Coordinated Hierarchical Co-Optimization of Speed Planning and Energy Management for Electric Vehicles Driving in Stochastic Environment

Abstract

Active co-optimization of future speed profiles together with powertrain control is the optimal solution to further exploiting the energy benefit of electric vehicles (EVs) in real-world operation. However, with uncertainties in driving conditions and concerns about driving safety, speed planning results are cautious and with frequent speed variations, which deteriorates the energy economy of EVs in turn. To comprehensively optimize the energy economy and driving safety of EVs in a stochastic driving environment, this paper develops a chance constraint model predictive control (CC-MPC) for co-optimizing the speed planning and powertrain control, which forms an advanced energy management method. To handle the instantaneous disturbance, a coordinated hierarchical method (CHM) is engineered for solving the CC-MPC. As suggested by simulation, the driving safety (measured by success rate) can be increased to 81% with the CC-MPC, which realizes a 62% improvement compared with situations without CC-MPC. Moreover, the proposed CC-MPC significantly mitigates the conflict between driving safety and the energy economy, and the worst deterioration of the energy economy is only 9.3%. Sacrificing merely 2.1% sub-optimality, CHM removes 86% computation loads, and the median of CPU time is merely 0.58s at each computation step (control interval 1s), which makes the CC-MPC promising for online implementation.