论文部分内容阅读
为了得到构架式空间可展开天线结构优化中目标函数与设计变量的解析表达式,基于BP神经网络建立一种天线结构优化参数的预测模型.根据天线背架的结构及神经网络的训练原理,构建对优化参数进行预测的网络模型;应用有限元软件ANSYS对优化参数进行数值计算,通过正交试验设计,得到BP神经网络的训练样本;调整传递函数、隐层节点数及训练算法,建立满足误差要求的优化参数的预测模型,利用检验样本对预测模型进行泛化能力检验.结果表明:网络预测值与有限元计算结果吻合较好,整体预测误差≤10%;并且模型运行时间短,仅需0.13 s.该模型能够较准确地预测结构的优化参数,为结构的优化设计提供了理论参考.
In order to obtain analytic expressions of objective function and design variables in structural space optimization of antenna structure, a prediction model of antenna structure optimization parameters is established based on BP neural network.According to the structure of antenna back frame and the training principle of neural network, Optimize the parameters of the network model; the application of finite element software ANSYS numerical optimization of the parameters, through orthogonal experimental design, the BP neural network training samples; adjust the transfer function, the number of hidden nodes and training algorithm to establish to meet the error The results show that the predicted value of the network is in good agreement with the finite element calculation, the overall prediction error is less than or equal to 10%, and the running time of the model is short, only 0.13 s. The model can predict the optimal parameters of structures more accurately and provide a theoretical reference for the optimal design of structures.