Self-adaptive vibration control of simply supported beam under a moving mass using self-recurrent wavelet neural networks via adaptive learning rates

Abstract

A self-recurrent wavelet neural network (SRWNN) is used to control suppression of vibration of an Euler-Bernoulli beam under excitation of a moving mass traveling along a vibrating path. The proposed control structure uses one SRWNN as an identifier and one as a controller. The SRWNN identifier is trained to model the dynamic behavior of the process and provide the SRWNN controller with information about system sensitivity. The SRWNN controller uses the sensitivity information provided by the SRWNN identifier to update weights and produce a signal that controls beam vibration. The gradient descent method and adaptive learning rates (ALRs) are used to update all SRWNN weights. The ALRs are obtained using the discrete Lyapunov stability theorem which guarantees the convergence of the proposed control structure. The performance and robustness of the proposed controller are evaluated at different mass ratios of moving mass to beam and for dimensionless velocity of a moving mass. The simulations verify the effectiveness and robustness of the controller.

Trapez Hız Profilinin Ayarlanması Yoluyla Hareketli Bir Kütle Altındaki Kirişin Titreşim Kontrolü

“…Rezaei and Porseifi (2018) used on-line neural network controller for vibration suppression of a simply supported beam under a moving mass. Ganjefar et al (2015) used selfrecurrent wavelet neural networks as an identifier and as a controller to suppress the vibration of a beam under a moving mass excitation. Zrnic et al (2013) considered the theoretical studies of moving loads on crane structures and discussed how to convert theoretical ideas into designing realistic mega quayside cranes.…”

A wavelet neural control scheme for a quadrotor unmanned aerial vehicle

“…WNN-based control systems have been adopted widely for control of complex dynamical systems owing to their fast learning properties and good generalization capability. Moreover, recurrent wavelet neural networks (RWNNs), which combine properties such as dynamic response of recurrent neural networks and the fast convergence of WNNs, have been proposed to identify and control nonlinear systems [19][20][21][22][23][24][25][26]. An intelligent control system using a recurrent wavelet-based Elman neural network for position control of a permanent magnet synchronous motor servo drive was proposed in [27].…”