Research On Intelligent Control Algorithm Based On MRE Shock Absorbing Structure

Building structure vibration control is not only an effective method, but also in line with the performance-based seismic concept now. Exploring the control devices and algorithm for civil engineering is the key to solve the structure vibration control. The damper of MRE is the ideal intelligent control device, and the MRE (short of Magnetorheological Elastomers) is in variation range of the elastic modulus and reversible rapidly. The intelligent control algorithm is suitable to solve the uncertain, nonlinear, complex problems, so it has a great theoretical value and practical significance of engineering application to research the MRE damping structure intelligent control.The parameter model of MRE damper in this article is based on the theoretical analysis and experimental test, and it is applied in the construction of structure vibration control. This article proposed a fuzzy-neural network intelligent control method based on MRE damping structure and its control effect is simulation analyzed by MATLAB/SIMULINK. Results are as follows:(1) Prepared magnetorheological elastomers whose relative MR effect can be achieved 175.2%. This elastomer is applied to make extrusion MRE damper. And the paper confirmed that the parameter model of MRE damper is to attach a viscoelastic model of magnetic-induced modulus to the Calvin model. All the work above is prepared for the seismic performance analysis of MRE damping structure.(2)The article explained the principles that how the fuzzy control algorithm improved by neural network in the aspects of fuzzy, defuzzification, membership function, fuzzy reasoning. And the paper proposed a fuzzy-neural network intelligent control method based on the MRE damping structure.(3) As an engineering example of nine steel structure, fuzzy control, neural network control and fuzzy-neural network control were used to analyse the seismic response of MRE damping structure under frequent and rare earthquake. Displacement and acceleration are the control index, and the article made a comparative analysis of fuzzy control effect, neural network control effect and fuzzy-neural network control effect. The results showed:the fuzzy-neural network control effect is the best, then the fuzzy control, and the neural network control effect is the last one.