基于DBO-BP神经网络的锂电池双层均衡控制研究OA

In allusion to the issues of slow balancing speed and low efficiency in traditional single-layer balancing topologies,a dual-layer balancing topology is proposed.At the bottom layer of the topology,an improved Buck-Boost circuit is adopted for equalization,while a flyback transformer circuit is applied at the top layer.This structure can optimize the energy transmission path and realize rapid equalization between any single cell and any battery pack.The dung beetle optimization(DBO)algorithm is used to optimize the traditional BP neural network,accelerating the convergence speed of the neural network and global optimization capability.The maximum error of SOC estimation is reduced to 0.49%,which effectively improves the accuracy of SOC estimation.The intra-group extreme range method and the inter-group mean difference comparison method are used in the equalization control strategy to simultaneously realize the equalization control.A simulation model of nine series-connected lithium-ion batteries is built in Matlab/Simulink.Comparative simulations are conducted between the traditional single-layer equalization topology and the proposed double-layer equalization topology under both charging and discharging equalization modes.The experimental results show that the proposed dual-layer equalization method can effectively shorten the equalization time and improve the equalization efficiency.