机械系统智能化设计的研究——用神经网络实现系统特性的预测

Research of Intelligent Design of Mechanical Structure System——Predict of Characters of Mechanical System via ANN

本文从系统设计学的角度，探讨采用神经网络的方法在机械系统设计过程中，实现对机械系统特性的预测。其目的在于在机械系统设计的方案设计阶段，相对于新产品的初步设计，设计系统基于现有产品的信息，简单、而且具有继承性地为系统提供具有反馈性的系统特性表达方法。这一研究对建立高效、短周期、智能化设计系统过程是有益的。文中以雷达的结构系统固有频率预测为例，提出运用神经网络对机械结构系统特性进行预测的方法。本文的主旨在于推介在机械系统设计的方案设计过程中，应用神经网络方法进行系统特性预测的方法。

The paper deals with predicting the characters of mechanical system by means of neural networks during the design of mechanical system from the point of system design.Its purpose is to provide the initial design of system characters with the inheritance from the existing products in the course of initial design of new products.The research is valuable for constructing an efficient,fast and intelligent design system.In the paper the method of predicting the inherent frequency of structure system of radar by means of neural networks is presented,which can be extend to the field of design of mechanical system.