Analysing traffic capacity at signalised intersections with structural equation modelling

Traffic capacity is influenced by numerous variables with complex interdependencies. Effectively characterising the intricate relationships among diverse factors and their impact on traffic capacity poses a formidable challenge. Therefore, this study introduces structural equation model (SEM), departing from previous research that focused on statistical characteristics between individual influencing factors and capacity (traffic: proportion of heavy vehicles; control: signal phase, signal cycle, all red time; geometry: road type, lane width, number of lanes, lane location). The methodology examines the combined effects of various factors on capacity at the macro level and elucidates each factor’s influence. Subsequently, variables are selected, and their associations are ascertained through correlation and analysis of variance. Furthermore, IBM AMOS 25.0 was used for SEM development and validation. Parameters were estimated using maximum likelihood estimation. Model fit was assessed using fit indices. Path analysis was employed to quantify the extent to which key factors influence capacity. Findings revealed that signal phase (standard load = −0.646), number of lanes (0.224), and proportion of heavy vehicles (−0.185) had the most substantial impact on traffic capacity. This study can contribute to improving intersection capacity from optimising signal phase design, increasing lane group numbers in congested districts, and regulating heavy vehicle access during peak hours.