Evaluation of Preferred Applications of Traffic Signal Algorithms Using Traffic Delay Index for an Isolated Intersection

本研究比較數種不同型式之常用交通號誌控制邏輯，包括定時控制、全觸動控制、運用卡曼濾波之動態計算號誌控制，以及改良OPAC(optimization policies adaptive control)。利用微觀車流模擬方式模擬號誌化獨立四叉路口，以指數分配函數產生不同交通車流情境，並以微觀車流模擬方式累計車輛停等延滯，分析比較此四種號誌控制邏輯在不同情境下之總車輛停等延滯時間，獲得在不同交通車流情境下較佳實施之號誌控制邏輯。研究結果將可作為選擇號誌控制邏輯之判斷標準，依此訂定客觀的號誌控制邏輯切換準則，對未來發展智慧型號誌系統將能更有效應付各種不同車流情境之變化。
This study aims to evaluate several common traffic signal control strategies, which include pre-timed, full-actuated, dynamic calculated with Kalman filter, and improved optimization policies for adaptive control. A four-leg signalized intersection is used as a typical case for the analysis. On account of the exponential distribution of generating different traffic scenarios, the cumulative delay time is obtained from microscopic traffic flow simulation experiments. By comparing total delay time among these strategies, the conclusion implies that the preferred situations for utilizing each strategy. The empirical results can be adopted as the criteria for choosing a desirable signal control strategy. Based on the proposed criteria, more efficient intelligent traffic control systems could be developed for managing various traffic situations.