An efficient approach for improving the predictive understanding of dynamic mechanical system variability is developed in this work. The approach requires low model assessment time through the fitting of surrogate models. ML-based surrogate algorithms for finite element analysis (FEA) are developed in this study to accelerate FEA and prevent rerunning complex simulations. The research begins with an overview of the recent novelties in ML algorithms applied to finite element (FE) and other physics-based computational schemes. To predict the time-varying response variables, that is, the displacement of a two-dimensional truss structure, a surrogate FE technique based on ML algorithms is developed. In this work, several ML regression algorithms, including decision trees (DTs) and deep neural networks, are developed, and their efficacies are compared. In this study, the ML-based surrogate FE models are able to effectively predict the response of the truss structure in two dimensions over the entire structure. Extreme gradient-boosting DTs provide more precise outcomes and outperform other ML algorithms.
16 February 2017 in Structural Control and Health Monitoring
Street defects, such as potholes and sunken manholes, in general develop quickly compared to other pavement distresses, such as cracking and rutting. Those street defects can result in vehicle damage. This paper proposes an automated and innovative method to obtain up-to-date information about those street defects with the use of a mobile data collection kit mounted on vehicles. In each mobile data collection kit, a triaxial accelerometer and global positioning system sensor collect data for the detection of street defects. A local algorithm is embedded in the mobile data collection kit to increase the efficiency of a local data logging process and to perform a preliminary detection of street defects. At a back-end server, a more precise street defect detection algorithm enhances the performance of the proposed monitoring system by integrating data collected from multiple sensor-equipped vehicles. The street defect detection algorithm at the back-end server relies on a supervised machine learning ...
A smart active vibration control (AVC) system containing piezoelectric (PZT) actuators, jointly with a linear quadratic regulator (LQR) controller, is proposed in this article to control transverse deflections of a wind turbine (WT) blade. In order to apply controlling rules to the WT blade, a state-of-the-art semi-analytical solution is developed to obtain WT blade lateral displacement under external loadings. The proposed method maps the WT blade to a Euler–Bernoulli beam under the same conditions to find the blade’s vibration and dynamic responses by solving analytical vibration solutions of the Euler–Bernoulli beam. The governing equations of the beam with PZT patches are derived by integrating the PZT transducer vibration equations into the vibration equations of the Euler–Bernoulli beam structure. A finite element model of the WT blade with PZT patches is developed. Next, a unique transfer function matrix is derived by exciting the structures and achieving responses. The beam structure is projected to the blade using the transfer function matrix. The results obtained from the mapping method are compared with the counter of the blade’s finite element model. A satisfying agreement is observed between the results. The results showed that the method’s accuracy decreased as the sensors’ distance from the base of the wind turbine increased. In the designing process of the LQR controller, various weighting factors are used to tune control actions of the AVC system. LQR optimal control gain is obtained by using the state-feedback control law. The PZT actuators are located at the same distance from each other an this effort to prevent neutralizing their actuating effects. The LQR shows significant performance by diminishing the weights on the control input in the cost function. The obtained results indicate that the proposed smart control system efficiently suppresses the vibration peaks along the WT blade and the maximum flap-wise displacement belonging to the tip of the structure is successfully controlled.