光电跟踪架传递函数辨识方法及应用
|
摘要
鉴于传统的光电跟踪伺服控制系统频率响应测试方法复杂的缺点,采用了一种全数字化频率响应测量方案对光电跟踪架的频率特性进行测试。提出了一种递阶辨识法与传递函数参数辨识算法相结合进行传递函数辨识的新方法,并推导出了适用于不同阶次传递函数的辨识算法数学模型。在使用递阶辨识法辨识出系统阶次的基础上,利用测试得到的数据,分别采用最小二乘法和以新推导出的数学模型为基础的Levy法、Sanko法和Vinagre法辨识出了光电跟踪架的传递函数。辨识结果表明,新推导出来的数学模型正确,以其为基础的辨识算法误差均小于最小二乘法;3种算法中,Sanko法在整个频域内的辨识效果最好;采用递阶辨识原理与参数辨识算法相结合的方法可行,且精度较高。
Transfer function is of great importance to the design and analysis of photoelectrical tracking servo system.However,as the basement of transfer function identification,traditional frequency response test method is too complex.To solve this problem,a digital test method was adopted.In order to identify accurate transfer function relatively,a new method which was based on hierarchical identification method and parameter identification algorithms,was proposed,and the identification algorithm mathematic model for different rank transfer function was also deduced.Using data from frequency response test,transfer function identification algorithms of photoelectrical tracking system were identified by using the least square method,Levy method,Vinagre method and Sanko method.Results show that the deduced model is correct,the performance of the three methods based on it is better than that of least square method.Among Levy method,Vinagre method and Sanko method,Sanko method has the best performance.The
Transfer function is of great importance to the design and analysis of photoelectrical tracking servo system.However,as the basement of transfer function identification,traditional frequency response test method is too complex.To solve this problem,a digital test method was adopted.In order to identify accurate transfer function relatively,a new method which was based on hierarchical identification method and parameter identification algorithms,was proposed,and the identification algorithm mathematic model for different rank transfer function was also deduced.Using data from frequency response test,transfer function identification algorithms of photoelectrical tracking system were identified by using the least square method,Levy method,Vinagre method and Sanko method.Results show that the deduced model is correct,the performance of the three methods based on it is better than that of least square method.Among Levy method,Vinagre method and Sanko method,Sanko method has the best performance.The
引文
[1]Song Tingsong,Tong Shoufeng,Jiang Huilin,et al.Variablestructure control technology of the fine tracking assembly inairborne laser communication system[J].Infrared and LaserEngineering,2010,39(5):934-938.(in Chinese)宋延嵩,佟首峰,姜会林,等.机载激光通信系统精跟踪单元变结构控制技术[J].红外与激光工程,2010,39(5):934-938.
[2]Lu Ning,Ke Xizheng,Zhang Hua.Research on APT coarsetracking in free-space laser communication[J].Infrared andLaser Engineering,2010,39(5):934-938.(in Chinese)卢宁,柯熙政,张华.自由空间激光通信中APT粗跟踪研究[J].红外与激光工程,2010,39(5):934-938.
[3]Hu Haojun,Ma Jiaguang,Wang Qiang,et al.Transferfunction identification in a fast steering mirror system[J].Opto-Electronic Engineering,2005,32(7):1-3.(in Chinese)胡浩军,马佳光,王强,等.快速控制反射镜系统中的传递函数辨识[J].光电工程,2005,32(7):1-3.
[4]Cascone E,Mancini D,Schipani P.Galileo telescope modelidentification[C]//SPIE,1997,3112:343-350.
[5]Min Yuejun,Xue Xu.Parameter identification method basedon frequency domain of flexure hinged pendulumaccelerometer[J].Journal of Chinese Inertial Technology,2008,16(1):94-103.(in Chinese)闵跃军,薛旭.基于频域的挠性摆式加速度计表头参数辨识方法[J].中国惯性技术学报,2008,16(1):94-103.
[6]Zhao Xia,Yao Yu,Fang Qiang.The study on thehierarchical identification method in the debugging ofturntable servo system[J].Control Theory and Applications,2002,19(2):229-234.(in Chinese)赵霞,姚郁,方强.递阶辨识方法在转台伺服系统调试中的应用研究[J].控制理论与应用,2002,19(2):229-234.
[7]Wang Shuai,Chen Tao,Li Hongwei,et al.Frequencycharacteristic test and model identification for O-E trackingservo system[J].Opt and Precision Eng,2009,17(1):79-82.(in Chinese)王帅,陈涛,李洪文,等.光电跟踪伺服系统的频率特性测试与模型辨识[J].光学精密工程,2009,17(1):79-82.
[8]Li Hongwen,Li Yuanchun.Neural network PID controlbased on model identifier for the odolite[J].Infrared andLaser Engineering,2006,35(S):442-446.(in Chinese)李洪文,李元春.基于模型辨识的神经网络PID控制在经纬仪中的应用[J].红外与激光工程,2006,35(S):442-446.
[9]Levy E C.Complex curve fitting[J].IRE Transaction onAutomatic Control,1959,4(3):37-43.
[10]Vinagre B M,Podlubny I,Hern譧ndez A,et al.Someapproximations of fractional order operators used in controltheory and applications[J].Fractional Calculus&AppliedAnalysis,2000,3:231-248.
[2]Lu Ning,Ke Xizheng,Zhang Hua.Research on APT coarsetracking in free-space laser communication[J].Infrared andLaser Engineering,2010,39(5):934-938.(in Chinese)卢宁,柯熙政,张华.自由空间激光通信中APT粗跟踪研究[J].红外与激光工程,2010,39(5):934-938.
[3]Hu Haojun,Ma Jiaguang,Wang Qiang,et al.Transferfunction identification in a fast steering mirror system[J].Opto-Electronic Engineering,2005,32(7):1-3.(in Chinese)胡浩军,马佳光,王强,等.快速控制反射镜系统中的传递函数辨识[J].光电工程,2005,32(7):1-3.
[4]Cascone E,Mancini D,Schipani P.Galileo telescope modelidentification[C]//SPIE,1997,3112:343-350.
[5]Min Yuejun,Xue Xu.Parameter identification method basedon frequency domain of flexure hinged pendulumaccelerometer[J].Journal of Chinese Inertial Technology,2008,16(1):94-103.(in Chinese)闵跃军,薛旭.基于频域的挠性摆式加速度计表头参数辨识方法[J].中国惯性技术学报,2008,16(1):94-103.
[6]Zhao Xia,Yao Yu,Fang Qiang.The study on thehierarchical identification method in the debugging ofturntable servo system[J].Control Theory and Applications,2002,19(2):229-234.(in Chinese)赵霞,姚郁,方强.递阶辨识方法在转台伺服系统调试中的应用研究[J].控制理论与应用,2002,19(2):229-234.
[7]Wang Shuai,Chen Tao,Li Hongwei,et al.Frequencycharacteristic test and model identification for O-E trackingservo system[J].Opt and Precision Eng,2009,17(1):79-82.(in Chinese)王帅,陈涛,李洪文,等.光电跟踪伺服系统的频率特性测试与模型辨识[J].光学精密工程,2009,17(1):79-82.
[8]Li Hongwen,Li Yuanchun.Neural network PID controlbased on model identifier for the odolite[J].Infrared andLaser Engineering,2006,35(S):442-446.(in Chinese)李洪文,李元春.基于模型辨识的神经网络PID控制在经纬仪中的应用[J].红外与激光工程,2006,35(S):442-446.
[9]Levy E C.Complex curve fitting[J].IRE Transaction onAutomatic Control,1959,4(3):37-43.
[10]Vinagre B M,Podlubny I,Hern譧ndez A,et al.Someapproximations of fractional order operators used in controltheory and applications[J].Fractional Calculus&AppliedAnalysis,2000,3:231-248.