用遗传算法改进的BP模型在刹车系统诊断中的应用研究

利用神经网络进行故障诊断是近些年来所兴起的一种重要的智能诊断方法。我们就是根据航空维修的特点及神经网络的优点, 提出了应用BP网络对飞行器的故障进行学习、训练和诊断的想法。同时, 为了克服单独应用BP算法时存在的缺陷, 我们利用遗传算法(GA)对其进行了改进。我们利用面向对象的计算机编程技术, 开发了一个软件程序, 初步试验, 取得良好效果。最后讨论了在航空维修方面智能诊断的未来发展趋势。

A Study of BP Neural Network Model Based on GA in Fault Diagnosis of Wheel Brake System

Fault diagnosis using neural network is an important intelligence fault diagnosis technique that emerged in recent years. In view of features of aircraft maintenance and advantages of neural networks, a concept of using the BP network to learn, train and diagnose aircraft faults is proposed. In order to overcome defects resulting from separate use of the neural network, a genetic algorithm is utilized to improve the BP algorithm. A BP genetic algorithm software is developed using the Object-Oriented Programming (OOP) technique and the priliminary test is successfully conducted. Finally, Future trends in intelligence diagnosis for aircraft maintenance are discussed.