自主移动农用机器人导航系统故障诊断分析

Fault Diagnosis Analysis of Navigation System in Automobile Agricultural Robot

通过对自主移动农用机器人运行过程数据序列的统计辨识,逐步累积故障识别能力,建立有效的故障诊断体系。运用时间序列分析法对正常和故障情况下导航系统中定位部件GPS输出随机信号进行分析,得到判断GPS性能故障的方法。建立GPS随机漂移数学模型AR(p)模型,并采用K-L(Kullback-Leibler)信息测度检测法构造相应的信息距离差别函数进行系统状态检测。利用该时序分析法能够有效地检测出系统故障。

Through statistical identification to the data series obtained in the course of unit running of automobile agricultural robot,as well as the gradual accumulation of the ability to identify the fault,an effective fault diagnosis system was set up.By means of time series analysis method,the random drift signal output by orientation GPS in navigation system was analyzed under normal and faulty conditions,and a fault identification method for GPS performance was obtained.By establishing mathematic AR(p) model and using K-L(Kullback-Leibler) measuring method based on information range difference function,the system status was tested.The method of time series analysis could effectively diagnose the faults in the system.