Taking the domestic waste reverse logistics network in Hefei as the research object, considering the multi frequency recycling and vehicle sharing scheduling strategy, a reverse logistics network optimization scheme based on the operation cost of the reverse logistics network is proposed. Firstly, a reverse logistics recycling operation cost model is constructed, which includes the construction cost of reverse logistics infrastructure, equipment maintenance cost, the sum of the transportation cost of municipal solid waste and the treatment cost of non-recyclable waste at all levels, and a product maximization revenue model of the recycling center is established. Secondly, according to the characteristics of the model, a k-means spatio-temporal clustering algorithm considering the geographical location of waste transfer stations, recycling frequency and recycling time window is designed, and then an improved GA-PSO hybrid algorithm is proposed. The effectiveness of the model and algorithm is verified by comparing with HGA, GA-TS and HACO algorithms. Finally, according to the actual data of Hefei domestic waste reverse logistics network, the optimization research was carried out, and the recycling frequency and vehicle scheduling of transfer stations in different locations were analyzed. The results show that the model and algorithm proposed in this paper can carry out the optimal selection of recycling cost, resource sharing of recycling vehicles and reasonable vehicle routing optimization scheduling, and the operation cost of reverse logistics network is reduced by 8% compared with the previous one; As the transportation distance is reduced, the transportation cost is greatly reduced, and the treatment efficiency of municipal solid waste is improved.