Abstract
To improve the performance of vehicle suspension, this paper proposes a semi-active vehicle suspension with a magnetorheological fluid (MRF) damper. We designed an optimized fuzzy skyhook controller with grey wolf optimizer (GWO) algorithm base on a new neuro-inverse model of the MRF damper. Because the inverse model of the MRF damper is difficult to establish directly, the Elman neural network was applied. The novelty of this study is the application of the new inverse model for semi-active vibration control and optimization of the semi-active suspension control method. The calculation results showed that the new inverse model can accurately calculate the required control current. The fuzzy skyhook control method optimized by the grey wolf optimizer (GWO) algorithm was established based on the inverse model to control the suspension vibration. The simulation results showed that the optimized fuzzy skyhook control method can simultaneously reduce the amplitude of vertical acceleration, suspension deflection, and tire dynamic load.
Highlights
Energies 2021, 14, 1674. https://Magnetorheological fluid (MRF) is an intelligent material that can transform between a Newtonian fluid state and a semi-solid state according to the strength of the magnetic field
The inverse model of the magnetorheological fluid (MRF) damper and the fuzzy skyhook control optimized by the grey wolf optimizer (GWO) algorithm were investigated through experiment with numerical simulation, aiming to improve the ride comfort and handling stability of the vehicle
Compared with the back propagation (BP) neural network model, the Elman neural network model reduced the local detail error of the inverse model and improved the prediction accuracy of the control current; 2
Summary
Magnetorheological fluid (MRF) is an intelligent material that can transform between a Newtonian fluid state and a semi-solid state according to the strength of the magnetic field. When the MRF damper is applied to vehicle suspension, a precise mechanical model of the damping force is essential for vibration control. Two kinds of models—the physical model and the phenomenological model—are used to describe the working characteristics of the MRF damper [6] The latter is usually applied to vibration control, for which parameters are identified from the relationships between the damping force produced by the MRF damper and the piston velocity, current, and the piston displacement. The BP neural network does not have a feedback structure and cannot obtain a good prediction effect when dealing with unstable and irregular signals To solve this problem, the Elman neutral network was used to build the inverse model of the MRF damper.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
