Optimal Strategies for Disjunctive Sensing and Control

TL;DR

This paper develops optimal strategies for disjunctive sensing and control in linear stochastic systems, ensuring bounded states and satisfying probabilistic constraints, with applications demonstrated in spacecraft control.

Contribution

It introduces a novel method for designing disjunctive sensing and actuation sequences that guarantee system stability and probabilistic constraint satisfaction.

Findings

Ensures bounded states and estimation errors with disjunctive strategies.

Extends to probabilistic chance constraints satisfaction.

Demonstrates effective performance in spacecraft control simulations.

Abstract

A disjunctive sensing and actuation problem is considered in which the actuators and sensors are prevented from operating together over any given time step. This problem is motivated by practical applications in the area of spacecraft control. Assuming a linear system model with stochastic process disturbance and measurement noise, a procedure to construct a periodic sequence that ensures bounded states and estimation error covariance is described along with supporting analysis results. The procedure is also extended to ensure eventual satisfaction of probabilistic chance constraints on the state. The proposed scheme demonstrates good performance in simulations for spacecraft relative motion control.