Multi mode resource constrained multi stochastic project scheduling

Nowadays, the business organizations in many sectors like aviation, construction and defense industry carry out multiple projects simultaneously by taking into account the usage of scarce resources. By adapting Resource Constrained Project Scheduling Problem (RCPSP) in a multi project environment, a research topic in the literature which is called Resource Constrained Multi-Project Scheduling Problem (RCMPSP) is created. The scheduling of activities in a multi project environment by considering precedence relationship and resource constraints is an important business management problem. Besides, business organizations have many different machines and equipment to do the same jobs. Since these machines and equipment can have new/old/various technologies, the same jobs can be completed with different cost rate and in various time frames. In real life, the duration of the activities in a project is not certain. Therefore, in this thesis, Multi-mode Resource Constrained Multi-project Scheduling Problem (MMRCMPSP) with stochastic activity duration is considered as a special case of RCPSP and modeled as Chance Constrained Programming (CCP). The goal is to determine the sequence and the schedule of all activities by keeping precedence relations and providing resource constraints under resource dedication (RD) policy so that the tardiness for each project is minimized. Since MMRCMPSP is NP–hard, a problem-specific Simulated Annealing (SA) algorithm is proposed to provide solutions for large-scale problem with no solution. SA algorithm is a stochastic search method that provides good solutions for optimization problems. Shortest Processing Time rule which is one of the sequencing rule based heuristic is suggested to generate initial solution. Besides, Combinatorial Auction (CA) method is used to solve RD problem. The effectiveness of our solution approaches is analyzed by computational analyses and the results are evaluated.