To address the issues of instantaneous power fluctuations and tight power supply resources caused by concentrated charging of electric vehicles, this paper proposes a charging station power dynamic allocation algorithm based on weighted maximum fairness. This method achieves real-time power allocation and resource optimization under limited power supply capacity by comprehensively considering the charging needs, charging duration, and priority weights of each vehicle. The system is deployed using the open-source energy management platform OpenEMS, establishing a control architecture where cloud-based backend and on-site edge nodes operate in tandem. Experimental evaluations were conducted to assess the algorithm’s real-time performance and scalability. Experimental results demonstrate that this method effectively enhances the fairness and response efficiency of power allocation, showcasing significant engineering application value.