【作者】 史鹏飞；

【导师】 吴斌； BillieF.Spencer；

【作者基本信息】 哈尔滨工业大学 ， 防灾减灾工程及防护工程， 2011， 博士

【摘要】 磁流变阻尼半主动控制能够取得与主动控制相当的控制效果,却不需要大量的能源输入,在很长一段时间内是结构振动控制研究的热点。拟负刚度控制方法的控制力由阻尼力和“负刚度”控制力组成,其中,阻尼力部分可由磁流变阻尼器出力中的粘滞阻尼力部分实现,“负刚度”控制力可以通过调节磁流变阻尼器的驱动电压实现。磁流变阻尼器为速度相关型控制装置且具有非常强的非线性特性,拟动力试验和地震模拟振动台试验等试验方法均很难满足检验磁流变阻尼控制系统性能的要求。作为检验磁流变阻尼控制系统性能的一种重要手段,实时混合试验方法引起了很多学者的关注。但是,对实时混合试验来说,很难在一个时间步长内实现结构响应的计算、驱动试验子结构达到预定的速度以及对试验子结构的反力进行测量和反馈。通常,实时混合试验中存在的时滞会引起试验结果的不准确甚至系统的不稳定。因此,实时混合试验时滞补偿方法的研究具有重要的意义。本文对采用拟负刚度控制的结构的动力特性和减振效果、多自由度结构拟负刚度控制及其控制效果、测量位移对等效力控制方法的影响及基于试验系统模型的等效力控制方法进行了研究。1.证明了拟负刚度阻尼减振结构和拟负刚度与粘滞阻尼混合减振结构均为齐次非线性结构,其齐次性保证了可以通过位移响应系数、传力系数及反应比谱对两种结构的动力特性及减振效果进行分析。位移响应系数、传力系数和反应比谱的分析结果表明拟负刚度控制能够延长结构的等效周期；结构周期较长时,与不同阻尼比的结构相比,拟负刚度阻尼减振结构的加速度更小而位移较大；拟负刚度与粘滞阻尼混合减振结构的位移和加速度均要小于不同阻尼比的结构。2.以一栋安装磁流变阻尼器的隔震结构为研究对象,对拟负刚度控制进行了数值和实时混合试验研究。证明了采用拟负刚度与粘滞阻尼混合控制的隔震结构同样为齐次非线性结构,拟负刚度与粘滞阻尼混合控制能够在不增加底部剪力的同时减小结构的位移。稳定性和时滞影响分析表明实时混合试验能够用于拟负刚度与粘滞阻尼混合控制系统性能的研究。试验结果表明：拟负刚度与粘滞阻尼混合控制对结构位移和加速度的控制效果均好于Passive-off控制；对加速度的减振效果好于Passive-on控制,而对位移的控制效果差于Passive-on控制。渤海JZ20-2NW海洋平台拟负刚度与粘滞阻尼混合控制的数值分析同样表明拟负刚度与粘滞阻尼混合控制系统具有较好的减振效果。3.实时混合试验等效力控制方法以力反馈控制环代替隐式积分算法的迭代计算,该控制环还可以对实时混合试验系统中的时滞进行补偿。但是,还有其他一些因素会影响到等效力控制方法的时滞补偿效果,例如采用作动器位移命令或响应计算数值子结构恢复力和拟动力、试验子结构位移中存在的测量噪声等。本文对分别采用作动器位移命令和响应计算数值子结构恢复力和拟动力、测量噪声对等效力控制方法的影响进行了分析。为了保证试验子结构恢复力、数值子结构恢复力和拟动力的同步,必须采用作动器位移响应计算数值子结构恢复力和拟动力。PID等效力控制器的比例增益较大时,作动器位移响应中的测量噪声会导致试验子结构反力中存在不可忽略的高频成分。为了减小测量噪声的影响,本文采用Kalman滤波器对测量噪声进行滤波,从而提高PID控制器比例增益的取值,改善等效力控制方法的时滞补偿效果。实时混合试验结果表明采用Kalman滤波器的等效力控制方法能够减小测量噪声对试验子结构反力的影响并有效地补偿时滞,其补偿效果与基于模型的时滞补偿方法基本相同。4.提出并研究了基于试验系统模型的等效力控制方法。该方法在试验系统模型的基础上,利用开环补偿或者闭环补偿方法,减小等效力命令与响应之间的时滞,使等效力响应能够更好地跟踪等效力命令,从而提高等效力控制方法的时滞补偿效果。本文分别以弹簧和磁流变阻尼器为试验子结构研究了单自由度和多自由度结构基于试验系统模型的等效力控制方法的时滞补偿效果,并与等效力控制方法的时滞补偿效果进行了比较。分析结果表明：基于试验系统模型的等效力控制方法的时滞补偿效果更好。200kN磁流变阻尼器的实时混合试验同样表明基于试验系统模型的等效力控制方法能够有效的对时滞进行补偿。更多还原

【Abstract】 Semi-active control with magnetorheological (MR) fluid damper can obtain a similar control effect with active control without huge power input. It has been a hotspot of structural control researches for a long period. The control force of pseudo-negative stiffness control can be represented by damping force and "negative stiffness" control force. The damping force can be achieved by the viscous damping element of MR damper force, while the "negative stiffness" control force can be achieved through changing the drive voltage of MR damper. Due to the velocity-dependent and strong nonlinear properties of MR damper, it is difficult for pseudo-dynamic test and shaking table test to verify the performances of control systems with MR dampers. As an important method to evaluate the performance of the control system with MR damper, the Real-time Hybrid Testing (RHT) has got a lot of attentions. However, it is difficult to calculate the response of the structure, drive physical substructure to reach the desired velocity, and measure and feedback the restoring force of physical substructure in a single time step for RHT. Usually, time delay exists in the RHT can introduce inaccurate result even instability to the system. Therefore, it is significant to investigate the time delay compensation for RHT.In this study, the dynamic characteristics and control effects of structures with pseudo-negative stiffness dampers, pseudo-negative stiffness control and its control effect for MDOF structure, influence of measured displacement on Equivalent Force Control (EFC) method and model-based EFC method were studied.1. It has been proved that structures with pseudo-negative stiffness dampers and pseudo-negative stiffness and viscous dampers are nonlinear homogeneity structures. The homogeneity ensures that dynamic characteristics and control effects can be analyzed through transmissibility, deformation response factor and response ratio spentrum. The analyses of transmissibility, deformation response factor and response ratio spectrum show that pseudo-negative stiffness control can extend the equivalent period of structure. For structures with longer periods, acceleration of structure with pseudo-negative stiffness dampers is smaller than that of structures with different damping ratios, while the displacement is larger. Both of displacement and acceleration of structure with pseudo-negative stiffness and viscous dampers are smaller than that of structures with different damping ratios.2. Pseudo-negative stiffness control with a five-story isolated structure incorporating MR damper was studied through analytical and RHT methods. It is proved that isolated structure with pseudo-negative stiffness and viscous dampers is also nonlinear homogeneity structure. Pseudo-negative stiffness and viscous damper control can reduce displacement without increasing the base shear. The analyses of stability and influence of time delay show that RHT technique is sufficient for the study of the performance of pseudo-negative stiffness and viscous damper control system. The test results show that control effects of displacement and acceleration with pseudo-negative stiffness and viscous damper controller are better than that of passive-off controller. The pseudo-negative stiffness and viscous damper controller can reduce the acceleration more than passive-on controller while control effect of displacement is worse than passive-on controller. The numerical analysis of Bohai JZ20-2NW offshore platform with pseudo-negative stiffness and viscous dampers also show that the pseudo-negative stiffness and viscous damper control system has good control effect.3. The EFC method has been developed for RHT to replace the numerical iteration for implicit integration with a force-feedback control loop. With this control loop, the EFC method can also compensate the time delay in RHT. However, some other factors may influence the delay compensation effect of EFC method, such as calculating restoring force of numerical substructure and pseudo-dynamic force using actuator displacement command or response, measured noise in the displacement of physical substructure. This study analyzes the influence of calculating restoring force of numerical substructure and pseudo-dynamic force using actuator displacement command and response, respectively, and the influence of measured noise on EFC method. To ensure the synchronization of restoring forces of numerical substructure and physical substructure and psudo-dynamic force, actuator displacement response have to be used to calculate restoring force of numerical substructure and pseudo-dynamic force. With a higher proportional gain of PID equivalent force (EF) controller, measured noise on actuator displacement response can introduce component with high frequency, which can not be disregarded, into the restoring force of physical substructure. To reduce the influence of the measured noise, a Kalman filter was employed to filter the noise in this study. A higher proportional gain of PID controller can be obtained with the Kalman filter, which improves the effect of time delay compensation of EFC method. The results of RHTs demonstrate that EFC method with Kalman filter can reduce the influence of measured noise on resoring force of physical substructure and effectively compensate the time delay, and its effect is similar with that of model-based compensation method.4. The model-based EFC method was proposed and studied. This method reduces the time delay between the EF command and EF response using the open-loop or closed-loop compensations based on the model of experimental system. Then, the EF response can track the EF command well which can improves time delay compensation effect of EFC method. Time delay compensation effects of SDOF and MDOF model-based EFC methods were studied with spring and MR damper specimens and compared with that of EFC method, respectively. The results show that the time delay compensation effect of model-based EFC method is better than that of EFC method. The test results of RHTs with a 200kN MR damper specimen also show that the model-based EFC method can effectively compensate the time delay.更多还原