A graphical processing unit‐based parallel hybrid genetic algorithm for resource‐constrained multi‐project scheduling problem

Author(s):  
Furkan Uysal ◽  
Rifat Sonmez ◽  
Selcuk Kursat Isleyen
Author(s):  
Wenjian Liu ◽  
Jinghua Li

In multi-project environment, multiple projects share and compete for the limited resources to achieve their own goals. Besides resource constraints, there exist precedence constraints among activities within each project. This paper presents a hybrid genetic algorithm to solve the resource-constrained multi-project scheduling problem (RCMPSP), which is well known NP-hard problem. Objectives described in this paper are to minimize total project time of multiple projects. The chromosome representation of the problem is based on activity lists. The proposed algorithm was operated in two phases. In the first phase, the feasible schedules are constructed as the initialization of the algorithm by permutation based simulation and priority rules. In the second phase, this feasible schedule was optimized by genetic algorithm, thus a better approximate solution was obtained. Finally, after comparing several different algorithms, the validity of proposed algorithm is shown by a practical example.


2013 ◽  
Vol 483 ◽  
pp. 607-610 ◽  
Author(s):  
Chun Jie Zhong ◽  
Ying Yu ◽  
Yun Lang Jia

A resource-constrained project scheduling problem with stochastic resource-dependent activity durations is presented in this paper,and the two-point method is employed to simulate the uncertain property.Furthermore a genetic algorithm combined with this method is provided to solve the problem. Compared with the results from the genetic with Monte Carlo simulation, the proposed method is verified to be effective and more efficient.


Sign in / Sign up

Export Citation Format

Share Document