机载光电稳定平台的模型辨识研究

Study on Model Identification of Airborne Electro- Optical Stabilized Platform

为了提高机载光电稳定平台伺服控制系统的稳定精度和设计效率,针对平台的数字化特点,提出一种数字化频率特性测试方法,并根据频率特性辨识出系统的数学模型。首先利用DSP控制器产生测试信号,自动采集测试数据并上传到计算机,然后利用数据处理减小摩擦对线性模型辨识的影响,最后分别利用Levy法和递推levy法进行参数辨识。实验结果表明:该数字化频率特性测试法自动化水平高;Levy法和递推Levy法在辨识精度和速度上各有优缺点。在实际应用中可以根据要求,选择合适的参数辨识方法。

In order to improve the stabilization precision and design efficiency of airborne Electro-Optical （EO） stabilized platform＆#39;s servo control system,a new method for frequency characteristic measurement was proposed based on the digitization characteristics of the EO stabilized platform,and the system model was identified by the measured frequency characteristics .A DSP controller was used to automatically generate input signal,collect testing data and transmit the tested data to the PC .The effect of friction on linear model identification was reduced by using data processing .Then,Levy and recursive Levy parameter identification algorithms were used for parameter identification .The results of experiment show that： 1） The frequency characteristic test method can improve the level of automation;and 2） Levy algorithm and recursive Levy algorithm have different advantages and disadvantages on parameter identification precision and speed .The parameter identification algorithm should be selected properly according to practical requirements .