.

[0026] Compared with the prior art, the advantages of the present invention are:

[0027] The present invention obtains historical data of coalfield fire areas and extracts historical monitoring data of each fire in the historical data, obtains the interaction between various monitoring data before the occurrence of fire in the coalfield fire area, determines the corresponding parameter association model, and then determines the fire area prediction value of the coalfield fire area according to the theoretical change value and the real-time change parameter calculated in the parameter association model, and then generates an abnormal risk signal from the fire area prediction value, so as to timely respond to the abnormal signal of the coalfield fire area in the early stage of the fire, thereby improving the prediction and detection capabilities of the coalfield fire area abnormal situation;

[0028] The present invention sets a simulation operation model and sets the three-dimensional coordinate position of the monitoring equipment. When the monitoring equipment generates an abnormal risk signal, the three-dimensional coordinate position of the corresponding monitoring equipment is identified, and the position where the abnormality occurs in the coalfield fire area is directly determined, thereby improving the fire extinguishing efficiency and reducing economic losses. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a schematic diagram of the method flow structure of the present invention. DETAILED DESCRIPTION

[0030] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0031] Reference Figure 1 The present embodiment adopts a method for monitoring a coalfield fire zone, which specifically includes the following steps:

[0032] Step 1: Collect regional information involved in the coalfield fire area, and mark the coalfield fire area as the target fire area, wherein the regional information includes the surface area range, regional area, coal seam depth, thickness and distribution form of the coalfield fire area;

[0033] It should be further explained that the coalfield fire area in this embodiment refers to the fire area that has been controlled. At this time, there is no obvious fire in the fire area, but continuous observation is still required.

[0034] Then, based on the regional information of the coalfield fire zone, the monitoring equipment and the position of the monitoring equipment in the target fire zone are obtained, and a simulation monitoring model of the target fire zone is constructed, wherein the simulation monitoring model is a three-dimensional stereoscopic model and is provided with a three-dimensional spatial coordinate system, and then the three-dimensional coordinate position of the monitoring equipment is determined according to the position of the monitoring equipment in the target fire zone, wherein the simulation monitoring model restores the actual coalfield fire zone according to a preset ratio size, and the preset ratio size in this embodiment is set to 1:1;

[0035] The monitoring equipment in this embodiment includes a video monitoring device and a sensor monitoring device. The video monitoring device is used to perform all-round monitoring of the target fire area. In this embodiment, the video monitoring device uses an infrared imager, which can simultaneously collect regional images and surface temperature distribution of the target fire area. The sensor monitoring device is used to perform range monitoring of the monitoring points of the target fire area. The sensor monitoring device in this embodiment includes a temperature sensor, a humidity sensor, a gas sensor, and a wind speed sensor, which correspond to the temperature, humidity, gas concentration, and wind speed in the monitoring parameters respectively;

[0036] Step 2: Randomly select a monitoring parameter as the main analysis parameter, and use the remaining monitoring parameters as influencing parameters, and establish a parameter association model. The specific calculation method of the parameter association model includes:

[0037] S1: According to the fires that have occurred in the historical data, the historical monitoring data of each fire is obtained, wherein the historical monitoring data includes the monitoring data before the fire occurs and the monitoring data when the fire occurs. At the same time, the monitoring data before the fire occurs is marked as the pre-data, and the monitoring data when the fire occurs is marked as the post-data;

[0038] It should be further explained that when selecting the monitoring data before the fire occurs, the fire occurrence time is taken as the benchmark, and the Ty time period is subtracted from the fire occurrence time to obtain the leading time point, and then the corresponding monitoring data is obtained at the position of the leading time point, and used as the monitoring data before the fire occurs, wherein the Ty time period is a fixed time value, and the specific value is obtained by technicians in this field after big data calculation;

[0039] S2: Subtract the pre-data from the post-data in each fire to obtain the data change value FGj, where j represents different monitoring parameters. When subtracting the pre-data from the post-data, only the monitoring data of the same data type are subtracted;

[0040] Extract the data value corresponding to the main analysis parameter in the data change value FGj, and mark the remaining monitoring data in the data change value FGj as influencing parameters. Then mark the data change values ​​in the same fire as a data group, obtain all the data groups in the historical data, and divide the data groups into training data and test data;

[0041] S3: First, a simulation operation model is constructed based on the training data, and then the simulation operation model is evaluated using the test data, and the final parameter association model is determined based on the evaluation data;

[0042] It should be further explained that, in this embodiment, when constructing the simulation operation model, a neural network algorithm is selected for operation, by taking the influencing parameter as the input layer of the neural network algorithm, the hidden layer is a plurality of neurons, the main analysis parameter is taken as the output layer, the training data is input into the neural network, and the simulation operation is performed to obtain the parameter association model GCj, wherein GCj represents the parameter association model with the monitoring parameter j as the main analysis parameter. The specific use of the neural network algorithm to construct the simulation operation model is a prior art, which will not be described in detail here;

[0043] Each monitoring parameter is set as a main analysis parameter in turn, and a parameter association model of each main analysis parameter is determined according to the method in step S3 above;

[0044] Step 3: Obtain the monitoring parameters collected in real time, and identify the monitoring parameters before the Ty time period based on the Ty time period, taking the real time time as the time node, and then subtract the monitoring parameters before the Ty time period from the monitoring parameters collected in real time to obtain the real-time change parameter BHj;

[0045] Obtain the parameter association model GCj of the monitoring parameter j, then obtain the real-time change parameter corresponding to the influencing parameter in the parameter association model GCj, and input the real-time change parameter corresponding to the influencing parameter into the parameter association model GCj for calculation, and mark the value obtained after the calculation as the theoretical change value LBj;

[0046] Obtain the theoretical change value LBj of all monitoring parameters, and then use the formula The fire zone prediction value FIR is obtained, where J represents the total amount of monitoring parameters, a1 is a constant coefficient, and 0<a1<1;

[0047] It should be further explained that for the fire zone prediction value FIR, when When the value is larger, the deviation between the theoretical change value LBj of the monitoring parameter and the real-time change parameter BHj is greater. At this time, the smaller the fire zone prediction value FIR is, the smaller the risk value of fire in the coalfield fire zone is. On the contrary, when The smaller the value, the smaller the deviation between the theoretical change value LBj of the monitoring parameter and the real-time change parameter BHj. At this time, the larger the fire zone prediction value FIR is, the greater the risk value of fire in the coalfield fire zone is.

[0048] The fire zone prediction value FIR is compared with the risk coefficient. If the fire zone prediction value FIR is less than the risk coefficient, a normal monitoring signal is generated. When the monitoring system detects a normal monitoring signal, a normal monitoring instruction is transmitted to the monitoring device at this location. On the contrary, if the fire zone prediction value FIR is greater than or equal to the risk coefficient, an abnormal risk signal is generated. The specific value of the risk coefficient is obtained by technicians in this field after big data calculation.

[0049] Step 4: When the monitoring system detects an abnormal risk signal, the three-dimensional coordinate position of the monitoring device that generates the abnormal risk signal is obtained, and the three-dimensional coordinate position and the abnormal risk signal are simultaneously transmitted to the terminal device of the relevant management personnel. When the relevant management personnel receive the abnormal risk signal, they conduct an emergency inspection of the coalfield fire area according to the corresponding three-dimensional coordinate position, and then control the fire in the coalfield fire area in a timely manner.

[0050] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for monitoring a coalfield fire zone, characterized in that: The method specifically comprises the following steps: Step 1: Collect regional information of the coalfield fire area, set up a simulation monitoring model based on the regional information, and then determine the three-dimensional coordinate position of each monitoring device in the simulation monitoring model, and use the monitoring equipment to monitor the coalfield fire area in real time; Step 2: arbitrarily select a monitoring parameter as the main analysis parameter, and use the remaining monitoring parameters as influencing parameters, obtain the historical monitoring data of each fire, determine the pre-data and post-data according to the time of fire occurrence, and use the pre-data and post-data to respectively determine the data change value of the monitoring parameter, the data change value of the influencing parameter and the data change value of the main analysis parameter to construct a simulation operation model; train the simulation operation model to obtain a parameter association model of the monitoring parameter; Step 3: Obtain the monitoring parameters collected in real time, and calculate the real-time change parameters by combining them with the monitoring parameters before the preset time period. Then, according to the parameter association model, take the influencing parameters in the real-time change parameters, and input them into the corresponding parameter association model. Then, mark the operation output result as the theoretical change value, obtain the theoretical change values ​​of all monitoring parameters, combine the theoretical change values ​​with the corresponding real-time change parameters, determine the fire zone prediction value, and then determine the abnormal risk signal based on the fire zone prediction value. Step 4: Detect abnormal risk signals and identify the monitoring equipment that generates abnormal risk signals, obtain the three-dimensional coordinate position of the corresponding monitoring equipment, and transmit the three-dimensional coordinate position and the abnormal risk signal to the terminal device of the relevant management personnel at the same time. The management personnel confirm the abnormality of the coalfield fire area based on the three-dimensional coordinate position.

2. A method for monitoring a coalfield fire zone according to claim 1, characterized in that: The simulation monitoring model is a three-dimensional model and is equipped with a three-dimensional spatial coordinate system. Based on the regional information of the coalfield fire area, the monitoring equipment and the position of the monitoring equipment in the target fire area are obtained. According to the position of the monitoring equipment in the target fire area, the three-dimensional coordinate position of the monitoring equipment in the simulation monitoring model is determined. At the same time, the simulation monitoring model restores the actual coalfield fire area according to the preset scale.

3. A method for monitoring a coalfield fire zone according to claim 1, characterized in that: The parameter association model setting method includes: S1: According to the fires that have occurred in the historical data, the historical monitoring data of each fire is obtained, wherein the historical monitoring data includes the monitoring data before the fire occurs and the monitoring data when the fire occurs. At the same time, the monitoring data before the fire occurs is marked as the pre-data, and the monitoring data when the fire occurs is marked as the post-data; S2: Subtract the pre-data from the post-data in each fire to obtain the data change value. When subtracting the pre-data from the post-data, only the monitoring data of the same data type are subtracted; Extract the data value corresponding to the main analysis parameter in the data change value, and mark the remaining monitoring data in the data change value as the influencing parameter, then mark the data change value in the same fire as a data group, obtain all the data groups in the historical data, and divide the data groups into training data and test data; S3: First, a simulation operation model is constructed based on the training data, and then the simulation operation model is evaluated using the test data, and the final parameter association model is determined based on the evaluation data.

4. A method for monitoring a coalfield fire zone according to claim 3, characterized in that: When selecting monitoring data before a fire occurs, the fire occurrence time is used as a reference, and the fire occurrence time is subtracted from the preset time period to obtain the leading time point. The corresponding monitoring data is then obtained at the location of the leading time point and used as the monitoring data before the fire occurs.

5. A method for monitoring a coalfield fire zone according to claim 4, characterized in that: The process of constructing the simulation operation model is as follows: a neural network algorithm is selected for operation, the influencing parameters are used as the input layer of the neural network algorithm, the hidden layer is a number of neurons, and the main analysis parameters are used as the output layer; The process of obtaining the parameter association model is: inputting the training data into the neural network, and performing simulation operations to obtain the parameter association model.

6. A method for monitoring a coalfield fire zone according to claim 1, characterized in that: The calculation method of the fire area prediction value described in step 4 includes: Acquire the monitoring parameters collected in real time, and at the same time, according to the preset time period, use the real time time as the time node to identify the monitoring parameters before the preset time period, and then subtract the monitoring parameters collected in real time from the monitoring parameters before the preset time period to obtain the real-time change parameters; Obtain a parameter association model of the monitoring parameter, then obtain the real-time change parameter corresponding to the influencing parameter in the parameter association model, and input the real-time change parameter corresponding to the influencing parameter into the parameter association model for calculation, and mark the value obtained after the calculation as the theoretical change value; Obtain the theoretical change value LBj of all monitoring parameters, and then use the formula The fire zone prediction value FIR is obtained, where J represents the total amount of monitoring parameters, a1 is a constant coefficient, and 0<a1<1.

7. A method for monitoring a coalfield fire zone according to claim 6, characterized in that: The method for determining the abnormal risk signal in step 4 includes: The fire zone prediction value FIR is compared with the risk coefficient. If the fire zone prediction value FIR is less than the risk coefficient, a normal monitoring signal is generated. When the monitoring system detects a normal monitoring signal, a normal monitoring instruction is transmitted to the monitoring equipment at this location. Conversely, if the fire zone prediction value FIR is greater than or equal to the risk coefficient, an abnormal risk signal is generated.

Images