遗传小波神经网络仿真技术在复合管铸造中的应用  

Application of the Genetic Wavelet Neural Networks on the Solidification Process of Composite Pipe

作　　者：张俊[1] 张海燕[1] 

机构地区：[1]襄樊学院机械系,湖北襄樊441053

出　　处：《铸造技术》2008年第3期400-403,共4页Foundry Technology

摘　　要：提出了一种基于遗传算法学习的小波神经网络(GAWNN),它继承了小波分析良好的局部性及其神经网络的学习和推广能力,又具有遗传算法全局性优化搜索的特点,是多层前向神经网络学习的一种理想算法。将该算法应用于双金属复合管铸造过程的数值仿真,用热电偶对铸造温度场进行实测,并以实测数据为样本,仿真双金属复合管充型、凝固过程的温度分布。与实测数据比较表明,仿真数值的最大相对误差为1.3%,为双金属复合管的设计和工艺制订提供了理论依据。A new type of wavelet neural networks of genetic algorithm （GAWNN） is proposed, which not only has the local property of the wavelet analysis and the generalization capability of the artificial neural networks, but has the advantages of the fast global searching of the genetic algorithm as well. It is an ideal algorithm for the training of the multi-layer feeding forward neural networks. The GAWNN is used to simulate the casting process of composite pipe, the solidifying temperature field is measured by thermocouples, the temperature distribution during filling and solidification is simulated based on those measured data. By contrasting the data of simulation with those of testing, the maximum relative error for simulation is 1. 3% in comparison with the measured data. The investigation can provide the reference for the design of bi-metal composite pipe.

关 键 词：遗传算法 小波神经网络 双金属复合管 数值仿真 

分 类 号：TG249[金属学及工艺—铸造]