Using Neural Networks to construct the Estimating models of the Productivity for household connection.
碩士 === 國立高雄第一科技大學 === 營建工程所 === 97 === Estimating productivity of household connection activities is according to a multiple regression model nowadays. However, there are two shortcomings as follow: (1) The accuracy of linear equation does not perform algorithms as the non-linear one. (2) Using re...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Published: |
2009
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/77203486706296115127 |
| Summary: | 碩士 === 國立高雄第一科技大學 === 營建工程所 === 97 === Estimating productivity of household connection activities is according to a multiple regression model nowadays. However, there are two shortcomings as follow: (1) The accuracy of linear equation does not perform algorithms as the non-linear one. (2) Using reaching 12 linear equations to estimate productivity of household connection is too complicated. In short, the purpose of this research is to develop one kind of neural network model in order to estimate productivity of household connection.
This research describes a statistical model developed to forecast the productivity of household connection activities. The model is a non-linear multiple regression model, developed by observed and interviewed information on-site. Model coefficients in regard to influencing factors, and have set up six kinds of models of household connection productivity.
The productivity of each type of household connection are: (1) Productivity of front-lane excavation is 0.95( wh/m). (2) Productivity of back-lane excavation is 2.48 (wh/m3). (3) Productivity of front-lane connection-pipe is 1.89 (wh/unit). (4) Productivity of back-lane connection-pipe is 0.61(wh/unit). (5) Productivity of front-lane recovery is 0.10 (wh/m). (6) Productivity of back-lane recovery is 0.17 (wh/m).
|
|---|