大數據分析於智慧運輸系統之應用與發展：以國道五號及蘇花公路為例

「一條安全回家的路」將隨著「臺9線蘇花公路山區路段改善計畫」於2020年的全線竣工及通車，進一步實現花蓮、臺東居民的交通保障與社會正義。此等發展，亦是繼2006年國道5號通車後，另一項大幅改善臺灣東部地區交通的重大公路建設。國道5號在「蘇花改」通車後，已有6家業者經營共32條客運路線。臺灣北部、宜蘭等地往返花蓮的公路交通便捷化，勢必為「國道5號」與「蘇花公路」帶來更大的車流量與艱難挑戰：「國道５號」在現有交通壅塞問題尚未解決，又面對用路需求大幅增加的挑戰；「蘇花公路」則是面臨道路容量不變，車流量卻大幅成長的窘境。民眾從臺北地區往返花蓮、臺東地區，更可能要連闖「國道5號」與「蘇花公路」兩大交通壅塞關卡。顯而易見的，前述議題已成道路管理機關之重要課題，也為眾多平行機關間，政策制定之協調與整合帶來難題。　　本研究立基於「人工智慧」、「演算法」、「資料探勘」與「大數據」之電腦科學，發展「交通大數據」、「運輸與物流規劃及管理」及「交通政策決定與執行」之應用，利用現有「智慧化公路運輸系統」在硬體的建設上已近完備的優勢，提出透過Dijkstra演算法建構「即時決策執行成效回饋與即時決策調整架構」，以經由電腦科技的高速運算，來達到「交通政策決定與執行」之重要管理目的。換言之，本研究即是使用「交通資訊蒐集系統」取得即時且充分的交通數據，再經由「大數據分析及運算」輸出至「交通控制系統」，以有效利用道路與管理車流，同時藉由「大量的數據收集」與「快速的資訊分析及決策執行」，達到減少決策時間與降低決策成本，用以整合分析與處理東部路廊的交通議題，並達成政策分析。　　本研究的貢獻在於建構出「即時決策執行成效回饋與即時決策調整架構」與「大數據分析之智慧運輸系統」，藉由「大數據」的基礎，應用Dijkstra演算法，發展「交通決策」之「智慧系統」設計與開發。基此，本研究所關切的課題，已能以「智慧運輸系統」獲得完整的交通資訊，再由其系統自動演算，提出即時且能確實執行的最佳解決方案，進言之，更可依循此等途徑，建構出「整合政策決定考量因素」並符合各方需求的「智慧化公路運輸系統」，以做成最適政策組合。

關鍵字

Dijkstra演算法；政策分析；大數據；人工智慧；智慧運輸系統；國道５號；蘇花改

並列摘要

With the completion and opening of the “Provincial Highway No. 9 Suhua Highway (Suao-Hualien) Mountainous Section Improvement Project” in 2020, the “One safe road home” transportation and social justice objective for Hualien and Taitung residents will be achieved. This development is a major public road construction project for eastern Taiwan that followed the opening of Freeway No. 5 in 2006. After the opening of “Suhua Highway Improvement Project”, six bus operators opened a total of 32 bus lines in the Freeway No. 5. The convenience of travel between northern Taiwan and Yilan and Hualien will bring higher traffic volume and greater challenges for Freeway No. 5 and the Suhua Highway. The congestion of Freeway No. 5 has not yet been solved and now it must face a significant increase in traffic. On the other hand, the road capacity for the Suhua Highway has not changed, but the traffic volume has significantly grown. People traveling back and forth from Taipei and Hualien/Taitung may have to pass the major Freeway No. 5 and the Suhua Highway gridlock. As this clearly demonstrates, the aforementioned items have become important issues for road management agencies. These issues have also created difficulties for policy formulation and integration among many parallel agencies. 　This study uses computer science such as artificial intelligence, algorithms, data exploration, and big data to develop traffic big data, transportation and logistics planning and management, and traffic policy decision and implementation applications. We propose using Dijkstra’s algorithm to construct a “real-time decision implementation and result feedback and real-time decision adjustment architecture” and utilizing the “Intelligent public transportation system” hardware that is nearing completion to achieve important management objectives such as traffic policy decision and implementation through high-speed calculations. In other words, this study uses the traffic information collection system to obtain sufficient real-time traffic data, and then uses the big data analysis and calculation to output the results to the traffic control system. The process effectively uses road and manages traffic flow; meanwhile, big data collection and fast information analysis and decision implementation can reduce decision time and decision costs. Thus, the system can be used to integrate, analyze, and process traffic issues on the eastern corridor. This study has the significance of policy analysis. 　The contribution of this study is in constructing a “real-time decision implementation and result feedback and real-time decision adjustment architecture” and a “big data analysis intelligent transportation system”. Big data is used as the basis with Dijkstra’s algorithm to design and develop a “Intelligent System” for traffic decisions. In this study’s topic, the intelligent transportation system can obtain complete traffic information. The system then automatically analyzes the information and provides an immediate and implementable optimal solution. In other words, this process can construct an intelligent public road transportation system that conforms to “integrated policy decision factors” and other requirements and therefore can formulate the best policy.

並列關鍵字

Dijkstra’s Algorithm ； Policy Analysis ； Big Data ； Artificial Intelligence ； Intelligent Transportation System ； Freeway No.5 ； Suhua Highway Improvement Project