新数据环境下街道空间与行为关联机制研究.

宣 蔚; 彭 康; 臧传云

The new data environment lays a foundation for quantitative analysis of correlation between street space and behaviours. With the continuous urbanisation of China, the vehicle-oriented urban spatial expansion mode leads to a lack of pedestrian space for residents. This issue triggered rethinking and improvement of research into street spaces. The topics of improvement and design of street spaces are attracting increased academic attention. As a new type of special research into road spaces, street design guidelines are increasingly prominent in China and foreign cities. Exploring the correlation mechanism between urban street spaces and the travel behaviours of residents has important significance for the reasonable and effective allocation of street facility resources, for promoting health activities and green travel, and for enriching the depth of street guidelines. 　　Qiguitang Street in the old urban area of Hefei was chosen to collect empirical data. Firstly, OSM was used to crawl street images through Python. Components of street space quality were quantified by combining machine learning algorithms, space syntax, and the ArcGIS platform. The construction status of urban street spaces was analysed. Secondly, data about travel behaviours was collected through behaviour observation and the spatial-temporal characteristics of behaviours in streets were analysed. Finally, a multivariate linear regression model to study correlations between four spatial qualities (comfort, safety, convenience, and richness) and pedestrian behavioural characteristics was constructed. From this, a corresponding index system with 12 variables was determined. The degree of interaction and interaction methods between influencing factors was summarised according to data processing results. The research demonstrated a mathematical influencing relationship between travel behaviours and street space elements. The specific behaviours of pedestrians may be influenced by specific environmental features, and environmental features may also guide pedestrians towards specific behaviours. Environmental features such as functional types, walking width, and interface openness have a greater influence on overall travel behaviour. Therefore, strategies for street optimisation should be determined according to local conditions that are based on street space quality characteristics favoured by different behaviour. On this basis, some strategies were proposed for street space optimisation based on space-behaviour correlation mechanisms. Due to limitations in multisource data acquisition, quality characteristics of some elements were not considered due to significant challenges in acquisition and transformation. As a result, variables affecting travel behaviours are not comprehensive; for example, sense of harmony, microclimate, soundscape, traffic flow, travellers' social factors, etc. are not included. Additionally, as behavioural observation methodology has limitations in collecting a range of travel behaviours, it is impossible to determine largescale and high-quantity statistical data, thus influencing sample size and data quality. Follow-up studies should acquire travel data characteristics by using machine learning algorithms and WIFI data to improve richness and efficiency of data analysis. The correlation mechanisms between street space quality and travel behaviours provide significant reference for scientific diagnosis of urban environmental problems and optimisation of pedestrian travel experiences. They can provide new solutions for effective facility allocation, low-carbon travel, and guidelines for street design in Hefei.

HEFEI Shi (China); CHINA; URBAN growth; CITIES & towns; PUBLIC spaces; URBAN policy; TRAFFIC flow; PEDESTRIANS; STREETS

South Architecture / Nanfang Jianzhu, 2023, Issue 4, p49

1000-0232

Academic Journal

10.3969/j.issn.1000-0232.2023.04.006