Approximating Constrained Minimum Input Selection for State Space Structural Controllability

TL;DR

This paper introduces a graph-theoretic approach and polynomial approximation algorithms for solving the NP-hard problem of constrained minimum input selection in structured systems, ensuring controllability with minimized cost.

Contribution

It provides a new necessary and sufficient graph condition for controllability, reduces the problem to a known NP-hard flow problem, and offers a polynomial approximation algorithm.

Findings

Polynomial reduction of input selection to minimum cost fixed flow problem

A $ ext{Δ}$-approximate algorithm based on bipartite matching

Efficient greedy scheme for flow network approximation

Abstract

This paper looks at two problems, minimum constrained input selection and minimum cost constrained input selection for state space structured systems. The input matrix is constrained in the sense that the set of states that each input can influence is pre-specified and each input has a cost associated with it. Our goal is to optimally select an input set from the set of inputs given that the system is controllable. These problems are known to be NP-hard. Firstly, we give a new necessary and sufficient graph theoretic condition for checking structural controllability using flow networks. Using this condition we give a polynomial reduction of both these problems to a known NP-hard problem, the minimum cost fixed flow problem (MCFF). Subsequently, we prove that an optimal solution to the MCFF problem corresponds to an optimal solution to the original controllability problem. We also show that approximation schemes of MCFF directly applies to minimum cost constrained input selection problems. Using the special structure of the flow network constructed for the structured system, we give a polynomial approximation algorithm based on minimum weight bipartite matching and a greedy selection scheme for solving MCFF on system flow network. The proposed algorithm gives a $\Delta$-approximate solution to MCFF, where $\Delta$ denotes the maximum in-degree of input vertices in the flow network of the structured system.