Topology-Agnostic Planning and Graph-Theoretic Power Allocation in Modern Distribution Systems Using Cloudpss

Under the “carbon emission peak and carbon neutrality” goals, traditional distribution networks are evolving into modern new type distribution systems challenged by stochastic renewables, flexible storage, and adaptive loads. This study proposes a topology-agnostic planning framework integrated with graph-theoretic power allocation to address highpenetration distributed renewable energy and flexible load challenges. The framework first conducts resource aggregation and topology-independent capacity planning for distributed PV, wind turbines, and energy storage systems, minimizing annualized costs while maintaining power balance and equipment constraints. Subsequently, the distribution network is abstracted as a weighted graph where Dijkstra's algorithm identifies optimal paths and convex optimization allocates branch power to minimize transmission losses. Implemented on CloudPSS digital twin platform, an 8760 -hour planning case for AC/DC hybrid distribution systems demonstrates balanced model accuracy and computational efficiency. The results validate the method's rationality, providing a practical paradigm for long-term planning of modern AC/DC distribution systems under renewable energy integration.