Model Predictive Control for Speed and Charge Planning of Electric Vehicles in Minimizing Trip-time and Energy-consumption

초록

This paper presents different methods of model predictive control(MPC) for optimal decision-making on EV charging and speed planning. The goal of high-level planning is to rninirnize trip time and energy consumption for which EV charging planning, i.e., knowing where it will be charged and how much charge it needs, is explicitly considered, as well as energy-efficient speed planning. We propose both non-convex and convex optimization problem formulations for MPC-based, high-level planning of vehicle speed and charging in a spatial domain. A problem in linear MPC is presented as a convex quadratic programming approximation(i.e., convexified quadratic program) of the original nonlinear MPC. In the proposed linear MPC, the square of the vehicle speed is considered as a state variable, battery dynamics is simplified, and the powertrain constraint is convexified. To assess the conservatism of a convexified MPC problem and its solution, we will compare the performances of nonlinear and linear MPC solutions in a driving simulation of a Munich-Cologne trip (573 km) with four charging stations. The optimality of a linear MPC is shown as comparable to nonlinear counterparts, whereas its computation speed is one order of magnitude faster. This implies that the proposed linear MPC can be also used in short-term replanning, in which improved energy efficiency and increased reduction in trip time could be achieved in real driving conditions. Copyright © 2023 KSAE.

키워드