Abstract:Magnetorheological dampers (MRDs) are intelligent devices for semi-active control and are widely applied in vibration isolation. A high-fidelity modeling method is necessary to take full advantage of the controllable properties of MRDs. Therefore, a nested long short-term memory (NLSTM)-convolutional neural network-efficient channel attention (NLCE) modeling method based on a dual-flow neural network architecture is proposed herein. It uses the time, current, amplitude, frequency, displacement, and velocity as inputs and the damping force as the output. Extensive sinusoidal excitation experiments were conducted using a materials test system and two datasets (large and small sample numbers) were obtained. Five testing sets with different emphases were obtained from different experimental series. Four evaluation indexes were used for a quantitative comparison. First, after training with the large sample dataset, network ablation and comparison experiments were conducted based on a testing set-1. The mean absolute relative error (MARE) evaluation index decreased by 2.290% relative to that of the NLSTM (baseline), indicating that the NLCE method is optimal for predicting the motion characteristics of MRDs. Furthermore, after training with the small sample dataset, comparison experiments were conducted based on testing set-1 and testing set-2. The MAREs decreased by 3.984% and 0.871% relative to that of the NLSTM (baseline), respectively, indicating that the NLCE is also the best modelling method for small sample dataset. The visualization results from the above experiments verified the abilities of the NLCE modeling method for small sample-adaptation, fighting randomness, and identifying similarities. Finally, based on testing set-3, testing set-4 and testing set-5, the NLCE model trained with small sample datasets has high prediction accuracy in predicting the peak damping force (MAREs = 1.456%, 0.880%, and 1.482%, respectively), indicating a high prediction accuracy in the non-hysteretic region. Combining all of the experimental results shows that the NLCE is an effective method for predicting the motion characteristics of MRDs.

Method for improving the neural network model of the magnetorheological damper

Research on a New Mechanical Model for MR Damper Based on Neuron Sigmoid Transfer Function

Experimental Investigations on the Nonlinear Dynamics of Magnetorheological Dampers

Dynamic Characteristics and Modeling of A New Magnetorheological Damper for Broadband Vibration Control

Modeling of Magnetorheological Damper Using ANFIS

Test of a MR damper and its modeling using neural network

A Novel Parametric Model for Magnetorheological Dampers Considering Excitation Characteristics

Frequency-Dependent Bouc–Wen Modeling of Magnetorheological Damper Using Harmonic Balance Approach

Design, Testing and Modeling of a Magnetorheological Damper with Stepped Restoring Torque

An invertible hysteresis model for magnetorheological damper with improved adaption capability in frequency and amplitude

A Unified MR Damper Model and Its Inverse Characteristics Investigation Based on the Neuro-Fuzzy Technique

A General Inverse Control Model of a Magneto-Rheological Damper Based on Neural Network

A new invertible model of magnetorheological damper based on sigmoid function

Bp Neural Network Modeling Of A Magnetorheological Damper

Modelling Of A Self-Sensing Magnetorheological Damper Using Bayesian Regularized Narx Neural Network

Modeling of magnetorheological dampers based on a dual-flow neural network with efficient channel attention

Experimental Study on Inverse Model-Based Force Tracking Control of MR Damper

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

Non-Parametric Model of Magneto-Rheological (MR) Damper Based on Adaptive Neuro-Fuzzy Inference System (ANFIS)

Self-updating inverse model for magnetorheological dampers

Enhanced Generalization of Nonparametric Model for Magnetorheological Dampers