Research On Prediction And Control Of Gas Emission Quantity With Sub Sources Based On PSOBP-adaboost

Gas is a great challenge to the safety production of coal mine enterprise.When gas disasters happen,the safety of people's lives and property will suffer a huge threat.The size of the gas emission quantity is the basis for the design of the mine ventilation designing,the design of the gas drainage engineering,and the work of the gas preventing and controlling.So the high accurate prediction of gas emission quantity is the important measure to improve the gas prevention and control,and also improve the safety production situation of mining enterprises and people's life safety has important practical significance.Therefore,it is necessary to study the prediction method of gas emission quantity improving the prediction accuracy of gas emission quantity.In this paper,on the basis of studying some classical forecasting methods and some hot spot prediction methods,a new method for predicting gas emission quantity was discussed,in which BP neural network with the characteristics of nonlinear mapping,PSO optimization algorithm,adaboost iterative algorithm and gas emission prediction method with different sources were combined,and a gas emission prediction model with different sources based on PSOBP-adaboost algorithm was established.On the basis of in-depth understanding the principle of BP neural network,PSO optimization algorithm and adaboost algorithm,the PSOBP-adaboost algorithm was proposed,and the application of this algorithm to the predicting identification of nonlinear systems.By using the MATLAB to carry out the simulation experiment,according to a classical data set of UCI database,the corresponding nonlinear forecasting system is established,and the accuracy of the model based on the PSOBP-adaboost algorithm was tested.On the basis of summarizing the classical methods of gas prediction,the theory of the method of the sub source prediction was introduced,and three gas emission sources,such as the coal seam,the adjacent coal seam and the goaf,were selected.In view of the characteristics and influence factors of the three gas emission sources,the relevant factors were determined,and the main components were extracted from the influence factors by the method of principal component analysis.According to the theory of PSOBP-adaboost algorithm,a gas emission prediction model based on PSOBP-adaboost algorithm was established.On this basis,by using MATLAB simulation software and a coal mine gas emission actual samples,the gas emission was predicted with the established model.The results show that the method has high prediction accuracy and it can meet the requirements of production.In order to further explain the feasibility and high precision of the method,the results are compared with the results of the sub source prediction model based on the BP neural network.By the error analysis,it shows that the prediction accuracy of the prediction model based on the PSOBP-adaboost algorithm is better than that of the BP artificial neural network model.It is proved that it has good prediction performance and generalization performance to predict gas emission quantity by using the PSOBP-adaboost algorithm,and it also proves the rationality and superiority of the theory.