Power parameter allocation of electric vehicles by integrating optimized AFSA and system parameter classification

Introduction Traditional methods for allocating power parameters in electric vehicles are prone to becoming trapped in local optima, which hinders their ability to meet rising performance demands. To enhance the accuracy of this allocation, this study proposes a novel power parameter allocation method for new energy vehicles. Methods The proposed method integrates an optimized Artificial Fish Swarm Algorithm (AFSA) with system parameter classification. First, a parameter classification model is constructed based on sensitivity analysis. Subsequently, the improved AFSA is employed to perform the parameter optimization. Results The parameter allocation based on sensitivity analysis demonstrated strong performance. Specifically, with a main reduction ratio of 11:1, the high-speed re-acceleration time was 5.7 s. When the coolant flow rate was set to 7 L/min, the peak power duration reached 30.1 s. Compared to other methods, the comprehensive energy consumption achieved by the improved AFSA was the lowest, recording 13.8 kWh/100 km with a battery capacity of 80 kWh. Discussion The developed method effectively overcomes the tendency of traditional approaches to fall into local optima. It significantly improves both the dynamic performance and energy efficiency of electric vehicles, offering a more effective solution for power parameter allocation.