Identification Of Model-free Nonlinear Behaviors Of Base-Isolated/MR Dampers And Structures

Under strong external excitation,most engineering structures exhibit nonlinearities such as base-isolated structures,vibration reduction structures with dampers,damage structures,high-rises,and super-tall buildings.The base-isolated structure has been widely used in seismic response control of major infrastructure and engineering structures.The towering structure and the cable-stayed bridge structure are soft and have small damping,and the wind-induced vibration response is very large.Vibration reduction needs to be achieved by installing dampers.The engineering structure will be severely damaged under strong external excitation,such as steel yield,cracking or crushing of concrete.At this time,the relationship between the restoring force and deformation of the nonlinear component is nonlinear.Considering the structural safety,reliability,and economical durability,the information of the nonlinear characteristics of these special structures not only help to accurately reflects the mechanical properties,but also can be used in assessment of mechanical properties,degree of damage,and service life of the entire process of operation,maintenance and repair.At present,the main limitations of the researches about nonlinear behavior identification are as follows:1)It's needed to know the hysteretic nonlinear model or need to expand the nonlinear model approximately;2)the locations and quantity of the sensor are more demanding.In the actual projects,the hysteretic nonlinear model of the system is very complex and it is difficult to accurately describe it with the theoretical model.Considering engineering economics,the use of sensors should be minimized on the premise of ensuring good identification effect;in addition,the sensors installed in key locations(such as isolation layer,damper installation layer,and structural damage layer)may breakdown.At this time,it is necessary to identify nonlinear characteristics without the response data of these sensors.This thesis summarizes the current research progress of nonlinear system identification,and studies the method of identifying the model-free nonlinear characteristics of base-isolated structures,vibration reduction structures with dampers,damage structures and large-scale high-rise building structures.The proposed methods only need partial measurements of structural responses,and overcome the limitations of the need to observe the acceleration response at key locations.Chapter 1 of this thesis summarizes the research progress of nonlinear system identification.A series of Kalman filter methods which this thesis is based on are briefly described,including classic Kalman filter method(KF)and classic extended Kalman filter method(EKF),extended Kalman filter with unknown input(EKF-UI),Kalman filter with unknown input(KF-UI),generalized Kalman filter with unknown input(GKF-UI),generalized extended Kalman filter with unknown input(GEKF-UI).In Chapter 2,an identification method is proposed for model-free nonlinear characteristics of base-isolated structural rubber bearings.The main idea is as follows:First,under the weak ground motion,the overall structure with the isolation rubber bearings is in a linear state,the linear stiffness and viscous damping parameters of the isolation rubber bearing and the linear stiffness and viscous damping of the superstructure are identified by the EKF method.Then,under strong ground motion,the seismic isolation bearing enters into a nonlinear state,and the superstructure is still in a linear state.The plastic restoring force generated by the isolation bearing is considered as an unknown 'additional virtual force' imposed on the overall linear system,and the unknown 'additional virtual force' is identified by GKF-UI method.The sum of plastic restoring force and linear elastic restoring force restoring forces is the total restoring force of the isolation bearing in the non-linear state.The proposed method only need partial measurements of structural responses,and it is applicable to cases that either acceleration responses at the locations of nonlinear forces are measured or not.In Chapter 3,an identification method is proposed for the model-free nonlinear characteristics of MR dampers.The main idea is:First,under the weak external excitation,the overall system with dampers is in a linear state,and the linear stiffness,Rayleigh damping coefficient,linear stiffnesses and viscous damping of the dampers are identified by using EKF method or EKF-UI method.Then,under the strong external excitation,the damper enters into the nonlinear state,the structure is still in a linear state,and the total restoring force generated by the damper is regarded as an unknown'additional virtual force' imposed on the linear structure.Then use KF-UI to identify the total restoring force generated by the damper.The proposed method can also identify the unknown external excitation at the same time even if the external excitation is unknown.The proposed method only need partial measurements of structural responses.In Chapter 4,an identification method is proposed for structural model-free nonlinear characteristics.The main idea is as follows:First,under the weak external excitation,the structure is not damaged and in a linear state.EKF method or EKF-UI method is used to identify the linear stiffnesses and damping parameters of the linear structure.Then,under the strong external excitation,the structure enters a nonlinear state due to the damaged or yielding components.Using the nonlinear elements locating method,the nonlinear elements can be located;finally,the nonlinear forces generated by the damaged or yielding components are regarded as the unknown 'additional virtual force' imposed on a linear structure,and the unknown forces are identified by GKF-UI method.The proposed method can also identify the unknown external excitation at the same time if the external excitation is unknown.The proposed method only need partial measurements of structural responses,and it is applicable to cases that either acceleration responses at the locations of nonlinear forces are measured or not.In Chapter 5,the identification method of structural model-free nonlinear characteristics is combined with the substructure method,which is applied to the large-scale structural nonlinear characteristics identification.Under weak external excitation,the overall structure is a linear structure,and it is divided into a series of linear substructures,the boundary forces of each substructure are regarded as unknown inputs imposed on the substructure,and GEKF-UI method is used to identify the linear stiffness and damping parameters of each substructure parallelly to obtain the linear stiffness and damping parameters of the entire structure.Under strong excitation,the structure partially enters nonlinear state.Using the nonlinear elements locating method to the overall structure,the nonlinear elements can be located.The nonlinear elements are divided into the target substructure,and the nonlinear forces generated by the nonlinear components and the boundary forces of the target substructure are regarded as the 'additional virtual force' imposed on the linear target substructure,which can be identified by GKF-UI method.The proposed method can also identify the unknown external excitation at the same time even if the external excitation is unknown.The proposed method only need partial measurements of structural responses.Chapter 6 summarizes the main work,the obtained conclusions and the innovation points of this thesis,looking forward to the research content in the future.