周期激励下的风机塔架模态识别方法

Modal Identification Method ofWind Turbine Towers under Periodic Excitations

辨识模态参数是准确获得风机塔架结构动态特性的基础。由于风力机叶片上的重力和风切变引起的气动载荷都会周期性地改变大小和方向及受到桨叶螺距等周期载荷的影响,运行中的风力机表现为线性时间-周期(Linear time-periodic)系统。提出一种基于响应信号的运行中风机塔架工作模态分析辨识方法。以长度为60 m的风机塔架结构为研究对象,利用Periodic past output multivariable output-error state space(简称Periodic PO-MOESS)算法,对受到周期激励信号作用的结构进行模态辨识,并比较辨识结果和仿真结果的差异。讨论周期激励下该算法辨识结构模态参数的可行性,并分析在白噪声工况下的辨识结果精度,结果表明该算法可以有效识别运行中风力机系统的模态参数,且具有良好抗噪性能。

Modal parameter identification is the basis for obtaining the dynamic characteristics of wind turbine tower structures.Due to the effects of periodical changes in both magnitudes and directions of the gravity and the wind shearing induced air-dynamic loads acting on the blades as well as the influences of the blade pitch and other periodic loads,the wind turbine in operation behaves as a linear time-periodic(LTP)system.In this paper,a modal analysis identification method for wind turbine structure in operation based on response signal is proposed.With a 60m high wind turbine structure as the research object,the periodic past output multivariable output-error state space(Periodic PO-MOESS)algorithm is used to identify the modals of the wind turbine structure under the periodic excitation.The errors of identification results and simulation results are compared.The feasibility of the algorithm is analyzed and the accuracy of the results under white noise condition is discussed.The results show that the algorithm can effectively identify the modal parameters of the wind turbine system in operation and has a good noise resistance performance.