State Estimation and Dynamic Control Allocation for Over-actuated Electric Vehicles with In-wheel Motors

The electrification of road vehicles is an attractive research area, as problems associated with greenhouse gas emission are now a global challenge. Particularly, electric vehicles with in-wheel steering and driving motors, (also called four-wheel-independent-steering (4WIS) and four-wheel-independent-driving (4WID) electric vehicles), can utilize redundant control actuators to achieve multiple control targets. Hence these vehicles are receiving considerable attention. This thesis aims to develop novel control allocation strategies for 4WIS-4WID electric vehicles to improve their performance and energy efficiency. A comprehensive literature review of vehicle dynamics and various control systems is the focus of the introduction. Then various real-time vehicle state estimators are proposed and a comprehensive vehicle dynamics model is suggested for 4WIS-4WID electric vehicles. Based on the state estimators and vehicle dynamics model, the linear feedback and non-linear over-actuated control allocation methods for the primary control targets, such as handling and stability control, of 4WIS-4WID electric vehicles are suggested. Next, the application of the over-actuated control allocation method is extended into the secondary control targets, such as energy-efficient control and fault-tolerant control. The autonomous vehicle trajectory control can be achieved by applying the over-actuated control allocation method. Finally, a scaled electric vehicle is utilized in the experiment to measure vehicle states and to verify the proposed vehicle mass estimator and road slope estimator. It can be concluded that the redundant control actuators provided by 4WIS-4WID electric vehicles are effective in achieving multiple control targets through the application of designed control allocation methods. These findings demonstrate the promising future of 4WIS-4WID electric vehicles.