Research On Multi-channel Adaptive Control Of Vibration

As the improved requirements of the precision instrument, spacecraftand ship to the vibration environments, isolation requirements areincreasingly stringent. Passive vibration isolation can no longer meet therequirements of the high-performance vibration isolation. Especially in thelow frequency range, the isolation result is not so desirable and passivemethod cannot adapt to the strongly changing external vibration which,however, can be solved by the active vibration isolation. Active vibrationisolation has relatively good isolation capacity at the middle and lowfrequency range. Usually, active vibration is applied combining withpassive vibration isolation. Passive method inhibits middle-high frequencyvibration, while active method inhibits middle-low frequency vibration.Compared with single-channel active vibration control, the multi-channelactive control way has greater flexibility and control ability. Multi-channelmethod is applied to control the whole system’s vibration by coordinationof multiple concurrent actuators. The control process is fast, smooth, and overall. Multi-channel control method has broad application prospect.However, in the practical applications of multi-channel activevibration isolation, some factors such as channel crosstalk restrict ultimatecontrol effects. Based on existing research, the four actuator coupledmulti-channel active vibration isolation system was discussed. And theactuator control failure problem being causing by channelcross-interference at specific frequencies was analyzed. By changing thecontrol focus and inhibiting channel crosstalk, adaptive control methodwas improved to solve the actuator control failure problem.In chapter I, the research background and current status of activevibration control are introduced. By comparing the vibration isolationcharacters, the relative advantages of active control are summarized. Andfinally based on developing situation of multi-channel active vibrationisolation, the research theme of this paper is given.In chapter II, the control algorithm for control model construction isdiscussed and selected. Adaptive filter is the application foundation foralgorithm. By the available information extraction of filters from the inputsignals, weight coefficients are updated and control instructions are got.Adaptive Control is realized through Wiener solving; after constructions ofseveral typical LMS algorithms, the algorithm meeting the vibration isolation requirements is selected.In chapter III, methods for active isolation model construction areresearched. Based on Markov model and frequency response functionmodel, two control models are summarized. Corresponding simulation forthe two control models are given to validate the effectiveness of the activecontrol. Based on normalized LMS algorithm, online frequency estimation,tracking filter, single-channel control model is constructed. Athree-degree-of-freedom spring oscillator model is presented to do thesimulation; the simulation results are given to demonstrate the practicalityof single-channel active control model. Markov control model and FRFmodel are built and multi-channel active control model with four actuatorsare presented to do the simulation; the simulation results are given todemonstrate the practicality of multi-channel active control model. Finallyfactors, such as algorithm defects, actuator delay, channel saturation andchannel crosstalk that impacting the control effects are discussed.In chapter IV, mechanism of control failure caused by channelcrosstalk in the active control process is discussed; from a mathematicalpoint, drawback of conventional distributed control method is analyzed.Because of the existing positive feedback crosstalk between channels,excessive active forces are outputted leading to the control effects weakening; control focus of improved method is transferred from localinterference suppression to both local interference and cross-interferencesuppression. And an improved control algorithm formula is presented.Simulation is based on multi-channel control model with two actuators andsimulation results are given to demonstrate the effectiveness of improvedcontrol method in the inhibition of positive feedback path crosstalk.In Chapter V, an active vibration isolation suspension system coupledwith four actuators is presented as simulation object; simulation results aregiven to discuss different influence effects under the differentchannel-coupling ways of the multi-channel control system. Through thesimulation, the active control effectiveness of the conventional distributedcontrol method is demonstrated; in the frequency band that existingpositive feedback channel crosstalk, control states of actuators in thevibration control deterioration situation are simulated; in the controldeterioration frequency, improved control method is applied and thesimulation results is given to justify that improved method has goodproperty on inhibiting path crosstalk to realize overall control goal.In chapter VI, experiment results are given to further validate the goodinhibition property on channel crosstalk. Improved control method isdemonstrated to have important practical application significance. An active vibration isolation suspension experiment system coupled with fouractuators is given; signals are picked by acceleration sensors, processed bysignal amplifiers and data acquisition module and finally are transferredinto controllers to calculate; output signals amplified by the signalamplifiers to drive the electromagnetic actuators which will generatedefault active forces to cancel the whole system’s vibration responses. Theapplication environment and practical limitation are considered in theentire experiment process and the improved method is proved to besuitable for engineering environment with mutation interferences.In chapter VII, the conclusions and contributions of the whole thesisare summarized and highlighted. Finally, some problems for further studyare suggested as well.