Nonlinear control and on-line optimization of biochemical and chemical reactors

该研究计划的主要目标是开发生物化学和化学反应器的优化和非线性自适应控制算法。研究中的反应堆模型分为两类：由常微分方程描述的系统（集中参数系统）和由偏微分方程描述的系统（分布参数系统）。这些方程是从非稳态质量和热量平衡得到的。该算法是基于建设性的方法的基础上发展的控制李雅普诺夫函数。这些功能保证了控制系统的稳定性。在参数不确定的情况下，算法将不得不被修改，以包括参数自适应。化学反应器的算法的发展将集中在甲烷后燃烧的自循环反应器。这些反应器由偏微分方程描述。生化反应器算法的发展将集中在由常微分方程描述的系统上。将为植物细胞和动物细胞培养开发基于代谢工程的集总参数模型。这项工作的主要新奇是基于后集总的分布参数系统的自适应非线性控制算法的发展。后集总方法与更常见的前集总方法的不同之处在于，在设计控制器之后，将偏微分方程（包括控制器）的系统离散化以获得数值解。

The main objective of the research program is to develop optimization and nonlinear adaptive control algorithms for biochemical and chemical reactors. The reactors models under study are divided into two categories : systems described by ordinary differential equations (lumped parameter systems) and systems described by partial differential equations (distributed parameter systems). These equations are obtained from the unsteady state mass and heat balances. The algorithms are to be based on constructive approaches based on the development of control Lyapunov functions. These functions ensure stability of the control system. In the case of parameter uncertainty, the algorithms will have to be modified to include parameter adaptation. The development of algorithms for chemical reactors will focus on auto-cyclic reactors for methane post-combustion. These reactors are described by partial differential equations. The development of algorithms for biochemical reactors will concentrate on systems described by ordinary differential equations. Lumped parameter models based on metabolic engineering will be developed for plant cell and animal cell cultures. The main novelty of this work is on the development of adaptive nonlinear control algorithms for distributed parameter systems that are based on post lumping. The post lumping approach is different from the more common pre lumping approach in that the system of partial differential equations (including the controller) is discretized to obtain the numerical solution after the design of the controller.