Prediction Of Coal And Gas Outburst Based On Improved Fuzzy Support Vector Machine

Coal and gas outburst is a geology disaster that occurs during coal mining and is extremely destructive.If accurate predictions can be made in time,the corresponding protective measures can be taken before the disaster to ensure the safety of the workers underground.At present,the most widely used coal and gas outburst prediction is the support vector machine algorithm,with strong generalization ability.However,after testing the application,there are still some shortcomings in the algorithm: 1.The anti-dryness is not strong,and it is easy to be misjudged by the wrong sample in the prediction process;2.It is affected by the parameter,and the blind selection parameter will affect The classification effect of the algorithm.In terms of the selection of the influencing factors,coal and gas outburst are affected by many factors and there is a high degree of nonlinear relationship between them.In this paper,the gray correlation degree theory is combined with the mine's outstanding measured data to calculate from 8 influencing factors.In the above prominent prediction method,aiming at the shortcomings of the above two support vector machines this paper proposes a new particle swarm optimization fuzzy support vector machine for coal and gas outburst prediction.Firstly,the fuzzy membership degree of each sample is calculated by the fuzzy membership function.this method can be used to reduce the influence of noise points on the classification results.Secondly,the particle swarm optimization algorithm is used to optimize the parameters of the fuzzy support vector machine.Although it is simpler and easier to operate than other commonly used parameter optimization algorithms such as genetic algorithm and least squares method,the particle swarm algorithm still has the potential to fall into local optimum.The shortcomings of this paper are improved in this paper.Firstly,the particle swarm optimization algorithm is improved in the particle swarm optimization algorithm.In the particle swarm optimization algorithm,the optimal ability of the inertia weighting algorithm with nonlinear reduction of the number of iterations is introduced.Secondly,the simulated annealing algorithm is used to make the particle swarm.The particles in the algorithm forcibly jump out of the local optimal trap with a certain probability.The improved particle swarm optimization algorithm greatly improves the optimization efficiency and overcomes the blindness of parameter selection in the traditional classification model.Finally,a fuzzy support vector machine prediction model based on particle swarm optimization is constructed.The model firstly assigns the corresponding membership degree to the measured data,and reduces the influence of the error samples on the model prediction ability.Then the particle swarm optimization algorithm is used to find the optimal parameters and the parameters.The impact on the predictive model is minimized.In this paper,the particle swarm optimization support vector machine model,the fuzzy support vector machine model,the BP neural network model and the BP neural network model are compared with the performance of the model.The experimental results show that the particle swarm optimization fuzzy support vector machine model has fast training speed and the highest classification accuracy.The improved fuzzy support vector machine(SVM)prediction model proposed in this paper is simulated by using MATLAB software combined with the measured data of the mine.The output results show that the algorithm has the fastest training speed compared with other traditional prediction methods,and can be compared in a shorter time.Highlights make more accurate judgments.The method effectively solves the problems of poor anti-dryness,slow training speed and low prediction accuracy in the traditional prediction method,and has strong practicability.The paper has 22 pictures,11 tables,and 86 references.