论文部分内容阅读
将人工神经网络应用于飞机总体外形智能ＣＡＤ中，针对现有方法的局限性，研究了参数神经网络，提出了一种综合考虑影响神经网络学习３个主要因素（权值、激励函数和拓扑结构）的ＷＡＦＳ学习算法，并研究了隶属函数的神经网络表达和基于神经网络的并行推理，给出了有关应用实例。
In this paper, the artificial neural network (ANN) is applied to the general CAD of aircraft shape. In view of the limitation of the existing methods, the parametric neural network is studied and a comprehensive consideration is given to the three main factors affecting the neural network learning (weight, incentive function and topology ) WAFS learning algorithm, and studied the membership function of neural network expression and neural network-based parallel reasoning, given the application examples.