본문 바로가기
  • Home

Developing Trip Generation Models Using Spatial Regression Analysis : A Case for Seoul, Korea

Jin Changjong 1 Hyang sook Lee 1 Choo Sangho ORD ID 1

1홍익대학교

Accredited

ABSTRACT

The paper develops trip generation models using the 2006 Seoul Household Travel Survey data for trip production and attraction models by the total and purpose(commute, school, shopping, business, others). The existing models frequently use linear regression models assuming independence among the error terms. However, traffic analysis zones(TAZs) are not simply distributed at random while having autocorrelation each other. To control such autocorrelation, spatial regression models need to be considered rather than linear regression models. Therefore, the paper examines the autocorrelation for trip production and attraction, and verify if the spatial regression models are suitable. Then, SEM(Spatial Error Model) or SLM(Spatial Lagged Model) is estimated according to the spatial regression decision process. The results indicate that spatial regression model is more reliable for most trip generation models except the school attraction model and other production/attraction models. The paper provides the improved trip generation models by applying the spatial interactions among TAZs and land use data.

Citation status

* References for papers published after 2023 are currently being built.