Relative time‐averaged gain array (RTAGA) for distributed control‐oriented network decomposition

Abstract

Input-output partitioning for decentralized control has been studied extensively using various methods, including those based on relative gains and those based on relative degrees and sensitivities. These two concepts are characterizations of long-time and short-time input-output response, respectively. A unifying new input-output interaction measure, called relative time-averaged gain, which characterizes the input-output interactions during a time scale of interest for linear time-invariant systems is proposed. This measure is used as a basis for community detection in the input-output bipartite graph of a process network to produce subnetworks whose responses are weakly coupled in the time scale of interest. As such, the resulting decomposition accounts for both response characteristics and the network topology, and can be used efficiently for distributed control architecture design. In a case study, the proposed decomposition is applied to the distributed model predictive control of a reactor-separator benchmark process. V

Decomposition of control and optimization problems by network structure: Concepts, methods, and inspirations from biology

“…To account for response intensities or sensitivities , Tang & Daoutidis 102 defined an affinity measure to be used combined with the input-output structural interactions. An alternative measure, called relative time-averaged gain array, which captures output response interactions over a finite time scale, and can be seen as a hybrid of relative gains 103 (capturing steady-state output responses) and relative degrees 37 (capturing structural input-output interactions), was proposed in Tang et al 104 Pourkargar et al 105,106 performed simulation studies to compare different decompositions for distributed model predictive control of a benchmark reactor-separator process; it was shown that community-based decompositions, especially the one suggested by Jogwar and Daoutidis, 101 help to significantly reduce the computational cost while maintaining good performance for a wide range of simulation experiments.…”

An efficient algorithm for community detection in complex weighted networks

“…Moreover, modern manufacturing plants are increasingly integrated, 10,11 leading to structural and computational complexities. In recent years, the problem of decomposing a large complex network into a set of interacting small networks that adequately capture the interactions of the original large network has received great attention in many engineering and science fields 12‐15 …”

Non‐linear‐disturbance‐observer‐enhanced MPC for motion control systems with multiple disturbances

“…For largescale process networks, e.g. reactor-separator, distributed MPC methods also manage to position themselves due to the advantages in decomposition and computational efficiency [4][5][6]. As its name suggested, the nominal model plays a fundamental role in the philosophy of MPC methods, especially in the phase of the model prediction.…”