Nature-inspired and teaching-learning-based methods for improving convergence speed in multi-agent systems

This paper suggests a novel method for inverse optimal control of the multi-agent systems (MAS) via a linear quadratic regulator (LQR) based on meta-heuristic algorithms. In this regard, first, the consensus protocol is designed and then the cost function is optimized via Jaya algorithm (JA), teaching-learning algorithm (TLBO), a novel meta-heuristic algorithm called advanced teaching-learning (ATLBO) and water cycle algorithm (WCA). ATLBO consists of two phases with two random values in both phases which affect the convergence rate. The optimal value of the controller’s parameter is obtained via these algorithms. Simulation outputs show the usefulness of nature-inspired and learning-based methods to calculate the cost with a better convergence rate. This research consists of an inverse optimal control approach and meta-heuristic algorithms for solving the consensus problem with the least cost.