基于PSO-RBF算法的锂电池SOH研究与预测

PSO-RBF Neutral Network in the Lithium Battery SOH Research and Prediction

为了寻找锂电池充电的最优策略,采用建立模型的方法进行研究与预测充电策略的优劣。电池健康管理状态(State ofHealth)反映了锂电池的剩余寿命,一般作为锂电池充电策略优劣的一个评判标准。在实际应用中,不同的充放电策略对锂电池的SOH有不同的影响,由于对锂电池SOH影响因子很多,各影响因子之间相互耦合,实验验证极其复杂。RBF神经网络是一种比较常用的预测性神经网络,PSO算法是一种较为先进的优化网络参数的算法,将PSO算法和RBF神经网络融合,借助大量实验数据,训练RBF神经网络,使用PSO算法优化其网络参数,建立基于PSO-RBF算法的锂电池SOH预测模型,再将不同充电策略进行仿真验证。仿真结果表明,该模型预测能力优于普通RBF模型,可作为锂电池最优充放电策略验证的最优模型。

In order to find the optimal strategy for lithium battery charging,the method of establishing a model was used to study and predicted the pros and cons of the charging strategy.The State of Health reflects the remaining life of the lithium battery,and is generally used as a criterion for judging the quality of the lithium battery charging strategy.In practical applications,different charging and discharging strategies have different effects on the SOH of lithium batteries.Because there are many influence factors on the SOH of lithium batteries,and the influence factors are coupled with each other,the experimental verification is extremely complicated.RBF neural network is a relatively commonly used predictive neural network.PSO algorithm is a more advanced algorithm for optimizing network parameters.PSO algorithm and RBF neural network were merged,and a large amount of experimental data was used to train RBF neural.The PSO algorithm was used to optimize the network parameters,a lithium battery SOH prediction model was established based on the PSO-RBF algorithm,and then different charging strategies were simulated and verified.The simulation results show that the predictive ability of the model is better than the ordinary RBF model,and it can be used as the optimal model for the verification of the optimal charging and discharging strategy of lithium batteries.