Decentralised Co-operative control of Mobile Robots

1.    

Advances in embedded systems and associated technologies, have paved the way for research on multi-agent robot systems to effectively co-ordinate assigned tasks. In this regards decentralized cooperative control schemes are a prime research focus due to their resemblance to biologically systems and many advantages over centralized schemes. This thesis presents a Decentralized Cooperative Control Scheme for Mobile Robots, with focus on formation control and obstacle avoidance.  The scheme is designed for differential drive robots in a decentralized setting and composes of a navigation layer and a tracking controller. The navigation layer generates individual robot trajectory for formation controlled swarming and obstacle avoidance based on the initial position of the robots and obstacle locations. This is achieved using virtual and behavioral structures along with artificial potential field functions using non-linear systems theory. The non-holonomic nature of the differential drive robots makes it very challenging to track the generated trajectories with linear controllers. Hence a robust sliding mode controller is employed (in polar coordinates due to its advantage of sliding surface selection and improved performance). Moreover three different sliding mode controllers are employed; the first-two minimize the position tracking errors for positive and negative velocity directions respectively. The third one corrects the heading error when position tracking errors are limited to a dead band and position reference is stagnant.

Simulations are carried out for the navigation layer in order to investigate for the scenarios of: 1) Stagnant formation control, 2) Formation control with swarming and change in no. of robots and 3) Formation control with obstacle avoidance. Next simulations are carried out to investigate the performance of the sliding mode controller for position stabilization and trajectory tracking. Finally the complete system is tested on an integrated MATLAB/Simulink® based platform, for formation control with obstacle avoidance in a practical scenario with reasonable assumptions; and simulation results validate the working of the proposed scheme.