基于ARIMA和神经网络的单脉冲雷达设备故障预测研究

Research on technology of fault prediction for Monopulse Radar based on ARIMA and Neural Network

摘要:

单脉冲雷达设备目前采用的计划式维护模式过于依靠人力资源，也不能完全避免设备故障造成的巨大损失，已经越来越难以满足航天测控领域的实际应用需求。基于单脉冲雷达设备使用维护现状，本文研究了故障预测技术在单脉冲雷达设备维护的应用可行性，总结出了单脉冲雷达设备故障预测的基本流程与方法，并结合工程实际分别使用ARIMA时间序列分析和神经网络验证了单个参数劣化过程的预测与系统级劣化过程的预测，预测结果满足设备使用维护需求。

Abstract:

The equipment maintenance of monopulse radar is planed, which relies on manpower, and also cannot avoid the huge loss caused by equipment failure. It`s more and more difficult to supply the actual demand of TT&C system in spacecraft launching site. Based on the current situation about the maintenance of monopulse radar equipment in spacecraft launching site, the feasibility of fault prognosis technology applies in equipment maintenance of monopulse radar was researched, and the basic procedure and method of fault prognosis for monopulse radar was summarized. Combined with engineering practice, the degradation process of individual parameter was predicted by using ARIMA time series analysis, and the degradation process of system parameter was predicted by using Neural network analysis. The result was verified to Meet the demand of the monopulse radar equipment in space launch site.