公車電子票證資料與公車動態系統應用在發車時刻表產製之研究

Abstract

搭乘公車是民眾使用大眾運輸的主要服務，而公車的發車時刻以及班次多寡會影響民眾搭乘公車的意願，因為若是民眾搭乘需求多的時段公車班次供給不夠多，便會產生民眾等候公車時間過長而有抱怨之現象，若是民眾搭乘需求少的時段公車班次卻仍密集發車，則會給人浪費公車資源的觀感且令營運公司營運效率下降。在過去的研究中，公車班表的制定對於民眾搭乘需求分析是基於假設的需求分佈做推估而非應用精準的搭乘需求資料，而且大部份文獻求解的方式為使用啟發式演算法來求解，此方法只能證明在當下情況是最好的解但是不一定是最佳解。

為解決這些問題，本研究利用電子票證資料搭配公車動態系統資料整理出每站公車離站後車內人數的乘載比率與表示班次間密集程度的班距資料，然後應用本研究提出的非線性數學規劃模型以Cplex OPL軟體求解得到最佳發車的時刻表。本研究使用豐原客運55路線5月30日一天營運時間之實際資料驗證最佳化數學模型，結果顯示本研究提出之數學模型所得到的最佳發車時刻表可以改善上述問題且新時刻表也符合合理使用範疇。

Bus services are the main public transportation services for the majority of people. The timetable of bus departing from the terminal station and the extent of intensiveness trips assigned may prominently affect passengers’ intention of using bus services. If the quantity of bus trips departing from the terminal are less than passengers’ need in the high bus services demand period, the un-satisfaction and inconvenience of passengers will arise. On the contrary, if the amount of bus trips departing from the terminal are more than passenger’s need in the low bus services demand period, the operating efficiency of the bus company will decline. However, earlier studies merely utilize demand distribution as hypothesis to estimate bus services demand rather than sophisticated demand data and, moreover, process heuristic algorithms to solve the timetabling problems, which not always acquire the optimized solution.

Hence, to improve these problems, this study establishes a non-linear programming model and solve it with Cplex OPL software based on dynamic bus system data and smartcard data. These data give the information regard the ratio of passengers to the maximum capacity in a bus and the ratio of headway to entire operating time. This study use the data of 55 route in May 30 from Fengyuan Bus Transportation Company to verify the availability of the model. The results show the model is concrete and feasible and indeed enhance the terminal departing timetable problems.