Hybrid genetic algorithm-based optimization of powertrain and control parameters of plug-in hybrid electric bus

Abstract

This paper proposes a novel hybrid genetic algorithm for the simultaneous optimization of the powertrain and control parameters in plug-in hybrid electric bus (PHEB) with trade-off between economy and dynamic performance. PHEBs are potential public transportations to alleviate energy shortages and urban environment pollution. The PHEB powertrain and control parameters significantly impact the vehicle performance and economy, and an optimization process is needed to design a set of optimized parameters for a given driving route. A novel hybrid genetic algorithm (HGA) which combines an enhanced genetic algorithm (EGA) with simulated annealing (SA) is proposed in this paper. By merging EGA with SA, simulated annealing process is applied to the better half population after EGA operations, and then an adaptive cooling schedule is introduced. In addition, several techniques are implemented to achieve the goals of sustaining the convergence capacity and maintaining diversity in the population, such as orthogonal design method, adaptive mechanisms of crossover and mutation probabilities. A solution relative error distance is defined to express the performance of standard genetic algorithm (SGA), EGA, and HGA. The optimization is performed over the following two driving cycles: (1) a driving cycle CYC_873 collected from a real bus route; and (2) Urban Dynamometer Driving Schedule+China Typical Urban Driving Cycle (UDDS+CTUDC). Simulation results indicate that the convergence speed and global searching ability of HGA are significantly better for optimal PHEB powertrain and control parameters design. And the optimal parameters might obtain the best comprehensive performance of PHEB for the given Chinese urban driving cycles.