Name
Abstract | ||
|---|---|---|
Scheduling trains is a complex problem that until recently has been a slow and inflexible process requiring the precise knowledge of experts to perform, often needing constant recalculation in response to real world changes such as train delays, extra trains and so forth. This paper derails the application of a Genetic Algorithm (GA) to this problem. Using a two dimensional representation scheme, both mutation and crossover are used to guide the GA to a valid solution. Good results are presented for scheduling multiple trains on a real world large complex railway. An analysis of the effects of varying a number of CA parameters such as pool size, percentage of mutation, percentage of crossover and mutation sizes is presented to give some insight into how to optimise the performance of the algorithm and to give an idea of the geography of the solution space. |
Year | Venue | Field |
|---|---|---|
1997 | PROGRESS IN CONNECTIONIST-BASED INFORMATION SYSTEMS, VOLS 1 AND 2 | Computer science,Scheduling (computing),Artificial intelligence,Train,Machine learning,Genetic algorithm |
DocType | Citations | PageRank |
Conference | 0 | 0.34 |
References | Authors | |
0 | 2 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Ariel E. Bud | 1 | 0 | 0.34 |
| Ann E. Nicholson | 2 | 692 | 88.01 |