A Tractable Truthful Profit Maximization Mechanism Design with Autonomous Agents

Mina Montazeri,Hamed Kebriaei,Babak N. Araabi

DOI: https://doi.org/10.1109/tac.2024.3360335

IF: 6.549

2024-01-01

IEEE Transactions on Automatic Control

Abstract:In this paper, we propose a new profit maximization mechanism for the task allocation problem, where the task publisher seeks an optimal incentive function to maximize its own profit and simultaneously ensure the truthful announcing of the agent?s private information (type) and its participation in the task, while an autonomous agent aims at maximizing its own utility function by deciding on its participation level and announced type. Our mechanism stands out from the classical contract theory-based truthful mechanisms as it empowers agents to make their own decisions about their level of involvement, making it more practical for many real-world task allocation scenarios. It has been proven that by considering a linear form of incentive function consisting of two decision functions for the task publisher the mechanism?s goals are met. The proposed truthful mechanism is initially modeled as a non-convex functional optimization with the double continuum of constraints, nevertheless, we demonstrate that by deriving an equivalent form of the incentive constraints, it can be reformulated as a tractable convex optimal control problem.

automation & control systems,engineering, electrical & electronic