| Coal and gas outburst with characteristics of sudden occurrence,rapid process occurred,the severity of consequences,the mechanism of complexity and so on,are great dangers and destructive in coal mine production safety.In recent years,there has a strong growth trend with the continuous extension of the deep mine,mining conditions of the increasing complexity,greatly enhance the degree of mechanization of production,the strength of coal and gas outburst occurred and the casualties were caused.More accurately predict coal and gas outburst,the letter has become an important issue to be addressed.Therefore,in this paper,coal and gas outburst prediction warning system conducted in-depth research,and prediction model of GSO-ELM Network Control combined static and dynamic indicators is proposed in coal and gas control system.Coal and gas outburst control system has characteristics of complex control object,non-linear,model is difficult to establish and so on.General control strategy is not able to meet the requirements of coal and gas outburst control system and can not achieve the appropriate quality control.Therefore,the article adopts the generalized predictive control strategy of ELM neural network.For a number of factors in coal and gas outburst,use principal component analysis method to extract main factors from the static factors of coal and gas outburst.For the problem of slowly convergence and not strong generalization ability,ELM neural network forecasting model is optimizated by glowworm swarm optimization.This method not only improves the recognition speed and accuracy of the model,but also reduces online computation of predictive control,and improved control system's real-timeFinally,combined with GSO-ELM model predictive control,build and simulate coal and gas predictive control system.MATLAB simulation results show that the improved method to achieve the goal of early warning purposes,good flexibility,high accuracy in coal and gas outburst predictive control. |
PDF Full Text Request