Predictive Control and Signal on Noise Reduction for Semi-Active Hydraulic Damper

Abstract

In the implementation of active or semi-active control systems, it is necessary to process the measured signals because they are not perfect in reality. At present, the current energy-dissipating method for controlling semi-active dampers is flawed because of some restrictions on processing and measuring the signals. Thus, a detection methodology of signal control is proposed in this research based on the direction of structural motion; a velocity estimating calculator is developed by using the least-square polynomial regression. Comparison of the analytical results and experimental data confirms that the proposed calculator is effective in predicting when to switch the moving direction of a semi-active damper. It can detect when the direction of the structure motion reverses as well as when to compensate the poor influence on the performance of a semi-active damper caused by the delayed response. Additionally, the noise of displacement signal will not affect the phase difference of predictive signals.

A model for signal processing and predictive control of semi-active structural control system

“…In the velocity predictor method (Shih & Sung 2007), diminishing the time delay and recovering the capacity loss are proposed herein as the method of time compensation. The structural reaction signals including displacement, velocity and acceleration, in a previous step are used to establish the signals for the next step.…”

“…Recently, the optimal prediction theory (Ernst 1993;Schober et al 1998;Kjell 2002;Bolcskei & Hlawatsch 2001;Huang et al 2001;Janssens et al 2009;Sun et al 2009;Ciesielka & Golas 2006) and Kalman filter (Pavkovic et al 2009;Subrahmanyam et al 2008;Petersen et al 2008) provided signal-processing methodology to predict the response of structures with some restrictions. Therefore, Shih & Sung (2007) proposed a velocity predictor based on the least-square polynomial regression to improve the calculation of velocity for detecting the correct switch timing of semi-active dampers. Some characteristics of this velocity predictor are: (i) a linear acceleration regression module is used to successfully perceive the opportune moment for compensating the time delay; (ii) when the standard deviation of the predictive velocity noise is less than one-tenth of that of the original velocity, the probability of misjudging the control signal is only 10%; (iii) the standard deviation evaluated based on the viable velocity noise is less than the critical value regressed by displacement noise; (iv) for the actual design, the length of sampling history data is related to the natural frequency of the structure.…”

“…Some characteristics of this velocity predictor are: (i) a linear acceleration regression module is used to successfully perceive the opportune moment for compensating the time delay; (ii) when the standard deviation of the predictive velocity noise is less than one-tenth of that of the original velocity, the probability of misjudging the control signal is only 10%; (iii) the standard deviation evaluated based on the viable velocity noise is less than the critical value regressed by displacement noise; (iv) for the actual design, the length of sampling history data is related to the natural frequency of the structure. For example, with 1-Hz natural frequency, 0·4% noise of displacement signal, and 0·1-sec history, the length of sampling history data is 0·2 seconds (Shih & Sung 2007). Therefore, this new methodology, which predicts the control signals based on the direction of structural motions, is further studied in this research for effective control of the semi-active damper system.…”

Improvement of near-fault seismic isolation using a resettable variable stiffness damper

“…This concept was originated from a structural control system called the active variable stiffness (AVS) system proposed by Kobori et al [26,27], in which the variable stiffness was achieved by adding additional stiffness through an on/off switch. Later on, the concept of AVS evolved into a class of control devices commonly referred to as resettable semi-active stiffness dampers (RSASD) [28][29][30][31][32][33][34]. Generally, an RSASD can be modeled by a constant spring connected in series with a resetting device that functions as an on-off switch.…”

“…Generally, an RSASD can be modeled by a constant spring connected in series with a resetting device that functions as an on-off switch. The spring and the resetting device can be realized by using pressurized gas (or oil) and some gas (or hydraulic) valves and parts [25,33,34]. In order to effectively control the RSASD damper, Yang et al [29] derived a general resetting control law based on the Lyapunov theory, which resets the RSASD stiffness at each moment when the relative velocity cross the damper reaches zero.…”