New Vehicle Functionality Using Electric Propulsion

Abstract

More stringent emission and fuel efficiency requirements for cars are leading to increased electrification of drivetrains. There arises a need to be able to justify the significantly increased costs associated with the electrification. One way to do this is to achieve new or improved functions using the new electric actuators available at our disposal. Significant number of publications are available that study making use of electrification to do basic functions such as propulsion, regenerative braking and in some cases, direct yaw control. This project aims to go beyond such basic functions, identify new functions and develop one of them with high potential. First, brainstorming was done to identify new functions. These functions were subjected to detailed analysis and rating to select the one with the highest potential. The vehicle configurations (location of motors, engine and the wheels/axles they drive) were also analysed to identify the best among them. The selection process yielded \Customise feel" as the best function. Next, the function was developed to tune the transient yaw dynamics of the vehicle using an empirical feed forward controller and the steering torque feedback using a model based controller. The steering torque controller was rudimentary in design and was a proof of concept to show that effective control of steering torque can be done. The yaw response controller on the other hand was a more comprehensive controller. The controller parameters were determined through optimisation by Matlab's Genetic Algorithm. The controller optimisation was done for step steer manoeuvres at different speeds. The tuned controller was then validated for different manoeuvres and scenarios. The results of simulations run with these controllers on a comprehensive 9 degree of freedom vehicle model are presented and it is seen that a yaw response time reduction of between 30% to 60% and a yaw overshoot reduction of between 15% to 80% can be achieved in the step steer manoeuvres at different speeds. The steering torque was also reduced by the targeted 30% effectively. It is seen that the torque vectoring capability is very effective in controlling both the lateral dynamics and the steering torque of the vehicle. This capability of torque vectoring therefore opens up new possibilities in effective control of vehicle motion and hence in future active safety systems