Event-Triggered Output-Feedback H∞ Control with Minimum Directed Information

In this paper, the H∞ control problem with minimum information exchange between a plant and a remote controller is considered. For transmission of the measured outputs, an event-triggering mechanism is proposed. We measure the transferred information satisfying the causality condition by the directed information, and measure the H∞ control performance by the L2-gain. We achieve the minimum directed information required for a guaranteed level of control performance. In particular, corresponding to the worst case disturbances, we develop a filter with a recursive variable, and show that the optimal control policy is a linear function of this recursive variable. Finally, we define the value of information, and determine the structure of the optimal sampling policy based on the value of information.