Machine Tool Mechanics Correction Model Based on Gaussian Process Assisted Neural Network
Machine tool foundation has a serious impact on the dynamic characteristics of machine tool.Because the establish-ment of machine tool foundation system model needs accurate material property parameters,this paper proposes an adaptive neu-ral network model based on Gaussian process improvement,which takes advantage of the strong reasoning ability and fast calcu-lation efficiency of Gaussian process.The structure and training steps of neural network are improved by Gauissian process.Based on the improved neural network model,the former three natural frequencies are used as inputs,and the machine tool and founda-tion material properties are used as outputs to establish the mechanical correction model of the machine tool foundation system.The accuracy of the proposed model is verified by comparing the experimental and simulation values,which provides a theoretical basis for the design of machine tools.
Machine ToolFoundationGaussian ProcessNeural NetworkCorrection Model
万方数据
维普
