查詢結果分析
來源資料
頁籤選單縮合
題 名 | A Study on Automatic Incident Detection--Application of Fuzzy Sets and Neural Networks=應用模糊集與類神經網路理論於自動事件偵測之研究 |
---|---|
作 者 | 陳惠國; 蔣偉寧; 張美香; 曾信忠; | 書刊名 | 模糊系統學刊 |
卷 期 | 3:1 1997.01[民86.01] |
頁 次 | 頁91-108 |
分類號 | 557.3 |
關鍵詞 | 自動事件偵測; 模糊型態識別; 模糊系統鑑定; 後饋型類神經網路; Automatic incident detection; Fuzzy pattern recognition; Fuzzy system identification; Back propagation neural network; |
語 文 | 英文(English) |
中文摘要 | 自動事件偵測於高速公路管理系統中扮演舉足輕重的角色。近年來的研究大多屬 於明確型的演算法。其中以加州法( TSC7 )最為廣泛使用,也最常為新發展的自動事件偵 測演算法的比較基準。本研究發展三個意外事件自動偵測演算法:型態識別、系統鑑定、後 饋型類神經網路。採用中山高速公路離線事件資料,分別測試所提出演算法,其事件偵測率 均較加州演算法( TSC7 )高。證實此三個演算法似乎適用於高速公路之意外事件自動偵測 ,進而預期此三個演算法可應用於真實世界中,處理更複雜之線上資料。 |
英文摘要 | Automatic incident detection in general plays the key role in the Freeway Incident Management Systems (FIMS). Not until recent years, most articles appeared in the literature are deterministic in nature. Among those deterministic types of algorithms, the Technology Services Corporation 7 (TSC7) algorithm is the most widely used and usually deemed as the benchmark for the newly developed algorithms to compare with. In this paper, three newly developed algorithms are accommodated and applied to the automatic incident detection (AID); i.e., fuzzy pattern recognition (FPR), fuzzy system identification (FSI), and back propagation neural network (BPNN) algorithms. The three algorithms show their superiority over the TSC7 mainly in terms of the detection rate, according to our off-line case study on the Sun Yat-Sen National Freeway. Though the results are somewhat satisfactory, the application of the proposed algorithms to the real world is subject to more comprehensive on-line tests. |
本系統中英文摘要資訊取自各篇刊載內容。