ABSTRACT Displacement is a common occurrence in large‐span steel structures; however, the lack of standardized and targeted statistical analysis methods often leads to neglecting the characteristics of displacement and its variations, making it challenging to identify potential structural damage risks. This paper proposes a standardized data processing method using time‐series models of measured column displacement data for safety evaluation. The group of steel structures consists of 17 large‐span exhibition halls, and the column displacement has been continuously measured for a period of 14 years, resulting in an extensive dataset. A 12‐parameter tuple is introduced to store column information, including the displacement measured using prism‐free total stations. Modeling with a linear time‐varying model was applied to analyze the time series, establishing a correlation between statistical parameters and the nonlinear engagement mechanisms of the column nodes. By integrating various statistical methods, parameters highly relevant to the risk level of the columns were extracted, and the risk index was calculated to provide guidance for maintenance and management. The probability density distributions of the displacement over each 5 years indicate that the displacement is currently safe and controllable. Data analysis reveals that the displacement of the indoor columns along the long axis of the halls is not significant, while the displacement of the outdoor columns is very large. Correlation analysis indicates a relationship between glass curtain cracking and column displacement. The improved structural wind‐resistance bracing system has effectively reduced column displacement. The research highlights the importance of establishing a long‐term displacement database for structural safety evaluation and maintenance.
Ma et al. (2026) studied this question.