Parameter Optimization of Model Predictive Direct Motion Control for Distributed Drive Electric Vehicles Considering Efficiency and the Driving Feeling

This paper presents a novel motion control strategy based on model predictive (MPC) for distributed drive electric vehicles (DDEVs), aiming to simultaneously the longitudinal and lateral while considering efficiency driving feeling. Initially, we analyze vehicle’s dynamic model, vehicle body in-wheel motors, establish foundation control. Subsequently, propose direct (MPDMC) approach that utilizes single CPU directly follow driver’s commands by generating voltage references with minimum cost function. The function of MPDMC is constructed, incorporating factors such as velocity, yaw rate, displacement, efficiency. We extensively weighting parameters introduce an optimization algorithm particle swarm (PSO). takes into account aforementioned well feeling, which evaluated using trained long short-term memory (LSTM) neural network. LSTM network labels response under different in various working conditions, i.e., “Nor”, “Eco”, “Spt”. Finally, evaluate performance optimized through simulations conducted MATLAB CarSim software. Four typical scenarios are considered, results demonstrate outperforms baseline methods, achieving best performance.