摘要
论文综合利用有限元方法、正交试验法、人工神经网络以及遗传算法对龙门起重机结构系统进行优化研究。首先利用有限元模型对结构进行灵敏度分析,确定对结构系统特性敏感的设计变量作为神经网络的输入变量。然后利用正交试验法确定神经网络训练样本,并利用有限元模型计算出样本数据,建立人工神经网络模型。最后利用遗传算法对所建立的神经网络模型寻优。
The paper uses finite element method, orthogonal experiment method, neural network and genetic algorithm to optimize the structure of gantry crane. First of all, it uses finite element model to analyze the structural sensitivity and determines the design variables of the structural sensitivity as the input variables of neural network. Then it uses orthogonal experiment method to determine the training sample and uses the finite element model to calculate the sample data and to build the neural network model. Finally, it uses the neural network model built by genetic algorithm to optimize the structure.
出处
《起重运输机械》
2009年第7期52-55,共4页
Hoisting and Conveying Machinery
关键词
龙门起重机
有限元法
正交试验法
BP神经网络
遗传算法
gantry crane
finite element method
orthogonal experiment method
BP neural network
genetic algorithm
