本研究係應用專家系統，構建高速公路上匝道控制系統，以期能充分考慮高速公路主線﹑入口匝道及平行道路間之互動關係，同時，為使所構建專家系統具備豐富的知識庫及充足的控制功能，以適用各種交通狀況之需，本研究乃進行廣泛的模擬實驗及控制策略篩選，以作為系統構建之基礎。目前系統能彈性控制三個匝道並自動反應地區感應儀控、定時儀控、固定儀控、儀控終止、匝道關閉、整合感應儀控等六種控制策略。更重要的是，當系統執行地區感應儀控時，能隨交通狀況之變化及偵測器佈設之不同，而自動選擇需求-容量、ALINEA、佔有率等控制模式，最後，評估結果顯示，專家系統能整合運用各種模式，其控制績效比單獨使用需求-容量、ALINEA或佔有率等控制模式為佳，未來，仍須進一步考慮進行系統擴充，以利系統能執行運輸走廊及平面號誌之整合控制。

This paper presents the application of an expert system to tackle the sophisticated on-ramp control issues among freeway mainline, on-ramp, and surface street. To ensure the proposed expert system with sufficient knowledge to active its real time control functions under various traffic conditions, the extensive simulation experiments as well as the proper control strategy selections have been conducted for system constructions. In current stage, the proposed system can control more than3 ramp, and perform the following control strategies: local traffic responsive (LTR), pretimed, fixed rate, metering off, ramp closure, and coordinated traffic responsive metering. More importantly, under the LTR conditions, the system can respond to execute demand-capacity model, ALINEA model,or occupancy model based on the real time traffic conditions and the different surveillance deployments. At last, the testing results show that the proposed expert system with sufficient knowledge and inference capability can provide promising performance under different level of congestion and can also outperform the conventional system by solely use demand-capacity, ALINEA, or occupancy model. The ongoing research along the same line is: (1) to extend the proposed expert system with freeway corridor control function, and (2) to integrate freeway on-ramp metering with surface traffic signal control.