The Evaluation Of Rockburst Tendency In Deep Of Sanshandao Gold Mine

The phenomenon of rockburst threatens the safety of underground construction personnel and equipment in the process of deep mining of hard rock mines.Evaluating the tendency of rockburst in metal mines has become one of the research direction of deep mineral development.In order to effectively evaluate the rockburst tendency level in the deep part of the mine,this paper attempts to apply five machine learning algorithms to evaluate the rockburst tendency level in the deep part of the mine on the basis of collecting the data of mine rockburst cases at home and abroad,and analyze the evaluation results.The main contents and achievements are as follows:(1)Firstly,a large number of literature and rock burst databases were retrieved,from which the relevant data of rock burst cases at home and abroad were obtained,and on this basis,8 index data evaluating the propensity of rock bursts were obtained,and the relevant data were preprocessed,and finally 245 complete rock burst case data were obtained.(2)Secondly,Training the evaluation model.The first method,bring the collected rockburst case data into machine learning algorithms such as decision trees,random forests,gradient boosting trees,extreme random trees,and adaptive boosting trees,build rockburst propensity evaluation models,and test the model’s performance.Accuracy.At the same time,Try the second method,a rockburst propensity evaluation model based on principal component analysis was built to test the accuracy of the model.The rockburst propensity evaluation model has high accuracy.Among them,the adaptive boost tree algorithm built by applying the second method has an accuracy rate of 75%.(3)Then,combined with the actual situation of the underground metal mine,by collecting the mine in-situ stress data information,obtaining the mine rock samples on the spot,and performing rock mechanics experiments on the mine rock samples,the evaluation index of the rockburst tendency of the mine to be tested is finally obtained.(4)Finally,several rockburst tendency models are used to evaluate the rockburst tendency of the underground metal mine,and then the results are comprehensively analyzed by voting method,and finally the rockburst tendency grade of the underground metal mine area to be tested is obtained.The results show that there is the hign possibility of medium rockburst in several deep mine areas to be measured.