Coal pile detection safety early warning monitoring method

Abstract

The invention relates to the technical field of coal management, and provides a coal pile detection safety early warning monitoring method, which obtains coal data and coal pile data and performs training in combination with a random forest model to obtain a coal pile detection safety early warning monitoring model. The method comprises the steps of collecting coal pile image data, extracting coal pile shape features, matching the coal pile shape features with coal pile shape knowledge information to obtain coal pile shape information, and substituting the coal pile shape information into a model to obtain a monitoring result. And when the monitoring result is that the potential safety hazard exists, generating a verification task and sending the verification task to the inspection equipment. After being preprocessed, the feedback data of the inspection equipment is substituted into the inspection data analysis model to obtain an analysis result, error comparison is performed on the analysis result and a monitoring result, and the accuracy of an early warning result is improved. According to the invention, data analysis can be carried out according to the shape of the coal pile, the early warning accuracy and timeliness are improved, the loss of a coal yard is reduced, the safety is improved, and automatic and intelligent coal pile potential safety hazard identification and early warning are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of coal management, and particularly to a method for detecting and safely warning and monitoring coal piles. Background Art

[0002] The safety of coal pile detection is an essential part of industrial production. The safety of coal pile detection is related to the life safety of operators and the sustainable development of the environment. During the storage and use of coal piles, comprehensive safety inspections should be carried out regularly, including but not limited to temperature monitoring of coal piles, assessment of spontaneous combustion tendency, detection of harmful gas emissions, and inspection of pile stability. An abnormal increase in temperature may be a precursor to spontaneous combustion, and measures need to be taken promptly to cool down. The assessment of spontaneous combustion tendency can prevent the occurrence of fire accidents. At the same time, detecting harmful gases such as carbon monoxide and hydrogen sulfide released from coal piles is crucial for preventing personnel poisoning.

[0003] In the actual application process of existing coal pile detection methods, it is impossible to perform targeted data analysis on the shape of coal piles, resulting in the inability to give timely warnings according to the actual shape of coal piles, increasing the occurrence frequency of events such as coal heating, smoking, and spontaneous combustion, and increasing the loss of stored coal in the coal yard. To solve this technical pain point, the present invention provides a method for detecting and safely warning and monitoring coal piles. Summary of the Invention

[0004] In view of the above technical problems, the present invention provides a method for detecting and safely warning and monitoring coal piles, which is used to solve the problem that existing technologies cannot perform targeted data analysis on the shape of coal piles, resulting in the inability to give timely warnings according to the actual shape of coal piles.

[0005] A method for detecting and safely warning and monitoring coal piles according to the present invention includes:

[0006] Step S101, obtaining coal data and coal pile data. The coal data includes coal production information, coal attribute information, coal demand information, and coal transportation capacity information. The coal pile data includes used information of the coal yard, unused information of the coal yard, coal pile location information, coal pile shape knowledge information, and coal pile historical management data;

[0007] Step S102, dividing the coal data and coal pile data into a training set and a validation set, and using the training set and the validation set to train a random forest model to obtain a coal pile detection safety warning and monitoring model;

[0008] Step S103, collecting coal pile image data, extracting data features from the coal pile image data to obtain coal pile shape features, matching the coal pile shape features in the coal pile shape knowledge information to obtain coal pile shape information corresponding to the coal pile shape features, and substituting the coal pile shape information corresponding to the coal pile shape features into the coal pile detection safety warning and monitoring model to obtain a coal pile detection safety warning and monitoring result;

[0009] Step S104, the safety warning monitoring result of coal pile detection includes the risk of potential safety hazards and the absence of potential safety hazards. If the safety warning monitoring result of coal pile detection indicates the risk of potential safety hazards, a verification task for potential safety hazards is generated and sent to the inspection equipment.

[0010] Step S105, receive the inspection data feedback from the inspection equipment. The inspection data includes the image data of the target coal pile, the temperature data of the target coal pile, the gas detection data of the target coal pile, and the attribute data of the target coal pile. Substitute the inspection data feedback from the inspection equipment into the preset inspection data analysis model to obtain the inspection data analysis result. Compare the inspection data analysis result with the safety warning monitoring result of coal pile detection. If the error value of the comparison result between the inspection data analysis result and the safety warning monitoring result of coal pile detection exceeds the preset error range, the safety warning monitoring result of coal pile detection is incorrect. If the error value of the comparison result between the inspection data analysis result and the safety warning monitoring result of coal pile detection does not exceed the preset error range, the safety warning monitoring result of coal pile detection is accurate.

[0011] Furthermore, for a safety warning monitoring method for coal pile detection according to the present invention, the step S101 includes:

[0012] The coal production information includes the production date of coal, the production location of coal, the production batch of coal, and the production volume of coal.

[0013] The coal attribute information includes the type of coal, the coal quality, and the particle size distribution.

[0014] The coal demand information includes the market demand for coal and customer orders.

[0015] The coal transportation capacity information includes the transportation mode, the transportation capacity, and the transportation cost.

[0016] The information on the used area of the coal yard includes the used area, the amount of coal stacked, and the stacking location.

[0017] The information on the unused area of the coal yard collected includes the size, location, and terrain of the remaining available space in the coal yard.

[0018] The information on the coal pile location collected includes the location of the coal pile, the coordinates of the coal pile, the height of the coal pile, and the width of the coal pile.

[0019] The coal pile shape knowledge information includes the heat dissipation performance and the tendency of oxidation reaction of coal piles with different shapes.

[0020] The coal pile historical management data includes the historical coal pile temperature monitoring data, the oxidation reaction situation, and the records of spontaneous combustion events.

[0021] Further, for a coal pile detection safety early warning monitoring method according to the present invention, the step S102 includes:

[0022] Randomly divide the preprocessed data set into a training set and a validation set, select random forest as the model algorithm, set the model parameters, use the training set data to train the random forest model, and the random forest model learns the mapping relationship between the features in the coal data and the coal pile data and whether there are potential safety hazards in the coal pile, so as to obtain a coal pile detection safety early warning monitoring model.

[0023] Further, for a coal pile detection safety early warning monitoring method according to the present invention, the step S103 includes:

[0024] The inspection equipment includes drones, cameras, and inspection robots. Use the drones, cameras, and inspection robots to collect images of the coal pile from multiple angles and with high resolution;

[0025] Preprocess the collected coal pile image data, use image processing techniques, which include edge detection, contour extraction, and region segmentation, to extract the shape features of the coal pile from the image data. The shape features of the coal pile include the contour line, vertices, area, and perimeter of the coal pile;

[0026] Match the extracted shape features of the coal pile with the pre-established coal pile shape knowledge information, which includes the feature description and classification of the coal pile shape;

[0027] Use a matching algorithm to find the coal pile shape most similar to the extracted features and determine its corresponding coal pile shape information;

[0028] Take the matched coal pile shape information as input features and substitute them into the previously trained coal pile detection safety early warning monitoring model to obtain the coal pile detection safety early warning monitoring result.

[0029] Further, for a coal pile detection safety early warning monitoring method according to the present invention, the step S104 includes:

[0030] Judge the received coal pile detection safety early warning monitoring result. If the coal pile detection safety early warning monitoring result indicates no potential safety hazard risk, the process ends;

[0031] If the coal pile detection safety early warning monitoring result indicates a potential safety hazard risk, proceed to the next step to generate a potential safety hazard risk verification task;

[0032] Generate a potential safety hazard risk verification task according to the coal pile information indicating a potential safety hazard risk in the coal pile detection safety early warning monitoring result;

[0033] The safety hazard risk verification task includes verifying whether there are abnormal temperatures and gas releases in the coal pile.

[0034] Furthermore, for the coal pile detection safety warning and monitoring method described in the present invention, the step S105 includes:

[0035] Preprocess the received inspection data, extract the shape, color, and texture features of the coal pile from the preprocessed image data, extract the surface temperature and temperature distribution features of the coal pile from the temperature data, extract the concentration and type features of harmful gases from the gas detection data, and extract the type, stacking time, and stacking method features of the coal pile from the attribute data.

[0036] Furthermore, for the coal pile detection safety warning and monitoring method described in the present invention, the step S105 includes:

[0037] Substitute the extracted features into a preset inspection data analysis model. The preset inspection data analysis model outputs inspection data analysis results based on the input feature data. The inspection data analysis results include the safety status assessment, potential risk warning, and anomaly detection of the coal pile;

[0038] Align the inspection data analysis results with the coal pile detection safety warning and monitoring results in terms of time stamps and coal pile identification information;

[0039] Set the acceptable error range for each indicator, and compare each indicator in the inspection data analysis results with the corresponding indicator in the coal pile detection safety warning and monitoring results;

[0040] Calculate the error value for each indicator, that is, the difference between the inspection data analysis results and the coal pile detection safety warning and monitoring results;

[0041] For each compared indicator, determine whether its error value exceeds the preset error range;

[0042] If the error value of any one indicator exceeds the preset range, it is considered that there is an error in the coal pile detection safety warning and monitoring results;

[0043] If the error values of all indicators are within the acceptable range, it is considered that the coal pile detection safety warning and monitoring results are accurate, and the comparison results are output in the form of reports, charts, or alarms;

[0044] If the coal pile detection safety warning and monitoring results are judged to be incorrect, send the coal pile detection safety warning and monitoring results to the management operation terminal.

[0045] The present invention provides a coal pile detection safety warning and monitoring method, and its beneficial effects are mainly reflected in the following aspects:

[0046] The present invention obtains coal data and coal pile data, and conducts targeted analysis on these data. Especially considering the shape characteristics of the coal pile, it can more accurately evaluate the potential safety hazards of the coal pile. Compared with the prior art that cannot conduct data analysis on the shape of the coal pile, the present invention significantly improves the accuracy and timeliness of early warning.

[0047] Due to the ability to give early warnings based on the shape of the coal pile, the present invention effectively reduces the occurrence frequency of events such as coal combustion heating, smoking, and spontaneous combustion, thereby reducing the loss of stored coal in the coal yard and improving economic benefits. By giving early warnings and verifying the potential safety hazards of the coal pile in a timely manner, the present invention helps to prevent the occurrence of fire accidents, protects the lives of operating personnel, and reduces environmental pollution caused by coal pile safety problems.

[0048] The present invention not only considers the image data of the coal pile, but also combines various data sources such as the temperature data, gas detection data, and attribute data of the coal pile. By comprehensively analyzing these data, it provides a more comprehensive and accurate assessment of the safety status of the coal pile. Using inspection equipment such as unmanned aerial vehicles, cameras, and inspection robots, the present invention realizes the automatic acquisition and processing of coal pile image data, improving work efficiency. At the same time, through the preset inspection data analysis model and coal pile detection safety early warning monitoring model, the intelligent identification and early warning of the potential safety hazards of the coal pile are realized.

[0049] The present invention receives the inspection data fed back by the inspection equipment, and compares the data with the coal pile detection safety early warning monitoring results to verify the accuracy of the early warning results. This error verification mechanism improves the reliability of the early warning system and reduces the situations of false alarms and missed alarms. The coal pile detection safety early warning monitoring method provided by the present invention can be adjusted and optimized according to actual needs, such as adjusting model parameters, adding new data sources, etc., to adapt to the monitoring requirements of different coal yards and coal piles.

[0050] In summary, the present invention provides an efficient, accurate, and reliable coal pile detection safety early warning monitoring method, which is of great significance for improving the safety of the coal yard and reducing economic losses. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, for those of ordinary skill in the art, other drawings can also be obtained according to the drawings without creative efforts.

[0052] Figure 1 It is a schematic flow chart of a coal pile detection safety early warning monitoring method provided by the present invention.

[0053] Figure 2 It is a schematic diagram of the shape of an existing coal pile provided by the present invention.

[0054] Figure 3 The first schematic diagram of the new coal pile provided by the present invention.

[0055] Figure 4 The second schematic diagram of the new coal pile provided by the present invention. Specific embodiments

[0056] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with specific embodiments of the present invention and the corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of the present invention. The technical solutions provided by each embodiment of the present invention will be described in detail below with reference to the drawings.

[0057] To better understand the objectives of the present invention, the present invention will be further described in detail below.

[0058] Please refer to Figures 1 to 4 , a method for detecting and safely warning and monitoring a coal pile according to the present invention includes:

[0059] Step S101, obtaining coal data and coal pile data. The coal data includes coal production information, coal attribute information, coal demand information, and coal transportation capacity information. The coal pile data includes the used information of the coal yard, the unused information of the coal yard, the coal pile location information, the coal pile shape knowledge information, and the coal pile historical management data;

[0060] Step S102, dividing the coal data and the coal pile data into a training set and a validation set, and using the training set and the validation set to train a random forest model to obtain a coal pile detection safety warning and monitoring model;

[0061] Step S103, collecting coal pile image data, extracting data features from the coal pile image data to obtain coal pile shape features, matching the coal pile shape features in the coal pile shape knowledge information to obtain the coal pile shape information corresponding to the coal pile shape features, and substituting the coal pile shape information corresponding to the coal pile shape features into the coal pile detection safety warning and monitoring model to obtain a coal pile detection safety warning and monitoring result;

[0062] Step S104, the coal pile detection safety warning and monitoring result includes the risk of potential safety hazards and no risk of potential safety hazards. If the coal pile detection safety warning and monitoring result is the risk of potential safety hazards, a potential safety hazard risk verification task is generated and sent to the inspection equipment;

[0063] Step S105: Receive the inspection data fed back by the inspection device. The inspection data includes the image data of the target coal pile, the temperature data of the target coal pile, the gas detection data of the target coal pile, and the attribute data of the target coal pile. Substitute the inspection data fed back by the inspection device into the preset inspection data analysis model to obtain the inspection data analysis result. Compare the inspection data analysis result with the coal pile detection safety warning and monitoring result. If the error value of the comparison result between the inspection data analysis result and the coal pile detection safety warning and monitoring result exceeds the preset error range, the coal pile detection safety warning and monitoring result is incorrect. If the error value of the comparison result between the inspection data analysis result and the coal pile detection safety warning and monitoring result does not exceed the preset error range, the coal pile detection safety warning and monitoring result is accurate.

[0064] Specifically, for a coal pile detection safety warning and monitoring method according to the present invention, the step S101 includes:

[0065] The coal production information includes the production date of the coal, the production location of the coal, the production batch of the coal, and the coal production volume.

[0066] The coal attribute information includes the type of the coal, the coal quality, and the particle size distribution.

[0067] The coal demand information includes the market demand situation for the coal and the customer orders.

[0068] The coal transportation capacity information includes the transportation mode, the transportation capacity, and the transportation cost.

[0069] The used information of the coal yard includes the used area, the amount of coal stacked, and the stacking location.

[0070] The information on the unused part of the coal yard collected includes the size, location, and terrain of the remaining available space in the coal yard.

[0071] The information on the location of the coal pile collected includes the location of the coal pile, the coordinates of the coal pile, the height of the coal pile, and the width of the coal pile.

[0072] The knowledge information on the shape of the coal pile includes the heat dissipation performance and the tendency of oxidation reaction of coal piles with different shapes.

[0073] The historical management data of the coal pile includes the historical temperature monitoring data of the coal pile, the oxidation reaction situation, and the record of spontaneous combustion events.

[0074] Specifically, for a coal pile detection safety warning and monitoring method according to the present invention, the step S102 includes:

[0075] Randomly divide the preprocessed dataset into a training set and a validation set. Select the random forest as the model algorithm, set the model parameters, and use the training set data to train the random forest model. The random forest model learns the mapping relationship between the features in the coal data and the coal pile data and whether there are potential safety hazards in the coal pile, and obtains a safety warning monitoring model for coal pile detection.

[0076] Specifically, for a safety warning monitoring method for coal pile detection according to the present invention, the step S103 includes:

[0077] The inspection equipment includes drones, cameras, and inspection robots. Use drones, cameras, and inspection robots to collect images of the coal pile from multiple angles with high resolution;

[0078] Preprocess the collected coal pile image data. Use image processing techniques, including edge detection, contour extraction, and region segmentation, to extract the shape features of the coal pile from the image data. The shape features of the coal pile include the contour line, vertices, area, and perimeter of the coal pile;

[0079] Match the extracted shape features of the coal pile with the pre-established coal pile shape knowledge information, which includes the feature description and classification of the coal pile shape;

[0080] Use a matching algorithm to find the coal pile shape most similar to the extracted features and determine its corresponding coal pile shape information;

[0081] Take the matched coal pile shape information as input features and substitute it into the previously trained safety warning monitoring model for coal pile detection to obtain the safety warning monitoring result for coal pile detection.

[0082] Specifically, for a safety warning monitoring method for coal pile detection according to the present invention, the step S104 includes:

[0083] Judge the received safety warning monitoring result for coal pile detection. If the safety warning monitoring result for coal pile detection indicates no potential safety hazard risk, the process ends;

[0084] If the safety warning monitoring result for coal pile detection indicates a potential safety hazard risk, proceed to the next step to generate a verification task for potential safety hazard risk;

[0085] Generate a verification task for potential safety hazard risk according to the coal pile information indicating a potential safety hazard risk in the safety warning monitoring result for coal pile detection;

[0086] The verification task for potential safety hazard risk includes verifying whether there are abnormal temperatures and gas releases in the confirmed coal pile.

[0087] Specifically, for a coal pile detection safety early warning and monitoring method according to the present invention, the step S105 includes:

[0088] Preprocess the received inspection data, extract the shape, color, and texture features of the coal pile from the preprocessed image data, extract the surface temperature and temperature distribution features of the coal pile from the temperature data, extract the concentration and type features of harmful gases from the gas detection data, and extract the type, stacking time, and stacking method features of the coal pile from the attribute data.

[0089] Specifically, for a coal pile detection safety early warning and monitoring method according to the present invention, the step S105 includes:

[0090] Substitute the extracted features into a preset inspection data analysis model. The preset inspection data analysis model outputs inspection data analysis results according to the input feature data. The inspection data analysis results include the safety status assessment, potential risk early warning, and anomaly detection of the coal pile;

[0091] Align the inspection data analysis results and the coal pile detection safety early warning and monitoring results in terms of timestamp and coal pile identification information;

[0092] Set the acceptable error range for each indicator, and compare each indicator in the inspection data analysis results with the corresponding indicator in the coal pile detection safety early warning and monitoring results;

[0093] Calculate the error value of each indicator, that is, the difference between the inspection data analysis results and the coal pile detection safety early warning and monitoring results;

[0094] For each compared indicator, determine whether its error value exceeds the preset error range;

[0095] If the error value of any one indicator exceeds the preset range, it is considered that there is an error in the coal pile detection safety early warning and monitoring results;

[0096] If the error values of all indicators are within the acceptable range, it is considered that the coal pile detection safety early warning and monitoring results are accurate, and the comparison results are output in the form of reports, charts, or alarms;

[0097] If the coal pile detection safety early warning and monitoring results are judged to be incorrect, send the coal pile detection safety early warning and monitoring results to the management operation terminal.

[0098] The coal pile shape knowledge information includes:

[0099] The principle of coal pile heating: for different coal pile shapes, there are differences in the internal temperature rise rate. And in a closed coal yard, there is no rain or snow on the coal pile surface. At the same time, the air flow in the coal storage yard is restricted, and the convective heat transfer between the air and the coal pile surface is greatly weakened. The heat dissipation of the coal pile mainly depends on the heat conduction between the raw coal and the air convection and heat conduction in the gaps between the raw coal. Designing a reasonable pile shape enables the heat in the coal pile to be dissipated in a timely manner, reduces the possibility of forming a heat source, and at the same time can isolate the entry of internal oxygen and reduce the degree of oxidation reaction.

[0100] From the cross-sectional analysis, due to the stacking order of the incoming coal in the upper half, the coal storage time is generally shorter, and the sequence of phenomena such as oxidation heating and smoking occurs later than that in the lower half.

[0101] In the lower half, A is the central area of the coal pile. The coal in this area is compacted under the pressure of the upper half and is relatively tight. At the same time, it is located inside and has difficulty in contacting the air. The coal in this position has a lower degree of oxidation reaction due to the lack of oxygen participation. However, at position B in the lower half, that is, the inner and outer edge areas, when the large pile undergoes oxidation, phenomena such as temperature rise and smoking usually first appear in this area.

[0102] The main reason is: Please refer to Figure 2 , this position is at the bottom of the large pile, with a longer coal storage time. The coal in this position is on the outside and has convenient contact with oxygen. The heat generated by the oxidation reaction is released in large quantities, and the high temperature accelerates further oxidation until spontaneous combustion. This process is in an accelerating state.

[0103] For the new coal pile configuration, with the idea of inhibiting the oxidation process of raw coal, under the existing equipment conditions, the main solutions that can be adopted are: the covering inhibition type and the equal-position replacement type.

[0104] For the covering inhibition type, please refer to Figure 3 , which mainly covers a new layer of coal type Q outside the B area to make it more difficult to contact oxygen.

[0105] For the equal-position replacement type, please refer to Figure 4 , which mainly replaces the coal type in the B area with Q when stacking the coal, so that the most easily heated part of the overall coal type structure of the large pile is replaced with coal.

[0106] The historical management data of the coal pile includes the data of coal pile tests, including:

[0107] Before the start of the test, first, the coal management professional needs to clear an open space with a length of more than 30m in the coal yard. Before coal stacking, use a bulldozer to clean the site and let it air for 8 hours. After 8 hours, the program control operator uses an infrared camera to check that there is no obvious heat source on the site surface exceeding the ambient temperature, and records the photos and time.

[0108] The shift supervisor reports to the operation engineer and the test team to start piling coal. Leave a space of 4 - 5m on the inner and outer sides during coal piling. Record the start time of coal piling. After complete piling, record the time when the complete coal pile is formed.

[0109] Pile the coal to the inner and outer sides of the coal pile. During coal piling, the bucket wheel coal flow adheres to the edge of the coal pile to ensure that the coal completely covers the inner and outer sides of the large pile. Stop coal piling when the edge is about 1m away from the coal yard wall. The operation personnel record the start and stop times.

[0110] Regularly measure the temperature, measure and record it every 4 hours. The coal pipe personnel and the program control operator conduct daily inspections on the inner and outer sides of the coal pile. The program control operator is responsible for taking infrared photos of the temperature of the inner and outer sides of the coal pile.

[0111] When white hot air starts to emerge from the inner and outer sides, record the location and time.

[0112] And other position replacement type coal pile test data:

[0113] Before the test starts, first, the coal pipe professional needs to clear an open space with a length of more than 30m in the coal yard. Use a bulldozer to clean the site before coal piling and let it air for 8 hours. After 8 hours, the program control operator uses an infrared camera to check that there is no obvious heat source on the site surface exceeding the ambient temperature and records the photo and time.

[0114] The shift supervisor reports to the operation engineer and the test team to start piling coal. Pile the coal within the test area, about 1m away from the edge of the coal yard, and pile a long strip to cover the test area. Record the time after completion.

[0115] Pile the test coal type in the middle area of the long strip coal pile and record the start time of piling the test coal type.

[0116] Slowly raise the coal pile until the edge of the large pile coincides with the curve of the coal. At this time, the coal of the large pile is about to start sliding from above and covering the coal. Then stop coal piling, record the time, and the operation personnel record the time.

[0117] Regularly measure the temperature and record it. The coal pipe personnel and the program control operator conduct daily inspections on the inner and outer sides of the coal pile. The program control operator is responsible for taking infrared photos of the temperature of the inner and outer sides of the coal pile. When white hot air starts to emerge from the inner and outer sides, record the location and time.

[0118] The present invention provides a method for detecting and safely warning and monitoring coal piles, which solves the problem that the prior art cannot perform targeted data analysis on the shape of coal piles, resulting in the inability to give timely warnings according to the actual shape of coal piles through the following technical solutions:

[0119] The present invention obtains coal data and coal pile data, which include coal production information, attribute information, demand information, transportation capacity information, as well as the used and unused information of the coal yard, coal pile location information, coal pile shape knowledge information, and historical management data. Then, these data are divided into a training set and a validation set for training a random forest model.

[0120] The training set and the validation set are used to train the random forest model, enabling the model to learn the mapping relationship between the features in the coal data and coal pile data and whether there are potential safety hazards in the coal pile, thereby obtaining a coal pile detection safety warning and monitoring model. Image data of the coal pile is collected, and image processing techniques (such as edge detection, contour extraction, and region segmentation) are used to extract the shape features of the coal pile from the image data, such as contour lines, vertices, area, and perimeter.

[0121] The extracted shape features of the coal pile are matched with the pre-established coal pile shape knowledge information to determine the shape information of the coal pile. The obtained coal pile shape information is substituted into the trained coal pile detection safety warning and monitoring model to obtain the coal pile detection safety warning and monitoring result. If the monitoring result shows a risk of potential safety hazards, a verification task for potential safety hazards is generated and sent to inspection equipment (such as drones, cameras, inspection robots). The inspection equipment conducts detailed image, temperature, gas detection, etc. on the target coal pile and feeds back the inspection data.

[0122] The inspection data is preprocessed, key features are extracted, and substituted into a preset inspection data analysis model to obtain the inspection data analysis result. The inspection data analysis result is compared with the coal pile detection safety warning and monitoring result, and the accuracy of the monitoring result is judged by calculating the error value.

[0123] Through the above steps, the present invention can conduct detailed data analysis on the shape of the coal pile and give timely warnings according to the actual shape of the coal pile, which not only improves the accuracy and timeliness of the warning, but also effectively reduces the losses in the coal yard and improves the operation safety.

Claims

1. A coal pile detection safety early warning monitoring method, characterized in that: include: Step S101, obtaining coal data and coal pile data, wherein the coal data includes coal production information, coal attribute information, coal demand information and coal transportation capacity information, and the coal pile data includes coal yard used information, coal yard unused information, coal pile location information, coal pile shape knowledge information and coal pile historical management data; Step S102, dividing the coal data and the coal pile data into a training set and a validation set, using the training set and the validation set to train the random forest model to obtain a coal pile detection safety early warning monitoring model; Step S103, collecting coal pile image data, extracting data features from the coal pile image data to obtain coal pile shape features, matching the coal pile shape features with coal pile shape knowledge information to obtain coal pile shape information corresponding to the coal pile shape features, substituting the coal pile shape information corresponding to the coal pile shape features into a coal pile detection safety early warning monitoring model to obtain coal pile detection safety early warning monitoring results; Step S104: the coal pile detection safety early warning monitoring result includes the presence of potential safety hazards and the absence of potential safety hazards. If the coal pile detection safety early warning monitoring result is the presence of potential safety hazards, a potential safety hazards risk verification task is generated and sent to the inspection equipment; Step S105, receiving the inspection data fed back by the inspection equipment, including the inspection data including the image data of the target coal pile, the temperature data of the target coal pile, the gas detection data of the target coal pile and the attribute data of the target coal pile, substituting the inspection data fed back by the inspection equipment into the preset inspection data analysis model to obtain the inspection data analysis results, and performing data comparison between the inspection data analysis results and the coal pile detection safety early warning monitoring results. If the error value of the comparison result between the inspection data analysis results and the coal pile detection safety early warning monitoring results exceeds the preset error range, then the coal pile detection safety early warning monitoring result is wrong. If the error value of the comparison result between the inspection data analysis results and the coal pile detection safety early warning monitoring results does not exceed the preset error range, then the coal pile detection safety early warning monitoring result is accurate.

2. A coal pile detection safety early warning monitoring method as claimed in claim 1, characterized in that: The step S101 includes: Coal production information includes coal production date, coal production location, coal production batches and coal output; Coal attribute information includes coal type, coal quality and particle size distribution; Coal demand information includes market demand for coal and customer orders; Coal transportation information includes transportation mode, transportation capacity and transportation cost; The information of coal yard usage includes the area used, the amount of coal piled up and the location of the pile; Collect unused information of coal yards, including the size, location and topography of the remaining available space in the coal yards; Collecting coal pile location information including the location of the coal pile, the coordinates of the coal pile, the height of the coal pile, and the width of the coal pile; The coal pile shape knowledge information includes the heat dissipation performance and oxidation reaction tendency of coal piles of different shapes; Coal pile historical management data includes historical coal pile temperature monitoring data, oxidation reaction conditions and spontaneous combustion event records.

3. A coal pile detection safety early warning monitoring method as claimed in claim 1, characterized in that: The step S102 includes: The preprocessed data set is randomly divided into a training set and a validation set. Random forest is selected as the model algorithm, the model parameters are set, and the random forest model is trained using the training set data. The random forest model learns the mapping relationship between the features in the coal data and coal pile data and whether there are safety hazards in the coal pile, and obtains a coal pile detection safety early warning monitoring model.

4. A coal pile detection safety early warning monitoring method as claimed in claim 1, characterized in that: The step S103 includes: Inspection equipment includes drones, cameras and inspection robots, which are used to collect multi-angle, high-resolution images of coal piles; Preprocess the collected coal pile image data by using image processing technology, including edge detection, contour extraction and region segmentation, to extract the shape features of the coal pile from the image data. The shape features of the coal pile include the contour line, vertices, area and perimeter of the coal pile. Matching the extracted coal pile shape features with pre-established coal pile shape knowledge information, the coal pile shape knowledge information including feature description and classification of the coal pile shape; Use a matching algorithm to find the coal pile shape that is most similar to the extracted features, and determine the corresponding coal pile shape information; The matched coal pile shape information is used as input features and substituted into the previously trained coal pile detection safety early warning monitoring model to obtain the coal pile detection safety early warning monitoring results.

5. A coal pile detection safety early warning monitoring method as claimed in claim 1, characterized in that: The step S104 includes: The received coal pile detection safety warning monitoring results are judged. If the coal pile detection safety warning monitoring results show that there is no potential safety hazard risk, the process ends; If the coal pile detection safety warning monitoring result shows that there is a potential safety hazard risk, the next step is to generate a potential safety hazard risk verification task; Generate a safety hazard risk verification task based on the coal pile information with safety hazard risks indicated in the coal pile detection safety early warning monitoring results; The safety hazard risk verification task includes verifying whether there is abnormal temperature and gas release in the coal pile.

6. A coal pile detection safety early warning monitoring method as claimed in claim 1, characterized in that: The step S105 includes: The received inspection data is preprocessed, and the shape, color, and texture features of the coal pile are extracted from the preprocessed image data. The surface temperature and temperature distribution features of the coal pile are extracted from the temperature data. The concentration and type characteristics of harmful gases are extracted from the gas detection data. The type, stacking time, and stacking method characteristics of the coal pile are extracted from the attribute data.

7. A coal pile detection safety early warning monitoring method as claimed in claim 6, characterized in that: The step S105 includes: Substitute the extracted features into a preset inspection data analysis model, which outputs inspection data analysis results based on the input feature data. The inspection data analysis results include safety status assessment of the coal pile, potential risk warning, and anomaly detection; The inspection data analysis results and the coal pile detection safety warning monitoring results are aligned in terms of timestamp and coal pile identification information; Set the acceptable error range for each indicator, and compare each indicator in the inspection data analysis results with the corresponding indicator in the coal pile detection safety early warning monitoring results; Calculate the error value of each indicator, that is, the difference between the inspection data analysis results and the coal pile detection safety warning monitoring results; For each compared indicator, determine whether its error value exceeds the preset error range; If the error value of any indicator exceeds the preset range, it is considered that there is an error in the coal pile detection safety warning monitoring result; If the error values ​​of all indicators are within the acceptable range, the coal pile detection safety warning monitoring results are considered accurate, and the comparison results are output in the form of reports, charts or alarms; If the coal pile detection safety early warning monitoring result is judged to be wrong, the coal pile detection safety early warning monitoring result will be sent to the management personnel operation terminal.

Images