Mobile Objects Control in Three-Dimensional Area Using the Hybrid Decentralized Algorithm

摘要

The problem of mobile objects group motion in an uncertain three-dimensional area with obstacles is considered in the paper. The decision of the problem is carried out using a hybrid algorithm that combines the immediate reaction to the obstacles and the scheduling algorithms at the upper level. The reactive component is implemented on the basis of unstable modes, allowing to transform obstacles into repellers. The motion planning is based on heuristic analysis of the data received from the sensory systems and the clustering based on the Delaunay triangulation. The direction of the obstacles bypass and the position of the mobile object in the group are determined as a result of this analysis executing. The method of the trajectory correction by adding the component, generated by the dynamic system transforming into an unstable mode during the obstacles detection, into the upsetting control is proposed in the paper. The proposed algorithms can be implemented decentrally. The analysis of the proposed algorithms is considered and the numerical simulation is performed.