Distributed linear-quadratic control with graph neural networks
نویسندگان
چکیده
Controlling network systems has become a problem of paramount importance. In this paper, we consider distributed linear-quadratic and propose the use graph neural networks (GNNs) to parametrize design controller for systems. GNNs exhibit many desirable properties, such as being naturally scalable. We cast self-supervised learning problem, which is then used train GNN-based controllers. also obtain sufficient conditions resulting closed-loop system be input-state stable, derive an upper bound on how much trajectory deviates from nominal value when matrices that describe are not accurately known. run extensive simulations study performance controllers show they computationally efficient
منابع مشابه
Linear matrix inequality approach for synchronization of chaotic fuzzy cellular neural networks with discrete and unbounded distributed delays based on sampled-data control
In this paper, linear matrix inequality (LMI) approach for synchronization of chaotic fuzzy cellular neural networks (FCNNs) with discrete and unbounded distributed delays based on sampled-data controlis investigated. Lyapunov-Krasovskii functional combining with the input delay approach as well as the free-weighting matrix approach are employed to derive several sufficient criteria in terms of...
متن کاملRobust Linear Quadratic Control of Systems with Integral Quadratic Constraints
In this paper, we consider the robust linear quadratic (LQ) control problem for a class of uncertain linear systems which are subject to a general type of integral quadratic constraints (IQCs). Both analysis and synthesis problems are considered in this paper. For the analysis problem, we determine if the system satisses a desired linear quadratic performance index for all admissible uncertaint...
متن کاملLinear Quadratic Optimal Control
In previous lectures, we discussed the design of state feedback controllers using using eigenvalue (pole) placement algorithms. For single input systems, given a set of desired eigenvalues, the feedback gain to achieve this is unique (as long as the system is controllable). For multi-input systems, the feedback gain is not unique, so there is additional design freedom. How does one utilize this...
متن کاملLinear-quadratic Optimal Control with Integral Quadratic Constraints
We derive closed-form solutions for the linear-quadratic (LQ) optimal control problem subject to integral quadratic constraints. The optimal control is a non-linear function of the current state and the initial state. Furthermore, the optimal control is easily calculated by solving an unconstrained LQ control problem together with an optimal parameter selection problem. Gradient formulae for th...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Signal Processing
سال: 2022
ISSN: ['0165-1684', '1872-7557']
DOI: https://doi.org/10.1016/j.sigpro.2022.108506