非线性滑移滞回模型建模

目的为了对结构进行弹塑性动力分析以及评估其未来的性能，对结构的滞回特性需要建立合理的数学模型．方法利用三折线形式的函数关系来模拟试验滞回曲线的斜率与恢复力关系，并结合Duhem微分算子，建立描述滑移捏拢效应和刚度、强度退化的微分滞回模型．然后使用粒子群优化方法，基于最小二乘法准则识别出模型中的待定参数．结果用参数辨识结果得到的仿真滞回曲线与试验滞回曲线的对比表明，笔者建立的非线性滑移滞回模型与试验结果吻合较好，包括了强度退化、刚度退化以及捏拢效应．结论根据滞回曲线斜率与恢复力的关系并结合Duhem微分算子来构造滞回模型，比扩展Bouc—Wen模型具有更广泛的建模适应能力．对于试验滞回曲线的建模问题，当模型结构可辨识时，粒子群算法能够稳健地识别出模型中的参数．

Modeling of Nonlinear Hysteresis with Pinching

Based on the Duhem operator and the relationship between the slope of hysteresis loops and the restoring force,the modeling of the nonlinear hysteresis curves with pinching is studied.Firstly,a differential hysteresis model is established,and then the unknown parameters are identified by the Particle Swarm Optimization in accordance with the least square minimization of identification error.In comparison with the test loops,the proposed mathematical model can reproduce the hysteresis properties including strength degradation,stiffness degradation and pinching satisfactorily.The paper concludes that hysteresis model is constructed on the basis of the relationship of hysteresis curves with pinching,coupled with Duhem operator,which is suitable to be modeled,compared with the extended Bouc-wen model.When the model construction is identifiable,Particle Swam optimization can surely identify the parameters in the model.