Solving shortest path problem using gravitational search algorithm and neural networks
For many years, scientists have aware the importance of conducting research related to actual problems, and Shortest Path Problem (SPP) is one of such examples. SPP is meant to find the shortest path between two given cities or nodes. The travelled distance in such a path obviously depends on the or...
Saved in:
| 主要作者: | |
|---|---|
| 格式: | Thesis |
| 語言: | English |
| 出版: |
2010
|
| 主題: | |
| 在線閱讀: | http://eprints.utm.my/id/eprint/11402/5/SamanSinaieMFSKSM2010.pdf |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 總結: | For many years, scientists have aware the importance of conducting research related to actual problems, and Shortest Path Problem (SPP) is one of such examples. SPP is meant to find the shortest path between two given cities or nodes. The travelled distance in such a path obviously depends on the order in which the cities are visited. Hence, it is the problem of finding an optimal ordering of the cities. Therefore, SPP is commonly known as the combinatorial optimization problems. In this study, a hybrid method is proposed that can solve the SPP accurately and fast without being trapped in a local optimum during searching. Gravitational Search Algorithm (GSA), a new optimization algorithm is applied to solve the above problem. In this scenario, decision making is provided by the Neural Network within a short amount of time. The results illustrate that with the precision of the GSA and the relatively high speed of the Neural Network, a very efficient method is obtained accordingly. |
|---|