Abstract:Identification methods for dynamic networks typically require prior knowledge of the network and disturbance topology, and often rely on solving poorly scalable non-convex optimization problems. While methods for estimating network topology are available in the literature, less attention has been paid to estimating the disturbance topology, i.e., the (spatial) noise correlation structure and the noise rank in a filtered white noise representation of the disturbance signal. In this work we present an identification method for dynamic networks, in which an estimation of the disturbance topology precedes the identification of the full dynamic network with known network topology. To this end we extend the multi-step Sequential Linear Regression and Weighted Null Space Fitting methods to deal with reduced rank noise, and use these methods to estimate the disturbance topology and the network dynamics in the full measurement situation. As a result, we provide a multi-step least squares algorithm with parallel computation capabilities and that rely only on explicit analytical solutions, thereby avoiding the usual non-convex optimizations involved. Consequently we consistently estimate dynamic networks of Box Jenkins model structure, while keeping the computational burden low. We provide a consistency proof that includes path-based data informativity conditions for allocation of excitation signals in the experimental design. Numerical simulations performed on a dynamic network with reduced rank noise clearly illustrate the potential of this method.

Network Identification for Diffusively-Coupled Systems with Minimal Time Complexity

Identifiability and Identification of Switching Dynamical Networks: A Data-Based Approach

On Necessary and Sufficient Conditions for Identifiability and Identification of Switching Dynamical Networks

A scalable multi-step least squares method for network identification with unknown disturbance topology

Local Identifiability of Networks with Nonlinear Node Dynamics

Topology Identification in Two-Layer Complex Dynamical Networks

A New Method for Topology Identification of Complex Dynamical Networks

Identifying Topologies of Complex Dynamical Networks with Stochastic Perturbations

Topology Identification of A Class of Complex Spatio-Temporal Networks with Time Delay

Network Identifiability from Intrinsic Noise.

Nonlinear network identifiability: The static case

Optimization-Based Structure Identification of Dynamical Networks

Latent Dynamic Networked System Identification with High-Dimensional Networked Data

Spectral identification of networks with generalized diffusive coupling

Distributed Identification of Stable Large-Scale Isomorphic Nonlinear Networks Using Partial Observations

Identifying Structures of Continuously-Varying Weighted Networks.

Monitoring Link Faults in Nonlinear Diffusively-coupled Networks

Topology identification for stochastic multi-layer networks via graph-theoretic method

Interacting Particle Systems on Networks: joint inference of the network and the interaction kernel

Finite-time topology identification of complex spatio-temporal networks with time delay

Finite-Time Topology Identification of Delayed Complex Dynamical Networks and Its Application