This paper proposes a cooperative distributed model predictive control (DMPC) to control the constrained interconnected nonlinear large-scale systems. The main contribution of this approach is its proposed novel cooperative optimization that improves the global cost function of any subsystem. Each subsystem calculates its optimal control by solving the corresponding global cost function. For each subsystem, the global cost function is defined based on a combination of cost functions of all subsystems. If the sampling time is selected appropriately, then the feasibility of the proposed approach will be guaranteed. Furthermore, the sufficient conditions for stability and consequently, for the convergence of the whole system states towards the neighborhood of the origin’s positive region are provided. The effectiveness and performance of the proposed approach are demonstrated via applying it to a nonlinear quadruple-tank system for both minimum-phase and nonminimum-phase models.

本文提出了一种协作分布式模型预测控制（DMPC）来控制受约束的互连非线性大规模系统。这种方法的主要贡献是它提出的新颖的协同优化，它改进了任何子系统的全局成本函数。每个子系统通过求解相应的全局代价函数来计算其最优控制。对于每个子系统，全局成本函数是基于所有子系统的成本函数的组合来定义的。如果采样时间选择得当，则该方法的可行性将得到保证。此外，还提供了稳定的充分条件，从而提供了整个系统状态向原点正区域附近收敛的充分条件。