Generative adversarial networks-based data clustering in transportation-oriented energy interconnected system

Data clustering plays a critical role in processing massive data in transportation-oriented energy interconnected systems, where complex data distributions and limited network representation capability often lead to unsatisfactory clustering performance. To address these challenges, this paper proposes a generative adversarial networks (GANs)-based data clustering method that enhances feature representation and clustering effectiveness. The objective of this study is to improve clustering accuracy and robustness by overcoming the limitations of conventional network structures, while reducing computational burden during the training process. To achieve this, a GAN-based feature extraction framework is developed to learn discriminative representations through adversarial training. A progressive updating strategy is introduced to adaptively adjust network structure and parameter scale, improving training efficiency. Furthermore, a fuzzy clustering mechanism combined with an attraction-based assignment regularization term is designed to enhance both intracluster compactness and intercluster separability. Experimental results on a pressure dataset demonstrate that the proposed method achieves superior clustering performance compared with several representative approaches. The proposed method shows strong potential for practical applications in transportation-oriented energy interconnected systems, including pipeline monitoring, fault detection, and intelligent operation.