A Distributed Unscented Predictive Cooperation Approach for Networked Mobile Manipulators

Abstract

Cooperation control of networked mobile manipulators is a challenging task to be met. This paper presents a distributed unscented model predictive control method to address the problem of cooperation control for networked mobile manipulators with uncertain parameters. Based on the Euler-Lagrange approach, an auxiliary model with uncertain parameters and disturbances is obtained via the nonlinear feedback technique. Under the framework of distributed predictive control, a local optimization problem with chance constraints is established. An equivalent optimization problem based on the probability distribution predictive model is derived using the unscented transform to approximate the probability distributions for nonlinear models. Moreover, a sufficient condition guaranteeing the overall system's stability is developed. The performance of the proposed distributed unscented predictive control strategy is validated by simulation examples.