Particle Swarm Optimization-based Support Vector Regression for Prediction of Electrical Discharge Machining Process Parameters
In the process of EDM,there is a non-linear relationship between discharge parameters and surface roughness,which complicates the task of determining appropriate electrical parameters for machining.Therefore,a prediction model of EDM process parameters based on a support vector regression optimized by the particle swarm algorithm is proposed.The research results show that the root mean square error(RMSE)of PSO-SVR on the test set is 0.302,and the coefficient of determination(R2)is 0.994,significantly better than the traditional SVR model(RMSE of 0.577,R2 of 0.981),verifying the effectiveness of the PSO algorithm in optimizing SVR parameters.The original data is preprocessed,and the PSO-SVR model is trained based on the optimized data.The results show that the RMSE of the data-preprocessed PSO-SVR model on the test set is further reduced to 0.255,and R2 is improved to 0.996,enhancing the prediction accuracy and generalization ability.
support vector regressionparticle swarm optimizationEDMprocess parameterssurface roughness
万方数据
维普
