基于DBFNN的非线性自适应控制在CSTR中的应用

讨论了一类放射非线性系统的自适应控制问题.首先对于利用方向基神经网络(DBFNN)对系统的不确定性进行建模.所得到的系统模型作为对象的数学模型用来设计控制器.首先设计了反馈线性化控制器,为了克服建模误差的影响又引入了内模控制机制.证明了只要网络的学习精度足够高,所设计的闭环系统是最终一致有界的.把所设计的控制策略用于CSTR的控制中,仿真结果表明了所设计的控制器有效性.

Application of Nonlinear Adaptive control Based on DBFNN for CSTR

Adaptive control using DBFNN (Direction Basis Function Neural Network) of a class of nonlinear systems is discussed in this paper. First, DBFNN is used to model the uncertainties of nonlinear systems. Then the acquired NN model is used as system model to and a feedback linearization controller is designed based the trained well network. In order to overcome the model mismatch the internal model control (IMC) mechanism is introduced in. It is proved the closed-loop system is uniformly ultimately (UUB). The proposed control strategy is application one CSTR (continuous stirred tank reactor) system. Simulations show the proposed strategy is effective.