結合Network DEA與Russell測量方法之運輸系統營運績效評估

Combining Network DEA with Russell Measure for Evaluating Operation Performance of Transit System

楊志慶(Chih-Ching Yang) ； 魏慶地(Ching-Ti Wei)

《運籌與管理學刊》 8卷2期 (2009/12) Pp. 25-44

https://doi.org/10.7047/JLM.200912.0025

摘要

運輸系統的產能由於是無法儲存的，因此衡量其績效應包括系統的產能製造、產能銷售及整體績效。Fielding(1987)提出一個包含成本效率、服務效果及成本效果之評估架構，可充分反應系統的各項績效。資料包絡分析法(data envelopment analysis，DEA)是估計Fielding績效指標常用方法之一，但過去的文獻在使用此方法時，大多是以建立不同的模型分別估計。由於DEA所估計的效率是一種相對效率，因此不同模型所估計的績效值不宜比較。此外，以不同的模型分別估計，隱含有系統在不同的營運階段之決策是獨立的不合理假設。本文結合Network DEA與Russell距離函數(Russell distance function)，提出一個整合的模型，研究者無需設定任何的主觀權重參數，可同時估計系統的三項績效指標，並以桃園地區之桃園客運及中壢客運所經營共52條市區公車路線為對象進行評估。結果發現，中壢客運的成本效果優於桃園客運，且其優勢主要是來自成本效率的貢獻而不是服務效果。

關鍵字

運輸系統 ； 績效 ； 網絡 ； 資料包絡分析法 ； 公車路線

並列摘要

Since the produced capacity of a transit system is non-storable, one should take all the production of capacity, the sales of capacity, and the whole operation process into account when evaluating the system's performance. Fielding (1987) suggested an evaluating structure including cost efficiency, service effectiveness and cost effectiveness, which can fully cover all kinds of performance in system operation. Data envelopment analysis (DEA) is one of the common methods to evaluate performances suggested by Fielding. But most past studies using DEA applied different models to evaluate all three performance indicators, which incurred one problem-the comparison among different performances. In addition, evaluating performances by using different models implies one unreasonable hypothesis that the decisions between different operation stages are independent. This paper presents an integrated model by combining Network DEA with Russell measure. Without any prior weighting parameters, the model can simultaneously estimate all of the three indicators. The model is applied to an empirical study which is composed of 52 city bus routes operated by Taoyuan Transit Company and Jungli Transit Company. The results show that Jungli Company has the better cost effectiveness than Taoyuan Company, and which is contributed from cost efficiency rather than service effectiveness.

並列關鍵字

Transit system ； Performance ； Network ； Data envelopment analysis DEA ； City bus route