Research On Risk Assessment Model And CO Prediction Of Mine Poisoning And Suffocation Accidents Based On Multiple Algorithm

This thesis is based on the sub-project "Research on the safety monitoring and early warning technology of production process and disaster evolution" of the Yunnan Provincial Key Research and Development Programme,"Research on the key technology and application demonstration of hazardous air management in underground engineering in highland areas"(Project No.: 202003AC100002).202003AC100002).In order to effectively prevent the occurrence of poisoning and asphyxiation accidents in underground mines,this paper establishes an evaluation index system for the factors affecting poisoning and asphyxiation accidents in underground mines,constructs a combined assignment model using game theory to solve the problem of low accuracy of assignment results caused by a single algorithm,learns that CO poisoning is the root cause of poisoning and asphyxiation accidents through analysis of accident causes,so the CO concentration after blasting is predicted,explores the accuracy of different The accuracy of time series prediction methods is explored to project the required ventilation time after blasting and to ensure the safe operation of personnel.The main research of this paper is as follows:(1)Constructing an index system for assessing factors influencing poisoning and asphyxiation accidents.Using the hierarchical holographic modeling(HHM)method and the risk filtering,rating and management(RFRM)method,we analyzed and filtered the risk factors influencing the occurrence of poisoning and asphyxiation accidents,combined the accident investigation reports and previous research results,started from the four aspects of "human-machine,environment and management",introduced social factors,and divided the risk factors into five categories.The risk factors were divided into five categories,and the index system was constructed to evaluate the influence factors of poisoning and asphyxiation accidents,and the indicators were assigned based on the index system.(2)Constructing a combined assignment model.Based on the constructed evaluation index system,the subjective sequential relationship analysis method(G1method)and the objective rough set theory,entropy weight method and CRITIC method are selected,and game theory is used to combine the subjective and objective assignment methods in two combinations to construct the combination assignment model,and the best combination assignment model is selected by comparing the weight ranking results of different combination assignment models and comparing the weight ranking results of accident investigation reports and consulting experts.(3)Evaluate the risk level of the mine.The combination of the preferred combined assignment model and cloud model is used to construct an assignme nt-rating model to evaluate the risk level of the mine,and a comprehensive ev-aluation cloud diagram is drawn using Matlab software to clearly and intuitive ly determine the risk level of the mine.The traditional "human-machine-environment-management" assessment index system is also used to compare the asses-sment index system constructed after the introduction of social factors,which p-rovides more accurate results.(4)Prediction of CO concentration after blasting.The GM(1,1)model,exponential smoothing method and ARIMA model were used to predict the CO concentration after blasting,and then the posterior difference index and three error evaluation indexes(SMAPE,MSE,MAE)of the three prediction methods were calculated to compare the accuracy of the results of the three prediction methods,from which the method with the best prediction effect was selected,and the method was used to make subsequent predictions to obtain the time for the CO concentration after blasting in the digging The time for CO concentration to drop to the allowable value after blasting at the working face,so that reasonable and effective ventilation can be carried out.