On measuring walking accessibility: A link-based utility approach
提出一种基于随机效用理论的步行可达性测量方法，考虑七种街道步行属性，利用路段递归logit模型校准，无需路径集即可无偏评估，并以香港观塘地铁站为例验证。
This study formulates a measure of walking accessibility that takes into account the impacts of seven street-level walking attributes, including sidewalk length, footbridge, staircases, at-grade crosswalk, tunnel, ramp, and escalators based on the random utility theory. We integrate the measure with the link-based recursive logit model, which is calibrated with revealed route choice data collected through an unobtrusive pedestrian following survey. This link-based walking accessibility measure does not require knowledge of the path sets, and hence can provide an unbiased accessibility assessment independent of path choice generation. We apply this approach to analyze the performance of walking accessibility to the Kwun Tong Mass Transit Railway station in Hong Kong as a case study. The proposed approach can be used as a practical tool to identify areas with deficient walking accessibility that need improvement and to forecast the expected outcome of any improvement schemes without bias.