基于余弦型伪随机信号的伺服阀辨识研究

Non-parameter Identification of Servo-valve Based on Cos-pseudorandom

针对M序列伪随机信号在电液伺服阀在线非参数辨识中存在状态有限的问题,提出将具有连续状态的余弦型伪随机信号作为输入信号,证明在不同信号幅值下余弦型伪随机信号具有最佳自相关特性,用于系统辨识中精度不受测量误差信号、直流分量信号的影响,具有精度高计算量小的优点,在相同序列长度下可以测试更多频率点。研究了在电液伺服阀非参数辨识中的余弦伪随机信号参数的选择方法,试验证明了余弦型伪随机信号在系统辨识中的优点。

To the problem that the function-states of M-pseudorandom is limited in the non-parameter identification of servo-valve on line, the cos-pseudorandom signal with continuous states is taken as the input signal to testify that the correlation of cos-pseudorandom is better, the precision of identification is higher, and the computational workload is small. The signal time period and the length of cos-pseudorandom are studled in the identification of electric-hydraulic servo valve. The experiment shows that the cos-pseudorandom is the optimization signal in the system identification.