CN-122022475-A - Coal mine safety monitoring method and monitoring system

CN122022475ACN 122022475 ACN122022475 ACN 122022475ACN-122022475-A

Abstract

The invention relates to the technical field of coal mine safety, in particular to a coal mine safety monitoring method and a monitoring system, which solve the technical problem that a complex direct causal relationship and indirect association path between different parameters cannot be mined by a risk monitoring method in the prior art, so that the accuracy of risk prediction is reduced. According to the coal mine safety monitoring method provided by the invention, the evolution mechanism of the safety risk can be revealed essentially through the association strength, the time lag coefficient and the space influence range among the association parameter pairs, the intrinsic driving mechanism of the risk evolution can be identified, whether the monitoring parameter is abnormal or not can be monitored on the basis of the rule cognition level, and the accuracy of judging whether the abnormality is improved. In addition, when an abnormal risk occurs, the risk evolution trend of the associated parameter affected by the abnormal parameter can be determined by the predicted time of occurrence of the abnormality in the future of the associated parameter associated with the parameter.

Inventors

TU HUI
ZHANG BEINING
ZHANG YAJUN
SUN XIAOHU
ZHENG XU
WANG JIE
WANG YAOHUI
WANG ZHENG

Assignees

华能煤炭技术研究有限公司

Dates

Publication Date: 20260512
Application Date: 20260129

Claims (10)

1. The coal mine safety monitoring method is characterized by comprising the following steps of: Preprocessing initial historical time sequence parameters monitored by a plurality of sensors in a historical time period before the current sampling time to obtain a plurality of parameter packets, wherein each parameter packet represents the historical time sequence parameters monitored by the plurality of sensors with the same position coordinates in the historical time period, and the historical time sequence parameters comprise historical parameter values and historical monitoring time stamps corresponding to the historical parameter values; determining a risk early warning time starting point and effective time sequence data of parameters corresponding to each sensor and an association rule set among the parameters corresponding to the sensors according to the parameter packets, wherein the association rule set comprises association parameter pairs, association strength among the association parameters in the association parameter pairs, time lag coefficients among the association parameters and a space influence range; Calculating association influence coefficients between association parameters in the association parameter pairs according to the association rule set; Calculating a dynamic early warning threshold corresponding to the associated parameter of each parameter according to the abnormal threshold corresponding to each parameter and the associated influence coefficient between the associated parameters in the associated parameter pair comprising the parameter; Determining whether the parameter is abnormal according to the current parameter value of the parameter and the abnormal threshold value of the parameter to obtain a preliminary judgment result, wherein the preliminary judgment result comprises that the parameter is abnormal, the parameter is at abnormal risk and the parameter is normal; when the parameters are abnormal or abnormal risk exists, determining the predicted time of the abnormality of the association parameters according to the association rule set and the dynamic early warning threshold corresponding to the association parameters.
2. The coal mine safety monitoring method according to claim 1, wherein when the parameter is abnormal or is at risk of abnormality, determining the predicted time for abnormality of the associated parameter according to the association rule set and the dynamic early warning threshold corresponding to the associated parameter comprises: inquiring the association parameters associated with the parameters and association influence coefficients between the parameters and the association parameters in the association rule set when the parameters are at abnormal risk or abnormal; Calculating the risk evolution rate of the associated parameter according to the current parameter value of the parameter, the abnormal threshold value of the parameter and the associated influence coefficient between the parameter and the associated parameter; calculating the prediction time corresponding to the associated parameter according to the risk evolution rate of the associated parameter, the current parameter value of the associated parameter and the dynamic early warning threshold corresponding to the associated parameter, wherein the prediction time indicates that the associated parameter reaches the corresponding dynamic early warning threshold after the prediction time.
3. The coal mine safety monitoring method according to claim 1, wherein determining whether there is an abnormality in the parameter according to a current parameter value of the parameter and an abnormality threshold of the parameter to obtain a preliminary determination result includes: Determining that an abnormality exists in the parameter when the current parameter value of the parameter is greater than or equal to the abnormality threshold; when the current parameter value of the parameter is smaller than the abnormal threshold value and is larger than or equal to a preset threshold value, determining that the parameter has abnormal risk, wherein the preset threshold value is obtained by calculating according to a judgment coefficient and a parameter mean value of the effective time sequence data corresponding to the parameter; And when the current parameter value of the parameter is smaller than the preset threshold value, determining that the parameter is normal.
4. The method for coal mine safety monitoring according to claim 1, wherein determining a risk early warning time starting point and valid time series data of parameters corresponding to each sensor according to a plurality of parameter packets, and a set of association rules between the plurality of parameters corresponding to the plurality of sensors, comprises: constructing a multi-source data matrix corresponding to each position coordinate according to a plurality of parameter packets, wherein row vectors of the multi-source data matrix are time nodes, and column vectors of the multi-source data matrix are parameter values corresponding to parameter types; Inputting a plurality of multi-source data matrixes into a pre-constructed causal association analysis model for association calculation to obtain an association rule set, wherein the association rule set comprises association parameter pairs, association strength among association parameters in the association parameter pairs, time lag coefficients among the association parameters and a space influence range; The time series data corresponding to each parameter is extracted from the multi-source data matrix, and the parameter change rate distribution corresponding to each parameter is calculated according to the time series data corresponding to each parameter, wherein the parameter change rate distribution comprises parameter change rates corresponding to a plurality of continuous data; and determining a risk early warning time starting point and effective time sequence data corresponding to each parameter according to the parameter change rate distribution corresponding to each parameter and the time sequence data corresponding to each parameter.
5. The coal mine safety monitoring method of claim 4, further comprising: and calculating an abnormal threshold value corresponding to each parameter type according to the risk early warning time starting point and the effective time sequence data corresponding to each parameter type.
6. The coal mine safety monitoring method according to claim 5, wherein calculating the abnormality threshold value corresponding to each parameter type according to the risk early warning time starting point and the effective time series data corresponding to each parameter type comprises: Extracting data in a preset time period before the starting point of the risk early-warning time from the effective time sequence data corresponding to the parameters to serve as a target risk early-warning data set; and calculating an abnormal threshold corresponding to the parameter based on the 3 sigma criterion and the target risk early warning data set.
7. The method of claim 5, wherein determining the risk early warning time start point and the effective time series data corresponding to each parameter according to the parameter change rate distribution corresponding to each parameter and the time series data corresponding to each parameter, comprises: extracting a reference change rate with a parameter change rate greater than a preset change rate threshold value from parameter change rate distribution corresponding to the parameter, and determining sampling time corresponding to the reference change rate as reference sampling time; Sequencing a plurality of reference sampling times according to time sequence to obtain a reference sampling time sequence; calculating the time interval between two adjacent reference sampling times in the reference sampling time sequence; Determining a front-end reference sampling time corresponding to a time interval greater than a preset time interval threshold as a target time interval; Extracting data in each target time period from the time sequence data to obtain a plurality of target parameter sets; calculating the number of parameter values in the target parameter sets, determining the target parameter sets with the number of parameter values smaller than the preset number as target deletion parameter sets, and deleting the target deletion parameter sets in the time sequence data to obtain effective time sequence data corresponding to the parameters; Determining a target time period corresponding to a target parameter group with the number of parameter values being greater than or equal to the preset number as a target estimated time period; and determining the minimum sampling time in the target estimated time periods as a risk early-warning time starting point corresponding to the parameters.
8. The method of claim 5, wherein constructing a multi-source data matrix corresponding to each position coordinate according to a plurality of parameter packets, comprises: extracting initial historical time sequence parameters of the sensor corresponding to each position coordinate from a plurality of parameter packets; Taking a preset time interval as a time node, performing difference processing on initial historical time sequence parameters of the sensor corresponding to each position coordinate to obtain effective historical time sequence parameters of the sensor corresponding to each position coordinate; Constructing a multi-source data matrix corresponding to the position coordinates according to the effective historical time sequence parameters of the sensor corresponding to the position coordinates, wherein row vectors of the multi-source data matrix are time nodes, and column vectors of the multi-source data matrix are parameter values corresponding to a plurality of parameters.
9. The coal mine safety monitoring method according to claim 1, wherein after determining the predicted time for occurrence of abnormality of the association parameter according to the association rule set and the dynamic early warning threshold corresponding to the association parameter, the coal mine safety monitoring method further comprises: When the preliminary judgment result is that the parameters are abnormal or the parameters are at abnormal risk, generating risk early warning distribution at different positions in the building space according to first abnormal information corresponding to the abnormal parameters, second abnormal information corresponding to the abnormal risk parameters and a correlation rule set among a plurality of parameters; And mapping the position information and the risk early warning information according to the three-dimensional structure diagram of the building space and the risk early warning distribution, and displaying the risk early warning information on the three-dimensional structure diagram according to the mapping result so as to visualize the early warning space.
10. A coal mine safety monitoring system, comprising: The data preprocessing unit is used for preprocessing initial historical time sequence parameters monitored by the plurality of sensors in a historical time period before the current sampling time to obtain a plurality of parameter packets, wherein each parameter packet represents the historical time sequence parameters monitored by the plurality of sensors with the same position coordinates in the historical time period, and the historical time sequence parameters comprise historical parameter values and historical monitoring time stamps corresponding to the historical parameter values; The association rule mining unit is used for determining risk early warning time starting points and effective time sequence data of parameters corresponding to each sensor according to a plurality of parameter packets, and an association rule set among the plurality of parameters corresponding to the plurality of sensors, wherein the association rule set comprises association parameter pairs, association strength among the association parameters in the association parameter pairs, time lag coefficients among the association parameters and a space influence range; The early warning threshold value determining unit is used for calculating the association influence coefficient between the association parameters in the association parameter pair according to the association rule set, and calculating the dynamic early warning threshold value corresponding to the association parameter of each parameter according to the abnormal threshold value corresponding to each parameter and the association influence coefficient between the association parameters in the association parameter pair comprising the parameter; The risk early warning unit is used for determining whether the parameter is abnormal according to the current parameter value monitored by the sensor and the abnormal threshold value of the parameter corresponding to the sensor so as to obtain a preliminary judgment result, wherein the preliminary judgment result comprises that the parameter is abnormal, the parameter is at abnormal risk and the parameter is normal, and when the parameter is abnormal or at abnormal risk, the prediction time of the abnormality of the associated parameter is determined according to the association rule set and the dynamic early warning threshold value corresponding to the associated parameter.

Description

Coal mine safety monitoring method and monitoring system Technical Field The invention relates to the technical field of coal mine safety, in particular to a coal mine safety monitoring method and a monitoring system. Background The prevention and control of various disaster accidents are core tasks for guaranteeing continuous development of production activities and protecting life safety of operators in the field of coal mine safety production. Along with the continuous increase of coal exploitation depth and continuous expansion of exploitation range, underground geological conditions of the coal mine are more complex, the occurrence probability of safety risks such as gas accumulation, ultra-limit of ground pressure stress, microseismic disasters, ventilation system failure and the like is obviously improved, and various risks often show the characteristics of multifactor coupling and chained evolution. To reduce the probability of risk occurrence, monitoring systems are employed to predict risk occurrence, e.g A gas monitoring system, an ore pressure monitoring system, a microseismic monitoring system, a ventilation monitoring system and the like. While the system can collect key monitoring data such as gas concentration, ground pressure stress, microseismic event parameters, wind speed and wind quantity in real time and determine whether risk occurs according to the data collected in real time, the system determines whether risk occurs according to a single parameter type. The evolution of coal mine safety risk is often the result of multi-parameter interaction, for example, the mutation of the ground pressure stress of a certain area may cause the increase of the frequency of microseismic events, thereby causing abnormal gas concentration gushing. Therefore, the risk monitoring method in the prior art cannot mine complex direct causal relationship and indirect association paths among different parameters, so that the accuracy of risk prediction is reduced. Disclosure of Invention The invention aims to provide a coal mine safety monitoring method and a monitoring system, which solve the technical problems that the risk monitoring method in the prior art cannot mine complex direct causal relationship and indirect association paths among different parameters, and further reduce the accuracy of risk prediction. As a first aspect of the present invention, there is provided a coal mine safety monitoring method comprising: Preprocessing initial historical time sequence parameters monitored by a plurality of sensors in a historical time period before the current sampling time to obtain a plurality of parameter packets, wherein each parameter packet represents the historical time sequence parameters monitored by the plurality of sensors with the same position coordinates in the historical time period, and the historical time sequence parameters comprise historical parameter values and historical monitoring time stamps corresponding to the historical parameter values; determining a risk early warning time starting point and effective time sequence data of parameters corresponding to each sensor and an association rule set among the parameters corresponding to the sensors according to the parameter packets, wherein the association rule set comprises association parameter pairs, association strength among the association parameters in the association parameter pairs, time lag coefficients among the association parameters and a space influence range; Calculating association influence coefficients between association parameters in the association parameter pairs according to the association rule set; Calculating a dynamic early warning threshold corresponding to the associated parameter of each parameter according to the abnormal threshold corresponding to each parameter and the associated influence coefficient between the associated parameters in the associated parameter pair comprising the parameter; Determining whether the parameter is abnormal according to the current parameter value of the parameter and the abnormal threshold value of the parameter to obtain a preliminary judgment result, wherein the preliminary judgment result comprises that the parameter is abnormal, the parameter is at abnormal risk and the parameter is normal; when the parameters are abnormal or abnormal risk exists, determining the predicted time of the abnormality of the association parameters according to the association rule set and the dynamic early warning threshold corresponding to the association parameters. In an embodiment of the present invention, when the parameter is abnormal or there is an abnormal risk, determining, according to the association rule set and a dynamic early warning threshold corresponding to the association parameter, a predicted time for the abnormality of the association parameter includes: inquiring the association parameters associated with the parameters and association influence coefficients between th