RESEARCH ON REMAINING LIFE PREDICTION METHOD OF LITHIUM-ION BATTERY FOR ENERGY STORAGE BASED ON HYBRID MODEL
Aiming at the safety problems of battery aging in energy storage systems under peak and frequency regulation conditions,a method for predicting the remaining useful life(RUL)of lithium-ion batteries for energy storage is proposed.The method extracts short-time voltage and current fragment data from the constant current and constant voltage charging process of lithium-ion batteries as a characteristic to characterize the battery aging by studying the characteristics of the charging curve of lithium-ion batteries,proposes a B-spline curve method for noise reduction of the obtained fragment data,and obtains the new health factor(HF)which is highly correlated with the capacity,and then applies the sparrow search algorithm improved by the butterfly optimization strategy to optimize the parameters of the convolutional neural network-Long and short-term memory network hybrid model parameters are optimized,and finally the RUL of lithium-ion battery is predicted by NASA dataset and experimental platform actual operation data and verified by comparing with other methods.The results show that the predicted results of the NASA dataset RMSE and MAE are controlled within 3.45%and 2.42%,and the predicted results of the battery aging dataset of the experimental platform are controlled within 0.96%and 2.42%,and the predicted results of the experimental platform are controlled within 0.96%and 0.96%.The range of RMSE and MAE is controlled within 0.96%and 0.83%,and the method proposed in this paper has high accuracy in predicting the RUL of lithium-ion batteries,which can effectively solve the problem of safety warning of energy storage batteries,and provide a guarantee for the stable control of power grids.
lithium-ion batteryenergy storageremaining useful lifeB-spline curvenew health factorbutterfly optimization strategy
万方数据
维普
