本論文係建構一個路堤填築工程可施工時間推估模式。首先進行文獻探討目前之相關文獻皆僅針對天候因素對工程可施工時間進行研究，經由對路堤填築工程之施工特性分析，發現影響路堤填築工程可否施工之重要因素為土壤種類及降雨量。同時，由工地實地訪談，路堤填築施工可簡化為整日施工及整日不可施工。本論文將路堤填築可施工時間推估之問題，視為一分類問題，而採用倒傳遞類神經網路來建構一路堤填築可施工時間推估模式。最後以南部第二高速公路路堤填築工程加以測試，以驗證本模式之適用性。

The aim of this paper is to build an estimation model in working time for embankment works. First, as for the related literatures discussing in this issue, only the weather factor is considered in these studies for working time. In the present study, after analyzing the characteristic of embankment works, we attempt to explore whether the types of the soil and rainfall are also the important factors for the working time of the works. At the same time, by investigating the spot coverage through designed questionnaire, we concluded that working can be briefly divided into working all day and non-work all day.
In this thesis, estimation in working of time embankment works is regarded as an issue of the classification. Thus, the back-propagation neural network will be employed to build an estimation model in working time for embankment works. Finally, in order to verify the suitability of this model, the test of the embankment works of the second highway in the south area of Taiwan will also be executed in this study.

第一章、 緒論
1.1 研究動機………….……………………………………………….…1
1.2 研究目的………….………………………………………………….1
1.3 研究範圍與內容….………………………………………………….2
1.4 研究流程………….………………………………………………….2
第二章、 文獻探討
2.1 作業可施工時間之推估………….………………………………….4
2.2 倒傳遞類神經網路…………….…………………………………….5
第三章、 路堤填築工程可施工時間推估模式之建構
3.1 路堤填築工程施工特性之分析………………………………...….10
3.2 模式建構方法之選擇……………………………………………....14
3.3 類神經網路架構……………….…………………………………...15
第四章、 模式之測試及應用
4.1 南二高路堤填築工程可施工時間之推估…………………………23
4.2 台灣地區路堤填築工程可施工時間之推估………………………27
4.3 台灣地區路堤填築工程可施工時間變異之推估…….…………...29
第五章、 結論與建議
5.1 結論………………………………………………………………....33
5.2 建議…………………………………………………………………34
參考文獻………………………………………………………………………35
附錄A…………………………………………………………………….……36
附錄B………………………………………………………………….………41
附錄C………………………………………………………………….………53

1. Carr, R. I., Simulation of construction project duration, Journal of the Construction Division, CO2, pp.117-128, 1979.
2. AbouRizk, S. M. and Wales, R. J., Combined discrete-event/continuous simulation for project planning, Journal of Construction Engineering and Management, 123(1), pp.47-64, 1997.
3. Dawood, N., Estimating project and activity duration: a risk management approach using network analysis, Construction Management and Economics, 16, pp. 41-18, 1998.
4. Shi, J. J., A neural network based system for predicting earthmoving production, Construction Management and Economics, 17, pp.463-471, 1999.
5. J.H.M TAH and V.CARR, A proposal for construction project risk assessment using fuzzy logic, Construction Management and Economics,18, pp.491-500, 2000.
6. 郭斯傑，李岳能，「降雨因素影響下之工期預測」，中國土木水利工程學刊，第十卷，第四期，10,4,pp.743-751,1998。
7. 張順傑，「以模糊推論解析降雨對道路施工作業時間之影響」，國立台灣科技大學營建工程研究所，民國89年。
8. Zhang, G. and Hu, M. Y., Neural network forecasting of the British Pound/US Dollar exchange rate, Omega, 26, pp. 495-506, 1998.
9. 「施工標準規範」，交通部台灣區國道高速公路局，民國76年3月。
10. Braja M. Das., Introduction to soil mechanics, Ames :Iowa State University Press, 1979.