机场助航灯光回路绝缘电阻预测方法研究  

Research of airfield lighting circuit insulation resistance prediction method

作　　者：王修岩[1] 李萃芳[1] 李宗帅[1] 林家泉[1] 

机构地区：[1]中国民航大学航空自动化学院,天津300300

出　　处：《计算机工程与设计》2013年第8期2882-2886,共5页Computer Engineering and Design

基　　金：中央高校基本科研业务费中国民航大学专项基金项目(ZXH2010C002)

摘　　要：为了更好地对助航灯光回路绝缘电阻值进行预测,采用了一种新型的灰色神经网络预测模型。建立灰色补偿RBF网络模型,并将遗传算法优化后的支持向量机引入到RBF神经网络中,得到改进后的灰色补偿RBF网络模型。根据上海浦东国际机场绝缘电阻的实际测量值进行实例仿真,仿真结果表明,改进后的灰色神经网络提高了系统稳定性和绝缘电阻值的预测精度。Airfield lighting circuit is an important part of airfield lighting system,in order to better predict airfield lighting circuit insulation resistance value,a new gray neural network model was adopted.Firstly,the gray compensating RBF network model is established,and the support vector machine optimized by genetic algorithm is introduced into the RBF network,the improved gray compensating RBF network model is obtained,then the resistance value is simulated and forecasted based on the actually measured value of the insulation resistance at Shanghai Pudong International Airport.The results show that the improved gray neural network improves the system stability and prediction accuracy of the insulation resistance value.

关 键 词：绝缘电阻 GM(2 1)模型 灰色神经网络 支持向量机 遗传算法 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]