Hierarchical predictive optimal control for range extension of EV with ANN based torque control for IPMSM drives

• A predictive control scheme where the road characteristics and braking dynamics which are not discussed in existing literature are considered for the control design. • The impact of regenerative braking and the resultant energy recuperation is addressed in this study, a factor typically overlooked in previous optimal control designs. • The control torque signal is realised in the vehicle drives using a novel AI based control scheme. • Existing control algorithms lacks on-line optimization corresponding to variations in driving commands, which is considered in this work. This study presents a novel methodology to enhance the energy efficiency of Electric Vehicles (EVs) while maintaining dynamic stability and driving comfort, even on uneven roads, using onboard vehicle measurements. The approach involves a hierarchical control scheme that integrates a model predictive controller with a torque vectoring algorithm to optimize torque demand and accurately anticipate road-specific torque requirements. The proposed control is designed and implemented on an EV with Interior Permanent Magnet Synchronous Motors (IPMSM) on four wheel drives. The optimal torque control is realised through an Artificial Neural Network (ANN) - based motor torque control scheme. The design is validated through tests in real-world driving scenarios. In comparison with conventional methods, the proposed method shows a 37% increase in energy efficiency across different test conditions, thereby resulting in an increase in EV driving range. These advancements are realised without substantial modifications to the EV’s drivetrain, representing a significant step forward in sustainable and efficient electric mobility.