自适应逆控方法的无人艇航向控制

Heading Control of Unmanned Surface Vehicle Based on Adaptive SVR Inverse Method

无人水面艇(Unmanned Surface Vehicle,USV)的复杂性、非线性和多变量等特征使得其控制问题一直是难点和热点。支持向量回归机(Support Vector Regression,SVR)作为模型控制问题的一种新方法,能避免过学习、陷入局部极小点,获得全局最优解,有很大的发展前景。将SVR引入USV的航向控制,提出自适应SVR逆控制方法,以输入输出反馈线性化理论为基础,进行逆动态模型和逆误差补偿项的离线辨识,并将辨识的逆模型作为控制器,构造出直接自适应逆控制系统模型。最后通过仿真实验表明该控制方案具有良好的动态响应性能和控制效果。

The features of complexity,nonlinearity and multivariable of unmanned surface vehicle（USV） model have made its control the difficult and hot point.The support vector regression（SVR）,as a promising new method for solving model control problem,can effectively give a global optimum solution avoiding over-learning and local minimum trapping.An adaptive SVR algorithm for heading control of USV is proposed.Based on input–output feedback-linearization theory,the off line identification of inverse dynamic model,as well as the compensation term for the inversion error,is conducted.The identified inverse model is used as a controller to construct a model of direct adaptive inverse control system.Numerical simulations indicate that the proposed control scheme has good dynamic response performance and control effectiveness.