Abstract: Travel preference of passenger plays an important role in passenger’s travel behaviour analytical theory. Passenger’s choice of travel modes is based on the comprehensive analysis of evaluating indexes of each plan. However, the commonly-used methods of determining the attribute weights, which denote the passenger’s preference for evaluating indexes, have their limitations. Based on the preference of passenger’s travel and choice behaviour, this paper analyzes and determines the weights of evaluating indexes, which was ruled by the appraisal criterion of multiple attribute decision making method TOPSIS and deduced with the large-scale original data from passenger’s questionnaires. The weight determining method provides important, objective, and reliable data for passenger’s travel choice behaviour analysis. The effectiveness of the method has been verified by an example. And the rationality of the result has been testified by the Multinomial Logit Model. Furthermore, the presented method is easy to calculate, convenient to implement and universally applicable in passenger’s preference choice problem under every traffic modes.

Key words: dedicated passenger railway lines, passenger’s travel preference, TOPSIS (technique for order preference by similarity to ideal solution)

摘要： 旅客出行偏好选择问题是旅客出行行为分析理论的一个重要部分，而旅客对出行方案的选择是基于对各种方案的评价指标的综合分析。但旅客对各评价指标的侧重程度，即属性权重的计算，目前常用方法都有一定的局限性。本文以旅客出行偏好选择为背景，借助多指标评价方法TOPSIS的评价准则，以实际的大规模旅客调查数据为基础，分析、确定旅客对出行方案各属性指标的权重，为旅客出行行为分析提供重要、客观、可靠的依据。经过实例验证，证明了该方法的有效性，并将计算结果代入多项Logit模型，进一步验证了计算结果的合理性。此外，该方法计算简便、易于实现，对于各种交通方式下的旅客出行偏好选择问题具有普遍适用性。

关键词: 客运专线, 旅客出行偏好, TOPSIS