SIMULATION OF OPTIMAL TWO-PHASE OPERATIONS IN RANDOM OPERATING ENVIRONMENTS
Ya. A. Mostovoi
Korolev Samara State Aerospace University, Moskovskoe Shosse 34, Samara, 443086 Russia
Keywords: статистическое моделирование, большие сети, кластеры объектов, порог перколяции, управляемая перколяция, навигация по визуальным ориентирам, statistical simulation, large networks, clusters of objects, percolation threshold, controlled percolation, landmark navigation
Abstract
Statistical simulation of distributed sets of objects integrated into a large network and forming an operating environment is considered. The simulation is carried out on square matrices with random filling using percolation theory. Statistical features of the clusters of objects that arise in this case are analyzed. Along with well-known stochastic percolation, the concept of a controlled percolation of the operating environment is introduced, which is implemented in two phases: in the first phase, a stochastic basis with a relatively low concentration of objects is created, and in the second phase, additional objects are introduced in the intercluster intervals in order to obtain the shortest percolation path while minimizing the total cost of such two-phase operation. This model is used to consider vehicle navigation by visual landmarks randomly distributed in the operating environment.
|
