Distributed Cooperative Object Transport of Mobile Manipulator

During the past few years, object transport, especially by mobile manipulators, has become one of the emerging andimportant research topics in robotic control systems. Most previous studies on object transport of multiple mobile manipulators focus on centralized methods for its control accuracy and robustness to external disturbances, however they don’t match the requirements and expectations for largescale complex tasks with multi-robot coordination in the industry. Therefore the proposed research seeks a distributed way to achieve cooperative object transport and to make full use of the redundancies of mobile manipulator by optimizing the manipulability. In this project, the main goal of object transport is to control the robot arms to keep the object’s weighted centroid tracking a time-varying reference, while adjusting the position of the mobile platforms such that the robot arms can be placed in a way with good manipulability. This goal is described by a distributed optimization problem, that is, both the manipulability of robot arms and desired platform positions are considered in the non-convex cost function and the requirements of weighted centroid tracking are satisfied in time-varying coupled equality constraints. In this way, the weighted centroid tracking and the control of arms and platforms can be solved in a unified distributed optimization framework.