摘要
为了解决WIFI指纹定位技术受到环境干扰大以及PDR(Pedestrian Dead Reckoning)导航定位技术存在累积误差的问题,提出基于BP神经网络的WIFI辅助IMU(Inertial Measurement Unit)室内联合定位系统。模型先设定一个校正周期,再使用基于BP神经网络的WIFI指纹定位算法周期性纠正改进PDR导航定位算法的定位结果,并更新PDR导航定位算法的初始位置坐标,进而削弱因长时间定位而产生的累积误差。通过仿真结果分析,可以看出定位精度有了明显提高,证明了本文所提方案的有效性。
In order to solve the problem that WIFI fingerprint positioning technology suffering from environmental interference and the cumulative error of PDR(Pedestrian Dead Reckoning) navigation and positioning technology, the paper proposes a WIFI-assisted IMU(Inertial Measurement Unit) indoor joint positioning system based on BP neural network. The model first sets a correction period, and then uses the WIFI fingerprint location algorithm based on BP neural network to periodically correct the positioning result of the improved PDR navigation and positioning algorithm, and update the initial position coordinates of the PDR navigation and positioning algorithm, thereby weakening the cumulative error due to long-term positioning. The simulation experiment and the result analysis show that the positioning accuracy has been significantly improved, which proves the effectiveness of the proposed scheme.
作者
李彪
袁国良
朱若琪
谢奎
LI Biao;YUAN Guo-liang;ZHU Ruo-qi;XIE Kui(College of Information Engineering,Shanghai Maritime University,Shanghai 201306,China)
出处
《计算机仿真》
北大核心
2021年第7期442-446,共5页
Computer Simulation
