| 研究実績の概要 |
The algorithmic foundations and applications of learning and optimization of graphs via enumerative queries in computer-aided design and robotics were proposed: (1) the research on minimal enumerative representation of graphs, implementing the efficient framework to generate graph and hypergraph representations with minimal space, and enabling to store, communicate, and sample the search space of graphs and networks while meeting user-defined criteria via efficient enumerative queries (both CPU and GPU-enabled). Our algorithms implement both the gradient and gradient-free optimization algorithms suitable for utmost efficiency; as such, our computational experiments using graph instances with varying degrees of sparsity have shown the merit of attaining minimal numbers for graph encoding/representation rendered by exploration-oriented strategies within few function evaluations. (2) learning the optimal coordination networks for robotics planning problems, implementing the optimal and efficient coordination of multi-robot navigation problems using graphs and lattice-based roadmap configurations, enabling the collision-free multi-robot coordinated path planning algorithms. Our results have the potential to elucidate new enumerative sample-based algorithms for graph representation, network design and network optimization.
|
