Abstract
In view of the low formation redundancy in the traditional rigid formation algorithm and its difficulty in dynamically adapting to the external environment, this study considers the use of the CVT (centroidal Voronoi tessellation) algorithm to control multiple robots to form the desired formation. This method significantly increases the complexity of the multi-robot system, its structural redundancy, and its internal carrying capacity. First, we used the CVT algorithm to complete the Voronoi division of the global map, and then changed the centroid position of the Voronoi cell by adjusting the density function. When the algorithm converged, it could ensure that the position of the generated point was the centroid of each Voronoi cell and control the robot to track the position of the generated point to form the desired formation. The use of traditional formations requires less consideration of the impact of the actual environment on the health of robots, the overall mission performance of the formation, and the future reliability. We propose a health optimization management algorithm based on minor changes to the original framework to minimize the health loss of robots and reduce the impact of environmental restrictions on formation sites, thereby improving the robustness of the formation system. Simulation and robot formation experiments proved that the CVT algorithm could control the robots to quickly generate formations, easily switch formations dynamically, and solve the formation maintenance problem in obstacle scenarios. Furthermore, the health optimization management algorithm could maximize the life of unhealthy robots, making the formation more robust when performing tasks in different scenarios.
Highlights
A total of 1000 × 1000 virtual sensors were evenly placed in the area, and the values of the sensors were fitted to the density function of the centroid Voronoi partition (CVT) algorithm
When the number of iterations of the CVT algorithm k = 30, the controller is affected by the integral saturation to produce overshoot, and the cost function increases to 3.3 × 103
In MATLAB, it was assumed that the robot was a first-order dynamic model, the formation scenario map was 30 × 30 m2, 3000 × 3000 virtual sensors were evenly placed on the map, and the values of the sensors were fitted to the density function of the CVT algorithm
Summary
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. A large number of research results have been produced for the multi-robot formation control problem, and a variety of formation control methods represented by the follower method, the behavior-based control method, the virtual structure method, and the artificial potential field method have been proposed. The impact of the environment on the health of robots must be taken into account, in order to be able to make real-time dynamic adjustments to the robots and to improve the performance and robustness of the robot formation. On this basis, this study combines the CVT (centroidal Voronoi tessellation) algorithm with multi-robot formation control.
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call