PSO-based fuzzy logic optimization of dual performance characteristic indices for fast charging of Lithium-ion batteries

Abstract

Efficacies of the charging strategy and electrified pattern dominate the functionality and lifespan of Lithium-ion (Li-ion) batteries intensely. In order to maximize the available performance of the Li-ion batteries, in this paper, a searching strategy based on the particle swarm optimization (PSO) conducted with a fuzzy-deduced fitness evaluator (FDFE) is proposed to find the best multistage charging current pattern. The objective function is to maximize the cost benefit of the applied prospective charging pattern based on the nonlinear weightings' allocation of the charge time (CT) and normalized discharge capacity (NDC) in the objective function. Rules of the weighting allocation are derived from the fuzzy logic inference. The fitness value of each candidate particle (charging pattern) is figured out by the FDFE to guide the searching path of the PSO algorithm for finding the optimal solution out. Experimental findings show that the resulted pattern is capable of charging the batteries to over 90% available capacity within 50 minutes. Comparing with the conventional constant current-constant voltage (CC-CV) method, the devised scheme has the performance enhancements of more than 80% charging time reduction, 21% more life cycles, and over 0.4% charging efficiency increase.