Real-valued quantum-inspired evolutionary algorithm for multi-issue multi-lateral negotiation

In this paper, a new variant of a quantum-inspired evolutionary algorithm is proposed, which is characterised by a population-based elitism, a resetting mutation for the qubits, and an evolutionary hill-climbing phase at the end of the main search, meant to further improve the quality of the solution. The algorithm was applied for finding near-optimal outcomes for multi-issue multi-lateral negotiation in a multiagent system, and it was designed for situations where the agents involved are cooperative and are willing to reveal their private information regarding their utilities to an external, impartial mediator. The proposed algorithm is shown to outperform two other classical techniques and also two other variants of quantum-inspired evolutionary algorithms, especially for large optimization problems. An evolutionary optimization approach is particularly useful in negotiation settings where the utility functions of the agents are non-linear.