Formation control of mobile robot systems incorporating primal-dual neural network and distributed predictive approach

This paper addresses the formation problem for multiple mobile robots with velocity mismatch and system constraints by a distributed model predictive control (DMPC) strategy and a modified virtual structure method. Firstly, a desired virtual structure is employed to generate a set of reference paths for the formation robots. By including approaching angle and path parameter synchronization constraints into cost function, a Nash-based DMPC strategy is presented, where a velocity integral controller is developed to solve the velocity mismatch. Further, consider state and input constraints, the distributed optimization problem is rewritten as a constrained quadratic programming (QP) problem. A PDNN is used to obtain the optimal control input increments, and the stability of the proposed algorithm is analyzed. Moreover, the dynamic formation control is achieved by a modified virtual structure method, and simulation examples are given to verified the effectiveness of the proposed strategies.