Parameter identification for a complex lead-acid battery model by combining fuzzy control and stochastic optimization

Abstract

The accurate simulation of lead-acid batteries requires the use of sophisticated models based on first principles containing many parameters. Existing methods for parameter identification often fail due to many local minima of the error function or the high computational needs to cover the parameter space. Therefore, a novel approach for parameter identification with complex physical models containing many unknown parameters is presented. It is based on the utilization of available expert knowledge regarding specific model parameters. The expert knowledge is integrated through fuzzy control and combined with stochastic optimization algorithms for solving the battery identification problem.