Automatic Synthesis of Dynamic Systems Based on Hungarian Algorithm and Genetic Programming
| Autor: | Wei Jie Pan, Yong Zhong, Guanci Yang, Shaobo Li |
|---|---|
| Rok vydání: | 2011 |
| Předmět: |
Scheme (programming language)
education.field_of_study Computer science business.industry Mechanical Engineering Population Genetic programming Base (topology) Machine learning computer.software_genre Mechanics of Materials Hungarian algorithm General Materials Science Artificial intelligence business Design methods education computer Bond graph Eigenvalues and eigenvectors computer.programming_language |
| Zdroj: | Key Engineering Materials. :160-163 |
| ISSN: | 1662-9795 |
| DOI: | 10.4028/www.scientific.net/kem.467-469.160 |
| Popis: | Genetic programming is an effective way to generate design candidates in an open-ended, but statistically structured, manner. A critical aspect of the procedure is a fitness measure, which guides candidate designs toward an optimal scheme in reasonable time. This paper has suggested a new definition of fitness base on Hungarian algorithm for automatically synthesizing designs for multi-domain, lumped parameter dynamic systems, and uses a type of embryo bond graph model with three modifiable sites to initialize population. Although the experiments run to date are not sufficient to allow making strong statistical assertions, it shows that the search capability of genetic programming combining Hungarian algorithm is good enough to make feasible automated design methodology proposed here for multi-domain systems. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|