Study Of Health Performance Supervision And Prediction For Airfield Lighting

Airfield lighting is an important visual equipment to protect the plane smooth taking-off,landing,taxiing at night or the bad weather conditions for airport.Timely and effective maintenance is the necessary premise for ensuring the normal running of the airfield lighting,and automatic monitoring and fault prediction technology are two important links of the airfield lighting health management.Using fault prediction technology can understand the potential failure in advance,and then workers can make maintenance strategy,therefore,the method not only can obtain more maintenance time for the workers,but also improve the support capability of the airfield lighting system.Using monitoring system real-time supervision can obtain more accurate fault information on airfield lighting,in addition,the basic experimental data for fault prediction can be obtained.In the paper,the approach area airfield lighting is used as the research object,and the health performance of the airfield lighting is studied from two aspects of single-lamp monitoring and fault prediction.A single-lamp monitoring system was designed and combination forecasting model was set up based on DGM(1,1)and RVM model to monitor and predict the airfield lighting.The supervisory system design was mainly carried on the hardware and the software.In order to detect short circuit,open circuit in both ends of the lamp and the light box into water,the status information acquisition circuit was designed.Then the hardware and software design of the WSNs each node was completed and the reliable transmission of the light status information was realized by using WSNs as the technical support,taking CC1101 as the wireless radio frequency communication module of the WSNs node,taking GPRS and SMS as the communication mode of between sink node of the WSNs and monitoring center.At the monitoring center,the monitoring platform design of the airfield lighting was completed,and then the functions of the lighting system inspection in the approach area and historical data inquiry were realized.Finally,Ordos International Airport was used as the test base of the system,and data transmission and the function of monitoring platform was tested.The result shows that the monitoring system basically achieves the design goal.For researching on the fault prediction of the airfield lighting,in order to be able to accurately and stably predict the lights fault,combination prediction model was built based on DGM(1,1)and RVM by improving the existing prediction models,and the kernel parameter in the combination forecasting model was optimized by DE algorithm.Simulation results show that the combination model has higher accuracy and stability for the fault prediction of the lamp,and so the method is feasible and effective compared with the single predictionmodels and other classical prediction models.At the end of this paper,the maintenance strategy was proposed by using the prediction results of the combination model.The calculation result verifies that condition-based maintenance strategy based on the prediction model can reduce maintenance cost rate of the airport airfield lighting system in theory under the premise to meet availability.