Strategy Based on Multiple Objectives and Null Space for the Formation of Mobile Robots and Dynamic Obstacle Avoidance

One of the current challenges in robotics is the cooperative control of multiple robots with a common goal. Thepossibility of making several smaller capacity robots to perform tasks that would be impossible or inefficient individually, has motivated the scientific community to encourage the development of innovative control strategies. The use of multiple robots versus one offers several advantages, which could include reduced costs, greater robustness, improved performance and efficiency [15]. Instead of designing a single specialized powerful robot, a multi-robot system can be simpler and less costly [12]. This feature has allowed formation control to be successfully used in various applications: military [18], civil [17] and service robotics. There are three basic structures in the bibliography. The control of multi-robot systems: leader-follower strategy, methods based on behavior and virtual structures, each with their respective advantages and disadvantages. In the leaderfollower structure, an agent is considered the leader and the remaining agents are considered followers of the designated leader [7] and [8]. In this structure, only the follower has information about the leader, so if it fails there is no possible mechanism that would ensure the compliance of the control target. However, this structure is easy to understand and implement.In the structure based on behavior, group behavior is defined as a combination of individual behavior of its members [5]. The main problem with this approach, is the difficult mathematical formalization and therefore it is not easy to ensure the convergence of the formation to the desired setting. In virtual structures geometry maintains a rigid connection between the robots and the reference system, which can be a virtual point or a virtual agent. One advantage of this method is that the virtual leader would never fail, so training will be maintained during the execution of the task.