Abstract: Addressing the issues of widely distributed network nodes and heterogeneous power consumption in distributed communication networks, this paper aims to overcome key challenges of traditional centralized power supply modes, such as high line losses, large construction costs, and system fragility. The main contributions are: development of a photovoltaic–wind–storage hybrid power supply system, design of an intelligent power forecasting model and a fuzzy coordinated control strategy, proposal of an energystorage capacity optimization method that accounts for battery aging. Innovations include the establishment of a Gaussian mixture model for communicationnetwork power consumption, the development of an SVR–LSTM hybrid forecasting algorithm, the design of a multilayer hierarchical fuzzy coordination control strategy, and the formulation of a lifecyclecostoriented storage optimization approach. Experimental results indicate high accuracy in power prediction, significantly improved renewable energy utilization, and substantially higher operational efficiency compared with conventional modes. The system maintains very high supply reliability under prolonged extreme climatic conditions. The proposed intelligent power supply architecture effectively ensures continuous power for communication networks and promotes the development of green communications.