• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 1
  • 1
  • Tagged with
  • 2
  • 2
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
1

Implementing an Intelligent Alarm System in Intensive Care Units

Kilinc, Derya, Ghattas, Mattias January 2016 (has links)
Today’s intensive care units monitor patients through the use of various medical devices, which generate a high ratio of false positive alarms due to a low alarm specificity. The false alarms have resulted in a stressful working environment for healthcare professionals that are getting more desensitized to triggered alarms and causing alarm fatigue. The patient safety is also compromised by having high noise levels in the patient room, which disturbs their sleep. This thesis has developed an intelligent alarm system with an improved alarm management and the use of 23 intelligent algorithms to minimize the number of false positive alarms. The suggested system is capable of improving the alarm situation and increasing the patient safety in critical care. The algorithms were modeled with fuzzy logics consisting of delays and multi parameter validation. The results were iteratively developed by having focus groups with various experts.
2

Processador inteligente de alarmes e modelos de programação matemática para diagnóstico de faltas em sistemas elétricos de potência / Intelligent alarm processor and mathematical programming models for fault diagnosis in electrical power systems

Oliveira, Aécio de Lima 24 June 2016 (has links)
Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / This thesis proposes an Intelligent Alarm Processor for fault diagnosis in electrical power systems. The objective is to develop a methodology for automatic fault analysis using reported alarms from Supervisory Control and Data Acquisition (SCADA) to allow the use of diagnosis systems in large power systems. The proposal can be used in real-time decision support systems to assist control center‟s operators during the decision-making after unscheduled contingencies with relevant information to power system restoration. This work expects to contribute to the development of advanced alarm management logics that allow modifying the chronological sequence of reported alarms, event mapping and the generation of operating patterns of protection systems according to topology network. Still, mathematical programming models have been formulated as a parsimonious set covering problem to fault section estimation and identification of protective devices with improper operation. Among these models, it stands out the model that deals with integrated analysis of reported alarms, events and diagnosis that better explain the alarms. The proposed approach has been tested in different portions of the Southern Brazilian power system. The results show that alarm processing allows the practical implementation of intelligent diagnosis methods in existing supervisory systems. The proposed diagnosis methods show better performance and accurate solutions than other methods presented in literature. / Esta tese propõe um Processador Inteligente de Alarmes para diagnóstico de faltas em sistemas elétricos de potência. O objetivo é desenvolver uma metodologia para a análise automática de faltas a partir dos alarmes reportados no sistema de supervisão e aquisição de dados (SCADA) que possibilite o uso de métodos de diagnóstico em sistemas de potência de grande porte. Essa proposta pode ser empregada em sistemas de apoio à decisão em tempo real, que auxiliem operadores de centros de controle do sistema (COS) na tomada de decisão após desligamentos não programados, com informações pertinentes para o restabelecimento do sistema. O trabalho espera contribuir com o desenvolvimento de lógicas avançadas de gerenciamento de alarmes que possibilitem a reordenação cronológica dos alarmes reportados, o mapeamento dos eventos e a geração de padrões de funcionamento de sistemas de proteção de acordo à topologia da rede. Além disso, os modelos de programação matemática foram formulados como um problema de recobrimento de conjuntos parcimonioso, para estimação da seção em falta e identificação dos dispositivos de proteção com atuação indevida. Dentre esses modelos, destaca-se o modelo que analisa, de forma integrada, os alarmes reportados e determina os eventos e diagnósticos que melhor explicam os alarmes. A abordagem proposta foi testada em diferentes porções do sistema sul do sistema interligado nacional (SIN). Os resultados mostram que as rotinas desenvolvidas para o processamento de alarmes permite a implantação prática de métodos inteligentes de diagnóstico em sistemas supervisórios existentes. Os métodos propostos para diagnóstico de faltas mostraram desempenhos e precisão nos resultados superiores a outros métodos presentes na literatura.

Page generated in 0.1132 seconds