Decentralized nonzero-sum game-based optimal control for reconfigurable robot via value iteration

A decentralized approximate optimal control method based on nonzero-sum game (NZSG) is introduced in this study for reconfigurable robots (RRs) in coordinated operation tasks (COTs). The RR system’s dynamic model is built by means of JTF method. The RR system’s optimal control problem, aimed at COTs, is restated as an NZSG within subsystems. By making use of ADP, the control policy is obtained via NZSG control approach. The value iteration is employed to solve Hamilton-Jacobi (HJ) equation and help in getting equilibrium. Error of RR under COTs is uniformly ultimately bounded (UUB).