Semi-active control of structures incorporated with magnetorheological dampersusing neural networks

Magnetorheological (MR) dampers are devices that can be used for structural vibration reduction under seismic excitation. These devices are used in semi-active control which require less power compared to active devices and offer high reliability compared to passive devices. Despite the advantages of MR damper, use of these dampers in an effective way in a structure is highly challenging and a precise modelling is required as these dampers are highly non-linear. Among the parametric models available, Bouc–Wen model is widely used because of its effective modelling of the hysteretic force–velocity curve of MR damper. The parameters of Bouc–Wen model are damper dependent and hence need to be identified before utilising the damper for further simulation studies. In this work, the parametric identification of Bouc–Wen model for commercially available long stroke and short stroke MR damper (RD 8040-1 and RD 8041-1) is done. For this, experimental characterization of the dampers are carried out using hydraulic actuators mounted on a self-restraining frame. The damper is driven harmonically in the testing setup at various combinations of frequency, amplitude, current and displacement. Using the experimental characterization, parameters of Bouc–Wen model are identified by Levenberg–Marquardt optimization Algorithm (LMA). The identified parameters are validated by comparing with the experimental results. The identified parameters are believed to be worthwhile for the use of these MR dampers in further studies of real-time semi-active vibration control of structures.

Integrated fuzzy logic and genetic algorithms for multi-objective control of structures using MR dampers

The jacket structure is one of the most important structures in the offshore industry. It is mainly installed in relatively shallow water conditions. The jacket structures are subjected to continuous wave forces, which severely affect the life of the offshore platform. Generally, the jacket structures are designed such that their response is small under its environmental loads, but it also increases the cost. So, controlling its response is a good alternative to increase its life span. In this study, a 2D jacket structure in a water depth of 120m is studied for its response control using Magneto-Rheological damper (MR damper). MR dampers have been successfully implemented in semi-active control of civil structures. They have been well proven in controlling the structural responses subjected to multiple frequency excitations. The MR damper's semi-active control algorithm requires the calculation of optimal force by assuming the structure to be linear. This optimal force can be obtained using various methods, and one such widely used method is the Linear Quadratic Regulator (LQR) strategy. The ocean waves are random, and hence the stochastic LQR Linear Quadratic Gaussian (LQG) has to be used to compute the optimal control force on the structure. This optimal force calculated can further be used by the MR damper's semi-active control algorithm to obtain the controlled response. From the theory of the LQG method, the input excitation has to be a white noise process, but the ocean waves that excite the offshore platforms are a narrow band process. So, the spectral content of the input excitation is captured using the shaping filters. The shaping filters are designed to obtain the required narrow banded spectrum of the water particle kinematics at different points of the structure. The filters are designed such that the cross-correlation of the excitation at various offshore platform points is maintained. The Cholesky factorization of the spectral matrix at various grid points degenerates the problem into a single input multiple output problem. Thus, individual fitting of the factorized spectral matrix components is sufficient to obtain the correct cross-correlation along with the depth. The 2D jacket requires the water particle kinematics to be calculated at different horizontal locations also. This horizontal propagation problem requires developing an ‘All-Pass’ filter, which allows one to design the phase shift at different frequency components while maintaining the unit magnitude transfer function. The developed filter equations are then augmented to the original jacket structure, and LQR gains are computed for the stochastic regulator problem. A Kalman filter is designed, and the state estimate is obtained for each time step, which is further used in the calculation of the optimal force. It is shown that stochastic structural control works for an offshore structure like the jacket, and control efficiencies and limitations under various conditions are discussed.

Magnetorheological (MR) dampers have become an effective means for the semiactive vibration control of civil structures. For achieving good control performance, it is necessary to make the accurate characterization of the civil structures to be controlled, the sensors, the actuators and some other systems integrated. Nowadays, the modeling and identification of the dynamics of an MR damper is still a challenging task due to its highly hysteretic nonlinearities. This paper presents experimental results on the modeling and identification of a type of MR damper prototype whose MR fluid is contained in a foam. A comparative study is made among three parametric models (Bingham, Bouc-Wen and Hyperbolic tangent).

A promising solution for semi-active vibration control of different dynamic systems is to use magnetorheological (MR) dampers. In the present study, the investigation focuses on developing a novel single and multi-coil self-powered magnetorheological (MR) damper system using electromagnetic induction (EMI) for seismic mitigation. The conventional MR dampers, which rely on external power sources, can be unreliable and impractical in earthquake-prone locations. Thus, the EMI device connected to the MR damper can be used as an effective and alternative power source for the MR damper, results in a self-powered system. The proposed energy-harvesting system is an MR damper placed above the top of the piston. The coil is wound around the piston, and the outer casing with a ferrite magnet is fixed. As the mechanical energy of the piston is converted into electrical energy, its self-tuning capacity is a perfect fit for structural vibration control applications. The MR damper is subjected to cyclic and time-history loading. At a maximum amplitude of 15 mm, the damper generated 1767.8 N for cyclic loading and -1780 N for earthquake loading, the El Centro earthquake 1940 is considered for the study. By placing EMI, the mechanical energy is converted into electrical energy and powers the damper to avoid external power. The experimental results showcase enhanced damping forces and adaptability, thereby establishing it as an innovative and effective it can be alternative to conventional MR dampers for vibration control in future

Magnetorheological (MR) damper is a prominent semi-active control device for earthquake response mitigation of structures. The most important topic for the intelligent MR structures is choosing the control current of MR damper quickly and accurately. The typical control strategy is on-off control strategy, while inherent time-delay and coarse control precision lie in this strategy. This article proposes a semi-active control strategy for restricting the MR structures efficiently, in which the genetic algorithm (GA) is adopted to search for the optimal feedback control gain in time-delayed dynamic equations forming a GA controller, and next optimize the control current of the MR damper accurately by a series of GA procedure. The GA strategy has the capability to effectively control the MR structure involving time delay and adjusting for controlled current of the MR damper timely and accurately. Accordingly, numerical examples are used to testify that the feedback gain optimized by GA has the capability to reduce a given objective response dramatically. In addition, a comparison between GA controller and LQR controller indicates that the proposed approach performs much better than LQR method, especially in time-delay systems. Finally, simulation results for a seismically excited structure subjected by the MR damper are presented to verify that the semi-active strategy using GA is predominant in mitigating the dynamic structure efficiently. The strategy and the results presented will also throw light on the development and control of other novel structures.

In this paper is addressed the problem of semiactive balancing control of a Jeffcott-like rotor system on journal bearings, one of them supported on radial Magneto-Rheological (MR) dampers. The mathematical model of the rotor system results from a Jeffcott model and the dynamics associated to the MR dampers, whose properties depend on the current inputs (control actions). There are different models for MR dampers proposed in the literature, most of them experimentally validated. For control purposes we use the Choi-Lee-Park polynomial model, which is quite consistent with the nonlinear and complex hysteresis damper models and also simplifies the physical implementation. The semiactive control scheme is then analyzed and synthesized to manipulate the unbalance response of the rotor system, by means of a proper modification of the rotordynamics coefficients (damping and stiffness) using. sliding-mode control techniques.

Coupled building control is a viable method to protect tall buildings from seismic excitation. In this study, the semi-active control of a building complex is investigated for mitigating seismic responses. The building complex is formed of one main building and one podium structure connected through Magneto-Rheological (MR) Dampers and Tuned Mass Damper. The conventional semi-active control techniques require a primary controller as a reference to determine the desired control force, and modulate the input voltage of the MR damper by comparing the desired control force. The fuzzy logic directly determines the input voltage of an MR damper from the response of the MR damper. The control performance of the proposed fuzzy control technique for the MR damper is evaluated for the control problem of a seismically-excited building complex. In this paper, a building complex that include a 14-story main building and an 8-story podium structure is applied as a numerical example to demonstrate the effectiveness of semi-active control with Magneto-Rheological dampers and its comparison with the passive control with the Tuned Mass Damper and two uncoupled buildings and hybrid semi-active control including the Tuned Mass Damper and Magneto-Rheological dampers while they are subject to the earthquake excitation. The numerical results show that semi-active control and hybrid semi-active control can significantly mitigate the seismic responses of both buildings, such as displacement and shear force responses, and fuzzy control technique can effectively mitigate the seismic response of the building complex.

Magneto-rheological (MR) fluid dampers can be used to reduce the traffic induced vibration in highway bridges and protect critical structural components from fatigue. Experimental verification is needed to verify the applicability of the MR dampers for this purpose. Real-time hybrid simulation (RTHS), where the MR dampers are physically tested and dynamically linked to a numerical model of the highway bridge and truck traffic, provides an efficient and effective means to experimentally examine the efficacy of MR dampers for fatigue protection of highway bridges. In this paper a complex highway bridge model with 263 178 degrees-of-freedom under truck loading is tested using the proposed convolution integral (CI) method of RTHS for a semiactive structural control strategy employing two large-scale 200 kN MR dampers. The formation of RTHS using the CI method is first presented, followed by details of the various components in the RTHS and a description of the implementation of the CI method for this particular test. The experimental results confirm the practicability of the CI method for conducting RTHS of complex systems.

To explore the addressed issue of closed-loop vibration control using semi-active magneto-rheological (MR) fluid damper, an innovative control algorithm based on feedback of acceleration response at several points along stay cable for active/semi-active control of stay cables is firstly proposed in the present paper. Theoretical analysis, laboratory investigation, and in-situ tests are conducted in this study to verify the proposed control algorithm. Moreover, this paper presents the negative stiffness characteristic of the combined stay cable/semi-active MR damper system. These investigations indicate that the semi-active MR damper can achieve much better mitigation efficacy than the other passive MR dampers and the proposed control algorithm is proved to be practical. The effect of the negative stiffness on response reduction is demonstrated by the improved energy dissipation ability of the damper due to the enhanced displacement of the stay cable at the location attached with the damper. Introduction In the past few decades, many control methods have been proposed and some have been implemented to mitigate the unaccepted stay cable vibration (Johnson, 2003). At present, semi-active dampers have been proven to be effective in many applications and can potentially achieve performance levels that are almost the same as comparable active devices. For stay cables vibration control incorporated with semi-active dampers, most of the existing control strategies are based on full-state feedback of generalized coordinates. However, the generalized displacements and velocities of the stay cable cannot be measured directly. Since accelerometers can readily provide reliable and inexpensive measurements of the stay cable accelerations at strategic points on a stay cable, development of control method based on acceleration feedback at limited locations along stay cable is an ideal solution to solve the addressed practical problem. Moreover, even though the characteristic of negative stiffness of active control of stay cable vibration has been observed by some researchers(Johnson, 2003), the effects of negative stiffness on the stay cable vibration control using active/semi-active dampers 2 has not been fully researched in the present literatures. In this paper, a control algorithm based on feedback of acceleration response at several points for active and semi-active control of mass-distributed dynamic systems, e.g. stay cables is firstly proposed. Using the proposed control algorithm, the vibration control for the stay cable model attached with one MR damper and a series of field tests are carried out to investigate the control efficacy achieved by different control strategies. These investigations address that the semi-active MR damper can achieve much better mitigation efficacy than other passive MR dampers and the proposed control algorithm is proved to be practicable and available. Moreover, the effect of the negative stiffness provided by semi-active MR dampers on response reduction of cables is theoretically and numerically demonstrated by the improved energy dissipation ability of the damper due to the enhanced displacement of the cable at the location attached with the damper. Control Algorithm Based on Acceleration Feedback at Several Locations Governing equation of the combined stay cable/MR damper system shown in Figure 1 is given as (Irvine, 1981) ) ( ) ( ) , ( ) , ( ) , ( ) , ( d s x x t u t x f t x v c t x v T t x v m − + = + ′ ′ − δ & & & (1) with the boundary conditions v(0,t)=v(l,t)=0 for all t .where v(x,t) is the transverse deflection of the cable, l is the length of the cable, T is the cable tension, m and c are the cable mass and viscous damping per unit length, respectively, xd is the distance of the installed MR damper position from the left end, us(t) is the transverse control force provided by the MR damper at location x= xd, f(x,t) is the distributed load along the cable. δ(.) is the Dirac delta function. Figure1. Combined stay cable/MR damper system The transverse deflection could be approximated using a finite series ) ( ) ( ) , ( ∑ = = r j j j x t q t x v 1 φ (2) where the φj(x) is a set of shape modal functions and continuous with piecewise continuous slope to satisfy the geometric boundary conditionsφj(0)= φj(l)=0. Utilizing the approximate series solution and a standard Galerkin approach, the motion equation of the cable-MR damper system in matrix form is written as ) ( ) ( f Kq q C q M q t u x s d φ + = + + & & & (3)

Neuro‐Fuzzy Control of Railcar Vibrations Using Semiactive Dampers

Seismic control performance of the magnetorheological (MR) damper, which has strong nonlinearity, varies with the dynamic characteristics of the structure and ground motion such as natural period, peak ground acceleration. The MR damper has a property of the friction damper so that the relative magnitude of the friction force over excitation intensity affects the control performance. In this study, through numerical analyses of single-degree-of-freedom (SDOF) structures, design spectra for the passive MR damper and optimal non-dimensional friction forces are suggested. Additionally, simple semi-active control algorithms modulating the friction force of the MR damper in accordance with structural responses are proposed and compared with each other and the passive MR damper. Finally, the condition that the semi-active control is more appropriate than the passive one is identified for reasonable selection of the control strategy.

The use of magnetorheological (MR) dampers for mitigating vibrations caused by seismic motions in civil engineering structures has attracted much interest in the scientific community because of the advantages of this class of device. It is known that MR dampers can generate high damping forces with low energy requirements and low cost of production. However, the complex dynamics that characterize MR dampers make difficult the control design for achieving the vibration reduction goals in an efficient manner. In this article, a semiactive controller based on the backstepping technique is proposed. The controller was applied to a three-story building with an MR damper at its first floor subjected to seismic motions. The performance of the controller was evaluated experimentally by means of real time hybrid testing.

This paper presents the procedure used for design, fabrication, testing, and numerical modeling of a magnetorheological (MR) damper that is to be applied for vibration control in a 70-ton railcar. MR dampers are semiactive vibration control devices whose damping characteristics can be modified in real time by varying an applied current. Design parameters for the MR damper are estimated from those exhibited by a linear viscous damper that exerts the necessary force required to limit vertical vibrations of the rail truck within acceptable limits. An MR damper is fabricated by modifying the piston of a standard hydraulic damper to function as a solenoid. The assembled MR damper is tested in a uniaxial testing machine by subjecting it to sinusoidal and random displacements while simultaneously varying the current flowing in the solenoid. A variable magnetic field is applied to the MR fluid that fills the damper cavity and the resisting force exerted by the damper is recorded. Data collected in the laboratory are used to train a fuzzy model of the MR damper that characterizes its behavior. Results indicate that a fuzzy model of the MR damper can predict its behavior with a sufficient degree of accuracy while requiring minimal computational time.

Magnetorheological (MR) dampers are used for semi-active vibration control of various dynamic systems. Existing MR dampers are usually cylinder-piston based design, which may limit the shapes and have constraints to the design of MR devices. In this paper, we propose a new MR-fluid working operation, in which MR fluids work between gear transmissions. This operation could provide more design flexibility. A prototype of the regenerative damper with MR fluids working between gear transmissions was designed, fabricated, and tested. This MR damper has the capability of power generation and velocity sensing. The feasibilities of the controllable MR damping force, power generation and velocity sensing are experimentally verified. The results of this research would be beneficial to advance the design and multiple functions of MR dampers while not limited to traditional piston-type design.