Net Present Value Maximization For The Multi-mode Resource-Constrained Project Scheduling Problem Under Uncertainty

With the development of the economy globalization and the fierce competition in marketing,the project schedule management technology began to develop and become mature gradually.The traditional project schedule management techniques and methods which consider the project can not adapt to the uncertain environment.Therefore,this paper studies the multi-mode resource-constrained project scheduling problem under uncertain environment.The main research contents of this paper are as follows:Firstly,aiming at the problem of maximizing the net present value of project,the net present value optimization model of multi-mode resource-constrained project scheduling under uncertain environment is proposed.Then,according to the model properties and computational complexity,a double-loop genetic algorithm is designed to solve the problem.Subsequently,the project generator Pro Gen is used to generate the data sets of four different project scales,and two genetic algorithms are proposed as the comparison algorithm.Then the three algorithms are tested and compared.Experimental results show that the proposed two-loop genetic algorithm performs well in solving efficiency and stability,and can help the contractor to maximize the benefits in the shortest time.Finally,the sensitivity analysis of the key parameters is carried out.The net present value of the project is proportional to the discount coefficient,and the complexity of the network structure has different effects on the net present value of projects of different sizes.Secondly,aiming at the problem of minimizing the conditional net present value at risk of project,a conditional net present value at risk optimization model for multi-mode resource-constrained project scheduling under uncertain environment is proposed.Additionally,a genetic algorithm with local search strategy is designed,which can efficiently select the execution mode with the maximum net cash flow discount value for each activity,and obtain a satisfactory scheduling strategy.The four data sets of the net present value optimization problem are used for numerical experiments,and the results of the proposed genetic algorithm with local search strategy and the three genetic algorithms mentioned above are compared to verify the superiority and effectiveness of the genetic algorithm with local search strategy,that is,it can help the contractor reduce the potential loss of the project net present value to the greatest extent.At the same time,the algorithm can get higher returns.In addition,through the sensitivity analysis of key parameters,it is found that the conditional net present value at risk is proportional to the discount coefficient and confidence level.Thirdly,the proposed net present value maximization model and conditional net present value at risk minimization model for multi-mode resource-constrained projects under uncertain environment are applied to engineering examples,and the net present value is optimized by double-loop cyclic genetic algorithm and the conditional net present value at risk is optimized by genetic algorithm with local search strategy.The scheduling strategies obtained by the two algorithms and the net present value and construction period under different scenarios are compared,and the effects of key parameters on the two objective functions are analyzed.eventually,the proposed models and algorithms are further proved to be of practical value by providing some decision support suggestions for the progress management and risk control of actual engineering projects.