Research On Vehicle Location Method Based On Multi-sensor Fusion

At this stage,the mainstream vehicle positioning methods are mainly GPS positioning and GPS/INS combined positioning,etc.However,the ordinary GPS positioning error is relatively large,and the high-precision GPS is expensive.For combined positioning such as GPS/INS,it is limited by GPS output frequency,when the vehicle speed is too fast,the improvement of the vehicle position accuracy by the combined positioning is limited.In view of this situation,this paper designs a vehicle positioning embedded system to improve the positioning accuracy of ordinary GPS,so as to realize the precise positioning of the vehicle.The system uses the extended Kalman filter algorithm to fuse the data of the three sensors of the vehicle OBD interface communication module,the MPU6050 module,and the ordinary GPS module and uses the constant rotation rate and speed model to design the prediction equation of the extended Kalman filter algorithm to predict the next step of the vehicle to improve the positioning accuracy of the GPS module.In order to overcome the problem of accumulated errors in the MPU6050 module,this paper first uses the MPU6050 module and the GPS module for the first fusion during the fusion process and then selects the MPU6050 module and the vehicle OBD interface communication module for secondary fusion outside the working cycle of the GPS module.This method can not only reduce the accumulated error of the MPU6050 module but also further improve the positioning accuracy of the vehicle.The work done in this paper is divided into three parts:(1)According to the driving environment of the vehicle on the road,two sensors,the MPU6050 module,and the vehicle OBD interface communication module are selected to improve the positioning accuracy of the GPS module,and the hardware circuit design of the three modules and the main control chip is carried out,and the corresponding principle is drawn that figure and PCB,hardware debugging after soldering is completed.(2)The data collected by the three sensors are processed accordingly,and then the prediction equation of the extended Kalman filter algorithm is designed using the constant rotation rate and speed model and the data of the MPU6050 module,and the update equation is designed using the data of the GPS module and the vehicle OBD interface communication module.(3)Based on the RT-Thread operating system,the software part of the vehicle positioning embedded system is written,and the completed vehicle positioning embedded system is mounted on the real vehicle for experiments.Referring to the positioning results of the Beidou probe,the vehicle positioning embedded system is tested for GPS The degree of improvement of the module positioning accuracy.The vehicle positioning embedded system designed in this paper has low hardware costs and greatly improves the positioning accuracy of the GPS module.The experimental results show that when the vehicle speed is 39km/h,in the east and north directions,the system can improve the positioning accuracy of the GPS module.are 29% and 30% respectively;when the vehicle speed is 70km/h,in the east and north directions,the system improves the positioning accuracy of the GPS module by 25% and 20% respectively.