Control Channel Identification Method and Experimental Research on FXLMS Vibration Control

Abstract

Control channel model parameters of adaptive filter feedforward control for piezoelectric smart flexible structures should be predicted, as not only they are the key factors of system stability and control effect but also they have immediate impacts on system convergence and the validity of control strategy. This paper takes the piezoelectric flexible structure as research object. According to least mean square (LMS) algorithm, control channel identification method and its implementation, applied in the MIMO adaptive filter feedforward vibration control, is deduced in detail. Parameter identification, based on which adaptive control for vibration suppression is developed, is performed for the controlled structure, a smart cantilever plate with piezoelectric patches. Experiment results demonstrate the correctness and availability of the control channel identification method and its implementation technique. Adaptive filter feedforward control for piezoelectric smart flexible structures arouses wide concern in the active vibration control field currently. However, identification of the input-output characteristic model H2 should to be done for such control algorithm, as system convergence and the validity of control strategy depends on the accuracy of control channel model parameters. Although several approaches are put forward for control channel identification already, only theoretical deducing and simulations were carried out. Thus, explorations for proper control channel identification methods and implementation techniques are of great significance in improving the robust and validity of the control algorithms (1-4). Based on the analysis of system behaviors, methods analysis of the MIMO control channel H2 identification and experiments were carried out for piezoelectric smart flexible structures. With active structural vibration control experiments to verify the methods proposed in this paper done, this paper explains the research approaches, the analysis of the identification methods, the structure design of experimental subject, the procedural details, the experimental results analysis, etc. II. ADAPTIVE FILTER FEEDFORWARD CONTROL APPROACH Active structural vibration suppression systems using the filtered-X least-mean-square (FXLMS) algorithm, comprise a Finite Impulse Response (FIR) horizontal filter used as a controller. The identification of control channel model H2 is carried out using LMS feedforward control algorithm. The weights of the controller update according to minimum mean-square error criterion, with the goal that response error of several measuring points is held to a minimum (5). Fig. 1 shows the block diagram of the MIMO adaptive filter feedforward control system.