Research On The Resource-Constrained Project Scheduling Problem And Disruption Management For Activity Disruption

The resource-constrained project scheduling problem studies project scheduling problem from the perspective of resource-constrained. It is to schedule the duration and resource requirements of activities with precedence and resource constraints, while minimizing project duration or other managerial objectives. The problem is typically NP-hard problem and is hard to solve. Projects are often performed when occurring some disruptions, which make the original schedule disturbed and even infeasible. It is also a difficulty for a project scheduling problem that how to deal with disruption events effectively and minimize the deviation from an original plan. Disruption Management is a method that resolves interference incidents in real-time making local optimization and adjustment. When solving the recovery problem, the original objective and new recovery constrains must be taken into account. The new objective is not only to minimize the deviation from the original schedule but also to minimize the recovery cost, then emerging a new schedule in time. In this paper, a new way for the activity disruption of the multi-mode resource-constrained project scheduling problem is presented. The main contents of paper are as follows:(1)The paper introduces an improved genetic algorithm approach for solving MRCPSP, using dual-activity list for encoding and adopting crossover and mutation to obtain the optimal schedule.(2)Based on the analysis of disturbance in MRCPSP, the paper proposes a method of disturbance identification and measurement with the methodology of disruption management. According to relative factors such as client, duration and recovery cost, functions and disruption model are constructed. On the basis of disturbance recovery strategies, a new plan can be generated.(3)In the paper, a benchmark instance of PSPLIB is selected to illustrate the solution procedure of GA. When an activity disruption event appears, a new schedule is obtained in real-time with disruption management. The improved genetic algorithm and disruption management model presented in the paper are tested feasibly and effectively.The disturbance identification and measurement from the perspectives of customer satisfaction, project duration and recovery cost for MRCPSP disruption, and the disturbance recovery strategies presented is an important breakthrough compared to the previous studies. The model and solution of disruption management can effectively reduce the impact of disturbance on the project system, and generate an optimal recovery schedule in a timely manner. The presented disruption management model and solution for MRCPSP has a strong practical significance in a project and also provides new idea for research in future.