Development and Improvement of a Situation-Based Power Management Method for Multi-Source Electric Vehicles

This paper presents an improved situation-based optimal power management method for multi-source electric vehicles. The presented method implements a novel situation recognition principle, based on mapping the vehicle operating conditions into a multi-dimensional space referred to as: grid- space. In grid-space, multiple recognition variables are depicted as axes. These variables have been initially determined based on literature analysis to be namely: vehicle speed, power demand, and speed dynamics. Control parameters of the supervisory power management algorithm have been optimized offline for every vehicle situation (state) in the grid-space. Improvement of grid- space structure is investigated based on two concepts: different axes' discretization and implementation of further variables. For each new grid-space structure, offline optimization of control parameters is performed, solutions to vehicle states in grid-space are assigned, and real-time application is conducted to analyze the energy saving results. In this work, six grid- space structures have been considered and tested using four driving cycles representing different operating conditions. Results analysis revealed the positive impact of accurate discretization of power demand and speed dynamics to achieve a significant improvement of energy efficiency robustly for all tested driving cycles.