Research On Prediction Of Bottom Hole Flowing Pressure For Coal Bed Methane Well Based On SVM

There is a huge demand for energy because of the fast development of industrialization in our country.The demand of terminal energy gradually transforms from traditional energy to quality,efficient and clean energy.Coalbed methane,a new environmental energy,occupies an extremely important position in China's energy structure.However,although there is rich coalbed methane resources in China,more than half of the coalbed methane is low production wells because of the high rank the capacity constraints caused by poor management.so we are badly in need of a new type of coalbed methane mining technology and equipment to improve productivity,and achieve reasonable coalbed methane management and the improvement of production capacity.The main factors influence the effect of coalbed methane well,including discontinued mining factors,bottom hole flowing pressure control factors,mining intensity,etc.This paper is discussed the basic principle and the technological process of single well gas extraction system in the coalbed methane,Based on the above,this paper makes a deep analysis of the bottom hole flowing pressure prediction methods in China's coalbed methane,where has special resources condition and gas well engineering condition,and guarantee continuous,smooth,step by step to reduce the bottom hole flowing pressure,control of mining strength,improve the success rate of development well production release,increase the operating life of coalbed methane well,and fundamentally changes our country's mining technology,which is relied heavily on the individual experience.On account of the bottom hole flowing pressure in coalbed methane single well system changes quite gentle,this paper studies the bottom hole flowing pressure of coalbed methane well based on support vector machine forecasting model,which is solved the parameters optimization problems of support vector machine(SVM).Aiming at the influence of the parameters of SVM to predict performance,this paper introduces several kinds of frequently optimization methods,it contains of cross validation method,grid search method,genetic algorithm and particle swarm optimization.Through contrast,this paper finds the best method of the optimization parameters by the sample data in coalbed methane system,and displayed clearly in the form of chart through the visual programming.On the basis,this paper researches the method of fuzzy information granulation,and establish the model of fuzzy granulation for obtaining the range of the approximate goal combined information granulation method and support vector machine(SVM)method.The model adopts the best parameter optimization algorithm to optimize fuzzy graining model parameters,and uses the error to evaluate precision of the model.The simulation results show that the method has good practicability,can effectively predict the tendency of the bottom hole flowing pressure,and has the good effect of forecast and analysis.