CAJ | 학술논문

背景:塑料注塑成型技术可用于康复工程的多种实用矫形器中,遗传算法所表现出的易于实现及健壮性等特点,使它在许多领域,特别是近年来在机器学习、模式识别、智能控制和最优化等领域得到了广泛应用.目的:建立基于混合神经网络与遗传算法方法的注塑工艺参数优化系统.设计:开发性研究.单位:南昌工程学院机电工程系及南昌铁路中心医院干部病房.方法:用Matlab语言编制了应用程序,对神经网络的参数预测与遗传算法的优化过程进行求解.将网络预测结果与CAE(computer aided engineering)模拟结果进行比较和误差分析,显示出反向传播网络的稳定性和可靠性.结果:优化结果经CAE模拟和实验验证,证明是正确的,结论:基于混合神经网络与遗传算法方法的注塑工艺参数优化方法是可行的.该优化系统具有工程实用价值.
배경:소료주소성형기술가용우강복공정적다충실용교형기중,유전산법소표현출적역우실현급건장성등특점,사타재허다영역,특별시근년래재궤기학습、모식식별、지능공제화최우화등영역득도료엄범응용.목적:건립기우혼합신경망락여유전산법방법적주소공예삼수우화계통.설계:개발성연구.단위:남창공정학원궤전공정계급남창철로중심의원간부병방.방법:용Matlab어언편제료응용정서,대신경망락적삼수예측여유전산법적우화과정진행구해.장망락예측결과여CAE(computer aided engineering)모의결과진행비교화오차분석,현시출반향전파망락적은정성화가고성.결과:우화결과경CAE모의화실험험증,증명시정학적,결론:기우혼합신경망락여유전산법방법적주소공예삼수우화방법시가행적.해우화계통구유공정실용개치.
BACKGROUND: Injection molding can be used in producing a variety of orthotic devices of practical use in rehabilitation medicine. With its merits of convenient realization and stability, genetic algorism(GA) has found wide application in many fields, especially in machine learning, pattern recognition, intelligent control and optimization.OBJECTIVE: To establish an optimization system for injection molding parameters based on hybrid neural network and genetic algorithm.DESIGN: An open trial.SETTING: Department of Mechanical and Electrical Engineering, Nanchang Institute of Technology; Ward for Cadres, Nanchang Railway Central Hospital.METHODS: The application program was developed in Matlab language and used in the parameter prediction with the neural network and genetic algorithm optimization. The comparison and error analysis were made between the neural network-predicted results and those generated by simulation with computer aided engineering (CAE), which showed that the back propagation network was stable and reliable.RESULTS: The results of optimization, after verified by CAE simulation and tested by experiment, are proved to be correct.CONCLUSION: It is feasible to optimize the injection molding parameters based on hybrid neural network and genetic algorithm approach. The optimization system is of value in practical engineering applications.