考量跟車及變換車道行為之高快速公路微觀車流模式研究

Abstract

台灣車輛持有率逐年增加，交通問題益顯複雜，更顯示出衍生駕駛行為不確定性(Uncertainty)相對提高。然而改善現有交通問題，在觀念與作法上已逐漸由增加道路容量的供給方式，轉變成有效利用現有資源與交通設施的管理措施。智慧型運輸系統（Intelligent Transportation Systems, ITS）乃當前世界各國對改善交通問題之有效措施。ITS發展影響所及非常廣，其中車流模型之建立，可作為交通管理者的決策分析工具之一，並可提供即時的交通路況資訊之服務。交通管理首重車流行為特性之瞭解，透過分析以瞭解與掌握車流行為特性，進而加以模組化，將可作為交通管理者的決策分析工具，亦可提供即時的交通路況資訊，為實現智慧型運輸系統所不可或缺的重要環節。因此微觀車流模式開發後有助於建立即時控制策略，其中跟車行為及變換車道行為是微觀車流模式重要部份之一。 本研究係以國內高快速公路小客車為研究對象，並基於微觀理念、近微觀理念、刺激－反應原理、跟車理論、行為門檻、駕駛行為理論（變換車道）等理論基礎，採用文獻評析法、攝影調查法、實車測試法、統計分析法、及模式參數校估法等，構建與驗證高快速公路微觀車流模式，以供整體微觀車流模型之基礎。然而本研究之高快速公路微觀車流模式包括有跟車模式及變換車道模式，其中跟車模式主要係利用實車測試法蒐集駕駛者跟車資料，分析跟車行為特性，構建GM跟車行為門檻模式，了解GM跟車模式之影響範圍，改善GM跟車模式之缺失。變換車道模式則是以攝影調查法蒐集高速公路變換車道資料，分析變換車道行為特性，以構建變換車道偏向角模式以及加速度模式，並以所構建之變換車道偏向角模式及加速度模式為基礎，進一步建立變換車道決策模式。因此，經由本研究結果將可供高快速公路智慧型運輸系統中，控制與管理所需之微觀車流模式。

The intelligent transportation system(ITS) is the countries all over the world to improving the problem effective measure of the traffic at present. ITS development is influenced very widely in one's power. The traffic flow model can be as the traffic administrator's decision and analysis tools and can offer the service of the traffic road conditions information immediately. Hence, the microscopic traffic flow model is importance of intelligent transportation system to contribute to setting up instant control strategy. The car following and lane changing are one of the important parts of microscopic traffic flow model. The research object is a passenger car on freeway and highway in Taiwan, and the research’s theory is the microscopic concept (or approximate microscopic concept), stimulus-response principle, car follow theory, behavioral threshold model, drive behavior, etc. It’s to structure and verify the microscopic traffic flow model on the freeway and highway for the foundation of the whole microscopic traffic flow model with the literature review, video investigation method, car following experiment method, statistics analytic approach, and parameter calibration, etc., However, the microscopic traffic flow model on freeway and highway includes with the car follow model and lane change model. The data of car following for the car follow model is collected by the car following experiment method, and analyses the behavior characteristic of car following behavior to set up GM car following threshold model which could improve the defect of GM car follow model by understanding the coverage of GM car follow model. The data of lane changing on freeway is collected by the video investigation method, and analyses the behavior characteristic of lane changing to set up lane change angle model and acceleration model. Lane change angle model and acceleration model are the foundation to set up lane change decision model. So, the result of study for microscopic traffic flow model can be supported in the intelligent transportation system on freeway and highway for control and management.