Power management strategy of hybrid electric vehicles based on particle swarm optimization

The Power management strategy of HEV using global optimization techniques can achieve optimum control solution. However, the “a priori” nature of the trip information and heavy computational cost prohibit it from being utilized in real world application. In this paper, a power management strategy using particle swarm optimization (PSO) is proposed. The aim is to achieve real time implementation and sub-optimal control solution without requiring the “a priori” knowledge of the driving cycle. Using pricewise linearization, at each time step, normalized comprehensive energy loss for each power split scenario is obtained and normalized over the traveling distance. The power split strategy that minimizes the normalized comprehensive energy loss is considered optimal. However, searching for the optimal power split is mathematically challenging and time consuming. To address the real time implementation, PSO algorithm is employed as the global minima searching tool. Simulation study on a series-parallel configuration passenger vehicle has been performed. In addition, Dynamic Programming (DP) technique has also been implemented in the simulation for the comparison purpose. The simulation results demonstrated that the proposed control strategy is able to achieve comparable fuel economy with global optimization while feasible for real time implementation.