標題: | Using a hybrid approach to optimize experimental network design for aquifer parameter identification |
作者: | Chang, Liang-Cheng Chu, Hone-Jay Lin, Yu-Pin Chen, Yu-Wen 土木工程學系 Department of Civil Engineering |
關鍵字: | Groundwater;Experimental design;Genetic algorithm |
公開日期: | 1-十月-2010 |
摘要: | This research develops an optimum design model of groundwater network using genetic algorithm (GA) and modified Newton approach, based on the experimental design conception. The goal of experiment design is to minimize parameter uncertainty, represented by the covariance matrix determinant of estimated parameters. The design problem is constrained by a specified cost and solved by GA and a parameter identification model. The latter estimates optimum parameter value and its associated sensitivity matrices. The general problem is simplified into two classes of network design problems: an observation network design problem and a pumping network design problem. Results explore the relationship between the experimental design and the physical processes. The proposed model provides an alternative to solve optimization problems for groundwater experimental design. |
URI: | http://dx.doi.org/10.1007/s10661-009-1157-5 http://hdl.handle.net/11536/32155 |
ISSN: | 0167-6369 |
DOI: | 10.1007/s10661-009-1157-5 |
期刊: | ENVIRONMENTAL MONITORING AND ASSESSMENT |
Volume: | 169 |
Issue: | 1-4 |
起始頁: | 133 |
結束頁: | 142 |
顯示於類別: | 期刊論文 |