车载GNSS失锁下基于BPNN的微陀螺误差估计及定位方法研究

针对在全球导航卫星系统(global navigation satellite system, GNSS)失锁阶段将微陀螺惯性系统作为备用定位系统时,由于微陀螺误差引起的定位误差发散问题,提出了基于后向传播神经网络(back propagation neural network, BPNN)的微陀螺误差估计及定位算法。在GNSS有效阶段为车辆提供定位信息,同时对微陀螺误差进行估计,并利用后向传播神经网络BPNN建立微陀螺误差预测模型,为GNSS失锁阶段车辆定位做准备;在GNSS失锁阶段,利用已建立好的微陀螺误差预测模型估计微陀螺误差,对微陀螺输出信息进行补偿,以抑制由陀螺误差引起的定位误差。最后利用仿真与试验验证了此方法的正确性与有效性。

Research on error estimation of micro-gyroscope and location method based on BPNN for vehicle during GNSS outages

In the global navigation satellite system (GNSS) outages phase, the micro-gyroscope inertial system is used as a standby positioning system. In order to solve the problem of positioning divergence error caused by micro-gyroscope error, the error estimation of the micro-gyroscope and location method based on back propagation neural network (BPNN) is proposed in this paper. In the GNSS effective phase, it provides positioning information for vehicles and estimates the micro-gyroscope error. A prediction model of micro-gyroscope error that is prepared for vehicle positioning in the GNSS outages phase is established by the BPNN. In the GNSS outages phase, the micro-gyroscope error prediction model is used to estimate the micro-gyroscope error and compensate the output information of the micro-gyroscope to suppress the positioning error caused by the gyroscope error. Finally, the correctness and effectiveness of this method is verified by simulation and experiment.