Resource Constrained Project Scheduling using evolution strategies
- 148 Downloads
To confront the Resource Constrained Project Scheduling Problem (RCPSP), metaheuristics have been proved very good alternatives, especially for large complicated projects. In this class of algorithms, Evolutionary Computation has recently gained much attention, with most important representative the Genetic Algorithms. Following the mainstream, we stress our efforts on another evolutionary algorithm, the Evolution Strategies (ES). The application of ES takes place under two discrete solution encodings; one works on vectors of priority values and the other is based on convex combinations of priority rules. The analysis of the results, produced from tests on the PSPLIB, inspired the development of two extended algorithms. The first extension assumes that ES work on vectors of priority values but the underlying evolutionary operators are modified so as to allow a fast reordering of activities. The second extension concerns the construction of a novel solution encoding which combines the priority values and the convex combination of priority rules. Both proposals indicate a far better performance when compared with genetic algorithms, hence, open a new research direction in the domain of project scheduling with evolutionary algorithms.
KeywordsResource Constrained Project Scheduling Metaheuristics Evolution Strategies
Unable to display preview. Download preview PDF.
- Alvarez-Valdes R. and Tamarit J.M. (1989). Heuristic Algorithms for Resource Constrained Project Scheduling: A Review and an Empirical Analysis. in Advances in Project Scheduling, (R. Slowinski and J. Weglarz ed.) Elsevier, Amsterdam, 113–134.Google Scholar
- Back, T. (1996). Evolutionary Algorithms in Theory and Practice. Oxford University Press, New York.Google Scholar
- Blickle, T. (2000). Tournament Selection, in Evolutionary Computation: Basic Algorithms and Operations vol. 1, (Back, T., Fogel, D.B. and Michalewicz, T., ed.) Institute of Physics Publishing, Bristol, 181–186.Google Scholar
- Eshelman, L.J. (2000). Genetic Algorithms, in Evolutionary Computation: Basic Algorithms and Operations vol. 1, (Back, T., Fogel, D.B. and Michalewicz, T., ed.) Institute of Physics Publishing, Bristol, 64–80.Google Scholar
- Grefenstatte, J. (2000). Proportional Selection and Sampling Algorithms, in Evolutionary Computation: Basic Algorithms and Operations vol. 1, (Back, T., Fogel, D.B. and Michalewicz, T., ed.) Inst of Physics Publishing, Bristol, 172–180.Google Scholar
- Kolisch, R. and Hartmann, S. (1999). Heuristics Algorithms for Solving the Resource Constrained Project Scheduling Problem: Classification and Computational Analysis, in Project Scheduling: Recent Models, Algorithms and Applications, (J. Weglarz, ed.) Kluwer, Amsterdam, 147–178.Google Scholar
- Rocha, M., Vileda, C., Cortez, P. and Neves, J. (2000). Viewing Scheduling Problems through Genetic and Evolutionary Algorithms. in Lecture Notes in Computer Science vol. 1800, (Rolim, J. et al., ed.) Springer Verlag, Berlin, 612–619.Google Scholar
- Schwefel, H.P. and Rudolph, G. (1995). Contemporary Evolution Strategies. 3rd International Conference on Artificial Life, 893–907.Google Scholar