基于Hopfield神经网络的腐蚀失效模式识别

Local Corrossion Failure Mode Recognization Based on Hopfield Neural Network

Hopfield神经网络是具有记忆功能的反馈型神经网络 ,但存储记忆的能力受到神经元数量的限制。本文根据Hopfield神经网络的特点 ,采用多模块分级识别的方法 ,研究开发适用于腐蚀失效模式识别的学习、推理机模型。在此基础上 ,运用面向对象的编程技术及数据库技术 ,实现能够学习。

To develop the reasoning and learning model for identification of corrosion failure mode, the prototype of Hopfield neural network was utilized. The identification capacity of Hopfield neural network is limited by the number of neural cells in the network. A grading method was introduced to overcome the above limitation. With the procedure of corrosion failure analysis and the utilization of object oriented programming(OOP)object oriented programming and database technology, an intelligent expert system for corrosion failure identification was developed and presented, which is capable of learning from knowledge inputted.