초록

용어보기논문에서 용어와 풀이말을 자동 추출한 결과로, 시범 서비스 중입니다. AI-Helper

Abstract ▼ AI-Helper

This paper shows that a simple hybrid of genetic algorithm and local heuristic search provides a powerful means for solving a complex scheduling problem, by exploiting the global perspective of the genetic algorithm and the rapid convergence of the heuristic search. Unlike most other approaches whe...

This paper shows that a simple hybrid of genetic algorithm and local heuristic search provides a powerful means for solving a complex scheduling problem, by exploiting the global perspective of the genetic algorithm and the rapid convergence of the heuristic search. Unlike most other approaches where the hybridization is built into the genetic operations, our hybridization is built into the framework of our iterative improvement search method which works with hierarchical problem decomposition. The hierarchical problem decomposition enhances efficiency by significantly reducing the search space, and thus compromising the optimality of the solution. However, the possible suboptimality induced by such a search space reduction can be overcome by the iterative improvement search which is in fact a reactive process of repeated schedule modification using the genetic algorithm hybrid as an internal search engine. This additional iteration demands little extra processing time because the genetic algorithm hybrid takes only a few iterations to converge to a schedule of high quality. Test results using real world data show that our approach clearly outperforms various other search schemes.