Abstract
Abstract. Structural health monitoring (SHM) is often approached from a statistical pattern recognition or machine learning perspective with the aim of inferring the health state of a structure using data derived from a network of sensors placed upon it. In this paper, two SHM sensor placement optimisation (SPO) strategies that offer robustness to environmental effects are developed and evaluated. The two strategies both involve constructing an objective function (OF) based upon an established damage classification technique and an optimisation of sensor locations using a genetic algorithm (GA). The key difference between the two strategies explored here is in whether any sources of benign variation are deemed to be observable or not. The relative performances of both strategies are demonstrated using experimental data gathered from a glider wing tested in an environmental chamber, with the structure tested in different health states across a series of controlled temperatures.
Highlights
Sensor placement optimisation (SPO) is the technique by which the number and location of sensors is optimised for a specific objective to reduce the cost of a structural health monitoring (SHM) system without compromising on the effect of monitoring
The feature-bagging results of Euclidean squared distance (ESD)-based features and Mahalanobis squared distance (MSD)-based features for the normal condition case and three damage cases from 36 sensors at seven different temperatures are shown in Figs. 6 to 9
By analysing the positions of sensors providing comparatively large discordance values in different damage cases, it is apparent that data collected from sensors close to the added mass blocks are sensitive to this damage, as may have been expected
Summary
Sensor placement optimisation (SPO) is the technique by which the number and location of sensors is optimised for a specific objective to reduce the cost of a structural health monitoring (SHM) system without compromising on the effect of monitoring. Eshghi et al (2019) used the detectability of different health states as a criterion to design a sensor network optimally, and a surrogate model was applied to reduce the computational burden In addition to these non-Bayesian OFs, an approach that utilises an OF based on minimising Bayes risk has been proposed (Flynn and Todd, 2010). This paper develops two strategies for considering environmental variations in the optimum design of the sensor deployment in an SHM system This technique aims at maximising the damage detection ability of an SHM system by proposing an objective function using a supervised-learning algorithm, namely an SVM.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
