CN-121993247-A - Multi-target intelligent scheduling gas extraction safety monitoring method

CN121993247ACN 121993247 ACN121993247 ACN 121993247ACN-121993247-A

Abstract

The invention discloses a multi-target intelligent dispatching gas extraction safety monitoring method, which relates to the technical field of mine safety production and comprises the steps of collecting real-time monitoring data of a gas extraction process, carrying out unified time stamp alignment to generate a multi-target gas extraction real-time monitoring data set, carrying out multi-target optimization pretreatment on the multi-target gas extraction real-time monitoring data set to generate a gas disaster risk real-time evaluation index, a gas extraction purity and total amount real-time index and a gas extraction energy consumption real-time index, carrying out safety bottom line penetrating screening on a pareto optimal solution set, carrying out deduction through a counterfactual to obtain a robust feasible solution pair with a statement, carrying out preferential confirmation on the robust feasible solution pair with the statement to generate a gas extraction dispatching instruction set, and carrying out execution to obtain a gas extraction safety monitoring report. The invention achieves the closed-loop management effect of traceability, dialectical and sustainable optimization of the safety decision.

Inventors

LI QINGSONG
YU HUAJIN
ZHANG YONG
WEI YANJUN
ZHANG WEI
JIANG XINGXING
ZHENG LULIN
ZHANG SHUJIN
ZHANG PENG
SHEN ZHENHUA
ZHANG YIPING

Assignees

贵州省矿山安全科学研究院有限公司
贵州大学

Dates

Publication Date: 20260508
Application Date: 20260127

Claims (10)

1. A multi-target intelligent dispatching gas extraction safety monitoring method is characterized by comprising the steps of, Collecting real-time monitoring data in the gas extraction process, and performing unified time stamp alignment to generate a gas extraction multi-target real-time monitoring data set; Performing multi-objective optimization pretreatment on the multi-objective real-time monitoring data set for gas extraction to generate real-time assessment indexes of gas disaster dangerousness, real-time indexes of gas extraction purity and total amount and real-time indexes of extraction energy consumption; constructing a multi-objective real-time optimization problem mathematical model according to the real-time evaluation index of the gas disaster risk, the real-time index of the gas extraction purity and total amount and the real-time index of the extraction energy consumption, inputting the multi-objective real-time optimization problem mathematical model into a multi-objective evolutionary algorithm solver for optimization calculation, and outputting a pareto optimal solution set; performing safe bottom line penetrating screening on the pareto optimal solution set, and obtaining a robust feasible solution pair with a statement through inverse fact deduction; And preferentially confirming the robust feasible solution pair with the statement, generating a gas extraction scheduling instruction set, and acquiring a gas extraction safety monitoring report after executing.
2. The method for monitoring gas extraction safety of multi-target intelligent scheduling according to claim 1, wherein the real-time monitoring data comprises continuous time series data of gas concentration, continuous time series data of mixed flow, continuous time series data of negative pressure value, continuous time series data of temperature, continuous time series data of operation current of the extraction pump and continuous time series data of operation voltage of the extraction pump.
3. The method for monitoring gas extraction safety by multi-target intelligent scheduling according to claim 2, wherein the method for generating the multi-target real-time monitoring data set for gas extraction comprises the following steps of, Inversion calculation is carried out on the continuous time sequence data of the gas concentration, the continuous time sequence data of the mixed flow and the continuous time sequence data of the negative pressure value, and a dynamic inversion result of the coal seam air permeability coefficient is generated; Performing on-line estimation of the drilling influence range on the mixed flow continuous time series data and the negative pressure value continuous time series data to obtain an on-line estimation value of the effective influence radius of the drilling; And carrying out unified time stamp alignment processing on the real-time monitoring data of the gas extraction process, the dynamic inversion result of the coal seam permeability coefficient and the on-line estimated value of the effective radius of the drilled hole, and generating a gas extraction multi-target real-time monitoring data set.
4. The gas extraction safety monitoring method for multi-objective intelligent scheduling according to claim 3, wherein the method comprises the steps of generating a gas disaster risk real-time evaluation index, a gas extraction purity and total amount real-time index and an extraction energy consumption real-time index, Extracting a first-order difference maximum value of gas concentration, a standard deviation of a negative pressure value and a mixed flow attenuation slope from a gas extraction multi-target real-time monitoring data set; Inputting the first-order difference maximum value of the gas concentration, the standard deviation of the negative pressure value and the mixed flow attenuation slope into a mutation risk discrimination model, carrying out nonlinear mapping calculation, and outputting a real-time evaluation index of the gas disaster risk; according to the continuous time series data of the gas concentration in the multi-target real-time monitoring data set of gas extraction and the continuous time series data of the mixed flow, carrying out time integration to obtain real-time indexes of the purity and the total quantity of the gas extraction; based on continuous time series data of the operating current of the extraction pump and continuous time series data of the operating voltage of the extraction pump in the gas extraction multi-target real-time monitoring data set, the input electric power of the extraction pump is calculated, and time integration is carried out to obtain real-time indexes of the extraction energy consumption.
5. The gas extraction safety monitoring method for multi-objective intelligent scheduling according to claim 4, wherein the construction of the multi-objective real-time optimization problem mathematical model comprises the following steps of, Setting a real-time evaluation index of the risk of the gas disaster as a first optimization target and adopting a minimizing direction, setting a real-time index of the gas extraction purity and the total amount as a second optimization target and adopting a maximizing direction, setting a real-time index of the extraction energy consumption as a third optimization target and adopting a minimizing direction, and generating an objective function; defining a drilling negative pressure set value, a drilling start-stop state combination and a pipeline valve opening combination as decision variables according to the control capability of on-site extraction equipment and a gas extraction safe operation rule; and combining the objective function with the decision variable to generate the multi-objective real-time optimization problem mathematical model.
6. The gas extraction safety monitoring method of the multi-target intelligent scheduling according to claim 5, wherein the input multi-target evolutionary algorithm solver performs optimization calculation and outputs a pareto optimal solution set, specifically comprising the following steps of, Inputting the multi-target real-time optimization problem mathematical model into a multi-target evolutionary algorithm solver, and carrying out standard multi-target evolutionary iteration to obtain the current pareto front; And (3) performing super-volume index calculation on the current pareto front edge, detecting the variation of the super-volume index, and stopping calculation if the variation is smaller than the super-volume convergence threshold value to generate the pareto optimal solution set.
7. The method for monitoring gas extraction safety of multi-target intelligent scheduling according to claim 6, wherein the safety bottom line penetration screening is performed on the pareto optimal solution set, specifically comprising the following steps of, Calculating the maximum amplitude of negative pressure fluctuation and the maximum gradient of gas concentration burst based on the continuous time series data of the negative pressure value and the continuous time series data of the gas concentration; Constructing a disturbance intensity template according to the maximum amplitude of the negative pressure fluctuation and the maximum gradient of the gas concentration sudden increase; Applying a disturbance intensity template to a drilling negative pressure set value, a drilling start-stop state combination and a pipeline valve opening combination which correspond to the pareto optimal solution set, and obtaining a post-disturbance decision scheme; Based on the post-disturbance decision scheme, calling a mutation risk discrimination model to recalculate a gas disaster risk real-time evaluation index, and acquiring the post-disturbance gas disaster risk real-time evaluation index; and (3) incorporating the pareto optimal solution with the disturbed gas disaster risk real-time evaluation index lower than the safety threshold value into a robust feasible subset.
8. The method for monitoring gas extraction safety of multi-target intelligent scheduling according to claim 7, wherein the method for obtaining the robust feasible solution with the statement through the counterfactual deduction comprises the following steps of, Selecting a solution with the lowest gas disaster risk real-time evaluation index value from the robust feasible subset as a safety reference solution; calculating the difference value of the solutions except the safety reference solution in the robust feasible subset and the safety reference solution on the real-time indexes of the gas extraction purity and the total amount, and the difference value of the extraction energy consumption on the real-time indexes; and combining the gas extraction purity and total real-time index difference value and extraction energy consumption real-time index difference value into a liability pre-declaration text through inverse facts logic, and binding the liability pre-declaration text with a robust feasible subset to obtain a robust feasible solution pair with declaration.
9. The method for monitoring gas extraction safety of multi-target intelligent scheduling according to claim 8, wherein the method comprises the steps of preferentially confirming the robust feasible solution pair with the statement to generate a gas extraction scheduling instruction set, Screening the optimal solution from the robust feasible solution pairs with declarations according to the preference rule; And carrying out instruction frame format packaging on the drilling negative pressure set value, the drilling start-stop state combination and the pipeline valve opening combination in the optimal solution to obtain a gas extraction scheduling instruction set.
10. The method for monitoring gas extraction safety of multi-target intelligent scheduling according to claim 9, wherein the method comprises the following steps of, Updating a drilling negative pressure set value, a drilling start-stop state combination and a pipeline valve opening combination according to a gas extraction scheduling instruction set to acquire a parameter updating completion signal; and combining the parameter updating completion signal, the gas extraction scheduling instruction set and the responsibility pre-declaration text into a structured record to generate a gas extraction safety monitoring report.

Description

Multi-target intelligent scheduling gas extraction safety monitoring method Technical Field The invention relates to the technical field of mine safety production, in particular to a multi-target intelligent dispatching gas extraction safety monitoring method. Background The multi-target intelligent dispatching gas extraction safety monitoring technology occupies a vital position in the field of mine safety production and green low-carbon development at present. Along with the improvement of the utilization requirement of coal bed gas resources, gas extraction is gradually expanded from a single safety guarantee target to a multidimensional cooperative process with disaster prevention and control, energy recovery and energy efficiency optimization. The current technology system generally adopts a multi-source sensing network to collect dynamic parameters of gas concentration, negative pressure and flow in real time, and builds the pareto optimal solution set by means of multi-target evolutionary algorithm (such as NSGA-II and MOEA/D), so as to provide decision support for safety and benefit balance for a manager, and embody the deep fusion of intelligent optimization and engineering practice. In the field of multi-objective intelligent dispatching gas extraction safety monitoring, although the traditional safety monitoring method can output a pareto front solution set, the engineering robustness of the pareto front solution set depends on ideal working condition assumptions, coupling influence of field disturbance on risk evaluation is not fully considered, so that a manager is difficult to ensure the safety and reliability of a high-benefit solution, meanwhile, the pareto front solution set only presents objective numerical value differences, and inverse fact balance explanation and compliance evidence facing management decisions are lacking, so that intelligence is balanced on a calculation level and is difficult to convert into traceable and responsible dispatching decision behaviors. Disclosure of Invention The present invention has been made in view of the above-described problems occurring in the prior art. Therefore, the invention provides a multi-target intelligent dispatching gas extraction safety monitoring method which solves the problems of insufficient robustness and poor management decision interpretation of the pareto solution set engineering. In order to solve the technical problems, the invention provides the following technical scheme: The invention provides a multi-target intelligent dispatching gas extraction safety monitoring method which comprises the following steps of, Collecting real-time monitoring data in the gas extraction process, and performing unified time stamp alignment to generate a gas extraction multi-target real-time monitoring data set; Performing multi-objective optimization pretreatment on the multi-objective real-time monitoring data set for gas extraction to generate real-time assessment indexes of gas disaster dangerousness, real-time indexes of gas extraction purity and total amount and real-time indexes of extraction energy consumption; constructing a multi-objective real-time optimization problem mathematical model according to the real-time evaluation index of the gas disaster risk, the real-time index of the gas extraction purity and total amount and the real-time index of the extraction energy consumption, inputting the multi-objective real-time optimization problem mathematical model into a multi-objective evolutionary algorithm solver for optimization calculation, and outputting a pareto optimal solution set; performing safe bottom line penetrating screening on the pareto optimal solution set, and obtaining a robust feasible solution pair with a statement through inverse fact deduction; And preferentially confirming the robust feasible solution pair with the statement, generating a gas extraction scheduling instruction set, and acquiring a gas extraction safety monitoring report after executing. As a preferable scheme of the multi-target intelligent scheduling gas extraction safety monitoring method, the real-time monitoring data comprise gas concentration continuous time sequence data, mixed flow continuous time sequence data, negative pressure value continuous time sequence data, temperature continuous time sequence data, extraction pump running current continuous time sequence data and extraction pump running voltage continuous time sequence data. As a preferable scheme of the multi-target intelligent scheduling gas extraction safety monitoring method, the invention generates a multi-target real-time monitoring data set of gas extraction, and comprises the following specific steps of, Inversion calculation is carried out on the continuous time sequence data of the gas concentration, the continuous time sequence data of the mixed flow and the continuous time sequence data of the negative pressure value, and a dynamic inversion result of the coal seam air