Heuristic driven genetic algorithm for job-shop scheduling
Chang, Sau Leng.
Date of Issue2000
School of Electrical and Electronic Engineering
In this project, a local search algorithm has been implemented on a job-shop scheduling program[l] given at the initial stage of the project. The program was tested on 35 JSS benchmarks. The results show a range of performance that is 10% to 15% from the known optimum. Besides proposing and implementing the local search algorithm, the original JSS program is also ported to run on Win32 platforms(Win95 and NT). Its memory handling functions are modified to dynamically handle job-shop problems of any sizes from 5x5 up to 30x30. A Windows based graphical user interface(GUI) is written so that the user is able to modify benchmark specific parameters on the GUI and run several benchmarks in batch mode.
DRNTU::Engineering::Electrical and electronic engineering