Search results for: Optical Network Unit

Authors: Oluwaponmile D. Alao

Abstract:

In this paper, two models have been developed to ascertain the best network needed for diagnosis of breast cancer. Breast cancer has been a disease that required the attention of the medical practitioner. Experience has shown that misdiagnose of the disease has been a major challenge in the medical field. Therefore, designing a system with adequate performance for will help in making diagnosis of the disease faster and accurate. In this paper, two models: backpropagation neural network and support vector machine has been developed. The performance obtained is also compared with other previously obtained algorithms to ascertain the best algorithms.

Keywords: breast cancer, data mining, neural network, support vector machine

Procedia PDF Downloads 348

Authors: Sirilak Areerachakul, Nat Ployong, Supayothin Na Songkla

Abstract:

This research’s objective is to select the model with most accurate value by using Neural Network Technique as a way to filter potential students who enroll in IT course by electronic learning at Suan Suanadha Rajabhat University. It is designed to help students selecting the appropriate courses by themselves. The result showed that the most accurate model was 100 Folds Cross-validation which had 73.58% points of accuracy.

Keywords: artificial neural network, classification, students, e-learning

Procedia PDF Downloads 429

Authors: Ashkan Zakaryazad, Ekrem Duman

Abstract:

A typical classification technique ranks the instances in a data set according to the likelihood of belonging to one (positive) class. A credit card (CC) fraud detection model ranks the transactions in terms of probability of being fraud. In fact, this approach is often criticized, because firms do not care about fraud probability but about the profitability or costliness of detecting a fraudulent transaction. The key contribution in this study is to focus on the profit maximization in the model building step. The artificial neural network proposed in this study works based on profit maximization instead of minimizing the error of prediction. Moreover, some studies have shown that the back propagation algorithm, similar to other gradient–based algorithms, usually gets trapped in local optima and swarm-based algorithms are more successful in this respect. In this study, we train our profit maximization ANN using the Migrating Birds optimization (MBO) which is introduced to literature recently.

Keywords: neural network, profit-based neural network, sum of squared errors (SSE), MBO, gradient descent

Procedia PDF Downloads 479

Authors: Shuo Liu

Abstract:

In Chinese text classification tasks, redundant words and phrases can interfere with the formation of extracted and analyzed text information, leading to a decrease in the accuracy of the classification model. To reduce irrelevant elements, extract and utilize text content information more efficiently and improve the accuracy of text classification models. In this paper, the text in the corpus is first extracted using the TextRank algorithm for abstraction, the words in the abstract are used as nodes to construct a text graph, and then the graph attention network (GAT) is used to complete the task of classifying the text. Testing on a Chinese dataset from the network, the classification accuracy was improved over the direct method of generating graph structures using text.

Keywords: Chinese natural language processing, text classification, abstract extraction, graph attention network

Procedia PDF Downloads 107

Authors: Mansoureh Naderipour, Susan Bastani, Mohammad Fazel Zarandi

Abstract:

Predicting links that may occur in the future and missing links in social networks is an attractive problem in social network analysis. Granular computing can help us to model the relationships between human-based system and social sciences in this field. In this paper, we present a model based on granular computing approach and Type-2 fuzzy logic to predict links regarding nodes’ activity and the relationship between two nodes. Our model is tested on collaboration networks. It is found that the accuracy of prediction is significantly higher than the Type-1 fuzzy and crisp approach.

Keywords: social network, link prediction, granular computing, type-2 fuzzy sets

Procedia PDF Downloads 330

Authors: Shin Je Lee, Go Bong Choi, Jeong Cheol Seo, Jong Min Lee, Gibaek Lee

Abstract:

Water pipe network is installed underground and once equipped; it is difficult to recognize the state of pipes when the leak or burst happens. Accordingly, post management is often delayed after the fault occurs. Therefore, the systematic fault management system of water pipe network is required to prevent the accident and minimize the loss. In this work, we develop online fault detection system of water pipe network using data of pipes such as flow rate or pressure. The transient model describing water flow in pipelines is presented and simulated using Matlab. The fault situations such as the leak or burst can be also simulated and flow rate or pressure data when the fault happens are collected. Faults are detected using statistical methods of fast Fourier transform and discrete wavelet transform, and they are compared to find which method shows the better fault detection performance.

Keywords: fault detection, water pipeline model, fast Fourier transform, discrete wavelet transform

Procedia PDF Downloads 516

Authors: Beatriz Martinez-Pastor, Maria Nogal, Alan O'Connor

Abstract:

During the last years, the number of extreme weather events has increased. A variety of extreme weather events, including river floods, rain-induced landslides, droughts, winter storms, wildfire, and hurricanes, have threatened and damaged many different regions worldwide. These events have a devastating impact on critical infrastructure systems resulting in high social, economical and environmental costs. These events have a huge impact in transport systems. Since, transport networks are completely exposed to every kind of climatological perturbations, and its performance is closely related with these events. When a traffic network is affected by a climatological hazard, the quality of its service is threatened, and the level of the traffic conditions usually decreases. With the aim of understanding this process, the concept of resilience has become most popular in the area of transport. Transport resilience analyses the behavior of a traffic network when a perturbation takes place. This holistic concept studies the complete process, from the beginning of the perturbation until the total recovery of the system, when the perturbation has finished. Many concepts are included in the definition of resilience, such as vulnerability, redundancy, adaptability, and safety. Once the resilience of a transport network can be evaluated, in this case, the methodology used is a dynamic equilibrium-restricted assignment model that allows the quantification of the concept, the next step is its improvement. Through the improvement of this concept, it will be possible to create transport networks that are able to withstand and have a better performance under the presence of climatological hazards. Analyzing the impact of a perturbation in a traffic network, it is observed that the response of the different links, which are part of the network, can be completely different from one to another. Consequently and due to this effect, many questions arise, as what makes a link more critical before an extreme weather event? or how is it possible to identify these critical links? With this aim, and knowing that most of the times the owners or managers of the transport systems have limited resources, the identification of the critical links of a transport network before extreme weather events, becomes a crucial objective. For that reason, using the available resources in the areas that will generate a higher improvement of the resilience, will contribute to the global development of the network. Therefore, this paper wants to analyze what kind of characteristic makes a link a critical one when an extreme weather event damages a transport network and finally identify them.

Keywords: critical links, extreme weather events, hazard, resilience, transport network

Procedia PDF Downloads 290

Authors: Rupu Yang, Cong Toan Tran, Assaad Zoughaib

Abstract:

The work provides an iterative nonlinear programming method to synthesize a heat exchanger network by manipulating the trade-offs between the heat load of process heat exchangers (HEs) and utilities. We consider for the synthesis problem two cases, the first one without fixed cost for HEs, and the second one with fixed cost. For the no fixed cost problem, the nonlinear programming (NLP) model with all the potential HEs is optimized to obtain the global optimum. For the case with fixed cost, the NLP model is iterated through adding/removing HEs. The method was applied in five case studies and illustrated quite well effectiveness. Among which, the approach reaches the lowest TAC (2,904,026$/year) compared with the best record for the famous Aromatic plants problem. It also locates a slightly better design than records in literature for a 10 streams case without fixed cost with only 1/9 computational time. Moreover, compared to the traditional mixed-integer nonlinear programming approach, the iterative NLP method opens a possibility to consider constraints (such as controllability or dynamic performances) that require knowing the structure of the network to be calculated.

Keywords: heat exchanger network, synthesis, NLP, optimization

Procedia PDF Downloads 169

Authors: Lori W. Gordon, Karen A. Jones

Abstract:

Today’s high-speed data connectivity depends on a vast global network of infrastructure across space, air, land, and sea, with undersea cable infrastructure (UCI) serving as the primary means for intercontinental and ‘long-haul’ communications. The UCI landscape is changing and includes an increasing variety of state actors, such as the growing economies of Brazil, Russia, India, China, and South Africa. Non-state commercial actors, such as hyper-scale content providers including Google, Facebook, Microsoft, and Amazon, are also seeking to control their data and networks through significant investments in submarine cables. Active investments by both state and non-state actors will invariably influence the growth, geopolitics, and security of this sector. Beyond these hyper-scale content providers, there are new commercial satellite communication providers. These new players include traditional geosynchronous (GEO) satellites that offer broad coverage, high throughput GEO satellites offering high capacity with spot beam technology, low earth orbit (LEO) ‘mega constellations’ – global broadband services. And potential new entrants such as High Altitude Platforms (HAPS) offer low latency connectivity, LEO constellations offer high-speed optical mesh networks, i.e., ‘fiber in the sky.’ This paper focuses on understanding the role of submarine cables within the larger context of the global data commons, spanning space, terrestrial, air, and sea networks, including an analysis of national security policy and geopolitical implications. As network operators and commercial and government stakeholders plan for emerging technologies and architectures, hedging risks for future connectivity will ensure that our data backbone will be secure for years to come.

Keywords: communications, global, infrastructure, technology

Procedia PDF Downloads 91

Authors: Jinan Fiaidhi, Sabah Mohammed

Abstract:

Cross-domain learning in the medical field is a research challenge as many conditions, like in oncology imaging, use different imaging modalities. Moreover, in most of the medical learning applications, the sample training size is relatively small. Although few-shot learning (FSL) through the use of a Siamese neural network was able to be trained on a small sample with remarkable accuracy, FSL fails to be effective for use in multiple domains as their convolution weights are set for task-specific applications. In this paper, we are addressing this problem by enabling FSL to possess the ability to shift across domains by designing a two-layer FSL network that can learn individually from each domain and produce a shared features map with extra modulation to be used at the second layer that can recognize important targets from mix domains. Our initial experimentations based on mixed medical datasets like the Medical-MNIST reveal promising results. We aim to continue this research to perform full-scale analytics for testing our cross-domain FSL learning.

Keywords: Siamese neural network, few-shot learning, meta-learning, metric-based learning, thick data transformation and analytics

Procedia PDF Downloads 64

Authors: Kemka C. Ihemelandu, Chukwuemeka U. Ihemelandu

Abstract:

The incidence of melanoma has been increasing rapidly over the past two decades, making melanoma a current public health crisis. Unfortunately, even as screening efforts continue to expand in an effort to ameliorate the death rate from melanoma, there is a need to improve diagnostic accuracy to decrease misdiagnosis. Artificial intelligence (AI) a new frontier in patient care has the ability to improve the accuracy of melanoma diagnosis. Convolutional neural network (CNN) a form of deep neural network, most commonly applied to analyze visual imagery, has been shown to outperform the human brain in pattern recognition. However, there are noted limitations with the accuracy of the CNN models. Our aim in this study was the optimization of convolutional neural network algorithms for the automated diagnosis of melanoma. We hypothesized that Optimal selection of the momentum and batch hyperparameter increases model accuracy. Our most successful model developed during this study, showed that optimal selection of momentum of 0.25, batch size of 2, led to a superior performance and a faster model training time, with an accuracy of ~ 83% after nine hours of training. We did notice a lack of diversity in the dataset used, with a noted class imbalance favoring lighter vs. darker skin tone. Training set image transformations did not result in a superior model performance in our study.

Keywords: melanoma, convolutional neural network, momentum, batch hyperparameter

Procedia PDF Downloads 105

Authors: Dana M. Ragab, Jasim A. Ghaeb

Abstract:

The three-phase power system suffers from different challenging problems, e.g. voltage unbalance conditions at the load side. The voltage unbalance usually degrades the power quality of the electric power system. Several techniques can be considered for load balancing including load reconfiguration, static synchronous compensator and static reactive power compensator. In this work an efficient neural network is designed to control the unbalanced condition in the Aqaba-Qatrana-South Amman (AQSA) electric power system. It is designed for highly enhanced response time of the reactive compensator for voltage balancing. The neural network is developed to determine the appropriate set of firing angles required for the thyristor-controlled reactor to balance the three load voltages accurately and quickly. The parameters of AQSA power system are considered in the laboratory model, and several test cases have been conducted to test and validate the proposed technique capabilities. The results have shown a high performance of the proposed Neural Network Control (NNC) technique for correcting the voltage unbalance conditions at three-phase load based on accuracy and response time.

Keywords: three-phase power system, reactive power control, voltage unbalance factor, neural network, power quality

Procedia PDF Downloads 204

Authors: Xianglu Tang, Zhenxue Jiang, Zhuo Li

Abstract:

Characterization of shale multi scale heterogeneity is an important part to evaluate size and space distribution of shale gas reservoirs in sedimentary basins. The origin of shale heterogeneity has always been a hot research topic for it determines shale micro characteristics description and macro quality reservoir prediction. Shale multi scale heterogeneity was discussed based on thin section observation, FIB-SEM, QEMSCAN, TOC, XRD, mercury intrusion porosimetry (MIP), and nitrogen adsorption analysis from 30 core samples in Silurian Longmaxi formation. Results show that shale heterogeneity can be characterized by pore structure and mineral composition. The heterogeneity of shale pore is showed by different size pores at nm-μm scale. Macropores (pore diameter > 50 nm) have a large percentage of pore volume than mesopores (pore diameter between 2~ 50 nm) and micropores (pore diameter < 2nm). However, they have a low specific surface area than mesopores and micropores. Fractal dimensions of the pores from nitrogen adsorption data are higher than 2.7, what are higher than 2.8 from MIP data, showing extremely complex pore structure. This complexity in pore structure is mainly due to the organic matter and clay minerals with complex pore network structures, and diagenesis makes it more complicated. The heterogeneity of shale minerals is showed by mineral grains, lamina, and different lithology at nm-km scale under the continuous changing horizon. Through analyzing the change of mineral composition at each scale, random arrangement of mineral equal proportion, seasonal climate changes, large changes of sedimentary environment, and provenance supply are considered to be the main reasons that cause shale minerals heterogeneity from microcosmic to macroscopic. Due to scale effect, the change of shale multi scale heterogeneity is a discontinuous process, and there is a transformation boundary between homogeneous and in homogeneous. Therefore, a shale multi scale heterogeneity changing model is established by defining four types of homogeneous unit at different scales, which can be used to guide the prediction of shale gas distribution from micro scale to macro scale.

Keywords: heterogeneity, homogeneous unit, multiscale, shale

Procedia PDF Downloads 457

Authors: M. Khouil, N. Saber, M. Mestari

Abstract:

This study aimed to propose, a different architecture of a Path Planning using the NECMOP. where several nonlinear objective functions must be optimized in a conflicting situation. The ability to detect and avoid collision is very important for mobile intelligent machines. However, many artificial vision systems are not yet able to quickly and cheaply extract the wealth information. This network, which has been particularly reviewed, has enabled us to solve with a new approach the problem of collision detection between two convex polyhedra in a fixed time (O (1) time). We used two types of neurons linear and threshold logic, which simplified the actual implementation of all the networks proposed. This article represents a comprehensive algorithm that determine through the AMAXNET network a measure (a mini-maximum point) in a fixed time, which allows us to detect the presence of a potential collision.

Keywords: path planning, collision detection, convex polyhedron, neural network

Procedia PDF Downloads 443

Authors: M. S. Stepanova, V. V. Zakharov, P. D. Khavlyuk, I. D. Skurlov, A. Y. Dubovik, A. L. Rogach

Abstract:

Carbon dots are small carbon-based spherical nanoparticles, which are typically less than 10 nm in size that can be modified with surface passivation and heteroatoms doping. The light-absorbing ability of carbon dots has attracted a significant amount of attention in photoluminescence for bioimaging and fluorescence sensing applications owing to their advantages, such as tunable fluorescence emission, photo- and thermostability and low toxicity. In this study, carbon dots were synthesized by the solvothermal method from citric acid and ethylenediamine dissolved in water. The solution was heated for 5 hours at 200°C and then cooled down to room temperature. The carbon dots films were obtained by evaporation from a high-concentration aqueous solution. The increase of both luminescence intensity and light transmission was obtained as a result of a 405 nm laser exposure to a part of the carbon dots film, which was detected using a confocal laser scanning microscope (LSM 710, Zeiss). Blueshift up to 35 nm of the luminescence spectrum is observed as luminescence intensity, which is increased more than twofold. The exact value of the shift depends on the time of the laser exposure. This shift can be caused by the modification of surface groups at the carbon dots, which are responsible for long-wavelength luminescence. In addition, a shift of the absorption peak by 10 nm and a decrease in the optical density at the wavelength of 350 nm is detected, which is responsible for the absorption of surface groups. The obtained sample was also studied with time-resolved confocal fluorescence microscope (MicroTime 100, PicoQuant), which made it possible to receive a time-resolved photoluminescence image and construct emission decays of the laser-exposed and non-exposed areas. 5 MHz pulse rate impulse laser has been used as a photoluminescence excitation source. Photoluminescence decay was approximated by two exhibitors. The laser-exposed area has the amplitude of the first-lifetime component (A1) twice as much as before, with increasing τ1. At the same time, the second-lifetime component (A2) decreases. These changes evidence a modification of the surface groups of carbon dots. The detected effect can be used to create thermostable fluorescent marks, the physical size of which is bounded by the diffraction limit of the optics (~ 200-300 nm) used for exposure and to improve the optical properties of carbon dots or in the field of optical encryption. Acknowledgements: This work was supported by the Ministry of Science and Higher Education of Russian Federation, goszadanie no. 2019-1080 and financially supported by Government of Russian Federation, Grant 08-08.

Keywords: carbon dots, photoactivation, optical properties, photoluminescence and absorption spectra

Procedia PDF Downloads 170

Authors: Se-Mi Lim, Won-Jae Jung, Jin-Sup Kim, Jun-Seok Park, Hyung-Il Chae

Abstract:

This paper presents a current-mode vertical hall device (VHD) structure using metal layers in a CMOS process. The proposed metal layer based vertical hall device (MLVHD) utilizes vertical connection among metal layers (from M1 to the top metal) to facilitate hall effect. The vertical metal structure unit flows a bias current Ibias from top to bottom, and an external magnetic field changes the current distribution by Lorentz force. The asymmetric current distribution can be detected by two differential-mode current outputs on each side at the bottom (M1), and each output sinks Ibias/2 ± Ihall. A single vertical metal structure generates only a small amount of hall effect of Ihall due to the short length from M1 to the top metal as well as the low conductivity of the metal, and a series connection between thousands of vertical structure units can solve the problem by providing NxIhall. The series connection between two units is another vertical metal structure flowing current in the opposite direction, and generates negative hall effect. To mitigate the negative hall effect from the series connection, the differential current outputs at the bottom (M1) from one unit merges on the top metal level of the other unit. The proposed MLVHD is simulated in a 3-dimensional model simulator in COMSOL Multiphysics, with 0.35 μm CMOS process parameters. The simulated MLVHD unit size is (W) 10 μm × (L) 6 μm × (D) 10 μm. In this paper, we use an MLVHD with 10 units; the overall hall device size is (W) 10 μm × (L)78 μm × (D) 10 μm. The COMSOL simulation result is as following: the maximum hall current is approximately 2 μA with a 12 μA bias current and 100mT magnetic field; This work was supported by Institute for Information & communications Technology Promotion(IITP) grant funded by the Korea government(MSIP) (No.R7117-16-0165, Development of Hall Effect Semiconductor for Smart Car and Device).

Keywords: CMOS, vertical hall device, current mode, COMSOL

Procedia PDF Downloads 309

Authors: Danilo López, Nelson Vera, Luis Pedraza

Abstract:

This paper analyzes fundamental ideas and concepts related to neural networks, which provide the reader a theoretical explanation of Long Short-Term Memory (LSTM) networks operation classified as Deep Learning Systems, and to explicitly present the mathematical development of Backward Pass equations of the LSTM network model. This mathematical modeling associated with software development will provide the necessary tools to develop an intelligent system capable of predicting the behavior of licensed users in wireless cognitive radio networks.

Keywords: neural networks, multilayer perceptron, long short-term memory, recurrent neuronal network, mathematical analysis

Procedia PDF Downloads 424

Authors: Leonid Zhukov, Dmytro Petrenko

Abstract:

Optical thermometry has no alternative in many cases of industrial most effective continuous temperature control. Classical optical thermometry technologies can be used on available for pyrometers controlled objects with stable radiation characteristics and transmissivity of the intermediate medium. Without using temperature corrections, it is possible in the case of a “black” body for energy pyrometry and the cases of “black” and “grey” bodies for spectral ratio pyrometry or with using corrections – for any colored bodies. Consequently, with increasing the number of operating waves, optical thermometry possibilities to reduce methodical errors significantly expand. That is why, in recent 25-30 years, research works have been reoriented on more perfect spectral (multicolor) thermometry technologies. There are two physical material substances, i.e., substance (controlled object) and electromagnetic field (thermal radiation), to be operated in optical thermometry. Heat is transferred by radiation; therefore, radiation has the energy, entropy, and temperature. Optical thermometry was originating simultaneously with the developing of thermal radiation theory when the concept and the term "radiation temperature" was not used, and therefore concepts and terms "conditional temperatures" or "pseudo temperature" of controlled objects were introduced. They do not correspond to the physical sense and definitions of temperature in thermodynamics, molecular-kinetic theory, and statistical physics. Launched by the scientific thermometric society, discussion about the possibilities of temperature measurements of objects, including colored bodies, using the temperatures of their radiation is not finished. Are the information about controlled objects transferred by their radiation enough for temperature measurements? The positive and negative answers on this fundamental question divided experts into two opposite camps. Recent achievements of spectral thermometry develop events in her favour and don’t leave any hope for skeptics. This article presents the results of investigations and developments in the field of spectral thermometry carried out by the authors in the Department of Thermometry and Physics-Chemical Investigations. The authors have many-year’s of experience in the field of modern optical thermometry technologies. Innovative technologies of optical continuous temperature control have been developed: symmetric-wave, two-color compensative, and based on obtained nonlinearity equation of spectral emissivity distribution linear, two-range, and parabolic. Тhe technologies are based on direct measurements of physically substantiated and proposed by Prof. L. Zhukov, radiation temperatures with the next calculation of the controlled object temperature using this radiation temperatures and corresponding mathematical models. Тhe technologies significantly increase metrological characteristics of continuous contactless and light-guide temperature control in energy, metallurgical, ceramic, glassy, and other productions. For example, under the same conditions, the methodical errors of proposed technologies are less than the errors of known spectral and classical technologies in 2 and 3-13 times, respectively. Innovative technologies provide quality products obtaining at the lowest possible resource-including energy costs. More than 600 publications have been published on the completed developments, including more than 100 domestic patents, as well as 34 patents in Australia, Bulgaria, Germany, France, Canada, the USA, Sweden, and Japan. The developments have been implemented in the enterprises of USA, as well as Western Europe and Asia, including Germany and Japan.

Keywords: emissivity, radiation temperature, object temperature, spectral thermometry

Procedia PDF Downloads 101

Authors: Yifan Fan, Xudong Luo, Pingping Lin

Abstract:

Nowadays, dialogue systems increasingly become the way for humans to access many computer systems. So, humans can interact with computers in natural language. A dialogue system consists of three parts: understanding what humans say in natural language, managing dialogue, and generating responses in natural language. In this paper, we survey deep learning based methods for dialogue management, response generation and dialogue evaluation. Specifically, these methods are based on neural network, long short-term memory network, deep reinforcement learning, pre-training and generative adversarial network. We compare these methods and point out the further research directions.

Keywords: dialogue management, response generation, deep learning, evaluation

Procedia PDF Downloads 172

Authors: Jui-Pin Hung, Qiao-Wen Chang, Kung-Da Wu, Yong-Run Chen

Abstract:

This study was aimed to investigate the machining stability of a spindle tool with different preloaded amount. To this end, the vibration tests were conducted on the spindle unit with different preload to assess the dynamic characteristics and machining stability of the spindle unit. Current results demonstrate that the tool tip frequency response characteristics and the machining stabilities in X and Y direction are affected to change for spindle with different preload. As can be found from the results, a high preloaded spindle tool shows higher limited cutting depth at mid position, while a spindle with low preload shows a higher limited depth. This implies that the machining stability of spindle tool system is affected to vary by the machine frame structure. Besides, such an effect is quite different and varied with the preload of the spindle.

Keywords: bearing preload, dynamic compliance, machining stability, spindle

Procedia PDF Downloads 389

Authors: Jie Liu, Jinqu Cheng, Qiyuan Peng, Yong Yin

Abstract:

Urban rail transit (URT) plays a significant role in dealing with traffic congestion and environmental problems in cities. However, equipment failure and obstruction of links often lead to URT links’ capacities loss in daily operation. It affects the reliability and transport service quality of URT network seriously. In order to measure the influence of links’ capacities loss on reliability and transport service quality of URT network, passengers are divided into three categories in case of links’ capacities loss. Passengers in category 1 are less affected by the loss of links’ capacities. Their travel is reliable since their travel quality is not significantly reduced. Passengers in category 2 are affected by the loss of links’ capacities heavily. Their travel is not reliable since their travel quality is reduced seriously. However, passengers in category 2 still can travel on URT. Passengers in category 3 can not travel on URT because their travel paths’ passenger flow exceeds capacities. Their travel is not reliable. Thus, the proportion of passengers in category 1 whose travel is reliable is defined as reliability indicator of URT network. The transport service quality of URT network is related to passengers’ travel time, passengers’ transfer times and whether seats are available to passengers. The generalized travel cost is a comprehensive reflection of travel time, transfer times and travel comfort. Therefore, passengers’ average generalized travel cost is used as transport service quality indicator of URT network. The impact of links’ capacities loss on transport service quality of URT network is measured with passengers’ relative average generalized travel cost with and without links’ capacities loss. The proportion of the passengers affected by links and betweenness of links are used to determine the important links in URT network. The stochastic user equilibrium distribution model based on the improved logit model is used to determine passengers’ categories and calculate passengers’ generalized travel cost in case of links’ capacities loss, which is solved with method of successive weighted averages algorithm. The reliability and transport service quality indicators of URT network are calculated with the solution result. Taking Wuhan Metro as a case, the reliability and transport service quality of Wuhan metro network is measured with indicators and method proposed in this paper. The result shows that using the proportion of the passengers affected by links can identify important links effectively which have great influence on reliability and transport service quality of URT network; The important links are mostly connected to transfer stations and the passenger flow of important links is high; With the increase of number of failure links and the proportion of capacity loss, the reliability of the network keeps decreasing, the proportion of passengers in category 3 keeps increasing and the proportion of passengers in category 2 increases at first and then decreases; When the number of failure links and the proportion of capacity loss increased to a certain level, the decline of transport service quality is weakened.

Keywords: urban rail transit network, reliability, transport service quality, links’ capacities loss, important links

Procedia PDF Downloads 131

Authors: Firas M. Tuaimah

Abstract:

This paper presents an Artificial Neural Network based approach for short-term load forecasting and exactly for two days ahead. Two seasons have been discussed for Iraqi power system, namely summer and winter; the hourly load demand is the most important input variables for ANN based load forecasting. The recorded daily load profile with a lead time of 1-48 hours for July and December of the year 2012 was obtained from the operation and control center that belongs to the Ministry of Iraqi electricity. The results of the comparison show that the neural network gives a good prediction for the load forecasting and for two days ahead.

Keywords: short-term load forecasting, artificial neural networks, back propagation learning, hourly load demand

Procedia PDF Downloads 470

Authors: Xiaoxiao Li, Shuangcheng Jia, Qian Li

Abstract:

Keypoint detection has always been a challenge in the field of image recognition. This paper proposes a novelty keypoint detection method which is called Multi-Scale Feature Fusion Convolutional Network with Attention (MFFCNA). We verified that the multi-scale features with the attention mechanism module have better feature expression capability. The feature fusion between different scales makes the information that the network model can express more abundant, and the network is easier to converge. On our self-made street sign corner dataset, we validate the MFFCNA model with an accuracy of 97.8% and a recall of 81%, which are 5 and 8 percentage points higher than the HRNet network, respectively. On the COCO dataset, the AP is 71.9%, and the AR is 75.3%, which are 3 points and 2 points higher than HRNet, respectively. Extensive experiments show that our method has a remarkable improvement in the keypoint recognition tasks, and the recognition effect is better than the existing methods. Moreover, our method can be applied not only to keypoint detection but also to image classification and semantic segmentation with good generality.

Keywords: keypoint detection, feature fusion, attention, semantic segmentation

Procedia PDF Downloads 123

Authors: Saleh H. Abud, Z. Hassan, F. K. Yam

Abstract:

Characteristics of MSM photodetector based on a porous In0.08Ga0.92N thin film were reported. Nanoporous structures of n-type In0.08Ga0.92N/AlN/Si thin films were synthesized by photoelectrochemical (PEC) etching at a ratio of 1:4 of HF:C2H5OH solution for 15 min. The structural and optical properties of pre- and post-etched thin films were investigated. Field emission scanning electron microscope and atomic force microscope images showed that the pre-etched thin film has a sufficiently smooth surface over a large region and the roughness increased for porous film. Blue shift has been observed in photoluminescence emission peak at 300 K for porous sample. The photoluminescence intensity of the porous film indicated that the optical properties have been enhanced. A high work function metals (Pt and Ni) were deposited as a metal contact on the porous films. The rise and recovery times of the devices were investigated at 390 nm chopped light. Finally, the sensitivity and quantum efficiency were also studied.

Keywords: porous InGaN, photoluminescence, SMS photodetector, atomic force microscopy

Procedia PDF Downloads 493

Authors: Mohammad Izadkhah, Mojtaba Hoseini, Alireza Khalili Tehrani

Abstract:

In this paper we presented a new method for tracking flying targets in color video sequences based on contour and kernel. The aim of this work is to overcome the problem of losing target in changing light, large displacement, changing speed, and occlusion. The proposed method is made in three steps, estimate the target location by particle filter, segmentation target region using neural network and find the exact contours by greedy snake algorithm. In the proposed method we have used both region and contour information to create target candidate model and this model is dynamically updated during tracking. To avoid the accumulation of errors when updating, target region given to a perceptron neural network to separate the target from background. Then its output used for exact calculation of size and center of the target. Also it is used as the initial contour for the greedy snake algorithm to find the exact target's edge. The proposed algorithm has been tested on a database which contains a lot of challenges such as high speed and agility of aircrafts, background clutter, occlusions, camera movement, and so on. The experimental results show that the use of neural network increases the accuracy of tracking and segmentation.

Keywords: video tracking, particle filter, greedy snake, neural network

Procedia PDF Downloads 343

Authors: Cristian F.T. Montenegro, Luís F. N. Lourenço, Maurício B. C. Salles, Renato M. Monaro

Abstract:

More PV systems have been connected to the electrical network each year. As the number of PV systems increases, some issues affecting grid operations have been identified. This paper studied the impacts related to changes in solar irradiance on a distribution/sub-transmission network, considering variations due to moving clouds and daily cycles. Using MATLAB/Simulink software, a solar farm of 30 MWp was built and then implemented to a test network. From simulations, it has been determined that irradiance changes can have a significant impact on the grid by causing voltage fluctuations outside the allowable thresholds. This work discussed some local control strategies and grid reinforcements to mitigate the negative effects of the irradiance changes on the grid.

Keywords: reactive power control, solar irradiance, utility-scale PV systems, voltage fluctuations

Procedia PDF Downloads 464

Authors: Martins Y. Otache, John J. Musa, Abayomi I. Kuti, Mustapha Mohammed

Abstract:

The performance of an artificial neural network (ANN) is contingent on a host of factors, for instance, the network optimisation scheme. In view of this, the study examined the general implications of the ANN training optimisation algorithm on its forecast performance. To this end, the Bayesian regularisation (Br), Levenberg-Marquardt (LM), and the adaptive learning gradient descent: GDM (with momentum) algorithms were employed under different ANN structural configurations: (1) single-hidden layer, and (2) double-hidden layer feedforward back propagation network. Results obtained revealed generally that the gradient descent with momentum (GDM) optimisation algorithm, with its adaptive learning capability, used a relatively shorter time in both training and validation phases as compared to the Levenberg- Marquardt (LM) and Bayesian Regularisation (Br) algorithms though learning may not be consummated; i.e., in all instances considering also the prediction of extreme flow conditions for 1-day and 5-day ahead, respectively especially using the ANN model. In specific statistical terms on the average, model performance efficiency using the coefficient of efficiency (CE) statistic were Br: 98%, 94%; LM: 98 %, 95 %, and GDM: 96 %, 96% respectively for training and validation phases. However, on the basis of relative error distribution statistics (MAE, MAPE, and MSRE), GDM performed better than the others overall. Based on the findings, it is imperative to state that the adoption of ANN for real-time forecasting should employ training algorithms that do not have computational overhead like the case of LM that requires the computation of the Hessian matrix, protracted time, and sensitivity to initial conditions; to this end, Br and other forms of the gradient descent with momentum should be adopted considering overall time expenditure and quality of the forecast as well as mitigation of network overfitting. On the whole, it is recommended that evaluation should consider implications of (i) data quality and quantity and (ii) transfer functions on the overall network forecast performance.

Keywords: streamflow, neural network, optimisation, algorithm

Procedia PDF Downloads 158

Authors: Edward Guillén, Jeisson Sánchez, Carlos Omar Ramos

Abstract:

In recent years, a wide variety of applications are developed with Support Vector Machines -SVM- methods and Artificial Neural Networks -ANN-. In general, these methods depend on intrusion knowledge databases such as KDD99, ISCX, and CAIDA among others. New classes of detectors are generated by machine learning techniques, trained and tested over network databases. Thereafter, detectors are employed to detect anomalies in network communication scenarios according to user’s connections behavior. The first detector based on training dataset is deployed in different real-world networks with mobile and non-mobile devices to analyze the performance and accuracy over static detection. The vulnerabilities are based on previous work in telemedicine apps that were developed on the research group. This paper presents the differences on detections results between some network scenarios by applying traditional detectors deployed with artificial neural networks and support vector machines.

Keywords: anomaly detection, back-propagation neural networks, network intrusion detection systems, support vector machines

Procedia PDF Downloads 362

Authors: Niyada Naksuk, Thoetchai Peeraphatdit, Vitaly Herasevich, Peter A. Brady, Suraj Kapa, Samuel J. Asirvatham

Abstract:

Introduction: Little is known about the safety of antipsychotic therapy for delirium in the coronary care unit (CCU). Our aim was to examine the effect of delirium and antipsychotic therapy among CCU patients. Methods: Pre-study Confusion Assessment Method-Intensive Care Unit (CAM–ICU) criteria were implemented in screening consecutive patients admitted to Mayo Clinic, Rochester, the USA from 2004 through 2013. Death status was prospectively ascertained. Results: Of 11,079 study patients, the incidence of delirium was 8.3% (n=925). Delirium was associated with an increased risk of in-hospital mortality (adjusted OR 1.49; 95% CI, 1.08-2.08; P=.02) and one-year mortality among patients who survived from CCU admission (adjusted HR 1.46; 95% CI, 1.12-1.87; P=.005). A total of 792 doses of haloperidol (5 IQR [3-10] mg/day) or quetiapine (25 IQR [13-50] mg/day) were given to 244 patients with delirium. The clinical characteristics of patients with delirium who did and did not receive antipsychotic therapy were not different (baseline corrected QT [QTc] interval 460±61 ms vs. 457±58 ms, respectively; P = 0.57). In comparison to baseline, mean QTc intervals after the first and third doses of the antipsychotics were not significantly prolonged in haloperidol (448±56, 458±57, and 450±50 ms, respectively) or quetiapine groups (459±54, 467±68, and 462±46 ms, respectively) (P > 0.05 for all). Additionally, in-hospital mortality (adjusted OR 0.67; 95% CI, 0.42-1.04; P=.07), ventricular arrhythmia (adjusted OR 0.87; 95% CI, 0.17-3.62; P=.85) and one-year mortality among the hospital survivors (adjusted HR 0.86; 95% CI 0.62-1.17; P = 0.34) were not different in patients with delirium irrespective of whether or not they received antipsychotics. Conclusions: In patients admitted to the CCU, delirium was associated with an increase in both in-hospital and one-year mortality. Low doses of haloperidol and quetiapine appeared to be safe, without an increase in risk of sudden cardiac death, in-hospital mortality, or one-year mortality in carefully monitored patients.

Keywords: arrhythmias, haloperidol, mortality, qtc interval, quetiapine

Procedia PDF Downloads 377

Authors: Anastasiia Maklakova

Abstract:

The perovskite type oxides formed in the Ln-Me-Me/-O systems (where Ln – rare-earth, Me – alkaline earth metal, Me/ - 3-d metal) have potential applications as gas sensors, catalysts or cathode materials for IT-SOFCs due to the high values of mixed electronic -ionic conductivity and high oxygen diffusivity. Complex oxides in the Sr-(Pr,Gd)-Co-O systems were prepared via the glycerol-nitrate technique The phase composition was determined using a Shimadzu XRD-7000 diffractometer at room temperature in air. Phase identification was performed using the ICDD database. The structure was refined by the full-profile Rietveld method using Fullprof 2008 software. Gradual substitution of strontium by Pr or Gd leads to the decrease of unit cell parameters and unit cell volume that can be explained by the size factor. An introduction of Pr or Gd into the strontium cobaltite increases the oxygen content in samples.

Keywords: phase equilibria, crystal structure, oxygen nonstoichiometry, solid oxide fuel cell

Procedia PDF Downloads 121