Title: Automated Piecewise Linear Regression for Analyzing Structural Health Monitoring Data

Author(s): Harshita Garg, Kai Yang, Anthony G. Cohn, Duncan Borman, Sreejith V. Nanukuttan, P.A. Muhammed Basheer

Publication: Materials Journal

Volume: 121

Issue: 2

Appears on pages(s): 93-104

Keywords: artificial intelligence (AI); automated clustering-based piecewise linear regression (ACPLR); diffusion coefficient; electrical resistance; in-service performance; structural health monitoring (SHM)

DOI: 10.14359/51740370

Date: 4/1/2024

Abstract:
The recent increased interest in structural health monitoring (SHM) related to material performance has necessitated the application of advanced data analysis techniques for interpreting the realtime data in decision-making. Currently, an accurate and efficient approach for the timely analyses of large volumes of uncertain sensor data is not well-established. This paper proposes an automated clustering-based piecewise linear regression (ACPLR)-SHM methodology for handling, smoothing, and processing large data sets. It comprises two main stages, where the gaussian weighted moving average (GWMA) filter is used to smooth noisy data obtained from electrical resistance sensors, and piecewise linear regression (PLR) predicts material properties for assessing the performance of concrete in service. The obtained values of stabilized resistance and derived values of diffusion coefficients using this methodology have clearly demonstrated the benefit of applying ACPLR to the sensor data, thereby classifying the performance of different types of concrete in service environments.