A coal mine safety data processing method based on multi-source data fusion

Abstract

The present invention belongs to the field of data processing technology, and in particular, relates to a method for processing coal mine safety data based on multi-source data fusion. It includes collecting gas concentration, wind speed and other data that affect coal mine safety, using a hybrid filtering algorithm combined with multi-scale analysis to denoise, an improved weighted average fusion algorithm to fuse data, and using an improved Pearson correlation coefficient to extract key features and judge the safety status. The method denoises the data through multi-scale decomposition and filtering, establishes a weight evaluation index system to dynamically adjust the weights for data fusion, sets a correlation threshold to screen key features, and uses a support vector machine to predict the safety status and issue a timely warning. The present invention can effectively improve data quality, accurately judge the safety status of coal mines, and is of great significance to ensuring safe production in coal mines.

Description

Technical Field

[0001] The invention belongs to the technical field of data processing, and in particular relates to a coal mine safety data processing method based on multi-source data fusion. Background Art

[0002] With the continuous increase in the depth and intensity of coal mining, the safety problem of coal mines is becoming increasingly serious. Coal mine accidents often cause huge losses to people's lives and property. Therefore, it is very important to ensure the safety of coal mines. Traditional coal mine safety monitoring mainly relies on a single data source, such as gas monitoring systems or ventilation systems. These systems can only provide local safety information and cannot fully reflect the overall safety status of coal mines.

[0003] In order to more accurately judge the safety status of coal mines, it is necessary to integrate data from multiple data sources, including gas concentration data, wind speed data, tunnel pressure data, temperature data, dust concentration data, equipment operation status data, etc. However, these multi-source data often have noise and interference, and effective data processing and analysis are required.

[0004] At present, there are still some shortcomings in the application of data processing and analysis technology in the field of coal mine safety. For example, the accuracy and reliability of data fusion algorithms need to be improved, and the methods of key feature extraction and trend analysis are not perfect, resulting in inaccurate and timely judgment of the safety status of coal mines. Summary of the invention

[0005] In view of the above-mentioned technical problems, the present invention proposes a method for determining the safety status of a coal mine which is simple, easy to operate and can effectively solve the problem.

[0006] In order to achieve the above object, the technical solution adopted by the present invention comprises the following steps:

[0007] S1. First, it is necessary to collect various data that affect coal mine safety from coal mines;

[0008] S2. Then, data preprocessing is performed, and a hybrid filtering algorithm combined with multi-scale analysis is used to denoise the collected data to improve the data quality;

[0009] S3, then adopt the improved weighted average fusion algorithm to fuse the data from different data sources to obtain more comprehensive and accurate coal mine safety status information;

[0010] S4. Finally, the fused data is analyzed and the improved Pearson correlation coefficient is used to extract key features to determine the safety status of the coal mine;

[0011] The implementation process of the hybrid filtering denoising of the multi-scale analysis in step S2 is as follows:

[0012] S21, first perform multi-scale decomposition on the input coal mine data signal Perform wavelet transform to obtain sub-bands of different scales. The wavelet transform coefficients are expressed as: low-frequency sub-band ,in represents the low-frequency signal component at scale j and position k, is a scaling function used to construct an approximate signal, namely, a low-frequency sub-band and a high-frequency sub-band ,in represents the high-frequency signal component at scale j and position k, It is a wavelet transform function, which is used to perform multi-scale analysis on the signal and extract detailed information of different frequencies;

[0013] S22. Then, the low-frequency sub-band is subjected to preliminary denoising by using mean filtering. The signal after mean filtering is: ,in is the filter window size, Represents the signal value of the original low-frequency subband signal at position, Represents the center of the target position The signal values ​​of all the location points in the neighborhood are summed up;

[0014] S23, then for the high frequency sub-band, use the improved median filtering algorithm, first calculate the median M of all pixel values ​​in the window, compare the central pixel value with the median, if the difference between the central pixel value and the median is greater than a preset threshold, replace the central pixel value with the median, if the difference is less than or equal to the threshold, keep the central pixel value unchanged;

[0015] S24. Finally, the processed low-frequency sub-band and high-frequency sub-band are reconstructed to obtain the denoised coal mine data signal. The reconstruction process is achieved by combining the approximate coefficients and detail coefficients at each scale.

[0016] Preferably, the various data in step S1 include gas concentration data, wind speed data, tunnel pressure data, temperature data, dust concentration data, and equipment operation status data.

[0017] Preferably, the specific implementation of the improved weighted average fusion algorithm in step S3 is:

[0018] S31. First, establish a weight evaluation index system, set the historical accuracy of the data source, data update frequency, and equipment stability index;

[0019] S32. Secondly, quantify each indicator and assign corresponding weights according to its importance, and score each data source on each indicator;

[0020] S33, then dynamically adjust the weight according to the real-time performance of the data source and environmental changes;

[0021] S34, then for each data point, the fused data value is , where n is the number of data sources, is the data value of the ith data source, is the weight of the i-th data source;

[0022] S35. Finally, in order to make the fused data smoother, the moving average method is used to take the average value of the fused data within a certain time window as the final fusion result.

[0023] Preferably, in step S4, the improved Pearson correlation coefficient is used to extract key features, and the implementation steps for judging the safety status of the coal mine are:

[0024] S41. First, perform feature extraction and introduce a penalty coefficient into the traditional Pearson correlation coefficient. ,in represents the improved Pearson correlation coefficient, Represents the i-th observation value of two features respectively, Represent the means of the two features, n is the number of observations, is the weight coefficient of the penalty term, is the penalty function;

[0025] S42, then set a correlation threshold according to the actual situation, and filter out features with correlation coefficients greater than the threshold;

[0026] S43. Finally, the selected key features are used to construct a coal mine safety status model, and the support vector machine method is used to predict the coal mine safety status. According to the output results of the model, the safety status of the coal mine is judged. The output result probability value greater than 0.7 corresponds to a high risk score, 0.3 to 0.7 corresponds to a medium risk score, and less than 0.3 corresponds to a low risk score, and an early warning is issued in time.

[0027] Compared with the prior art, the advantages and positive effects of the present invention are that the use of multi-source data fusion can more comprehensively reflect the overall safety status of coal mines and avoid the limitations of a single data source. The hybrid filtering algorithm combined with multi-scale analysis is used for data denoising, which improves the quality of the data and provides a more reliable data basis for subsequent analysis and prediction. The improved Pearson correlation coefficient is used to extract key features, which can better screen out important features related to the safety status of coal mines and improve the accuracy of safety status judgment. The support vector machine method is used to predict the safety status of coal mines, and risk scoring and timely warning are performed based on the output results, which helps to discover safety hazards in advance and take corresponding measures to ensure safe production in coal mines. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0029] Figure 1 The comparison chart of the three groups of coal mine safety status prediction and actual value; DETAILED DESCRIPTION

[0030] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described below in conjunction with embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0031] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways than those described herein. Therefore, the present invention is not limited to the specific embodiments of the following disclosure.

[0032] Embodiment, in the context of the increasingly severe situation of coal mine production safety, in order to effectively prevent the occurrence of coal mine accidents, ensure the life safety of miners and the normal production and operation of coal mines, a coal mine safety data processing method based on multi-source data fusion is adopted to improve the safety management level of coal mines. First, various data affecting coal mine safety are collected from coal mines, including gas concentration data, wind speed data, tunnel pressure data, temperature data, dust concentration data, and equipment operation status data.

[0033] Considering that the quality problems of the collected data will inevitably affect the subsequent research, the present invention uses a hybrid filtering algorithm combined with multi-scale analysis to denoise the collected data when processing the data to improve the quality of the data. First, multi-scale decomposition is performed to the input coal mine data signal. Perform wavelet transform to obtain sub-bands of different scales. The wavelet transform coefficients are expressed as: low-frequency sub-band ,in represents the low-frequency signal component at scale j and position k, is a scaling function used to construct an approximate signal, namely, a low-frequency sub-band and a high-frequency sub-band ,in represents the high-frequency signal component at scale j and position k, is a wavelet transform function, which is used to perform multi-scale analysis on the signal and extract detailed information of different frequencies. Secondly, the low-frequency sub-band is preliminarily denoised by using mean filtering. The signal after mean filtering is: ,in is the filter window size, Represents the signal value of the original low-frequency subband signal at position, Represents the center of the target position The signal values ​​of all position points in the neighborhood are summed up; then for the high-frequency sub-band, an improved median filtering algorithm is used. First, the median M of all pixel values ​​in the window is calculated, and the central pixel value is compared with the median. If the difference between the central pixel value and the median is greater than the preset threshold, the central pixel value is replaced by the median. If the difference is less than or equal to the threshold, the central pixel value remains unchanged; finally, the processed low-frequency sub-band and high-frequency sub-band are reconstructed to obtain the denoised coal mine data signal. The reconstruction process is achieved by combining the approximate coefficients and detail coefficients at each scale.

[0034] After denoising the data, the next step is to consider data fusion technology. This invention adopts an improved weighted average fusion algorithm. First, a weight evaluation index system is established to set the historical accuracy of the data source, data update frequency, and equipment stability index; second, each index is quantified and assigned a corresponding weight according to its importance, and each data source is scored on each index; then, the weight is dynamically adjusted according to the real-time performance of the data source and environmental changes; then, for each data point, the fused data value is , where n is the number of data sources, is the data value of the ith data source, is the weight of the ith data source.

[0035] Considering that multiple features will affect the judgment of coal mine safety data, unreasonable features may even lead to misjudgment, thus posing a potential threat to coal mine safety. In actual situations, coal mine safety data often contains many features, which may come from different data sources and have different properties and characteristics. Unreasonable features will mislead information and make the data analysis results deviate from the actual situation. Therefore, the present invention uses an improved Pearson correlation coefficient to extract key features and judge the safety status of coal mines. First, feature extraction is performed, and a penalty coefficient is introduced into the traditional Pearson correlation coefficient. ,in represents the improved Pearson correlation coefficient, Represents the i-th observation value of two features respectively, Represent the means of the two features, n is the number of observations, is the weight coefficient of the penalty term, is a penalty function; then the correlation threshold is set according to the actual situation, and the features with correlation coefficients greater than the threshold are screened out; finally, the selected key features are used to construct a coal mine safety status model, and the support vector machine method is used to predict the coal mine safety status. According to the output results of the model, the safety status of the coal mine is judged, and the output result probability value greater than 0.7 corresponds to a high risk score, 0.3 to 0.7 corresponds to a medium risk score, and less than 0.3 corresponds to a low risk score, and an early warning is issued in time.

[0036] For a more specific description, three sets of test data are given below. First, after feature selection, the feature subsets retained by the present invention are gas concentration data, tunnel pressure data, dust concentration data, and equipment operation status data.

[0037] The first group, equipment status data, 1 represents normal, 0 represents abnormal: 1, gas concentration (%): 0.45, dust concentration (mg / m 3 ):45, tunnel pressure (MPa): 10. Group 2, equipment status data: 1, gas concentration (%): 0.58, dust concentration (mg / m 3 ):50, tunnel pressure (MPa): 11. Group 3, equipment status data: 0, gas concentration (%): 0.65, dust concentration (mg / m 3 ):55, tunnel pressure (MPa):10.5.

[0038] The support vector machine method is used to predict the safety status of coal mines. The prediction results are as follows: Figure 1 As shown in the figure, the horizontal axis represents different groups, and the vertical axis represents the size of the predicted value and the true value. It can be seen from the figure that there is a certain difference between the predicted value and the true value. For example, in the first group, the predicted value is 0.28, while the true value is 0.3; in the second group, the predicted value is 0.41, and the true value is 0.4; in the third group, the predicted value is 0.75, and the true value is 0.7. In general, the predicted value is close to the true value, but there is still a certain error. In practical applications, the parameters of the support vector machine model can be further optimized, or other methods can be combined to improve the accuracy of the prediction. In addition, more data can be analyzed and verified to evaluate the reliability and stability of the model.

[0039] The above description is only a preferred embodiment of the present invention and does not limit the present invention in other forms. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A coal mine safety data processing method based on multi-source data fusion, characterized in that: The following steps are involved: S1. First, it is necessary to collect various data that affect coal mine safety from coal mines; S2. Then, data preprocessing is performed, and a hybrid filtering algorithm combined with multi-scale analysis is used to denoise the collected data to improve the data quality; S3, then adopt the improved weighted average fusion algorithm to fuse the data from different data sources to obtain more comprehensive and accurate coal mine safety status information; S4. Finally, the fused data is analyzed and the improved Pearson correlation coefficient is used to extract key features to determine the safety status of the coal mine; The implementation process of the hybrid filtering denoising of the multi-scale analysis in step S2 is as follows: S21, first perform multi-scale decomposition on the input coal mine data signal Perform wavelet transform to obtain sub-bands of different scales. The wavelet transform coefficients are expressed as: low-frequency sub-band ,in represents the low-frequency signal component at scale j and position k, is a scaling function used to construct an approximate signal, namely, a low-frequency sub-band and a high-frequency sub-band ,in represents the high-frequency signal component at scale j and position k, It is a wavelet transform function, which is used to perform multi-scale analysis on the signal and extract detailed information of different frequencies; S22. Then, the low-frequency sub-band is subjected to preliminary denoising by using mean filtering. The signal after mean filtering is: ,in is the filter window size, Represents the signal value of the original low-frequency subband signal at position, Represents the center of the target position The signal values ​​of all the location points in the neighborhood are summed up; S23, then for the high frequency sub-band, use the improved median filtering algorithm, first calculate the median M of all pixel values ​​in the window, compare the central pixel value with the median, if the difference between the central pixel value and the median is greater than a preset threshold, replace the central pixel value with the median, if the difference is less than or equal to the threshold, keep the central pixel value unchanged; S24. Finally, the processed low-frequency sub-band and high-frequency sub-band are reconstructed to obtain the denoised coal mine data signal. The reconstruction process is achieved by combining the approximate coefficients and detail coefficients at each scale.

2. A method for processing coal mine safety data based on multi-source data fusion according to claim 1, characterized in that: The various data in step S1 include gas concentration data, wind speed data, tunnel pressure data, temperature data, dust concentration data, and equipment operation status data.

3. The method for processing coal mine safety data based on multi-source data fusion according to claim 1, characterized in that: The specific implementation of the improved weighted average fusion algorithm in step S3 is: S31. First, establish a weight evaluation index system, set the historical accuracy of the data source, data update frequency, and equipment stability index; S32. Secondly, quantify each indicator and assign corresponding weights according to its importance, and score each data source on each indicator; S33, then dynamically adjust the weight according to the real-time performance of the data source and environmental changes; S34, then for each data point, the fused data value is , where n is the number of data sources, is the data value of the ith data source, is the weight of the i-th data source; S35. Finally, in order to make the fused data smoother, the moving average method is used to take the average value of the fused data within a certain time window as the final fusion result.

4. The method for processing coal mine safety data based on multi-source data fusion according to claim 1, characterized in that: In step S4, the improved Pearson correlation coefficient is used to extract key features, and the implementation steps for judging the safety status of the coal mine are as follows: S41. First, perform feature extraction and introduce a penalty coefficient into the traditional Pearson correlation coefficient. ,in represents the improved Pearson correlation coefficient, Represents the i-th observation value of two features respectively, Represent the means of the two features, n is the number of observations, is the weight coefficient of the penalty term, is the penalty function; S42, then set a correlation threshold according to the actual situation, and filter out features with correlation coefficients greater than the threshold; S43. Finally, the selected key features are used to construct a coal mine safety status model, and the support vector machine method is used to predict the coal mine safety status. According to the output results of the model, the safety status of the coal mine is judged. The output result probability value greater than 0.7 corresponds to a high risk score, 0.3 to 0.7 corresponds to a medium risk score, and less than 0.3 corresponds to a low risk score, and an early warning is issued in time.

Images