Inhalt des Dokuments
Steer-by-Wire (SbW) is a highly prospective steering technology for intelligent vehicles. As the elimination of the mechanical connection between the steering wheel and the steered wheels, it becomes free to adjust steering characteristics for an SbW system, which contributes to improving steering sensitivity, steering stability, and handling performance. On the other hand, how to design the variable steering angle ratio and how to generate a proper artificial steering feedback torque come to be vital and challenging issues.
As artificial neural network (ANN) is capable of learning complex non-linear correlations without requiring specific mathematical models, it can be adopted to estimate the desired steering feedback torque in SbW system. As for the design of variable steering angle ratio, the multi-objective optimization can be used for the steering performance improvement. Since we already have a well-performed EPS steering testbench and an BMW-F01 SbW test vehicle available in our Department Automotive Engineering, it is likely to develop proper function and application for SbW system, which provides the driver with a realistic steering feel, the same as in an electrical power steering (EPS) system.
- Optimization of variable steering angular ratio under different vehicle velocity and hand wheel angle.
- Collection of Steering datas by imposing various steering maneuvers in IPG Carmaker to the vehicle simulator coupled with the well-performed EPS steering testbench.
- Training and validation of the designed ANN.
- Implementation and verification of the designed function and application in the BMW-F01 SbW test vehicle.