Quantifying perceptual imagery: A Kansei engineering-based multiple regression model for design quantification

Abstract

The rapid global urbanization has propelled Smart Street Lamps (SSLs) as essential smart city infrastructure, yet current designs rely on subjective aesthetic judgments, failing to systematically integrate user perceptual preferences. This study addresses this gap by developing a multiple regression model to quantify relationships between perceptual imagery dimensions and SSL design parameters. Using a mixed-method approach, it collects diverse SSL samples, establishes a perceptual lexicon, and employs cluster analysis, factor analysis, and partial least squares (PLS) regression to identify key perceptual factors and derive equations linking design attributes to perceptual keywords. The validated model bridges design elements with user perceptual experience, enabling data-driven optimization of SSL characteristics. By integrating Kansei engineering with technological innovation, this research enhances design rigor and provides a replicable methodology for smart city products, contributing to sustainable urban development.