非线性预测函数控制研究与应用

针对非线性预测函数控制,分析了基于智能模型的非线性预测模型的优缺点,提出基于特征模型的非线性预测模型。其结合被控对象的动力学特征和控制性能要求,用一个二阶的慢时变线性模型表征原先的被控对象,适用于高阶线性对象和一般的非线性对象。将特征模型作为非线性预测模型,即可获得基于特征模型的非线性预测函数控制算法。提出利用测试建模法获取特征模型,通过在不同的工作条件下测试非线性对象,获得一族曲线,利用系数区间表示特征方程中的系数,避免了特征模型慢时变参数的在线辨识问题,从而使该算法具有运算速度快、控制性能优、适用范围广的特点。结合连续搅拌式罐反应器的温度控制和电厂球磨机入口负压的控制开展了应用研究。

Study and application of nonlinear predictive functional control

The intelligent model-based nonlinear predictive models for nonlinear predictive functional control are analyzed,and a typical nonlinear predictive model based on characteristic model is presented.Combining the dynamical feature of object to be controlled with the requirements of control performance,it characterizes the object with a second-order slow-time-varying linear model, suitable for both high-order linear and general nonlinear objects.Take the characteristic model as the nonlinear predictive model to achieve the nonlinear predictive functional control arithmetic.The test way is used to acquire the characteristic model,i.e.,the nonlinear object is tested under different working conditions to obtain a cluster of curves.The coefficient intervals are used to represent the coefficients of characteristic formula to avoid the on-line identification of slow-time -varying parameters.The control arithmetic has high operation speed,excellent control performance and wide applicability.Its applications in the temperature control of continuous stirring tank reactor and the inlet minus pressure control of ball mill are provided.