Study On Control Strategy Of SBW Based System Front Steering And It’s Fault Tolerance Techlogy

Steer by wire (SBW) uses electronic method to transmit signal instead of traditionalmachinery. It cancels the mechanical connection between the steering wheel and thefront steering wheel so that the steering ratio can be designed freely to get a bettersteering performance by appropriate matching. SBW can generate an extra steeringangle besides the driver gives to promote the response or the stability under urgency.These advantages make SBW the necessary part of electric vehicle, hydrogen vehicle orintelligent car. So SBW is becoming one of the researching hot points. This dissertationbuilt a SBW vehicle model based on the structure and theory of SBW; designed a varingsteering ratio to solve the contradiction between ‘light’ and ‘swift’; designed AFS byLQR to improve the stability of vehicle; built a fault tolerance system based on theextended kalman filter and the residual theory. The main work of this dissertationincludes the following aspects:①Introduced the basic strcture and theory of SBW. Built a magic tire systembased on which a vehicle model was built which would be used to simulate.This modelignored the influence of the suspension.②Explained the shortcomings of constand steering ratio and varing mechanicalsteering ratio to introduce the concept of ideal steering ratio and then designed an idealsteering ratio which can keep the yawrate gain a constant. To solve the limitation of theideal steering ratio, this dissertation count in the influence of speed and handwheelangle, designed a varing steering ratio used in SBW by fuzzy theory.③Built a2-DOF vehicle model based on which a LQR controller is designed.Finished two condition simulation to confirm the effect of the LQR controller,compared the vehicle without controlling, vehicle with LQR and the reference.④Built the extended kalman filter state update equation and the time updateequation according to the vehicle model. Made one of front steering angel, sideaccelerationa yand yawrate as measuring value to get a estimation value of theother two. Then compare the system state and the estimation value to get a residual,which can be used to detect the sensor fault. Finally simulated the fault of front steeringangle sensor, side acceleration sensor and yawrate sensor, and got the fault vector whichmatched the fault detection logic and used the estimation value to instead the faultsensor.