Against the backdrop of global energy transition and the rapid development of the new energy vehicle industry, as critical refueling infrastructure, the optimal layout of battery swap stations is essential for enhancing both service efficiency and power infrastructure effectiveness. This study focuses on the operational passenger vehicle battery swap market in City B. Operational data of battery-swap taxis are obtained through market research. A hybrid queueing model is introduced to establish a saturation prediction model based on dynamic dilution effects. Additionally, a fusion algorithm coupling the Voronoi diagram with the particle swarm optimization algorithm is proposed. Based on the aforementioned methods, a “prediction-optimization-layout” collaborative planning framework is constructed, quantifying policy sensitivity and supply-demand interactions. The reliability of the prediction of 23 theoretically new battery swap stations by 2025 is further verified through Monte Carlo simulation. Through the coordinated allocation of battery swap stations and charging guns (65 stations + 4 charging guns), the average user waiting time is controlled within 10 min, and the actual station construction demand is optimized to 13 stations. By integrating dynamic spatial partitioning with global optimization, the challenge of site optimization in high-density urban areas is addressed. The research findings provide an implementable solution for battery swap network planning that balances service efficiency and investment costs and also offer valuable insights for optimizing distributed power infrastructure.