Research Of Different-Source Gas Emission Prediction Based On Genetic Neural Network

The gas to coal mine enterprises' safe production has enormous influence,once produce gas explosion, the consequence is unimaginable, therefore, it is necessary to study precision method of gas emission quantity. The high precision of gas emission quantity is an important measure gas outburst prevention and treatment, for improving the coalmine safety situation and people's life safety has the significant practical significance. In the paper, the study is focused on the new techniques of the prediction known as the nonlinear neural network, with global optimization of genetic algorithms and different-source gas emission prediction.First the realistic significance and the ongoing study of the gas emission quantity prediction in china and foreign will be presented, this paper research from improving prediction precision of gas emission quantity, aiming at the gas emission is a complex, multivariable,dynamic system, introduced the neural network for non-linear mapping and genetic algorithm for the global search ability, at the same time, according to the characteristics of gas emission quantity, in order to awoid the interference of the irrelevant factors, proposed a new method of different-source gas emission prediction based on genetic neural network, neural network and genetic algorithm and combined method different-source prediction applied to prediction of gas. In this paper introduced method of genetic algorithm to optimize neural network, paper dissertates genetic algorithm to optimize neural network weights and threshold, established the different-source gas emission prediction model based on genetic neural network. On this basis, through the simulation of MATLAB software and the sample examples of gas emission, examples sample into test sample and examples, using testing samples to establish the model of gas emission prediction, and the prediction results and test sample were analyzed, the results show that the proposed method has higher precision of prediction, meet the production requirement.At the same time, in order to further illustrate the feasibility and the precision of prediction method with neural network, and the forecast method of gas outburst prediction of source points were compared, the result can be found, the prediction accuracy of prediction model of gas, the prediction method applied in the actual production of coal mining enterprises, the outburst prevention has certain directive Abstract significance.