Improved chaos genetic algorithm based state of charge determination for lithium batteries in electric vehicles

Lithium batteries are developed rapidly in electric vehicles, and the accurate online evaluation of available capacity for ensuring their safety and functional capabilities is challenging due to the stability of initial value, extensive computational requirements and convergence issues. This paper proposes an improved chaos genetic algorithm based method to evaluate the state of charge of batteries with low computational complexity and high initial stability. Based on a combined state space model employed to simulate battery dynamics, an improved chaos genetic algorithm based method which comprises chaos genetic algorithm, Ampere hour approach and adaptive switch mechanism is advanced to predict the available capacity. The method is validated by the experiment data collected from battery test system. Results indicate that the improved chaos genetic algorithm based method shows great performance with low computational complexity and is little influenced by the given initial value.