GPS信号易受建筑物或树木遮蔽,而低成本的MEMS惯性元件随机漂移大,性能随温度急剧变化,采用一般的Kalman滤波器融合GPS和低成本MEMS惯性元件信号,构成GPS/DR组合导航系统很难满足导航系统的容错和精度要求。针对GPS和MEMS惯性元件构成的低成本GPS/DR组合导航系统,设计了容错UKF/KF联邦Kalman滤波算法,提高了组合导航系统的定位精度和抗干扰能力;针对MEMS惯性元件随机漂移大的缺点,采用零偏试探消减算法,抑制了MEMS惯性元件的随机零漂,提高了MEMS惯性元件的精度。仿真结果表明,基于该算法的GPS/DR组合导航系统的定位精度高,抗干扰能力强,在GPS信号中断的情况下导航系统仍可在短时间内保持较高的定位精度。
GPS signal is especially susceptible to being obstructed by dense trees or high buildings.Meanwhile the inertial devices of micro electro mechanical system(MEMS) have large random drifts,and its performance changes sharply with temperature.So the low cost GPS/DR system using ordinary Kalman filter(KF) to fuse GPS and MEMS inertial devices can hardly meet the accuracy and fault tolerant requirements.To solve these problems,a fault tolerant UKF(Unscented Kalman filter)/KF federated Kalman filter algorithm is designed,and the random drift of the MEMS inertial devices is restrained by heuristic drift reduction(HDR) method.The simulation results show that the accuracy and fault tolerant ability of the GPS/DR system are both improved,and the system can remain high positioning accuracy within relatively long time when GPS signal is lost.