Research And Implement Of Dynamic CPM Optimal Algorithm Based On Constraints

Schedule management is the most important part of project management. It generates the project plan through breaking down and making the network planning, balance and optimize the resource and costs of projects. Reasonable project schedule management is the bases of projects' perform tracking and control. Modern projects become more complex. Shorter periods, less cost and must be finished on time are required. So modern project schedule management should be more feasible, more doubtless and more stable, however, the traditional project schedule management methods cannot fulfill these demands.The thesis analyzes the disadvantages of traditional project schedule management systematically, and advantages and problems of applying Critical Chain to do project schedule management. As a result, a new method of using time constraint transformed from other constraints is proposed to solve the problems. Critical Chain takes resource constrains and psychology factors into consideration, and makes use of buffering to control the project scheduling. As collecting information is hard and project managers are familiar with different aspects of knowledge, this article presents how to execute the project schedule management only considering from the time aspect, while other resource constraints are transform into time constrains. In addition, Monte Carlo method is imported to solve the problems of instable of task's duration.A new system of project schedule management system software is designed and coded, and a list of schedule data from the project of Sinopec is used to check the validity of the proposed method while set project period as criterion. Through establishing the simulation model of the proposed planning and traditional schedule planning, article compares the new period distribution and statistics of the proposed planning with traditional method's distribution and statistics by making use of the new project management system software, we find out that the new scheduling method makes great improvement and the proposed method planning is proved to be a more stable and more feasible project schedule management method.