To address the problem that the passive bus priority method with the objective of "minimum passenger delay" may cause an increase in the number of non-priority phase stops, a two-stage improvement optimization model is developed, which is divided into two steps: the first step is to find the minimum intersection passenger delay, and then the delay is appropriately increased by a certain range on this basis, so as to determine the bus priority The first step is to find out the minimum delay of passengers at the intersection, and then the delay is appropriately increased by a certain range, thus determining the set of solutions for bus priority signal timing, and then selecting the solution that minimizes the number of non-priority phase stops from the set of solutions. The particle swarm algorithm is used to solve the model, and the specific ideas and steps are given. The intersection of Huanghe Road-Guangping Street in Dalian City, Liaoning Province, is selected as the study object, and the effectiveness of the improved optimization model is examined by using VISSIM. The results of the empirical simulations show that the improved optimization model can achieve bus priority and reduce the stopping rate of non-priority phases within an acceptable range of delay increase compared to the passenger delay minimum model.