A Neurogenetic approach for the resource-constrained project scheduling problem

  • Authors:

  • Anurag Agarwal;Selcuk Colak;Selcuk Erenguc

  • Affiliations:

  • Department of Information Systems and Decision Sciences, College of Business, University of South Florida, Sarasota, FL 34243, USA;Department of Business, College of Economics and Administrative Sciences, Cukurova University, Adana, Turkey;Department of Information Systems and Operations Management, Warrington College of Business Administration, University of Florida, Gainesville, FL 32611-7164, USA

  • Venue:
  • Computers and Operations Research
  • Year:
  • 2011

Quantified Score

Hi-index 0.01

Visualization

Abstract

A variety of metaheuristic approaches have emerged in recent years for solving the resource-constrained project scheduling problem (RCPSP), a well-known NP-hard problem in scheduling. In this paper, we propose a Neurogenetic approach which is a hybrid of genetic algorithms (GA) and neural-network (NN) approaches. In this hybrid approach the search process relies on GA iterations for global search and on NN iterations for local search. The GA and NN search iterations are interleaved in a manner that allows NN to pick the best solution thus far from the GA pool and perform an intensification search in the solution's local neighborhood. Similarly, good solutions obtained by NN search are included in the GA population for further search using the GA iterations. Although both GA and NN approaches, independently give good solutions, we found that the hybrid approach gives better solutions than either approach independently for the same number of shared iterations. We demonstrate the effectiveness of this approach empirically on the standard benchmark problems of size J30, J60, J90 and J120 from PSPLIB.