Resource-constrained Multi-project Scheduling And Controlling Under Uncertainty

The continuous development of social economy has spawned a large number of project construction needs,and multiple project management has become the norm.However,due to the large management interface and the finite resource contention between subprojects,multi-project scheduling becomes a difficult problem in research and practice.In addition,due to the acceleration of urbanization process,the project environment becomes more and more complicated and adds a lot of uncertainty to project practice.The existence of these uncertainties further increases the difficulty of project scheduling,leads to engineering delays and cost overruns often occur.Multi-project scheduling problem in uncertain environment is an urgent problem to be solved in the field of project scheduling.To solve this problem,the author builds a "plan-control" two-stage model using literature analysis method,qualitative analysis method and quantitative analysis method.Firstly,the author constructs a resource-constrained multi-project scheduling model based on genetic algorithm.the selection operator is the combination of elitist and linear ranking,this method ensures that the best individual will not be excluded accidentally,and avoids algorithm convergence prematurely;the crossover operator is precedence preservative crossover,this method ensures that the offspring still satisfies the constraint of the tight front relation,and solves the non-line solution problem.Secondly,the author constructs a multi-project controlling model under uncertainty based on system dynamics.This model contains four subsystems:process,performance,resource and duration,These subsystems simulates the important feedback structure in project practice.It is worth mentioning that process subsystem contains precedence relation and resource constraint,this model achieves the effective combination between project scheduling and project control in this way.Lastly,simulation and analysis shows that the "program control" two-stage model can effectively solve the resource-constrained multi-project scheduling and controlling problem.First of all,the author constructs a multi-project examples that contains 3 subprojects and 36 real tasks for simulation and analysis.And then uses the genetic algorithm to find the optimal scheduling plan under different circumstances:without resource constraint,with resource constraint and with project weights.At last,the author simulates the multi-project control under uncertainty using system dynamics model.Simulation results show that "plan-control" two stage model is able to deal with resource-constrained multi-project scheduling under uncertainty effectively.