柔性鋪面現況服務力指標與預測模式建立之研究; The Study on Establishing the Present Serviceability Index and Predictive Model of Flexible Pavement

NCUIR > Engineering College > Graduate Institute of Civil Engineering > Electronic Thesis & Dissertation > Item 987654321/470

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/470

Title:	柔性鋪面現況服務力指標與預測模式建立之研究;The Study on Establishing the Present Serviceability Index and Predictive Model of Flexible Pavement
Authors:	洪境聰;ching-Tsung Hung
Contributors:	土木工程研究所
Keywords:	鋪面現況服務力指標;路面破損調查;小組評分;模糊積分;模糊迴歸;類神經網路;Present serviceability;Pavement distress survey;Panel rating;Fuzzy integral;Fuzzy regression;Neural network
Date:	2000-06-30
Issue Date:	2009-09-18 17:06:05 (UTC+8)
Publisher:	國立中央大學圖書館
Abstract:	路面管理系統最主要的目的在提供使用者優良舒適的道路，但是道路的狀況十分複雜，要如何對於道路之行駛品質進行判斷，則鋪面現況服務力指標的建立為路面管理系統重要的一環。另路面惡化速率的預測模式為路面管理系統進行生命週期成本分析不可或缺的項目，惟國內對於路面監測缺乏長期之資料收集，無法建立本土化之實用指標與預測模式，故對於養護作業之排程未達到最有效的利用。本研究之目的在回顧國內外進行服務力指標建立方法之相關文獻，針對交通部公路局一區工程處所管轄之省、縣道之部分路段進行鋪面現況服務力評分試驗、糙度調查、車轍深度量測與鋪面損壞調查等研究，並考量鋪面變數其成因錯綜複雜，導致各參數因交互影響無法有效表現與鋪面現況服務力評分間的相關性，故採用模糊積分方法建立無交互作用之綜合評估值，另考量調查資料之模糊性，若採用傳統迴歸模型並不符合現實狀況，故另行考量模糊迴歸模型尋找主觀資料與客觀資料的關係，以建立鋪面現況服務力指標。並以類神經網路模型建立預測模式供鋪面工程師瞭解路面惡化的情形，作為養護決策之參考。本研究依現場試驗與調查結果，經資料整理與統計分析得到省縣道現況服務力指標，與 PSI可容忍極限值為2.19。另以類神經網路進行預測模式之建立，可得到良好的預測結果。 The primary purpose of pavement management system is to provide the good and comfortable roads with users. However, the actual conditions of roads are very complex. In order to evaluate the ride quality of roads, it plays an important role to establish the present serviceability index in pavement management system. Besides, the predictive model of pavement deterioration is also the indispensable element for analyzing the life cycle cost in pavement management system. However, it is very difficult to establish the practical index and predictive model for lack of long-term data collected in Taiwan. And then, the maintenance activities cannot be planned effectively. According to the on-site survey and investigation, the present serviceability index for provincial and county roads can be derived by statistical analysis. And, the critical value of PSI is 2.19. In addition, the good predictive results also can be obtained from the predictive model established by neural network.
Appears in Collections:	[Graduate Institute of Civil Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Size	Format

社群 sharing

Loading...