Research On Maximizing Net Present Value Of Resource Constrained Project Scheduling Problem

Previous research regarded shortest time limit or resource equilibrium as theobjective of the resource constrained project scheduling problem.While with thehigh interest rates and high costs in present, obtaining the maximum net presentvalue may be more reasonable in project.Resource constrained project scheduling problem with discounted cashflows has been studied in this dissertion.Important factors are considered in models,such as mode selection, payment scheduling,mode of payment, dynamic interestrates, new accouting method of renewable resource.Since resourceconstrained project schedulingproblem is the NP-hard problem, the research designefficient heuristic algorithm for it, use a large number of instances to testthe performance of the algorithm, provide the method to support decision making.This main work and innovations of the dissertation include:(1)Aiming at the single-mode resource-constrained project schedulingproblem with discounted cash flows,a heuristic optimization procedure isdeveloped to maximize the net present value of a project. Bi-directional generationscheme and a recursive forward/backward improvement method are embed inscatter search framework to expand the scope of the search spread and improve thequality of the solution. The algorithm is tested by data sets to illustrate theimprovement of results.(2) Multi-mode is introduced in resource constrained project schedulingproblem model. Since the mechanism of Ant colony Optimization enables the useof domain-based heuristics to accelerate the search,Eight domain-based heuristicsare designed to consider the factors of time, cost resources, and precedencerelations. Among these heuristics, the hybrid heuristic that combines differentfactors together performs well.(3)Payment scheduling is futher considered in multi-mode resourceconstrained project scheduling problem model. Four payment models arepresented,including: lump sum payment at the terminal event, payments atprespecifed event nodes, payments at equal time interval and progresspayments.Genetic Algorithm is employed, it uses a special crossover operator which can exploit the multi-component nature of the problem. The example aretested under four different payment scheduling with the GA approach.(4)Modes of payment are considered in multi-mode resource constrainedproject scheduling problem model, such as bill and the cash. How the bill effect thenet present value are also researched under the different cash ratio and interstrate.And multi-pass heuristic algorithm are proposed which consists of two-stagetabu search.Computational experiments are conducted and proved the effect of thealgorithm.