Rapid Optimization of Battery Charging- Discharging Profiles Using SOC-SOC Rate Domain for Cruising Hybrid Vehicles

Abstract

Intelligent use of batteries in hybrid electric vehicles (HEV) is the main factor for high-fuel economy. For a cruising HEV, the state-of-charge Pulse and Glide (SOC-PnG) is known to be the optimal charging-discharging pattern. The optimal charging-discharging profiles of the SOC-PnG strategy can be obtained using the optimization algorithms, such as dynamic programming and Pontryagin's minimum principle. These methods are general tools for optimal control problems but they demand too much computing resources. This paper proposes a problem-specific optimization method that can solve an optimal control problem of the SOC-PnG with less computing resources than the general optimization algorithms. The problem is formulated in the SOC-SOC rate domain, which simplifies the optimization problem and removes many constraints. This approach transforms the necessary conditions for the optimality from one with a path function to one with a point function. An optimization algorithm based on the point function can obtain the optimal solutions about 100 times faster than dynamic programming. This fast optimization is possible because the algorithm exploits the specific cost structure and problem properties. Therefore, this algorithm cannot be applied to all optimization problems but it can accelerate the recursive and time-consuming system design or control design processes in HEV research.

Keywords

• Fuel efficiency
• hybrid vehicle
• optimization
• state-of-charge Pulse and Glide (SOC-PnG)