An Optimization Approach for Improving Comprehensive Performance of PHET Based on Evolutionary Many‐Objective Optimization

Abstract

The parameter optimization coupled with the control strategy and target driving cycle directly affects the performance of vehicles. This paper proposes an optimization approach for a plug‐in hybrid electric truck (PHET), which considers comprehensive performances including fuel economy, emissions, vehicle drivability, safety, and dynamics. First, 10 initial design parameters are selected from powertrain components and a real‐time energy management strategy (EMS). Then, a definitive screening design (DSD) is proposed to simplify the design parameters. Finally, non‐dominated sorting genetic algorithm‐III (NSGA‐III) is proposed to solve a constrained many‐objective optimization problem with 9 objectives, and the design space is refined through a sensitivity analysis. Simulation results demonstrate that the proposed optimization approach can achieve significant improvements regarding both the comprehensive performances, power repartition, and system efficiency. The simulation is conducted both on Chinese Heavy‐Duty Commercial Vehicle Test Cycle (CHTC) and Urban Dynamometer Driving Schedule for Heavy‐Duty Vehicles (UDDSHDV). In addition, to guide a decision maker (DM) to make trade‐offs among many objectives, preferences are also incorporated into the solutions.

MPC-based energy management with short-term driving condition prediction for a plug-in hybrid electric truck

“…Regarding the control principle, energy management strategies have been classified as rule-based strategies and optimization-based strategies. The former, such as the charge-depleting and charge-sustaining (CD-CS) control strategy, has always been an industry approach due to a simple implementation on real vehicles, 3 and the latter is mainstream to be studied by scholars. As a benchmark, dynamic programming (DP) is a global optimization-based EMS therein.…”