Lagrange Neural Network for Solving CSP Which Has Objective Function
|
Accession number;05A0661025
|
| Title;Lagrange Neural Network for Solving CSP Which Has Objective Function |
| Author;
NAKANO TAKAHIRO
(Graduate School of Life Sci. and Systems Engineering, Kyushu Inst. Technol., JPN)
NAGAMATSU MASAHIRO
(Graduate School of Life Sci. and Systems Engineering, Kyushu Inst. Technol., JPN)
|
Journal Title;Proceedings of the Annual Conference on JSAI (CD-ROM)
|
Journal Code:X0580B
|
ISSN:
|
|
VOL.19th;NO.;PAGE.1A1-01(2005)
|
| Figure&Table&Reference;FIG.2, REF.4 |
| Pub. Country;Japan |
| Language;Japanese |
| Abstract;We proposed a Lagrange Neural Network called LPPH-CSP to solve the CSP. This method is never trapped by any point which is not a solution of the CSP. From experimental results of the LPPH-CSP, we confirmed that our method is as efficient as the GENET which is a famous CSP solver. In addition, unlike other conventional CSP solver, our method is a continuous-valued method and it can update all variables simultaneously. Therefore, we can expect the speed-up of the LPPH-CSP if it is implemented by the hardware such as FPGA. In this paper, we extend the LPPH-CSP to treat the linear inequality constraints. By using this type of constraint, we can represent various practical problems more briefly. In this paper, we also define the CSP which has an objective function, and we extend the LPPH-CSP to solve this problem. In the experiment, we apply our method and the OPBDP to the WLP and compare the effectiveness. (author abst.) |
|
|
|
Related Articles;
|
|
