Article: Optimisation design of electric vehicle charging stations based on GAN and reinforcement learning Journal: International Journal of Information and Communication Technology (IJICT) 2026 Vol.27 No.81 pp.1 - 21 Abstract: Designing the optimal parameters for an integrated photovoltaic-storage-charging (PVSC) station is a complex task. Traditional methods rely on simulating all possible parameter combinations, which is time-consuming and computationally intensive. To accelerate the optimisation process and reduce computational costs, this paper proposes a novel approach combining generative adversarial networks (GANs) and reinforcement learning. First, the performance of a subset of parameter combinations is evaluated through simulation of charging demand, and these data are used to train the GAN to generate additional synthetic samples, thereby expanding the search space. Then, an artificial neural network (ANN) surrogate model is constructed to predict the performance metrics for any given parameter combination. Finally, the surrogate model is used within a reinforcement learning framework to efficiently search for the optimal parameters. Experimental results show that this method effectively accelerates the design optimisation process, providing an efficient framework for quickly determining the optimal parameters of photovoltaic charging stations. Inderscience Publishers - linking academia, business and industry through research

Title: Optimisation design of electric vehicle charging stations based on GAN and reinforcement learning

Authors: Hao Huang; Wendi Wang; Gang Li; Ye Ji; Liuting Gu; Shufan Wang

Addresses: State Grid Jiangsu Electric Power Co., Ltd. Nanjing Power Supply Branch Company, Nanjing, 210008, Jiangsu, China ' State Grid Jiangsu Electric Power Co., Ltd. Nanjing Power Supply Branch Company, Nanjing, 210008, Jiangsu, China ' Nanjing Suyi Industrial Co., Nanjing, 210008, Jiangsu, China ' State Grid Jiangsu Electric Power Co., Ltd. Nanjing Power Supply Branch Company, Nanjing, 210008, Jiangsu, China ' Nanjing Suyi Industrial Co., Nanjing, 210008, Jiangsu, China ' Nanjing Suyi Industrial Co., Nanjing, 210008, Jiangsu, China

Abstract: Designing the optimal parameters for an integrated photovoltaic-storage-charging (PVSC) station is a complex task. Traditional methods rely on simulating all possible parameter combinations, which is time-consuming and computationally intensive. To accelerate the optimisation process and reduce computational costs, this paper proposes a novel approach combining generative adversarial networks (GANs) and reinforcement learning. First, the performance of a subset of parameter combinations is evaluated through simulation of charging demand, and these data are used to train the GAN to generate additional synthetic samples, thereby expanding the search space. Then, an artificial neural network (ANN) surrogate model is constructed to predict the performance metrics for any given parameter combination. Finally, the surrogate model is used within a reinforcement learning framework to efficiently search for the optimal parameters. Experimental results show that this method effectively accelerates the design optimisation process, providing an efficient framework for quickly determining the optimal parameters of photovoltaic charging stations.

Keywords: integrated photovoltaic-storage-charging; charging station; reinforcement learning; neural networks; design parameter optimisation.

DOI: 10.1504/IJICT.2026.155110

International Journal of Information and Communication Technology, 2026 Vol.27 No.81, pp.1 - 21

Received: 18 Nov 2025
Accepted: 04 Feb 2026
Published online: 27 Jul 2026 *

Title: Optimisation design of electric vehicle charging stations based on GAN and reinforcement learning

Keep up-to-date