Based On LS-SVR Model Coal And Gas Prominent Forecast System Research

This paper analyzes the coal and gas outburst mechanism and influencing the factors, in thebase of using the coal mine gas supervisory system in gas density data message foundation,applies the support vector machines method in the coal and the gas prominent predict,It hasprovided a new way for the solution damp prominent forecast question.This article mainly carries on the following work: First of all, the coal and gas prominentmonitor gas density sequence’s time-domain characteristics of the vector, the frequencycharacteristics of the vector and the wavelet domain feature vector, carries on the featureextraction and the choice analysis. Proposed the method of based on the difference evolutionalgorithm feature selection. Afterward, selects the appropriate nuclear function, constructs thecoal and the gas prominent forecast smallest two rides the support vector return machine(LS-SVR) model, Similarly carries on the nuclear parameters optimization using the differenceevolution algorithm, enhances the parameter to search for the element speed and theenhancement forecast model performance. Through quite forecast with other forecast algorithm,The result indicated: Smallest two rode the support vector return machine model recognizer tosolve the small sample, the misalignment, the high dimension, the partial minimum and so onactual problem. At last, designs the coal and gas prominent forecast system, realizescontinuously the non-contact coal and the gas forecast prominently, The forecasting result andactual tallies has the significant practical significance.