Smith Predictor-Taylor Series Based LQG Time Delay Compensation Control Of Magneto-Rheological Semi-Active Suspension

With the development of the science technology and society, people put forward higher and higher demand on vehicle ride comfort. Compared with the other types of semi-active suspension, magneto-rheological semi-active suspension has the advantages of quick response, easy to control, continuously controllable resistance and so on, so it has great potential to improve vehicle ride comfort. In the working process of the magneto-rheological semi-active suspension, measuring and tansporting system state, computations and the response of actuator will take some time, so in a magneto-rheological semi-active suspension system, time delay is usually inevitable. Time delay deteriorates real-time control, further worsens the performances of the magneto-rheological semi-active suspension system. In order to reduce the impact caused by time delay, a new Smith predictor- Taylor series based LQG(STLQG) control is developed and researched. The primary contents of this paper are as follows:Firstly, random input road surface model is established. On the basis of magneto-rheological damper’s Bingham model and the formula of coulomp dampering force and current, a quarter vehicle dynamical model with two degree of freedom is built. Besides, evaluation indexes and comprehensive performance index are given.Secondly, Taylor series is integrated with LQG control, and to solove the problem in the process of the above combination, a Taylor series based LQG(TLQG) control is presented, which is based on active control force’s approximate transformation in the state space equation. Contrastive analysis of control force shows: along time axis, the deviation between the coulomp dampering force and ideal semi-active control force(horizontal shift) is small, but the coulomp dampering force has magnification phenomenon caused by Taylor series, and this magnification will become obvious as time delay increases.Thirdly, to solve the problem of the coulomp dampering force’s magnification, Smith predictor is introduced, and TLQG control and Smith predictor-LQG(SLQG) control are combined to design Smith predictor-Taylor series based LQG(STLQG) controller so as to conduct segmented time delay compensation. Contrastive analysis of control force shows: STLQG control can use SLQG control to alleviate the magnification phenomenon caused by TLQG control, and utilize TLQG control to reduce the horizontal shift caused by SLQG control.Finally, simulations are conducted on LQG control without time delay control measures, SLQG control, TLQG control, STLQG control and passive suspension to compare and analyze suspension performance under above control. Results show that STLQG control can effectively improve magneto-rheological semi-active suspension performance.The research results of the proposed STLQG control have both theoretical value and practical value, which can provide reference in further control theory research and experimental research of the magneto-rheological semi-active suspension.