Identification method of control parameters of energy storage converter based on LSTM-IPSO

A control parameter identification method for energy storage converters based on long short-term memory (LSTM) and improved particle swarm optimization (IPSO) algorithm is proposed to address the problem that it is difficult to accurately reflect the output characteristics of energy storage converters under multiple operating conditions by identified parameters. Firstly, the Spearman rank correlation coefficient method is utilized to analyze the correlation between control parameters and the selected observed values, and the current, active power, and reactive power of the orthogonal and direct axes are selected as the observed values. Then, the LSTM neural network is trained under multiple operating conditions to obtain preliminary control parameter values without setting an optimization range to simplify the identification process. The IPSO algorithm is used to further optimize the control parameters to improve identification accuracy. Finally, the effectiveness and feasibility of the LSTM-IPSO control parameter identification method were verified through experiment. The results show that the weighted average total deviation of the LSTM-IPSO identification algorithm was about 3%. Compared with the LSTM identification algorithm, the LSTM-IPSO identification algorithm can reduce the error by 80%.