Usage of Partial Genome Fitness Evaluation Mechanism to Get Faster Results in Genetic Algorithms

Abstract

The aim of the research presented in this paper is to test the efficiency of the proposed genetic algorithm in the fast or real-time optimizations. Proposed partial genome fitness evaluation mechanism was tested on the wireless network topology optimization and job shop scheduling problem (JSSP). Wireless network topology optimization requires frequent and very fast (sometimes real time) adjustments. Job shop scheduling problem is one of the most general and difficult of all traditional scheduling problems. Search based on traditional genetic algorithms usually are not suitable in solving these problems. Long computational time and high memory usage if a large population and/or a large number of generations are used, on the other hand, a larger population and a larger number of generations usually provide better results. In comparison with two other similar and commonly used genetic the proposed algorithms have significant improvements in speed and some case solution quality.