Optimal control and stability of four-wheeled vehicles

Two vehicular optimal control problems are visited. The first relates to the minimum lap time problem, which is of interest in racing and the second the minimum fuel problem, which is of great importance in commercial road vehicles. Historically, minimum lap time problems were considered impractical due to their slow solution times compared with the quasi-steady static (QSS) simulations. However, with increasing computational power and advancement of numerical algorithms, such problems have become an invaluable tool for the racing teams. To keep the solution times reasonable, much attention still has to be paid to the problem formulation. The suspension of a Formula One car is modelled using classical mechanics and a meta-model is proposed to enable its incorporation in the optimal control problem. The interactions between the aerodynamics and the suspension are thereby studied and various related parameters are optimised. Aerodynamics plays a crucial role in the performance of Formula One cars. The influence of a locally applied perturbation to the aerodynamic balance is investigated to determine if a compromise made in design can actually lead to lap time improvements. Various issues related to minimum lap time calculations are then discussed. With the danger of climate change and the pressing need to reduce emissions, improvements in fuel consumption are presently needed more than ever. A methodology is developed for fuel performance optimisation of a hybrid vehicle equipped with an undersized engine, battery and a flywheel. Rather than using the widely used driving cycles, a three-dimensional route is chosen and the optimal driving and power management strategy is found with respect to a time of arrival constraint. The benefits of a multi-storage configuration are thereby demonstrated. Finally, the nonlinear stability of a vehicle model described by rational vector fields is investigated using region of attraction (RoA) analysis. With the aid of sum-of-squares programming techniques, Lyapunov functions are found whose level sets act as an under-approximation to the RoA. The influence of different vehicle parameters and driving conditions on the RoA is studied.