On-line support vector machine-based building structure earthquake response intelligent control design method

Abstract

The invention discloses an on-line support vector machine (SVM)-based building structure earthquake response intelligent control design method, which is implemented by the steps of: 1, selecting a structural model, acquiring earthquake wave, performing earthquake response analysis on the structure, substituting in an LQG (Linear Quadratic Gaussian) controller, and calculating required optimal active control force; 2, arranging a plurality of MR (Magnetic Resonance) dampers for each floor of the structure according to a predetermined control level, determining MR damper parameters by taking the optimal active control force as target force and designing a structural LQG MR damper control system; and 3, estalibhsing a core module of a structure-SA (Semi-Active)-SVM-MR damper intelligent control system, and executing learning, training, prediction and control functions of on-line SVM. Numerical verification shows that the intelligent control system is more effective than the LQG MR damper control system. According to the method, the problem of poor structure acceleration response control effect or amplification of the LQG MR damper control system is solved, and a better control effect on structure acceleration response is achieved; and along with increasing floors of the structure, the intelligent control system has more obvious control advantage for the structure acceleration response.

Description

technical field [0001] The invention relates to the field of anti-seismic technology of building structures, in particular, it is a design method for establishing an intelligent control system of seismic response of building structures by using artificial intelligence technology such as support vector machines. Background technique [0002] At present, the magnetorheological (MR) damper is a kind of damper with good application prospect among various energy dissipation dampers widely studied at home and abroad. The fluid in the MR damper consists of a non-magnetic liquid and tiny soft magnetic particles with high magnetic permeability and low hysteresis uniformly dispersed in it. Under the action of a magnetic field, the liquid can quickly and reversibly transform from a Newtonian fluid with good fluidity to a Bingham plastic solid with high viscosity and low fluidity within milliseconds. When the current in the coil increases, the magnetic field in the orifice will increas...

AI Technical Summary

Problems solved by technology

However, in actual engineering, due to the large number of degrees of freedom in the actual structure, it is very difficult to arrange sensors on all degrees of freedom. At the same time, sensors and data acquisition systems are always affected by ambient noise

Therefore, it is often unrealistic and uneconomical to obtain all the state variables of the measurement system. Therefore, it is becoming more and more important to use new control methods to solve the above problems and further optimize the control effect of structural vibration response.

Images