Battery SOC prediction method based on long and short term memory network

According to the battery SOC prediction method based on the long and short term memory network, the battery SOC prediction time sequence can be captured, and the prediction result is more accurate. The method comprises the following steps: constructing a three-layer radial basis function neural network according to the structure of a modified long-short term memory network; downloading a battery data set of the NASA, and organizing a data format from data to SOC according to information such as voltage, current, temperature, time and the like; then randomly initializing parameters of the improved long and short term memory network; secondly, inputting the data into the constructed long-short-term memory network to obtain output; thirdly, relevant parameters are updated according to a time sequence gradient descent method; thirdly, determining model parameters, and ending the model training process; and finally, inputting test data into the trained model to predict the SOC of the battery. 本发明提出一种基于长短时记忆网络的电池SO