Krashchenko D. Methodology for building an automated intelligent building management system based on stochastic optimization methods.

Українська версія

Thesis for the degree of Doctor of Philosophy (PhD)

State registration number

0823U100073

Applicant for

Specialization

  • 123 - Комп’ютерна інженерія

03-02-2023

Specialized Academic Board

ДФ 26.861.014

State University of Telecommunications

Essay

ANOTATION In order to achieve the goal of the dissertation work, namely the development of a methodology for building an automated intelligent building management system, the following tasks were solved: 1. The problem of placement of sensors for heterogeneous WSNs is defined and solved, which consists in choosing the optimal places for their placement, minimizing energy consumption by the entire system while simultaneously fulfilling restrictions on connections and resources. The problem is solved by introducing requirements for protection levels and creating conditions for providing a network of potential backup solutions in case of unexpected partial failures; 2. The problem of criticality for homogeneous WSNs, which consists in identifying the most important sensors or critical nodes in the network, is defined and solved. When defining the problem, the generally accepted definition of criticality was changed. In particular, in the proposed solution, a critical node is defined as a node, the removal of which most disrupts the network. The problem was solved by developing a method of a selective adaptive search procedure for calculating the "delay" and "lifetime" parameters of the network, based on which the criticality assessment function of a node of a homogeneous WSN was created; 3. A mathematical model of WSN, has been developed, which differs from known models by the introduction of clustering. In a given, concrete case, a cluster is defined as a set of sensor positions, in which a group of sensor types must be represented exhaustively. This addition allows you to create a variety of solutions depending on the technical need and provides the ability to change the final layout within one intelligent building. 4. The IBMS resource manager model, have been developed, which differ from the known ones by introducing the concept of the forecasting threshold. This made it possible to take into account not only the energy consumption indicator, but also add an important comfort indicator to the evaluation. Using this model, the IBMS monitors the state of various rooms and, according to the expected actions of the occupants, manages the subsystems to minimize energy consumption, maintaining an acceptable level of comfort and eliminating the possibility of conflicting activations. The results obtained in the course of the work were used in the research work conducted at the State University of Telecommunications. The theoretical and practical provisions of the dissertation work are used in the educational process of the State University of Telecommunications. Keywords: wireless sensor networks (WSN), Internet of Things, IBMS, Smart house, clustering, energy optimization, gateway, sensor, heterogeneous network, "cloud" architecture, stochastic optimization.

Files

Similar theses