Adaptive control method for composite energy storage in intelligent distribution network based on convolutional neural network

A proposal is put forward for an adaptive control method for composite energy storage in smart distribution networks, which utilizes a convolutional neural network to achieve accurate control over the bidirectional charging and discharging of energy. Utilizing convolutional neural networks to train composite energy storage parameters, introducing softmax classifiers to identify the discharge state of composite energy storage, simulating energy storage capacity, light intensity, and temperature as inputs to the convolutional neural network, and using genetic algorithms to solve the output value of composite energy storage control, achieving adaptive adjustment of composite energy storage in distribution networks. Based on the experimental findings, it has been demonstrated that the suggested approach can adaptively modify the temporal arrangement of power energy considering fluctuations in seasons and temperature. This enhancement significantly enhances the operational flexibility and dependability of composite energy storage in distribution networks.