Direct nonlinear controller design using virtual reference and support vector machine
用虚拟目标值和支持向量机直接设计非线性控制器

Indirect methods are usually employed to design control systems for complicated industrial processes, which is often time-consuming because the process model must be obtained first by identification methods. Motivated by the virtual reference feedback tuning (VRFT) method, we propose a new direct nonlinear controller design method using virtual reference (VR) and support vector machine (SVM). Firstly, the relation between VRFT and internal model control (IMC) is analyzed; and then, the structure and design procedure of the proposed nonlinear controller are given. Simulation results demonstrate that this method can effectively deal with the noise and nonlinearity, and can eliminate the steady-state error. Moreover, the amount of calculation is reduced significantly in comparison with the normal indirect model reference control method using neural networks.