Cumulative residual Kullback-Leibler divergence based sensor placement using reliability criteria

In the current work, we propose a novel, information theoretic based sensor placement design approach for placing sensors in a steady state linear flow process for reliable estimation of variables. In particu-lar, the optimal sensor placement minimizes the cumulative residual Kullback-Leibler divergence based objective function. Unlike the existing approaches that maximize system reliability corresponding to a fixed time, the proposed approach utilizes the time varying system reliability to compute the objective function. Further, it provides a mechanism to the end-user to tailor the optimal design by specifying her preference as the reference system reliability in the objective function. The sensor placement problem proposed in our current work is an integer non-linear programming problem. We also propose a greedy heuristic algorithm to solve this problem with low computational cost. Finally, we demonstrate the utility of the proposed design approach on a benchmark case study.(c) 2022 Elsevier Ltd. All rights reserved.

Automated summary

This study proposes an information theoretic based sensor placement design approach for reliable estimation of variables in a steady state linear flow process. The approach minimizes residual Kullback-Leibler divergence based objective function, utilizes time varying system reliability, and allows user customization.