A bi-level optimization approach for siting and sizing of distributed wind-storage power systems in rural areas based on MCTS and IPSO algorithms

With the escalating land scarcity caused by rapid wind power expansion, rural areas have emerged as strategic hubs for distributed wind power deployment due to their abundant wind resources and spatial advantages. This study addresses the expansion needs of wind turbine generator system (WTGS) in rural distribution network (RDN) by proposing a bi-level optimization framework synergizing Monte Carlo Tree Search (MCTS) and multi-objective Improved Particle Swarm Optimization (IPSO). The upper layer employs MCTS to optimize the siting and sizing of WTGS and energy storage systems (ESSs), minimizing annual configuration costs. The lower layer utilizes IPSO to coordinate the operation of WTGS and ESS, reducing annual operational costs while enhancing grid reliability. The proposed algorithm significantly enhances the optimization efficiency compared to conventional algorithms in large-scale variable optimization. Moreover, it incorporates a dynamic power allocation strategy to balance the state-of-charge (SOC) across all ESSs. Implemented in a typical RDN in Guangdong Province, the framework delivers optimized outcomes: an annual comprehensive cost of about 28.24 million CNY over the planning period, a node voltage violation probability of 0 %, a daily average load interruption rate of 0.523 %, a wind curtailment rate of 0.589 %, and an overall grid reliability of 99.739 %. This work advances renewable energy source integration in RDN, providing technical support for the China's latest policy and offering a replicable model for global rural decarbonization.