structural health monitoring has been widely used during the past decades to evaluate the safety of structural assets, detect damages at an early stage, and prevent unexpected and costly damages. The research works in this field are often concerned with bridges and buildings and little research has been conducted on structural health monitoring of power plant infrastructures such as machine foundations. The turbo generator, also referred to as the heart of the power plant, is supported by massive concrete foundations in the turbine hall of thermal power plants. Most TG foundations in thermal power plants are at or close to their design age. The reports on structural damages in thermal power plants show that cracks are frequently observed on the beams and columns of frame-type TG foundations. It is probably the most appropriate time for developing vigorous methods for health monitoring of power plant structures to extend their reliability and life span. This paper uses the vibration-based approach for damage detection of TG foundations. Analytical mode decomposition (AMD) and experimental mode decomposition (EMD) methods are both used for modal parameter identification. A genetic algorithm is further used for finite element model updating and damage detection. The performance of the method is investigated using a 3-D finite element model of a frame-type TG foundation.

In the past twenty-five years, structural health monitoring (SHM) has become an increasingly significant topic of investigation in the civil and structural engineering research community. An SHM schema involves three main steps: (a) measurement and acquisition of signals related to the structural response, (b) signal processing consisting of pre-processing and feature extraction employing nonlinear measurements, and (c) interpretation using machine learning. This article presents a review of recent journal articles on nonlinear measurements used for feature extraction in SHM of building and bridge structures. It also reviews three recently-developed nonlinear indexes with potential applications in SHM.

In this article, the effectiveness of support vector machine (SVM) is examined for health monitoring of beam-like structures using vibration-induced modal displacement data. The SVM is used to predict the intensity or location of damage in a simulated cantilever beam from displacements of the first mode shape. Twelve levels of damage intensities have been simulated at 12 locations, and six levels of white Gaussian noise have been added, thereby obtaining 1, 008 simulations. About 90% of these are used for training the SVM, and the remaining are used for testing. The trained SVM is able to predict damage intensity and location of all the training set data with nearly 100% accuracy. The test set data reveal that SVM is able to predict damage intensity and damage location with errors varying from 0.28% to 4.57% and 0% to 20.3%, respectively, when there is no noise in the data. Addition of noise degrades the performance of SVM, the degradation being significant for intensity prediction and less for damage location prediction. The results demonstrate the use of SVM as a powerful tool for structural health monitoring without using the data of healthy state.

The development of structural health monitoring algorithms for wind turbines is an emerging need because of the aging issue in wind farm facilities. An emerging field of data-driven machine learning schemes has resulted in the development of new means in structural health monitoring. Although, these approaches are inclined to errors in the absence of good insight into the physics of the system. Therefore, a comprehensive model of the structure, as well as its uncertainties, could be a good complement to these approaches. In the current article, an algorithm is developed for autonomous health monitoring of a wind turbine blade, which is one of the most expensive parts of the turbine, based on acceleration measurements taken from several points on the blade. The data are acquired based on a close-to-reality finite element model of the blade. The acceleration signals are gathered from five nodes along with the wind turbine model, which act as vibration sensors in a common similar test setup. Advanced algorithms of system identification are used for extracting damage sensitive features. Moreover, a one-class kernel support vector machine is trained to find the data associated with a damaged state of the structure. Finally, the success of the procedure in the detection of the existence and location of the damage is depicted.

In recent years, impedance measurement method by piezoelectric (PZT) wafer active sensor (PWAS) has been widely adopted for non-destructive evaluation (NDE). In this method, the electrical impedance of a bonded PWAS is used to detect a structural defect. The electro-mechanical coupling of PZT materials constructs the original principle of this method. Accordingly, the electrical impedance of PWAS can sense any change in the mechanical impedance of the structure. A thermal stress on a structure, which was generated by environmental temperature, could change the electrical impedance of PWAS. The thermal stress which affects the output impedance of PWAS is also developed. A temperature-dependent model, the temperature dependency of PWAS, and structure material properties are investigated for a PWAS bonded to an Euler Bernoulli clamped-clamped beam. The Rayleigh-Ritz and spectral element methods are studied and, then, verified by 3D finite element method (FEM).