基于LS—SVM的机载天线伺服机构自适应控制

摘要

The airborne antenna servo mechanism runs in the conditions that the external force is time-varying changes fast during the whole scope, and its own friction and other nonlinear characteristics cannot be ignored, the paper propose the adaptive control designing method based on Least Squares Support Vector Machine (LS-SVM). This method regarded the nonlinear charaeteristies such as the friction as an unknown item, and estimated the nonlinear unknown item of the controlled object by the LS-SVM. Then, the learning rule of the LS-SVM was designed. The results show that: the method can well solve the problem of system movement stability within the whole scope, and stable track any smooth input signal online. Meanwhile, the method can eompensate for the nonlinear characteristics such as friction, and improve the dynamic quality and static characteristics.%针对机栽天线伺服机构运行时所受外力是快时变、变化范围大、且自身摩擦等非线性特性不能忽略的问题，提出一种基于最小二乘支持向量机（LS—SVM）的自适应控制设计方法。此方法将摩擦等非线性特性视为未知项，并利用LS—SVM在线估计被控对象的非线性未知项，从而设计得到LS—SVM的学习规则。系统运行结果证明：该方法能够很好地解决全范围内的系统稳定问题，并能在线跟踪任何光滑的输入信号，同时补偿摩擦等非线性特性，改善系统的动态品质和静态特性。