论文部分内容阅读
考虑了具有执行器饱和的大规模离散时间线性系统分散控制器的设计。首先进行了在执行器幅值饱和的情形下的研究,然后延伸到执行器具有多层饱和的情况,例如,幅值和速率同时存在饱和或通过多层神经元网络近似的执行器非线性。在这2种情况下,给出了闭环系统在分散状态反馈律的作用下,椭球收敛不变性的条件。基于这些条件,可取得大吸引域的分散状态反馈控制律的设计可以归结为具有双线性矩阵不等式(BM I)约束的优化问题。对这些双线性约束优化问题提出了数值算法。数值算例显示了所提出的设计方法的有效性。
Consider the design of decentralized controllers for large-scale discrete-time linear systems with actuator saturation. The first study was carried out in the case of actuator amplitude saturation and then extended to cases in which the actuator had multiple layers of saturation, for example, with both amplitude and velocity saturation or actuator nonlinearities approximated by a multi-layer neural network. In these two cases, the condition of the invariance of the ellipsoid convergence under the decentralized feedback law of the closed-loop system is given. Based on these conditions, the design of decentralized state feedback control law for large attracting domains can be attributed to the optimization problem with bilinear matrix inequality (BM I) constraint. A numerical algorithm is proposed for these bilinear constrained optimization problems. Numerical examples show the effectiveness of the proposed design method.