Walkability evaluation and optimization based on pedestrian and environment interaction: an agent-based modeling approach

Abstract

Introduction

On one hand, this study proposes an evaluation framework for walkability in residential neighborhood based on interactions between pedestrians and the environment in four dimensions, and applies it to the measurement of walking activities on real streets using agent-based modelling (ABM). On the other hand, we designed multiple scenarios with different combinations of measures, and calculated their street evaluation index to determine the optimal optimization scheme.

Methods

This study employed a mixed-methods fieldwork approach, collecting data from 13 streets in Shenzhen and administering importance rating questionnaires (n = 238). Data collection took place on June 16, 2021, and June 19, 2021. Based on the research data and analysis results, we used the NetLogo simulation platform, combined with AutoCAD and ArcGIS, to construct the road network for simulation.

Results and conclusion

Most simulated streets exhibited varying characteristics of walking activity at different times of day. Streets with high assessed values showed minimal daily fluctuations, whereas those with medium to low values experienced greater fluctuations throughout the day. The design of street interfaces significantly impacts pedestrian walking activities, and the greening index affects different types of activities differently. It deters commercial and recreational activities but facilitates leisure activities. Scenario optimization results indicate that different combinations of optimization strategies yield different effects. This simulation-based method provides a nuanced understanding of how design changes can impact urban environments.