多变量生物发酵过程的解耦控制

把非线性系统的逆系统方法与神经网络非线性辨识技术相结合 ,提出了一种基于神经网络逆系统的发酵过程多变量解耦控制策略 .当过程模型缺乏足够的先验知识时 ,所提出的解耦控制策略对多变量非线性连续发酵过程取得了良好的控制性能 .仿真结果表明 ,提出的解耦控制方法能够适应过程模型的不确定性和参数的时变性 ,具有较强的鲁棒性 .克服了基于微分几何理论的逆系统解耦控制方案依赖于过程模型和对模型参数变化很敏感的缺点 .

Decoupling control of multivariable biologic fermentation process

Based on neural network inverse system a decoupling control strategy for multiva riable fermentation process is proposed, in which the inverse system linearizing method of the nonlinear system is combined with the nonlinear identifying techn ology of the neural networks. When adequate prior information concerning the dy namics of the fermentation processes is not available and parameters of the proc esses change with time, the good control performance can be obtained by using th e proposed strategy. The simulation experiments demonstrate that the presented d ecoupling approach is more effective than inverse system method based on dif ferential geometry, because in inverse system method the design of the controlle r relies on the exact process model and its control performance is sensitive to the parameters of the model, which results in low accuracy and poor robust.