Abstract:
Vehicle electrification aims at improving energy efficiency and reducing pollutant emissions which creates an opportunity to use the electric machines (EM) as Regenerative Braking System (RBS) to support the friction brake system. Anti-lock Braking System (ABS) is part of the active safety systems that help drivers to stop safely during panic braking while ensuring the vehicle’s stability and steerability. Nevertheless, the RBS is deactivated at a safe (low) deceleration threshold in favour of ABS. This safety margin results in significantly less energy recuperation than what would be possible if both RBS and ABS were able to operate simultaneously. Vehicle energy efficiency can be improved by integrating RBS and friction brakes to enable more frequent energy recuperation activations, especially during high deceleration demands. The main aim of this doctoral research is to design and implement new wheel slip control with torque blending strategies for various vehicle topologies using four, two and one EM. The integration between the two braking actuators will improve the braking performance and energy efficiency of the vehicle. It also enables ABS by pure EM in certain situations where the regenerative brake torque is sufficient. A novelmethod for integrating the wheel slip control and torque blending is developed using Nonlinear Model Predictive Control (NMPC). The method is well known for the optimal performance and enforcement of critical control and state constraints. A linear MPC strategy is also developed for comparison purpose. A pragmatic brake torque blending algorithm using Daisy-Chain with sliding mode slip control is also developed based on a pre-defined energy recuperation priority. Simulation using high fidelity model using co-simulation in Matlab/Simulink and CarMaker is used to validate the developed strategies. Different test patterns are used to evaluate the controllers’ performance which includes longitudinal and lateral motions of the vehicle. Comparison analysis is done for the proposed strategies for each case. The capability for real-time implementation of the MPC controllers is assessed in simulation testing using dSPACE hardware.
