交控任務邏輯及績效評估之研究

Abstract

近年來都市交通需求漸增與變化不定，在有限的都市空間與道路容量下，號誌控制改善可為最為經濟而有效又可行之方式。國內外漸有許多學者投入即時控制方式之號誌控制策略研究，過去由於傳統的交通號誌控制器受技術與硬體上限制，在控制及修改上都缺乏彈性，而無法發展，但近年來微電腦迅速發展，電腦技術得應用於號誌控制器，民國九十年交通部之委託案「台灣地區智慧型運輸系統實驗城規劃計劃」在國內樹立第一個ITS完整架構典範，並提出以交通管理資訊中心(TMIC)、交通資訊站(TCIS)及車載電腦(OBC)為核心之三層結構構想，以實現ITS之整體運作，其中TCIS將成為各ITS子系統應用面之中介者，對於交控系統而言，可成為智慧型號誌控制器，作為交控中心與路端設備之中介控管設備，控制策略可由智慧型控制器自行決策與執行，達到即時控制、反應交通變化之最佳控制。

雖然目前國內外現存多種號誌控制策略邏輯，但有些並不適合使用於即時控制上，且尚未有量化之客觀標準，因此邏輯本身以及各種控制策略之適用範圍與門檻，成為控制策略自動化之障礙。本研究採用觸動控制、動態計算以及適應性控制三種策略，在適當修改後，撰寫為可供號誌控制器執行之應用程式，並基於現有不同功能設計之控制邏輯，設計一模擬與評估績效平台，能在同樣的車流狀況下，模擬各種控制策略之績效值，最後訂出各種控制策略之切換門檻。此外，各種控制策略所需硬體設計與數量均不相同，因此除了考量績效最佳化的策略門檻外，亦考量效益成本與績效之間的影響，訂出成本效益考量下最佳之策略門檻，使其可應用在實務上。

As traffic volumes and driving conditions change throughout the day, different traffic signal control strategies and criteria are applicable. In order to reduce traffic jam and improve transportation efficiency, signal control is the fast and effective way and it is also the foundation of developing intelligent area signal controller. Intelligent area signal controller can be considered a traffic information center to share some tasks of traffic management information center (TMIC), and decision of signal control strategies is one of these tasks. To achieve the objective of real-time response, the signal control strategy must be simple and be carried out on microprocessors, so the program must be run on different platforms. (In this thesis, JAVA is used to be the programming language.) This study attempts to determine the best suitable signal control strategy which includes semi-actuated control, dynamic computation control, adaptive control and pre-time control produced by Synchro under different traffic conditions by micro-simulation of traffic flow. We take isolated intersection as an example, and the phase pattern is simple two-phase pattern to simplify the question. All chosen control strategies would be written application programs which can be implemented after they are installed in signal controllers independently. This study adopts average stopping delay as the performance index, and divides traffic conditions into nine levels in accordance with degree of saturation to simulate. The result of simulation could be for reference of signal control strategy implements in the future.