کنفرانس مهندسی برق ایران

صفحه اصلی / سی و سومین کنفرانس بین المللی مهندسی برق

Multi-Agents Gaussian Estimation and Coverage Control Client-Server Architecture

نویسندگان :

Mohammad َAzizian Shishavan¹ Mahdi Zeinali² Azam Salari³

1- دانشگاه صنعتی سهند تبریز 2- دانشگاه صنعتی سهند تبریز 3- دانشگاه صنعتی سهند تبریز

کلمات کلیدی :

Gaussian estimation،Coverage control،Bayesian prediction

چکیده :

In this paper, the problem of multi-agent input natural gas coverage of the East Azarbaijan in Iran when the sensory field used to approximate the density of event appearance is not known in advance is addressed. The problem by considering a communication architecture is investigated: client-server. For this architecture, a nonparametric Gaussian regression approach is used to estimate the unknown sensory field from a collection of noisy samples. A probabilistic control strategy is proposed that allows the agents to estimate the true sensory field with arbitrary accuracy while computing and exploiting the corresponding centroidal Voronoi partitions. It also presents a numerically efficient approximation based on spatial discretization to balance the accuracy of the estimated map and the required computational complexity.