A New High-Precision General Inverse Model Based on Deep Learning for Magnetorheological Damper

Abstract

There are several main factors affect damping characteristics of a magnetorheological damper, the influence laws of some ones are eager to change with varying working conditions. Therefore, it is challengable to establish the relationship between the two: influence factors and damping characteristics of the magnetorheological damper. This paper proposed a new deep learning-based general inverse model to predict the required current value of a magnetorheological damper under changing complex working conditions. Specifically, a fully-connected multilayer perceptron was chosen to train the general forward model. The complex and time-varying relationships between the main factors and damping characteristics were accurately characterized on use of the high nonlinear mapping ability of the multilayer perceptron. Compared to most of general forward models, the proposed model was more excellent with an accuracy of 99.73% based on ensuring generality. Besides, the network inversion method was used in the inverse solution of the multilayer perceptron general forward model, which simplified the process of inverse solution without requiring a mass of training data. Furthermore, a genetic algorithm was used to replace the gradient descent method which was commonly used in network inversion, with the benefit of solving the problem of selecting the initial value and step size, as well as improving the accuracy. Simultaneously, the experiments were conducted on the proposed general inverse model, the experimental results, being consistent with the simulation results, demonstrated that the proposed model can accurately predict the control current value and track the required damping force in changing complex working conditions.INDEX TERMS General forward model, general inverse model, genetic algorithm, multilayer perceptron, network inversion.

Multi-condition adaptive detail characterization model of magnetorheological dampers and experimental verification

“…Lv et al [44] used the nonlinear autoregressive with external input neural network to predict the MRD output damping force, realizing accurate characterization of nonlinear details of MRD damping characteristics. Han et al [45] employed the fully-connected multilayer perceptron (MLP) to train the general forward model, achieving characterizing the nonlinear features of MRD damping characteristics taking advantage of MLP's strong nonlinear mapping capability. Li et al [46] proposed a nested long short-term memory-convolutional neural network-efficient channel attention modelling method based on a dual-flow neural network architecture, to achieve accurate predictions of MRD damping characteristics under variable working conditions.…”

Sequential neural network model for the identification of magnetorheological damper parameters

“…Due to excellent nonlinear regression and learning capability, numerous studies have utilized neural networks to simulate the behavior of MR dampers in non-parametric models. Chang and Roschke [29], Ekkachai et al [30], Tudón-Martínez et al [31] and Han et al [32] used the multilayer perceptron (MLP) as a feed-forward neural network (FNN) approach to model the dynamic behavior of a MR damper. To improve the accuracy of the FNN model, Wei et al [33] applied a Hilbert transformation to develop the input layer of the network by extracting the instantaneous variables such as amplitude and frequency and by adding them to the piston displacement, velocity and current.…”