关于机场助航灯光系统故障预测研究  

Research on Airfield Lighting System with Fault Prediction

作　　者：王丙元[1] 田坤[1] 张丹丹[1] WANG Bing - yuan TIAN Kun ZHANG Dan - dan(Electronic Information and Automation College, Civil Aviation University of China, Tianjin 300300, Chin)

机构地区：[1]中国民航大学电子信息与自动化学院,天津300300

出　　处：《计算机仿真》2017年第4期453-459,共7页Computer Simulation

基　　金：国家自然科学基金(U1333102)

摘　　要：为了保障助航灯在发生故障情况下不影响机场的正常运营,提出一种改进预测模型对助航灯故障进行预测。以单灯作为研究目标,以灯具两端的电压值作为原始数据,建立了将相关向量机(RVM)模型融入于离散灰色模型(DGM)模型的联合预测模型(DGM-RVM),提高了数据预测的精确性与稳定性。仿真结果表明:与单模型预测方法相比,上述模型对助航灯故障预测具有更高的精确度与稳定性,方法的有效性。将预测模型应用于维修策略制定中,采用实际进近区灯光系统作为分析案例,验证了根据上述模型制定的视情维修策略,在满足可用度的前提下,能够极大的降低机场助航灯维修费率。In order to ensure the airfield lights don＇t effect airport security operation in case of a failure, this paper presents an improved prediction model for airfield lighting system failure forecasting. Taking single lamp as a research goal and the voltage cross the lamp as initial data, we established a combined prediction model（ DGM -RYM） based on relevance vector machine（RVM） with discrete grey model（ DGM）, then the accuracy and stability of data forecasting was improved. Simulation results show that, compared with the single model prediction method, the model has higher accuracy and stability in the fault prediction of airfield lighting system, and is feasible and effective. Meanwhile, the prediction method was applied in maintenance strategy formulation. Using the actual approach area lighting system as an analysis case, the calculation result verifies that, on the premise of meeting availability, condition - based maintenance strategy based on the model can greatly reduce maintenance cost rate of the airport lighting system.

关 键 词：助航灯光系统 灰色模型 相关向量机 视情维修 寿命预测 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]