Parameter Identification of Retired Batteries Based on Improved Adaptive Particle Swarm Optimization

To address the issue of traditional Particle swarm optimization algorithms easily falling into local extremes and exhibiting low identification accuracy when identifying parameters in decommissioned batteries, an improved adaptive Particle swarm optimization algorithm-based parameter identification model is proposed. This model is developed by analyzing the five-parameter model of decommissioned batteries. The adaptive strategy adjusts the inertia weight factor within the conventional Particle swarm optimization algorithm. Additionally, an asynchronous learning factor is introduced to balance the search relationship between global and local extremes. The constructed adaptive Particle swarm optimization model is then employed to identify parameters in three distinct types of decommissioned batteries and compared with the traditional Particle swarm parameter identification method. Results demonstrate that the improved adaptive Particle swarm optimization algorithm exhibits higher accuracy in parameter identification, with average relative errors ranging from 0.975% and 5.975% and an overall error below 6%. These findings validate the feasibility and effectiveness of the proposed adaptive Particle swarm optimization algorithm in identifying parameters for retired batteries.