In this paper, a novel approach to damage identification in structures using Particle Swarm Optimization (PSO) combined with Artificial neural network (ANN) is proposed. With recent substantial advances, ANN has been extensively utilized in a wide variety of fields. However, because of the application of backpropagation algorithms based on gradient descent techniques, ANN may be trapped in local minima when seeking the best solution. This may reduce the accuracy of ANN. Therefore, we propose employing an evolutionary algorithm, namely PSO to deal with the local minimum problems of ANN. PSO is employed to improve the training parameters of ANN consisting of weight and bias ratios by reducing the deviation between calculated and desired results. These training parameters are then used to train the network. Since PSO applies global search techniques to look for the best solution, it can assist the network in avoiding local minima by looking for a beneficial starting point. In order to assess the effectiveness of the proposed approach, both numerical and experimental models with different damage scenarios are employed. The results show that ANN -PSO not only significantly reduces computational time compared to PSO but also possibly identifies damages in the considered structures more accurately than ANN and PSO separately.

In long-distance railways, some particular spans of high-speed railway simply supported beam bridges (HSRSBs) are commonly selected as the target structure. The target structure is the part of interest for the study and intended to be analyzed. Due to longitudinal constraints of the track system, the target structure is tightly coupled with other spans within certain range, and is affected by the coupled spans under longitudinal earthquake condition. A massive amount of time-consuming computation is required to determine the coupling span number using current finite element models. In an effort to overcome this challenge, an equivalent method for the longitudinal constraints of the track system is proposed, which greatly reduces the complexity of finite element model while retaining calculation precision. The coupling span number was determined by seismic analyses of a large number of cases using equivalent finite element models. Moreover, the influence of pier height and bottom pier stiffness on coupling span number was studied. Based on the relationship between the equivalent boundary sensitivity critical point and coupling span number, a method to quickly obtain coupling span number of the target structure in arbitrary HSRSB was constructed.

To improve the energy conversion efficiency and working frequency bandwidth of a single frequency piezoelectric vibration energy harvester, a new type of hybrid vibration energy harvester is developed which is combined with the mechanism of piezoelectric and electromagnetic energy conversion. The system comprises of a PZT cantilever beam, an elastic suspended magnetic mass, a magnet block attached to the end of the cantilever beam and a resonator. The addition of resonator can not only increase the mode, but also adjust the frequency of harvester flexibly. Nonlinear magnetic force of magnet block not only broadens the frequency band and improves the output performance of the system, but also changes the resonant frequency to make the harvester have better adjustable performance. On this basis, an improved electromechanical coupled analytical model of continuum is proposed which can be solved by the Runge-Kutta algorithm and the influence of different factors (the mass and spring stiffness of the resonator, as well as the electromechanical coupling coefficient, electromagnetic coupling coefficient, magnet mass and magnetic flux) on the output are analyzed. According to the prototype of the vibration energy harvester developed, an experimental system was built. The performance of the independent and hybrid energy harvesters is evaluated by experimental and analytical methods. The peak output voltage of the piezoelectric part was about 4 times that of the electromagnetic part. The peak output current of the electromagnetic part is about 30 times that of the piezoelectric part. The study results show that the proposed new hybrid vibration energy harvester can achieve a wider frequency range and multimodal vibration energy harvesting. In addition, the bandwidth and power of the harvester can be dynamically adjusted by changing the resonator or electromechanical coupling coefficient, and the bandwidth of the harvester can also be adjusted by changing the quality and characteristics of the magnet.

In this study, the modulation of multiple frequency content of a single ultrasonic wave in nonlinear structures is investigated analytically, numerically and experimentally. An experimental technique is proposed based on nonlinear lamb wave propagation in aluminum bars using piezoelectric wafer active sensors (PWAS) to study intrinsic nonlinearity of structures. First, a one-dimensional analytical procedure is developed to study the modulation of one dimensional wave with multiple-frequency content in isotropic medium with quadratic nonlinearity. This procedure is implemented to study modulation of frequency contents of a well-known tone burst signal in nonlinear medium. Then, predictions obtained by the proposed analytical procedure are compared with the results of finite element model, which show strong correlations. The experimental and analytical results reveal that in excitation with a train of tone burst, due to frequency modulation, some new harmonics including a strong sub harmonic generation with frequency of f₀/N_p appear in the response. The amplitude of this harmonic is even higher than common second harmonic generation (2f₀). This can be seen in the experimental results when the excitation frequencies are correctly selected. Finally, it is explained that, why the new sub harmonic generation is less affected by the nonlinearity induced by the excitation system.

The falling offs of building decorative layers (BDLs) on exterior walls are quite common, especially in Asia, which presents great concerns to human safety and properties. Presently, there is no effective technique to detect the debonding of the exterior finish because debonding are hidden defect. In this study, the debonding defect identification method of building decoration layers via UAV-thermography and deep learning is proposed. Firstly, the temperature field characteristics of debonding defects are tested and analyzed, showing that it is feasible to identify the debonding of BDLs based on UAV. Then, a debonding defect recognition and quantification method combining CenterNet (Point Network) and fuzzy clustering is proposed. Further, the actual area of debonding defect is quantified through the optical imaging principle using the real-time measured distance. Finally, a case study of the old teaching-building inspection is carried out to demonstrate the effectiveness of the proposed method, showing that the proposed model performs well with an accuracy above 90%, which is valuable to the society.

The piezoelectric-based smart interface technique has shown promising prospects for electro-mechanical impedance (EMI)-based damage detection with various successful applications. During the process of EMI monitoring and damage identification, the operational functionality of the smart interface device is a major concern. In this study, common functional degradations that occurred in the smart interface are diagnosed using a deep learning-based method. Firstly, the effect of functional degradations on the EMI responses is analytically discussed. Secondly, a critical structural joint is selected as the test structure from which EM measurement using the smart interface is conducted. Thirdly, a numerical model corresponding to the experimental model is established and updated to reproduce the measured EMI responses. By using the updated numerical model, the EMI responses of the smart interface under the common functional degradations, such as the shear lag effect, the adhesive debonding, the sensor breakage, and the interface detaching, are simulated; then, the functional degradation-induced EMI changes are characterized. Finally, a convolutional neural network (CNN)-based functional assessment method is newly proposed for the smart interface. The CNN can automatically extract and directly learn optimal features from the raw EMI signals without preprocessing. The CNN is trained and tested using the datasets obtained from the updated numerical model. The obtained results show that the proposed method was successful to classify four types of common defects in the smart interface, even under the effect of noises.

This paper systematically investigates the effect of the inerter on the damping enhancement of a cable with a viscous damper (VD) installed close to the cable end. Three cases are considered, including the inerter installed parallel with the VD (PVID), the inerter placed in series with the VD (SVID), and the inerter installed at a higher location of the VD (HVID). The asymptotic solutions of the three cases are derived, which can predict the cable modal damping ratio when the inerter and the VD cause minimal perturbation in the undamped frequency of the cable. The effect of the inerter on the modal behavior of the cable with the VD is investigated. Based on the constrained static output LQR method, the effects of the inerter on the damping enhancement of the VD in mitigating cable multi-mode vibrations are further evaluated. The results show that the inerter can improve the control performance of the VD when the inertance is less than the optimum value. Further increasing the inertance beyond the optimum value, the optimum modal damping ratio of the cable decreases, and mode crossover is observed for the cable with PVID and HVID. Compared with the case where the VD and the inerter are located at the same location, the case of the HVID is more effective in mitigating cable multi-mode vibrations.

Experiments involving soil-structure interaction are often constrained by the capacity and other limitations of the shake table. Additionally, it is usually necessary to consider different types of soil in experiments. Real-time hybrid simulation (RTHS) offers an alternative method to conduct such tests. RTHS is a cyber-physical testing technique that splits the dynamic system under investigation into numerical and physical components, and then realistically couples those components in a single test. A limited number of previous studies involving soil-structure interaction have been conducted using RTHS, with a focus on linear models and systems. The presence of isolators was not considered in these studies to the authors' best knowledge. Herein, we aim to advance the understanding of the RTHS method by developing and demonstrating its use for nonlinear soil-isolator-structure systems. A sliding mode controller able to deal with both system nonlinearities and wide range of potential uncertainties in such tests is designed and validated using a nonlinear shake table with a nonlinear specimen. By simply changing the numerical model and using the same controller and experimental setup, different soil types and ground motions can readily be considered with this approach. Numerical and RTHS results are compared for verification purposes.

In this paper we study the static deflection, natural frequency, primary resonance of an electrostatically actuated cracked gas sensor. Besides, a novel hybrid metaheuristic algorithm is proposed to detect the location and depth of possible crack on the microcantilever systems. The gas sensor configuration consists of a microcantilever with a rigid plate attached to its end. The nonlinear effects of the electrostatic force and fringing field are taken into account in the mathematical model. The crack is represented by a rotational spring. In the first part, the effect of crack on the static and dynamic pull-in instability are studied. The equations of motion are solved by the application of the perturbation methods. Next, an inverse problem is formulated to predict the location and depth of the crack in the gas sensor. For that purpose, the weighted squared difference of the analytical and predicted frequency response is considered as the objective function. The location and depth of the crack in the microsystem are determined using the hybrid Harris Hawk and Nelder Mead optimization algorithms. The accuracy and efficiency of the proposed algorithm are compared with the HHO, DA, GOA, and WOA algorithms. Taguchi design of experiments method is used in order to tune the parameters of optimization algorithms systematically. It is shown that the proposed algorithm can predict the exact location and depth of the open-edge crack on an electrostatically actuated microbeam with proof mass.

The propagation mechanism of blast waves in rock materials is hard to test. This paper explores the propagation mechanism of blast-induced strain waves in coal and rock using physical modeling together with numerical modeling. The results show that the strain waves in coal blocks were weaker than that in mortar blocks under the same blast loading. With increasing distance, the strain waves induced by the shock wave show a slighter decrease in coal blocks in the radial direction, but show a stable tendency in coal blocks and a slight decrease in mortar blocks in the tangential direction. However, the strain waves induced by the explosion gas show a stable tendency in both coal and mortar blocks. The actuation duration of strain waves in coal blocks is longer than that in mortar blocks. The gap of the radial strain waves induced by shock waves is narrowed gradually and moved similarly equal to each other both in coal and mortar blocks with increasing distance. The simulated results show similar values in coal and mortar blocks as compared with the test results. The coal blocks have a better fracturing effect than that of the mortar blocks in the physical test.