Research On Information Fusion Technology Of Steer-by-wire(SBW) System Based On Multi-sensor

The steer-by-wire(SBW)system is a new research hotspot of the vehicle steering system,which realizes real-time steering of the steer wheel through by the electronic control system.Combined with the increasingly developed tracking navigation function,it can greatly improve the handling stability of the vehicle and resistance fatigue of driver,which can effectively avoid traffic accidents.More important,the vehicle active steering control is based on real-time and accurate acquisition of the vehicle driving state parameters.However,the existing vehicle state parameters are mostly derived from the information such as wheel speed and lateral acceleration that collected by the onboard sensors,and the inherent calibration error,drift error,the single information and other limitations have been increasingly unable to meet people’s needs,and it is difficult to directly measure those parameters such as the lateral slip angle of the mass and the road surface adhesion coefficient during in the driving process.Therefore,how to use variety and small number of sensors to obtain accurate vehicle driving status becomes the focus of the steer-by-wire system research.The technology of information fusion can organically synthesize local information,and we can use those information fusion can organically synthesize local information,and we can use those information to reduce uncertainty.In this paper,we propose a multi-sensor information fusion algorithm to improve the practicality of the SBW of vehicle.Firstly,aiming at the subject requirements,we analyze the structure of the steer-by-wire system,and set up the steering wheel and motor,gear and rack components models to build the SIMULINK model of steer-by-wire system,and embed it into Carsim to establish the vehicle model with SBW.Secondly,based on determined the information source,we establish the multi-sensor network of the SBW system,and design the plan of the network,and then select the sensors according to the plan,and then design the hardware interface circuits of the angle sensor,attitude sensor(main control board)and vehicle speed sensor(sub-control board).Determine the software development tools and design the overall program for software,we design the software of the above sensors acquisition module,after these work,we establish the CAN bus communication between the main control board and the sub-control board to realize the data collection function of each sensor,and combine the debugging through the serial port,use MATLAB to complete the hardware online testing.Finally,Kalman filter and extended Kalman filter algorithm are designed to improve the real-time performance,accuracy and stability of the steer-by-wire system control parameters.Build the simple linear tire model and three degrees of freedom nonlinear vehicle model which covers lateral,longitudinal and lateral motions,and establish the multi-sensor fusion information algorithm based on extended Kalman filter.The inputs ofthe algorithm including the longitudinal acceleration,lateral acceleration and thesteering wheel angle(were acquired with the attitude sensor,is collected by therotation angle sensor),which can accurately estimate the yaw rate,the slip angle,and the longitudinal speed when the vehicle is traveling.In order to verify the validityand accuracy of the algorithm,the simulation experiment was carried out with rapid double line change test.In order to verify the universality of the established multi-sensor information fusion algorithm,establish the simulation which used by double lane change test and sine wave steer input test,the outputs of the Carsim model with steer-by-wire system is used as the inputs of simulation,the measured values including,.Carsim and MATLAB/Simulink co-simulation tests were performed in two working conditions to compare and analyze the yaw rate and the vehicle slip angle.Experiments show that the multi-sensor information fusion algorithm studied can accurately estimate the driving state parameters of vehicle,and it is suitable for the vehicle with steer-by-wire system,which provides the theoretical basis and design ideas for the follow-up road surface adhesion coefficient estimation and control strategy research.