A Hybrid GA-BP Algorithm for the Predictive Model of Multi-Parameters of Liquid-Solid Extrusion Process
A hybrid GA-BP algorithm was proposed aiming to the deficiency of BP network in this paper. By use of the global search ability of genetic algorithm (GA) and the local search ability of artificial neural network (ANN), the both algorithm were organically integrated together to accelerate the convergence speed and the convergence precision. Based on the hybrid GA-BP algorithm, the nonlinear mapping relation between design variables and objective function was established to control deformation uniformity of composite in the forming process of liquid-solid extrusion and reduce inner damaging defects of products. The simulation results of FEM named virtual samples were selected as the networks training samples. By training the sample, the knowledge base of the muti-parameters for the liquid-solid extrusion process was set up. The influences of main parameters, such as infiltration time, die preheating temperature, pouring temperature, working table length, platform width and cone angle of guide plane on the deformation uniformity, had been studied using the predictive function of the model. They supply good instructions for the design and optimization of the liquid-solid extruding composites process.
liquid-solid extrusion genetic algorithm neural network deformation uniformity
Lizheng Su Lehua Qi Jiming Zhou Zhenjun Wang Fang Yang
School of Mechatronic Engineering, Northwestern Polytechnical University, Xi’an, 710072 China the School of Mechatronic Engineering, Northwestern Polytechnical University, Xi’an, 710072 China
国际会议
武汉
英文
2007-09-21（万方平台首次上网日期，不代表论文的发表时间）