Model Predictive Control of Priced Timed Automata Encoded with First-Order Logic

Efe C. Balta, Ilya Kovalenko, Isaac A. Spiegel, Dawn M. Tilbury, Kira Barton

Research output: Contribution to journalArticlepeer-review

8 Scopus citations

Abstract

Priced timed automata (PTA) are discrete-event system models with temporal constraints and a cost function and are used to pose optimal scheduling and routing problems. To date, solutions to these problems have been found offline and executed open loop.This open-loop control strategy makes it impossible to account for disturbances, i.e., changes in costs or scheduling constraints over time. To address this shortcoming, this work's first contribution is a closed-loop model predictive control (MPC) framework for PTA, enabling decision-making based on real-time model updates. To ensure the feasibility of an MPC problem, it is often desirable to soften constraints. However, the contemporary PTA theory does not consider soft constraints. Thus, this work's second contribution is to integrate constraint softening with PTA control by harnessing the capabilities of new solvers enabled by the recasting of the models and control problem into first-order logic by employing modified encoding schemes based on existing works. Finally, the proposed control framework and implementation are demonstrated in a simulation case study on the guidance of a product through a manufacturing system.

Original languageEnglish (US)
Pages (from-to)352-359
Number of pages8
JournalIEEE Transactions on Control Systems Technology
Volume30
Issue number1
DOIs
StatePublished - Jan 1 2022

All Science Journal Classification (ASJC) codes

  • Control and Systems Engineering
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Model Predictive Control of Priced Timed Automata Encoded with First-Order Logic'. Together they form a unique fingerprint.

Cite this