State-of-Charge Estimation of Energy Storage Batteries Based on Modified Probabilistic Neural Networks
State of charge(SOC)estimation technology for Li-ion battery is an important part of battery management system in energy storage power station.In order to achieve accurate SOC estimation,the paper proposes a modified probabilistic neural network(MPNN)to make an estimation of SOC.Compared with traditional neural networks,MPNN combined with probability function and compensation mechanism can not only avoid falling into local optimization,but also has better fitting ability,further improving SOC estimation accuracy.Simulation results show that the mean absolute error and mean square error of SOC estimation using the proposed MPNN method are both lower than 1％,and satisfactory performance is obtained.
energy storage stationLi-ion batteriesstate of chargeneural networkdata-driven
万方数据
维普
