Research On Problem Of Hierarchical Resource-Constrained Program Scheduling

As the glob economy develops,the scale of project management is increasing.Some studies show that nearly 90% of project management is multi-project management.Multiproject management has become a hot topic since the 1980 s.Most studies assumed that there was no precedence relationship between projects,all projects were executed in parallel and unified scheduling with common optimization goals under the resource constraint.However,it is difficult to execute projects in parallel due to precedence relationship constraints,such as the relationship between civil engineering projects and hydropower projects in construction projects,the relationship between product design and product manufacturing in customized production,and the relationship between the upstream and downstream of the supply chain.To make multi-project management more practical,this thesis selects program scheduling as the research object.The program is a cluster of interrelated projects whose relationship is closer.The main characteristics of program management include program integrated management of multiple projects,and each project has their own goals and comply with the program goals,and there may be a close relationship and shared resources between projects.For the program scheduling problem,this thesis optimizes the program with multidisciplinary complementary advantages.Combining genetic algorithm,analytical target cascading,and critical chain theory,this thesis systematically studied the program composed of different types of projects（RCPSP,MRCPSP,and DTCTP-D）.By reducing the dimension and complexity of the program,we proposed different solutions by combining genetic algorithm and analytical target cascading.The first is an improved bottom-up fast non-dominated algorithm based on the NSGA-Ⅱ algorithm.The second is a top-down algorithm combined with the analytical target cascading method.After the code is written by Python and DOcplex,the simulation experiments are carried out by using the data from the standard project database PSPLIB and OR ＆ S problem library.The scientificity of the proposed model and algorithm are verified by analyzing and comparing the simulation results.Through the hierarchical scheduling of each project and program,the proposed method realizes the unification of project individual optimization and program global optimization,and provides an effective integrated scheduling method for program managers at different levels.This will help to promote the integration of multi-project scheduling theory and program management theory systems and lay the foundation for the application of multi-project scheduling theory.