Wyniki wyszukiwania - BazTech

1

Algorytmy wykrywania struktury sieci Bayesa z danych w ocenie ryzyka powstawania uszkodzeń budynków na terenach górniczych

Rusek Janusz, Chomacki Leszek, Słowik Leszek, Firek Karol

Materiały Budowlane

|

2022

|

nr 11

66--69

PL

W artykule zaprezentowano wyniki badań, które podjęto w celu utworzenia modelu do oceny ryzyka powstania uszkodzeń budynków poddanych wpływom statycznych i dynamicznych oddziaływań górniczych. Uzasadniono przyjętą metodykę na kanwie metod uczenia maszynowego (ML - Machine Learning). Omówiono specyfikę zagadnienia i na tej podstawie przedstawiono główne założenia stosowanego podejścia, a przede wszystkim metodykę pozwalającą na samoistne wyłanianie struktury sieci Bayesa z danych (BSL - Bayesian Structure Learning). Zaprezentowano rezultaty otrzymane w ramach badań w odniesieniu do wielokondygnacyjnych budynków prefabrykowanych oraz murowanych zlokalizowanych na terenie LGOM oraz GZW. W artykule wskazano również możliwość uniwersalnego stosowania przyjętej metodyki w przypadku predykcji ryzyka powstania uszkodzeń i diagnozowania przyczyn zaistniałych szkód.

EN

The article presents the results of research that was undertaken to create a model to assess the damage risk of buildings subjected to static and dynamic mining impacts. The justification of the adopted methodology on the basis of machine learning (ML) methods is given. The specificity of the problem was discussed and, on this basis, the main assumptions of the applied approach were presented, especially the methodology allowing for autonomous extraction of the Bayesian network structure from data (BSL - Bayesian Structure Learning). The results obtained in the research were presented in relation to multi-storey prefabricated and masonry buildings located in LGDC and USB mining terrain. The paper also indicates the possibility of universal application of the adopted methodology in the case of damage risk prediction and diagnosis of the causes of damage.

2

Field investigation in the detonation behavior of emulsion explosive column induced with air gaps

Balakrishnan Vineeth, Pradhan Manoj, Dhekne P. Y.

Mining Science

|

2019

|

Vol. 26

55--68

EN

In blasting of soft to medium hard rock, the problem of high density resulting in excessive utilization of emulsion explosive is well known. The authors have conducted some experimental blasts to delve into the detonation behavior of conventional blasting and various other explosive consumption reduction techniques which induce air gaps using plastic tubes, plastic bottles or plastic balls in the explosive column. Resistance wire technique is used for gauging in-hole continuous velocity of detonation. The VOD varies from 5321.6 m/s to 4544.2 m/s and from 5123.4 m/s to 4274.2 m/s in conventional site mixed emulsion column and distributed spherical air gap column respectively. The detonation behavior is stable and similar in both these cases. While using plastic bottles or plastic tubes as air gaps, the VOD is fluctuating from 4636.3 m/s to 3268.4 m/s and from 4935.9 m/s to 3362.8 m/s respectively with a collapse of about 12 % from the average VOD of conventional SME column. The VOD falls abruptly when the detonation wave encounters large air gaps but it is successfully travelling through the air gaps making the detonation behavior more capricious.

3

Problem-Independent Approach to Multiprocessor Dependent Task Scheduling

Król D., Zydek D., Koszałka L.

International Journal of Electronics and Telecommunications

|

2012

|

Vol. 58, No. 4

369-379

EN

This paper concerns Directed Acyclic Graph task scheduling on parallel executors. The problem is solved using two new implementations of Tabu Search and genetic algorithm presented in the paper. A new approach to solution coding is also introduced and implemented in both metaheuristics algorithms. Results given by the algorithms are compared to those generated by greedy LPT and SS-FF algorithms; and HAR algorithm. The analysis of the obtained results of multistage simulation experiments confirms the conclusion that the proposed and implemented algorithms are characterized by very good performance and characteristics.