Design of a lithium-ion battery pack for PHEV using a hybrid optimization method

This paper outlines a method for optimizing the design of a lithium-ion battery pack for hybrid vehicle applications using a hybrid numerical optimization method that combines multiple individual optimizers. A gradient-free optimizer (ALPSO) is coupled with a gradient-based optimizer (SNOPT) to solve a mixed-integer nonlinear battery pack design problem. This method enables maximizing the properties of a battery pack subjected to multiple safety and performance constraints. The optimization framework is applied to minimize the mass, volume and material costs. The optimized pack design satisfies the energy and power constraints exactly and shows 13.9-18% improvement in battery pack properties over initial designs. The optimal pack designs also performed better in driving cycle tests, resulting in 23.1-32.8% increase in distance covered per unit of battery performance metric, where the metric is either mass, volume or material cost.