Inspired by the Linear Quadratic Regulator (LQR), ILQ-games seeks to avoid the computational and performance related shortcoming of current approaches to solving multi-agent games in real-time. Many such works rely on simplifying assumption that limit the scope of their use case or otherwise effect online performance.
ILQ games seeks to address the limitations of other works by leveraging the well-known properties of LQ-games, including their ease of calculation. By taking a general nonlinear game of the appropriate form, approximating it as an LQ-game and iteratively solving to find a converged set of player strategies, a solution to the problem can be determined online, enabling this algorithm to be used in a wide-variety of use-cases with minimal performance compromise.