Mariana Antonia Aguiar-Furucho | UTFPR - UNIVERSIDADE TECNOLOGICA FEDERAL DO PARANÁ (original) (raw)
Uploads
Papers by Mariana Antonia Aguiar-Furucho
marianrogers.com.br
... POLI/USP 3Centro Universitário das Faculdades Metropolitanas Unidas UniFMU * Autor Responsá... more ... POLI/USP 3Centro Universitário das Faculdades Metropolitanas Unidas UniFMU * Autor Responsável: Mariana Antonia Aguiar, +55 11 3091 ... Funçao Diagnostica (Azul), Funçao Treinamento (Vermelho) e Valores Preditivos Retornados pela RNA (Verde) Funçao Diagnostica ...
Objectives: Memory lose takes place in an artificial koniocortex network when calcium-homeostasis... more Objectives: Memory lose takes place in an artificial koniocortex network when calcium-homeostasis impairment is simulated. According to the calcium dysregulation hypothesis of Alzheimer Disease, calcium dysregulation affects intrinsic plasticity. Intrinsic plasticity is the property of real neurons of adapting their firing thresholds according to the neurons activation regime: when a neuron is intensely stimulated it increases its firing threshold so that it becomes less responsive. On the contrary, when a neuron is scarcely stimulated, it reduces its firing threshold so that it is more prone to fire. Methods: A computational model of the koniocotex was trained with alphanumerical patterns, exhibiting successful learning capabilities. Successful learning was confirmed by stimulating each of the spiny stellate neurons of the model, and checking whether original alphanumerical patterns were recovered. Results: Due to calcium dysregulation, intrinsic plasticity might be impaired in eld...
International Journal of Neural Systems, 2016
In this paper, we use the neural property known as intrinsic plasticity to develop neural network... more In this paper, we use the neural property known as intrinsic plasticity to develop neural network models that resemble the koniocortex, the fourth layer of sensory cortices. These models evolved from a very basic two-layered neural network to a complex associative koniocortex network. In the initial network, intrinsic and synaptic plasticity govern the shifting of the activation function, and the modification of synaptic weights, respectively. In this first version, competition is forced, so that the most activated neuron is arbitrarily set to one and the others to zero, while in the second, competition occurs naturally due to inhibition between second layer neurons. In the third version of the network, whose architecture is similar to the koniocortex, competition also occurs naturally owing to the interplay between inhibitory interneurons and synaptic and intrinsic plasticity. A more complex associative neural network was developed based on this basic koniocortex-like neural networ...
Neural Plasticity, 2020
Several research studies point to the fact that sensory and cognitive reductions like cataracts, ... more Several research studies point to the fact that sensory and cognitive reductions like cataracts, deafness, macular degeneration, or even lack of activity after job retirement, precede the onset of Alzheimer’s disease. To simulate Alzheimer’s disease earlier stages, which manifest in sensory cortices, we used a computational model of the koniocortex that is the first cortical stage processing sensory information. The architecture and physiology of the modeled koniocortex resemble those of its cerebral counterpart being capable of continuous learning. This model allows one to analyze the initial phases of Alzheimer’s disease by “aging” the artificial koniocortex through synaptic pruning, by the modification of acetylcholine and GABA-A signaling, and by reducing sensory stimuli, among other processes. The computational model shows that during aging, a GABA-A deficit followed by a reduction in sensory stimuli leads to a dysregulation of neural excitability, which in the biological brain is associated with hypermetabolism, one of the earliest symptoms of Alzheimer’s disease.
A computational model for self-recovery of electricity distribution network was developed to simu... more A computational model for self-recovery of
electricity distribution network was developed to simulate it,
emulated by the IEEE 123 nodes model. The electrical system
considered has automatic switches capable of identifying a
momentary fault in the line and finding the best reconfiguration
for its reclosing. An artificial neural network (ANN),
backpropagation, was used to classify the type of failure and
determine the best reconfiguration of the distribution network.
Initially, five power failure scenarios were simulated in certain
different parts of the power grid, and power flow analysis via
OpenDSS was performed. Following, the most suitable
switching was observed within the shortest time interval to
restore the power supply. In this way it is possible to identify the
faulted segment in order to isolate it, leaving the smallest
number of consumers in the shortest possible time without power
supply. With the results of the simulations, tests and analyzes
were performed to verify their robustness and speed, in the
expectation that the model developed, be faster than an
experienced Operator of a Distribution Center.
International Journal of Neural Systems, 2016
In this paper, we use the neural property known as intrinsic plasticity to develop neural network... more In this paper, we use the neural property known as intrinsic plasticity to develop neural network models that resemble the koniocortex, the fourth layer of sensory cortices. These models evolved from a very basic two-layered neural network to a complex associative koniocortex network. In the initial network, intrinsic and synaptic plasticity govern the shifting of the activation function, and the modification of synaptic weights, respectively. In this first version, competition is forced, so that the most activated neuron is arbitrarily set to one and the others to zero, while in the second, competition occurs naturally due to inhibition between second layer neurons. In the third version of the network, whose architecture is similar to the koniocortex, competition also occurs naturally owing to the interplay between inhibitory interneurons and synaptic and intrinsic plasticity. A more complex associative neural network was developed based on this basic koniocortex-like neural network, capable of dealing with incomplete patterns and ideally suited to operating similarly to a learning vector quantization network. We also discuss the biological plausibility of the networks and their role in a more complex thalamocortical model.
The cost of equipment maintenance represents an important budgetary item in industrial and commer... more The cost of equipment maintenance represents an important budgetary item in industrial and commercial applications. Smart machines are able to evaluate on line a number of its own vitalities helping operators to diagnose faults. Most of time the origins of the problems are buried into intractable and not usually relevant data. Some neural architectures are presented in this paper for recognizing those operational trajectories that are the early symptoms of faults in these smart machines. In order to cope with such classification problem, a neural architecture defined as PREMON (Predictive Maintenance Oriented Network) has been designed. The main advantage of the system is its brain-inspired philosophy that allow it to be applied to a great deal of systems that are degraded or damaged because of their interaction with its environments.
Um dos principais problemas enfrentados pelas empresas que utilizam equipamentos de automação ban... more Um dos principais problemas enfrentados pelas empresas que utilizam equipamentos de automação bancária e comercial é aumentar ou manter a disponibilidade dos mesmos e reduzir o seu custo de manutenção. Atualmente, os equipamentos de automação possuem como recursos módulos de diagnósticos que auxiliam os técnicos em sua tomada de decisão, porém esses módulos são geralmente construídos sobre códigos de erros que mapeiam problemas pontuais (troubleshooting), sem levar em consideração a síndrome apresentada. Neste trabalho é apresentado o desenvolvimento e o desempenho de uma Rede Neural Artificial (RNA) para a identificação de falhas em equipamentos que apresentam como síndrome a degradação de seus componentes. Para isso, o algoritmo de Levenberg-Marquardt é utilizado visando agilizar o processo de treinamento da Rede implementada, demonstrando-se totalmente adequado a este tipo de aplicação. Os resultados apresentados mostram que a abordagem aqui descrita é viável para a detecção de falhas em equipamentos eletro-eletrônicos, fornecendo erros de identificação inferiores a 3%.
Este trabalho introduz os conceitos dos Sistemas Genéricos para Diagnóstico de Falhas e suas apli... more Este trabalho introduz os conceitos dos Sistemas Genéricos para Diagnóstico de Falhas e suas aplicações na manutenção de equipamentos aeronáuticos. As noções de Sistemas Complexos Diagnosticáveis bem como o processo de diagnóstico e sua classificação são mostrados sempre tendo como foco sua complexidade. Utilizando técnicas de Redes Neurais Artificiais para o diagnóstico preditivo de componentes que apresentam degradação em suas funcionalidades, a manutenção preventiva pode agir no momento oportuno para evitar um excessivo desgaste e a eventual parada do sistema considerado. Neste artigo, um estudo de caso considerando o sensor de ângulo de ataque de uma aeronave genérica é apresentado como componente possível de ser aplicado um Sistema de Diagnóstico Automático (SDA).
This paper presents an introduction to the concepts of Systems for Fault Detection and Diagnosis ... more This paper presents an introduction to the concepts of Systems for Fault Detection and Diagnosis and its automation applications to ATM maintenance. The basics of the diagnosis processes and its classifications are presented focusing on their complexity. The application of artificial neural networks for prognosticating early wear and degradation of critical components is explored. A case study involving the diagnosis of MMD sensored of ATM is also developed.
marianrogers.com.br
... POLI/USP 3Centro Universitário das Faculdades Metropolitanas Unidas UniFMU * Autor Responsá... more ... POLI/USP 3Centro Universitário das Faculdades Metropolitanas Unidas UniFMU * Autor Responsável: Mariana Antonia Aguiar, +55 11 3091 ... Funçao Diagnostica (Azul), Funçao Treinamento (Vermelho) e Valores Preditivos Retornados pela RNA (Verde) Funçao Diagnostica ...
Objectives: Memory lose takes place in an artificial koniocortex network when calcium-homeostasis... more Objectives: Memory lose takes place in an artificial koniocortex network when calcium-homeostasis impairment is simulated. According to the calcium dysregulation hypothesis of Alzheimer Disease, calcium dysregulation affects intrinsic plasticity. Intrinsic plasticity is the property of real neurons of adapting their firing thresholds according to the neurons activation regime: when a neuron is intensely stimulated it increases its firing threshold so that it becomes less responsive. On the contrary, when a neuron is scarcely stimulated, it reduces its firing threshold so that it is more prone to fire. Methods: A computational model of the koniocotex was trained with alphanumerical patterns, exhibiting successful learning capabilities. Successful learning was confirmed by stimulating each of the spiny stellate neurons of the model, and checking whether original alphanumerical patterns were recovered. Results: Due to calcium dysregulation, intrinsic plasticity might be impaired in eld...
International Journal of Neural Systems, 2016
In this paper, we use the neural property known as intrinsic plasticity to develop neural network... more In this paper, we use the neural property known as intrinsic plasticity to develop neural network models that resemble the koniocortex, the fourth layer of sensory cortices. These models evolved from a very basic two-layered neural network to a complex associative koniocortex network. In the initial network, intrinsic and synaptic plasticity govern the shifting of the activation function, and the modification of synaptic weights, respectively. In this first version, competition is forced, so that the most activated neuron is arbitrarily set to one and the others to zero, while in the second, competition occurs naturally due to inhibition between second layer neurons. In the third version of the network, whose architecture is similar to the koniocortex, competition also occurs naturally owing to the interplay between inhibitory interneurons and synaptic and intrinsic plasticity. A more complex associative neural network was developed based on this basic koniocortex-like neural networ...
Neural Plasticity, 2020
Several research studies point to the fact that sensory and cognitive reductions like cataracts, ... more Several research studies point to the fact that sensory and cognitive reductions like cataracts, deafness, macular degeneration, or even lack of activity after job retirement, precede the onset of Alzheimer’s disease. To simulate Alzheimer’s disease earlier stages, which manifest in sensory cortices, we used a computational model of the koniocortex that is the first cortical stage processing sensory information. The architecture and physiology of the modeled koniocortex resemble those of its cerebral counterpart being capable of continuous learning. This model allows one to analyze the initial phases of Alzheimer’s disease by “aging” the artificial koniocortex through synaptic pruning, by the modification of acetylcholine and GABA-A signaling, and by reducing sensory stimuli, among other processes. The computational model shows that during aging, a GABA-A deficit followed by a reduction in sensory stimuli leads to a dysregulation of neural excitability, which in the biological brain is associated with hypermetabolism, one of the earliest symptoms of Alzheimer’s disease.
A computational model for self-recovery of electricity distribution network was developed to simu... more A computational model for self-recovery of
electricity distribution network was developed to simulate it,
emulated by the IEEE 123 nodes model. The electrical system
considered has automatic switches capable of identifying a
momentary fault in the line and finding the best reconfiguration
for its reclosing. An artificial neural network (ANN),
backpropagation, was used to classify the type of failure and
determine the best reconfiguration of the distribution network.
Initially, five power failure scenarios were simulated in certain
different parts of the power grid, and power flow analysis via
OpenDSS was performed. Following, the most suitable
switching was observed within the shortest time interval to
restore the power supply. In this way it is possible to identify the
faulted segment in order to isolate it, leaving the smallest
number of consumers in the shortest possible time without power
supply. With the results of the simulations, tests and analyzes
were performed to verify their robustness and speed, in the
expectation that the model developed, be faster than an
experienced Operator of a Distribution Center.
International Journal of Neural Systems, 2016
In this paper, we use the neural property known as intrinsic plasticity to develop neural network... more In this paper, we use the neural property known as intrinsic plasticity to develop neural network models that resemble the koniocortex, the fourth layer of sensory cortices. These models evolved from a very basic two-layered neural network to a complex associative koniocortex network. In the initial network, intrinsic and synaptic plasticity govern the shifting of the activation function, and the modification of synaptic weights, respectively. In this first version, competition is forced, so that the most activated neuron is arbitrarily set to one and the others to zero, while in the second, competition occurs naturally due to inhibition between second layer neurons. In the third version of the network, whose architecture is similar to the koniocortex, competition also occurs naturally owing to the interplay between inhibitory interneurons and synaptic and intrinsic plasticity. A more complex associative neural network was developed based on this basic koniocortex-like neural network, capable of dealing with incomplete patterns and ideally suited to operating similarly to a learning vector quantization network. We also discuss the biological plausibility of the networks and their role in a more complex thalamocortical model.
The cost of equipment maintenance represents an important budgetary item in industrial and commer... more The cost of equipment maintenance represents an important budgetary item in industrial and commercial applications. Smart machines are able to evaluate on line a number of its own vitalities helping operators to diagnose faults. Most of time the origins of the problems are buried into intractable and not usually relevant data. Some neural architectures are presented in this paper for recognizing those operational trajectories that are the early symptoms of faults in these smart machines. In order to cope with such classification problem, a neural architecture defined as PREMON (Predictive Maintenance Oriented Network) has been designed. The main advantage of the system is its brain-inspired philosophy that allow it to be applied to a great deal of systems that are degraded or damaged because of their interaction with its environments.
Um dos principais problemas enfrentados pelas empresas que utilizam equipamentos de automação ban... more Um dos principais problemas enfrentados pelas empresas que utilizam equipamentos de automação bancária e comercial é aumentar ou manter a disponibilidade dos mesmos e reduzir o seu custo de manutenção. Atualmente, os equipamentos de automação possuem como recursos módulos de diagnósticos que auxiliam os técnicos em sua tomada de decisão, porém esses módulos são geralmente construídos sobre códigos de erros que mapeiam problemas pontuais (troubleshooting), sem levar em consideração a síndrome apresentada. Neste trabalho é apresentado o desenvolvimento e o desempenho de uma Rede Neural Artificial (RNA) para a identificação de falhas em equipamentos que apresentam como síndrome a degradação de seus componentes. Para isso, o algoritmo de Levenberg-Marquardt é utilizado visando agilizar o processo de treinamento da Rede implementada, demonstrando-se totalmente adequado a este tipo de aplicação. Os resultados apresentados mostram que a abordagem aqui descrita é viável para a detecção de falhas em equipamentos eletro-eletrônicos, fornecendo erros de identificação inferiores a 3%.
Este trabalho introduz os conceitos dos Sistemas Genéricos para Diagnóstico de Falhas e suas apli... more Este trabalho introduz os conceitos dos Sistemas Genéricos para Diagnóstico de Falhas e suas aplicações na manutenção de equipamentos aeronáuticos. As noções de Sistemas Complexos Diagnosticáveis bem como o processo de diagnóstico e sua classificação são mostrados sempre tendo como foco sua complexidade. Utilizando técnicas de Redes Neurais Artificiais para o diagnóstico preditivo de componentes que apresentam degradação em suas funcionalidades, a manutenção preventiva pode agir no momento oportuno para evitar um excessivo desgaste e a eventual parada do sistema considerado. Neste artigo, um estudo de caso considerando o sensor de ângulo de ataque de uma aeronave genérica é apresentado como componente possível de ser aplicado um Sistema de Diagnóstico Automático (SDA).
This paper presents an introduction to the concepts of Systems for Fault Detection and Diagnosis ... more This paper presents an introduction to the concepts of Systems for Fault Detection and Diagnosis and its automation applications to ATM maintenance. The basics of the diagnosis processes and its classifications are presented focusing on their complexity. The application of artificial neural networks for prognosticating early wear and degradation of critical components is explored. A case study involving the diagnosis of MMD sensored of ATM is also developed.