Resource-Aware Asynchronous Multi-Agent Coordination via Self-Triggered MPC

Abstract

In many applications, resource-aware devices are connected through a network, such as in the Internet of Things and energy hubs. These devices require proper coordination to achieve a high performance without violation of their resource limits. In this paper, we propose an asynchronous resourceaware multi-agent model predictive control to cooperatively coordinate agents to conduct a common task, whose resource concern is handled by a self-triggered control scheme. The consistency and recursive feasibility of the proposed MPC scheme are investigated. A reliable numerical implementation is introduced to deal with non-constant sampling times among agents, which is subsequently validated by a numerical example.

Resource-Aware Stochastic Self-Triggered Model Predictive Control

“…Before t k+1 , none of the triggers can generate feedback with respect to the events in F k,k+1 because F k,k+1 happens later than {ti} k i=0 . These facts conclude the conditional control law in (15). Based on the system dynamics (6), the mean dynamics (12) and the conditional control inputs (15), the SDE of the conditional deviation dynamics of…”

“…These facts conclude the conditional control law in (15). Based on the system dynamics (6), the mean dynamics (12) and the conditional control inputs (15), the SDE of the conditional deviation dynamics of…”

“…However, to balance performance and resource consumption more effectively, the response of the resource is explicitly considered in the model predictive control (MPC) problem in [10], [17]. The former work solves a mixed-integer problem and is designed for discretetime systems, while the latter work solves a non-convex continuous-time optimal control problem, and has been later generalized to a distributed control scheme [15]. Running a triggered system within an uncertain environment while maintaining system performance is challenging.…”

Scheduling Delays and Curtailment for Household Appliances with Deterministic Load Profiles using MPC

“…The utilization of the absolute time in the numerical implementation is important, as the t ri , t si and t 0 uniquely defines all the time grids used in this problem. The online operation follows the receding horizon MPC scheme, details can be found in [6,Chapter 1.3], [19].…”

“…Self-triggered MPC can systematically optimize the active/triggered time of a device proactively with explicit constraint enforcement [15], and has been widely used in sensor networks and Internetof-Things [16]. As opposed to the mixed-integer formulation [17], [18], the triggering time can be a continuous decision variable [19]- [21] in this scheme. This paper interprets the time shift as a trigger time viewed from the current time step, and transcribes the load scheduling problem into a continuous optimization problem such that we can take full advantage of continuous numerical solvers.…”