In this paper, a novel cooperative constrained distributed model predictive control algorithm is proposed to control the nonlinear interconnected constrained large-scale systems. In this algorithm, a novel reduced-order cooperative optimisation approach is proposed which is its main contribution that reconstructs and improves the global cost function of any local controller. In proposed algorithm, each local controller computes its optimal control by minimising the corresponding global cost function which is a combination of its own and its neighbouring subsystems’ cost functions. The sufficient conditions are derived to guarantee the recursive feasibility and closed-loop stability specifications to ensure the convergence of the overall states into the positive region which is the neighbourhood of origin. The performance of the proposed algorithm is illustrated via simulation results of a nonlinear large-scale cart-spring-damper system.

本文提出了一种新的协作约束分布式模型预测控制算法来控制非线性互联约束大规模系统。在该算法中，提出了一种新颖的降阶协同优化方法，其主要贡献是重构和改进任何本地控制器的全局成本函数。在所提出的算法中，每个本地控制器通过最小化相应的全局成本函数来计算其最优控制，全局成本函数是其自身及其相邻子系统成本函数的组合。推导出了保证递归可行性和闭环稳定性规范的充分条件，以确保整体状态收敛到作为原点邻域的正区域。