| Summary: | 碩士 === 長庚大學 === 資訊管理學系 === 99 === Open shop scheduling is a problem of great complexity. Choosing effective approaches to solve this problem becomes an important issue. In this research, operations need to be scheduled with transportation times .
The objective is minimizing the makespan.
First, we construct the mathematical model for open shop scheduling with transportation times. Then, optimal solutions are obtained by using CPLEX software. Finally, this research applied a genetic algorithm with three crossover operators and a simulated annealing approach with three neighborhood generation mechanisms to find near optimums and computational times and compared the differences between various approaches. Processing times of experimental instances use Taillard's problems. We generate transportation times based on parameters used in Taillard's problems. There are six different types of problems.
Computational results indicate that GA2, GA3 and SA2 are better and genetic algorithm is better than simulated annealing approach in the makespan. In the computational time, GA2 and GA3 are much better than GA1 for large scale instances. With the growing of instances' scale, various approaches take a lot of computational times, but simulated
annealing approach is much better than genetic algorithm.
|
|---|
