Journal Article

A Multi-Objective Sensor Placement Method Considering Modal Identification Uncertainty and Damage Detection Sensitivity

Huang, Hai-Bin, Zheng, Jun-Xing, Hou, Yuan, Pei, Xue-Yang

2025

Overview

Structural Health Monitoring relies on accurate modal identification and effective damage detection to assess structural performance and safety. However, traditional sensor placement methods struggle to balance modal identification uncertainty, which arises from limited sensor coverage and measurement noise and damage detection sensitivity, which requires sensors to be optimally positioned to capture structural stiffness variations. To address this challenge, this study proposes a multi-objective sensor placement optimization method based on the Non-Dominated Sorting Genetic Algorithm. The method introduces two key objective functions: minimizing modal identification uncertainty by leveraging Bayesian modal identification theory and information entropy and maximizing damage detection sensitivity by incorporating an entropy-based measure to quantify the uncertainty in stiffness variation estimation. By formulating the problem as Pareto-based multi-objective optimization, the method efficiently explores a trade-off between the two competing objectives and provides a diverse set of optimal sensor placement solutions. The proposed approach is validated through numerical experiments on a simply supported beam and a benchmark bridge structure, demonstrating that different optimization objectives lead to distinct sensor placement patterns. The results show that solutions prioritizing modal identification distribute sensors across the structure to improve global response estimation, while solutions favoring damage detection concentrate sensors in critical areas to enhance sensitivity. The proposed method significantly improves sensor placement strategies by offering a systematic and flexible framework for SHM applications, enabling engineers to tailor monitoring strategies based on specific structural assessment needs.

• Publisher: MDPI AG
• Subject: Accuracy / Algorithms / Analysis / Bayesian analysis / Bridges / Damage assessment / Damage detection / damage detection sensitivity / Energy consumption / Entropy / Entropy (Information theory) / Genetic algorithms / Integrated approach / Methods / Modal identification / modal identification uncertainty / multi-objective sensor placement / Multiple objective analysis / Noise control / Noise measurement / Noise sensitivity / non-dominated sorting genetic algorithm / Optimization / Parameter estimation / Parameter identification / Pareto optimization / Pareto optimum / Placement / Sensors / Sorting algorithms / Stiffness / Structural health monitoring / Uncertainty