A comparative analysis of meta-heuristic methods for power management of a dual energy storage system for electric vehicles

• Two meta-heuristic approaches are evaluated for multi-ESS management in electric vehicles.
• An online global energy management strategy with two different layers is studied.
• Meta-heuristic techniques are used to define optimized energy sharing mechanisms.
• A comparative analysis for ARTEMIS driving cycle is addressed.
• The effectiveness of the double-layer management with meta-heuristic is presented.

This work is focused on the performance evaluation of two meta-heuristic approaches, simulated annealing and particle swarm optimization, to deal with power management of a dual energy storage system for electric vehicles. The proposed strategy is based on a global energy management system with two layers: long-term (energy) and short-term (power) management. A rule-based system deals with the long-term (strategic) layer and for the short-term (action) layer meta-heuristic techniques are developed to define optimized online energy sharing mechanisms. Simulations have been made for several driving cycles to validate the proposed strategy. A comparative analysis for ARTEMIS driving cycle is presented evaluating three performance indicators (computation time, final value of battery state of charge, and minimum value of supercapacitors state of charge) as a function of input parameters. The results show the effectiveness of an implementation based on a double-layer management system using meta-heuristic methods for online power management supported by a rule set that restricts the search space.

Figure optionsDownload as PowerPoint slide