基于Wiener模型的非线性预测函数控制

针对传统的预测函数控制对非线性强、结构多变的实际对象控制效果不佳问题,提出了一种基于Wiener模型的非线性预测函数控制(PFC)方法。通过引入支持向量机回归方法,建立Wiener模型中非线性部分的逆模型,在输出反馈和参考轨迹上分别添加一个Wiener模型中非线性部分的逆模型,将非线性预测函数控制转化为线性预测函数控制,用线性优化算法解决非线性预测函数控制问题,避免了复杂的非线性优化。pH中和过程的计算机仿真结果表明,该方法比PID控制效果更好,而且对模型失配具有很好的自适应性能。

Approach to nonlinear predictive functional control using wiener models

The traditional predictive function control(PFC) can't control the nonlinear and changeful object effectively;a nonlinear PFC approach based on Wiener model was developed.Support vector machine regression(SVMR) was applied to model the inverse of nonlinear part of Wiener models,add this model to output feedback and reference trajectory route.The nonlinear control problem was simplified to a linear one by using the inverse of nonlinearity,and complicated nonlinear optimization was avoided.The simulation results of pH neutralization process show that this approach is more effective than PID,and is robust to model mismatch.