基于分块Hankel矩阵的抽水蓄能电机扩展卡尔曼滤波模型子空间循环辨识

发布时间：2019-07-08 15:26

【摘要】：为提高抽水蓄能电机模型辨识结果的精度,提出一种基于分块Hankel矩阵的抽水蓄能电机扩展卡尔曼滤波(extended Calman filter,EKF)模型子空间循环辨识方法。首先,给出抽水蓄能电机磁场方程,建立基于EKF滤波器抽水蓄能电机模型,并利用电阻和电感的两组状态向量构建两组辨识模型,两组模型协同工作构建循环辨识算法;然后,引入子空间辨识策略,利用分块Hankel矩阵建立相关数据估算方程,利用零空间投影进行输入项删除,降低辨识复杂度,采用最小二乘实现对系统参数矩阵的有效辨识;通过在Matlab平台和硬件环境下的仿真实验,验证了所提算法在抽水蓄能电机转速、转子电阻、励磁电感等参数辨识上的有效性。
文内图片：

图片说明：转速误差结果
[Abstract]:In order to improve the accuracy of the model identification results of pumped storage motor, a subspace cycle identification method of pumped storage motor extended Kalman filter (extended Calman filter,EKF) model based on block Hankel matrix is proposed. Firstly, the magnetic field equation of pumped storage motor is given, and the model of pumped storage motor based on EKF filter is established, and two sets of identification models are constructed by using two sets of state vectors of resistance and inductance, and the cycle identification algorithm is constructed by working together between the two groups of models. Then, the subspace identification strategy is introduced, the related data estimation equation is established by using the block Hankel matrix, the input item is deleted by zero space projection, the identification complexity is reduced, and the least square is used to realize the effective identification of the system parameter matrix. Through the simulation experiments on Matlab platform and hardware environment, the effectiveness of the proposed algorithm in the parameter identification of pumped storage motor speed, rotor resistance, excitation inductance and so on is verified.
【作者单位】：哈尔滨理工大学电气与电子工程学院;
【基金】：国家自然科学基金项目(51407050)~~
【分类号】：TV734;TV743

【相似文献】