During the research of project scheduling problem, in order to achieve objective that included arranging the execution sequence of the activities and the execution mode reasonably, minimizing the project duration, an objective function of multi-objective scheduling was built and restriction conditions were defined. Then considering the specialty of this combinatorial optimization problem, the standard particle swarm optimization algorithm was improved, and a method presented the concepts of swap operator and swap sequence was designed for improving the optimizing speed and capacity of the algorithm. Finally the validity of such method has been proved with an example.