Search results for: neural networking algorithm

Authors: Gwanghee Heo, Geonhyeok Bang, Chunggil Kim, Chinok Lee

Abstract:

This paper attempts to develop a wireless feedback control system as a primary step eventually toward a bio-inspired structure system where inanimate structure behaves like a life form autonomously. It is a standalone wireless control system which is supposed to measure externally caused structural responses, analyze structural state from acquired data, and take its own action on the basis of the analysis with an embedded logic. For an experimental examination of its effectiveness, we applied it on a model of two-span bridge and performed a wireless control test. Experimental tests have been conducted for comparison on both the wireless and the wired system under the conditions of Un-control, Passive-off, Passive-on, and Lyapunov control algorithm. By proving the congruence of the test result of the wireless feedback control system with the wired control system, its control performance was proven to be effective. Besides, it was found to be economical in energy consumption and also autonomous by means of a command algorithm embedded into it, which proves its basic capacity as a bio-inspired system.

Keywords: structural vibration control, wireless system, MR damper, feedback control, embedded system

Procedia PDF Downloads 202

Authors: Tomas Vampola, Michael Valášek

Abstract:

On the basis of simulation-generated failure states of structural elements of a turboprop engine suitable for the busy-jet class of aircraft, an algorithm for early prediction of damage or reduction in functionality of structural elements of the engine is designed and verified with real data obtained at dynamometric testing facilities of aircraft engines. Based on an expanding database of experimentally determined data from temperature and pressure sensors during the operation of turboprop engines, this strategy is constantly modified with the aim of using the minimum number of sensors to detect an inadmissible or deteriorated operating mode of specific structural elements of an aircraft engine. The assembled algorithm for the early prediction of reduced functionality of the aircraft engine significantly contributes to the safety of air traffic and to a large extent, contributes to the economy of operation with positive effects on the reduction of the energy demand of operation and the elimination of adverse effects on the environment.

Keywords: detectability of malfunction, dynamometric testing, prediction of damage, turboprop engine

Procedia PDF Downloads 85

Authors: B. R. Campomanes-Alvarez, P. Quiros, B. Fernandez

Abstract:

Automatic Speech Recognition (ASR) is a machine-based process of decoding and transcribing oral speech. A typical ASR system receives acoustic input from a speaker or an audio file, analyzes it using algorithms, and produces an output in the form of a text. Some speech recognition systems use Hidden Markov Models (HMMs) to deal with the temporal variability of speech and Gaussian Mixture Models (GMMs) to determine how well each state of each HMM fits a short window of frames of coefficients that represents the acoustic input. Another way to evaluate the fit is to use a feed-forward neural network that takes several frames of coefficients as input and produces posterior probabilities over HMM states as output. Deep neural networks (DNNs) that have many hidden layers and are trained using new methods have been shown to outperform GMMs on a variety of speech recognition systems. Acoustic models for state-of-the-art ASR systems are usually training on massive amounts of data. However, audio files with their corresponding transcriptions can be difficult to obtain, especially in the Spanish language. Hence, in the case of these low-resource scenarios, building an ASR model is considered as a complex task due to the lack of labeled data, resulting in an under-trained system. Semi-supervised learning approaches arise as necessary tasks given the high cost of transcribing audio data. The main goal of this proposal is to develop a procedure based on acoustic semi-supervised learning for Spanish ASR systems by using DNNs. This semi-supervised learning approach consists of: (a) Training a seed ASR model with a DNN using a set of audios and their respective transcriptions. A DNN with a one-hidden-layer network was initialized; increasing the number of hidden layers in training, to a five. A refinement, which consisted of the weight matrix plus bias term and a Stochastic Gradient Descent (SGD) training were also performed. The objective function was the cross-entropy criterion. (b) Decoding/testing a set of unlabeled data with the obtained seed model. (c) Selecting a suitable subset of the validated data to retrain the seed model, thereby improving its performance on the target test set. To choose the most precise transcriptions, three confidence scores or metrics, regarding the lattice concept (based on the graph cost, the acoustic cost and a combination of both), was performed as selection technique. The performance of the ASR system will be calculated by means of the Word Error Rate (WER). The test dataset was renewed in order to extract the new transcriptions added to the training dataset. Some experiments were carried out in order to select the best ASR results. A comparison between a GMM-based model without retraining and the DNN proposed system was also made under the same conditions. Results showed that the semi-supervised ASR-model based on DNNs outperformed the GMM-model, in terms of WER, in all tested cases. The best result obtained an improvement of 6% relative WER. Hence, these promising results suggest that the proposed technique could be suitable for building ASR models in low-resource environments.

Keywords: automatic speech recognition, deep neural networks, machine learning, semi-supervised learning

Procedia PDF Downloads 328

Authors: Amir Mohammad Fathollahi Fard, Mostafa Hajiaghaei-Keshteli, Mohammad Mahdi Paydar

Abstract:

With increasing life expectancy in developed countries, the role of home care services is highlighted by both academia and industrial contributors in Home Health Care Supply Chain (HHCSC) companies. The main decisions in such supply chain systems are the location of pharmacies, the allocation of patients to these pharmacies and also the routing and scheduling decisions of nurses to visit their patients. In this study, for the first time, an integrated model is proposed to consist of all preliminary and necessary decisions in these companies, namely, location-allocation-routing model. This model is a type of NP-hard one. Therefore, an Imperialist Competitive Algorithm (ICA) is utilized to solve the model, especially in large sizes. Results confirm the efficiency of the developed model for HHCSC companies as well as the performance of employed ICA.

Keywords: home health care supply chain, location-allocation-routing problem, imperialist competitive algorithm, optimization

Procedia PDF Downloads 391

Authors: Zizouni Khaled, Fali Leyla, Sadek Younes, Bousserhane Ismail Khalil

Abstract:

In structural vibration caused principally by an earthquake excitation, the most vibration’s attenuation system used recently is the semi active control with a Magneto Rheological Damper device. This control was a subject of many researches and works in the last years. The big challenges of searchers in this case is to propose an adequate controller with a robust algorithm of current or tension adjustment. In this present paper, a linear controller is proposed to control the MR damper using to reduce a vibrations of three story structure exposed to El Centro’s 1940 and Boumerdès 2003 earthquakes. In this example, the MR damper is installed in the first floor of the structure. The numerical simulations results of the proposed linear control with a feedback law based on clipped optimal algorithm showed the feasibility of the semi active control to protecting civil structures. The comparison of the controlled structure and uncontrolled structures responses illustrate clearly the performance and the effectiveness of the simple proposed approach.

Keywords: MR damper, seismic vibration, semi-active control

Procedia PDF Downloads 274

Authors: M. Riazat, H. Abdolvand, M. Baniassadi

Abstract:

In this present work, 3D reconstruction of cathode of SOFC is developed with various volume fractions and porosity. Three Phase Boundary (TPB) of construction of such derived micro structures is calculated. The neural network is used to optimize the porosity and volume fraction of each phase to reach a structure with maximum TPB.

Keywords: fuel cell, solid oxide, TPB, 3D reconstruction

Procedia PDF Downloads 312

Authors: Salah Al-Din I. Badran, Samad Ahmadi, Ismail Shahin

Abstract:

A blind source separation method is proposed; in this method, we use a non-uniform filter bank and a novel normalisation. This method provides a reduced computational complexity and increased convergence speed comparing to the full-band algorithm. Recently, adaptive sub-band scheme has been recommended to solve two problems: reduction of computational complexity and increase the convergence speed of the adaptive algorithm for correlated input signals. In this work, the reduction in computational complexity is achieved with the use of adaptive filters of orders less than the full-band adaptive filters, which operate at a sampling rate lower than the sampling rate of the input signal. The decomposed signals by analysis bank filter are less correlated in each subband than the input signal at full bandwidth, and can promote better rates of convergence.

Keywords: blind source separation, computational complexity, subband, convergence speed, mixture

Procedia PDF Downloads 570

Authors: Ali A. Ukasha

Abstract:

Data security needed in data transmission, storage, and communication to ensure the security. The single step parallel contour extraction (SSPCE) method is used to create the edge map as a key image from the different Gray level/Binary image. Performing the X-OR operation between the key image and each bit plane of the original image for image pixel values change purpose. The Arnold transform used to changes the locations of image pixels as image scrambling process. Experiments have demonstrated that proposed algorithm can fully encrypt 2D Gary level image and completely reconstructed without any distortion. Also shown that the analyzed algorithm have extremely large security against some attacks like salt & pepper and JPEG compression. Its proof that the Gray level image can be protected with a higher security level. The presented method has easy hardware implementation and suitable for multimedia protection in real time applications such as wireless networks and mobile phone services.

Keywords: SSPCE method, image compression, salt and peppers attacks, bitplanes decomposition, Arnold transform, lossless image encryption

Procedia PDF Downloads 483

Authors: Nima Farshidfar, Navid Daryasafar

Abstract:

In this paper, the seismic stability of reinforced soil slopes is studied using pseudo-dynamic analysis. Equilibrium equations that are applicable to the every kind of failure surface are written using Horizontal Slices Method. In written equations, the balance of the vertical and horizontal forces and moment equilibrium is fully satisfied. Failure surface is assumed to be log-spiral, and non-linear equilibrium equations obtained for the system are solved using Newton-Raphson Method. Earthquake effects are applied as horizontal and vertical pseudo-static coefficients to the problem. To solve this problem, a code was developed in MATLAB, and the critical failure surface is calculated using genetic algorithm. At the end, comparing the results obtained in this paper, effects of various parameters and the effect of using pseudo - dynamic analysis in seismic forces modeling is presented.

Keywords: soil slopes, pseudo-dynamic, genetic algorithm, optimization, limit equilibrium method, log-spiral failure surface

Procedia PDF Downloads 331

Authors: Gustavo Funes, Eduardo Peters, Jose Delpiano

Abstract:

It is widely known that photovoltaic technology has been massively distributed over the last decade despite its low-efficiency ratio. Dust deposition reduces this efficiency even more, lowering the energy production and module lifespan. In this work, we developed an artificial vision algorithm based on CIELAB color space to identify dust over panels in an autonomous way. We performed several experiments photographing three different types of panels, 30W, 340W and 410W. Those panels were soiled artificially with uniform and non-uniform distributed dust. The algorithm proposed uses statistical tools to provide a simulation with a 100% soiled panel and then performs a comparison to get the percentage of dirt in the experimental data set. The simulation uses a seed that is obtained by taking a dust sample from the maximum amount of dust from the dataset. The final result is the dirt percentage and the possible distribution of dust over the panel. Dust deposition is a key factor for plant owners to determine cleaning cycles or identify nonuniform depositions that could lead to module failure and hot spots.

Keywords: dust detection, photovoltaic, artificial vision, soiling

Procedia PDF Downloads 42

Authors: Feng Huang, Carlos G. Ochoa, Haitao Zhao

Abstract:

Poyang Lake, the largest freshwater lake in China, is located at the middle-lower reaches of the Yangtze River basin. It has great value in socioeconomic development and is internationally recognized as an important lacustrine and wetland ecosystem with abundant biodiversity. Impacted by ongoing climate change and anthropogenic activities, especially the regulation of the Three Gorges Reservoir since 2003, Poyang Lake has experienced significant water regime evolution, resulting in challenges for the management of water resources and the environment. Quantifying the contribution of hydrologic and topographic changes to water regime alteration is necessary for policymakers to design effective adaption strategies. Long term hydrologic data were collected and the back-propagation neural networks were constructed to simulate the lake water level. The impacts of hydrologic and topographic changes were differentiated through scenario analysis that considered pre-impact and post-impact hydrologic and topographic scenarios. The lake water regime was characterized by hydrologic indicators that describe monthly water level fluctuations, hydrologic features during flood and drought seasons, and frequency and rate of hydrologic variations. The results revealed different contributions of hydrologic and topographic changes to different features of the lake water regime.Noticeable changes were that the water level declined dramatically during the period of reservoir impoundment, and the drought was enhanced during the dry season. The hydrologic and topographic changes exerted a synergistic effect or antagonistic effect on different lake water regime features. The findings provide scientific reference for lacustrine and wetland ecological protection associated with water regime alterations.

Keywords: back-propagation neural network, scenario analysis, water regime, Poyang Lake

Procedia PDF Downloads 124

Authors: Medjadj Tarek, Ghribi Hayet

Abstract:

This study aims to develop a flood sensitivity assessment tool using machine learning (ML) techniques and geographic information system (GIS). The importance of this study is integrating the geographic information systems (GIS) and machine learning (ML) techniques for mapping flood risks, which help decision-makers to identify the most vulnerable areas and take the necessary precautions to face this type of natural disaster. To reach this goal, we will study the case of the city of M'sila, which is among the areas most vulnerable to floods. This study drew a map of flood-prone areas based on the methodology where we have made a comparison between 3 machine learning algorithms: the xGboost model, the Random Forest algorithm and the K Nearest Neighbour algorithm. Each of them gave an accuracy respectively of 97.92 - 95 - 93.75. In the process of mapping flood-prone areas, the first model was relied upon, which gave the greatest accuracy (xGboost).

Keywords: Geographic information systems (GIS), machine learning (ML), emergency mapping, flood disaster management

Procedia PDF Downloads 81

Authors: Ju-Hong Lee, Ding-Chen Chung

Abstract:

The paper deals with the optimal design of two-channel linear-phase (LP) quadrature mirror filter (QMF) banks using a metaheuristic based optimization technique. Based on the theory of two-channel QMF banks using two recursive digital all-pass filters (DAFs), the design problem is appropriately formulated to result in an objective function which is a weighted sum of the group delay error of the designed QMF bank and the magnitude response error of the designed low-pass analysis filter. Through a frequency sampling and a weighted least squares approach, the optimization problem of the objective function can be solved by utilizing a particle swarm optimization algorithm. The resulting two-channel QMF banks can possess approximately LP response without magnitude distortion. Simulation results are presented for illustration and comparison.

Keywords: quadrature mirror filter bank, digital all-pass filter, weighted least squares algorithm, particle swarm optimization

Procedia PDF Downloads 504

Authors: Miguel Contreras, David Long, Will Bachman

Abstract:

Background: Microscopy plays a central role in cell and developmental biology. In particular, fluorescence microscopy can be used to visualize specific cellular components and subsequently quantify their morphology through development of virtual-cell models for study of effects of mechanical forces on cells. However, there are challenges with these imaging experiments, which can make it difficult to quantify cell morphology: inconsistent results, time-consuming and potentially costly protocols, and limitation on number of labels due to spectral overlap. To address these challenges, the objective of this project is to develop an automatic computational machine learning pipeline to predict cellular components morphology for virtual-cell generation based on fluorescence cell membrane confocal z-stacks. Methods: Registered confocal z-stacks of nuclei and cell membrane of endothelial cells, consisting of 20 images each, were obtained from fluorescence confocal microscopy and normalized through software pipeline for each image to have a mean pixel intensity value of 0.5. An open source machine learning algorithm, originally developed to predict fluorescence labels on unlabeled transmitted light microscopy cell images, was trained using this set of normalized z-stacks on a single CPU machine. Through transfer learning, the algorithm used knowledge acquired from its previous training sessions to learn the new task. Once trained, the algorithm was used to predict morphology of nuclei using normalized cell membrane fluorescence images as input. Predictions were compared to the ground truth fluorescence nuclei images. Results: After one week of training, using one cell membrane z-stack (20 images) and corresponding nuclei label, results showed qualitatively good predictions on training set. The algorithm was able to accurately predict nuclei locations as well as shape when fed only fluorescence membrane images. Similar training sessions with improved membrane image quality, including clear lining and shape of the membrane, clearly showing the boundaries of each cell, proportionally improved nuclei predictions, reducing errors relative to ground truth. Discussion: These results show the potential of pre-trained machine learning algorithms to predict cell morphology using relatively small amounts of data and training time, eliminating the need of using multiple labels in immunofluorescence experiments. With further training, the algorithm is expected to predict different labels (e.g., focal-adhesion sites, cytoskeleton), which can be added to the automatic machine learning pipeline for direct input into Principal Component Analysis (PCA) for generation of virtual-cell mechanical models.

Keywords: cell morphology prediction, computational machine learning, fluorescence microscopy, virtual-cell models

Procedia PDF Downloads 194

Authors: Bry X., Trottier C., Mortier F., Cornu G., Verron T.

Abstract:

We address component-based regularization of a Multivariate Generalized Linear Model (MGLM). A set of random responses Y is assumed to depend, through a GLM, on a set X of explanatory variables, as well as on a set T of additional covariates. X is partitioned into R conceptually homogeneous blocks X1, ... , XR , viewed as explanatory themes. Variables in each Xr are assumed many and redundant. Thus, Generalised Linear Regression (GLR) demands regularization with respect to each Xr. By contrast, variables in T are assumed selected so as to demand no regularization. Regularization is performed searching each Xr for an appropriate number of orthogonal components that both contribute to model Y and capture relevant structural information in Xr. We propose a very general criterion to measure structural relevance (SR) of a component in a block, and show how to take SR into account within a Fisher-scoring-type algorithm in order to estimate the model. We show how to deal with mixed-type explanatory variables. The method, named THEME-SCGLR, is tested on simulated data.

Keywords: Component-Model, Fisher Scoring Algorithm, GLM, PLS Regression, SCGLR, SEER, THEME

Procedia PDF Downloads 388

Authors: K. R. Roopesh Bharatwaj, Avinash Maharana, Favour Tobi Aborisade, Roger Young

Abstract:

Presently, there are few cases of complete automation of drones and its allied intelligence capabilities. In essence, the potential of the drone has not yet been fully utilized. This paper presents feasible methods to build an intelligent drone with smart capabilities such as self-driving, and obstacle avoidance. It does this through advanced Reinforcement Learning Techniques and performs object detection using latest advanced algorithms, which are capable of processing light weight models with fast training in real time instances. For the scope of this paper, after researching on the various algorithms and comparing them, we finally implemented the Deep-Q-Networks (DQN) algorithm in the AirSim Simulator. In future works, we plan to implement further advanced self-driving and object detection algorithms, we also plan to implement voice-based speech recognition for the entire drone operation which would provide an option of speech communication between users (People) and the drone in the time of unavoidable circumstances. Thus, making drones an interactive intelligent Robotic Voice Enabled Service Assistant. This proposed drone has a wide scope of usability and is applicable in scenarios such as Disaster management, Air Transport of essentials, Agriculture, Manufacturing, Monitoring people movements in public area, and Defense. Also discussed, is the entire drone communication based on the satellite broadband Internet technology for faster computation and seamless communication service for uninterrupted network during disasters and remote location operations. This paper will explain the feasible algorithms required to go about achieving this goal and is more of a reference paper for future researchers going down this path.

Keywords: convolution neural network, natural language processing, obstacle avoidance, satellite broadband technology, self-driving

Procedia PDF Downloads 235

Authors: Kelemewerk Destalem, Joongjae Cho, Jaeseong Lee, Ju H. Park, Joonhyuk Yoo

Abstract:

Background subtraction and temporal difference are often used for moving object detection in video. Both approaches are computationally simple and easy to be deployed in real-time image processing. However, while the background subtraction is highly sensitive to dynamic background and illumination changes, the temporal difference approach is poor at extracting relevant pixels of the moving object and at detecting the stopped or slowly moving objects in the scene. In this paper, we propose a moving object detection scheme based on adaptive background subtraction and temporal difference exploiting dynamic background updates. The proposed technique consists of a histogram equalization, a linear combination of background and temporal difference, followed by the novel frame-based and pixel-based background updating techniques. Finally, morphological operations are applied to the output images. Experimental results show that the proposed algorithm can solve the drawbacks of both background subtraction and temporal difference methods and can provide better performance than that of each method.

Keywords: background subtraction, background updating, real time, light weight algorithm, temporal difference

Procedia PDF Downloads 327

Authors: A. Hassoune, M. Khafallah, A. Mesbahi, T. Bouragba

Abstract:

This work presents an improved strategy of control for charging a lithium-ion battery in an electric vehicle charging station using two charger topologies i.e. single ended primary inductor converter (SEPIC) and forward converter. In terms of rapidity and accuracy, the power system consists of a topology/control diagram that would overcome the performance constraints, for instance the power instability, the battery overloading and how the energy conversion blocks would react efficiently to any kind of perturbations. Simulation results show the effectiveness of the proposed topologies operated with a power management algorithm based on voltage/peak current mode controls. In order to provide credible findings, a low power prototype is developed to test the control strategy via experimental evaluations of the converter topology and its controls.

Keywords: battery storage buffer, charging station, electric vehicle, experimental analysis, management algorithm, switches control

Procedia PDF Downloads 157

Authors: Khaled M. EL-Naggar

Abstract:

Synchronizing and damping torque coefficients of a synchronous machine can give a quite clear picture for machine behavior during transients. These coefficients are used as a power system transient stability measurement. In this paper, a crow search optimization algorithm is presented and implemented to study the power system stability during transients. The algorithm makes use of the machine responses to perform the stability study in time domain. The problem is formulated as a dynamic estimation problem. An objective function that minimizes the error square in the estimated coefficients is designed. The method is tested using practical system with different study cases. Results are reported and a thorough discussion is presented. The study illustrates that the proposed method can estimate the stability coefficients for the critical stable cases where other methods may fail. The tests proved that the proposed tool is an accurate and reliable tool for estimating the machine coefficients for assessment of power system stability.

Keywords: optimization, estimation, synchronous, machine, crow search

Procedia PDF Downloads 125

Authors: Nadeem Akhtar, Hira Javed

Abstract:

An important aspect of research is literature survey. Availability of a large amount of literature across different domains triggers the need for optimized systems which provide relevant literature to researchers. We propose a search system based on keywords for text documents. This experimental approach provides a hierarchical structure to the document corpus. The documents are labelled with keywords using KEA (Keyword Extraction Algorithm) and are automatically organized in a lattice structure using Formal Concept Analysis (FCA). This groups the semantically related documents together. The hierarchical structure, based on keywords gives out only those documents which precisely contain them. This approach open doors for multi-domain research. The documents across multiple domains which are indexed by similar keywords are grouped together. A hierarchical relationship between keywords is obtained. To signify the effectiveness of the approach, we have carried out the experiment and evaluation on Semeval-2010 Dataset. Results depict that the presented method is considerably successful in indexing of scientific papers.

Keywords: formal concept analysis, keyword extraction algorithm, scientific documents, lattice

Procedia PDF Downloads 320

Authors: Xiaolei Hu, Enrico Ferrera, Riccardo Tomasi, Claudio Pastrone

Abstract:

Load Forecasting plays a key role in making today's and future's Smart Energy Grids sustainable and reliable. Accurate power consumption prediction allows utilities to organize in advance their resources or to execute Demand Response strategies more effectively, which enables several features such as higher sustainability, better quality of service, and affordable electricity tariffs. It is easy yet effective to apply Load Forecasting at larger geographic scale, i.e. Smart Micro Grids, wherein the lower available grid flexibility makes accurate prediction more critical in Demand Response applications. This paper analyses the application of short-term load forecasting in a concrete scenario, proposed within the EU-funded GreenCom project, which collect load data from single loads and households belonging to a Smart Micro Grid. Three short-term load forecasting techniques, i.e. linear regression, artificial neural networks, and radial basis function network, are considered, compared, and evaluated through absolute forecast errors and training time. The influence of weather conditions in Load Forecasting is also evaluated. A new definition of Gain is introduced in this paper, which innovatively serves as an indicator of short-term prediction capabilities of time spam consistency. Two models, 24- and 1-hour-ahead forecasting, are built to comprehensively compare these three techniques.

Keywords: short-term load forecasting, smart micro grid, linear regression, artificial neural networks, radial basis function network, gain

Procedia PDF Downloads 455

Authors: Zulkifli Zainal Abidin, Ahmad Shahril Mohd Ghani

Abstract:

Agent-based systems technology has been addressed as a new paradigm for conceptualizing, designing, and implementing software systems. Agents are sophisticated systems that act autonomously across open and distributed environments in solving problems. Nevertheless, it is impractical to rely on a single agent to do all computing processes in solving complex problems. An increasing number of applications lately require multiple agents to work together. A multi-agent system (MAS) is a loosely coupled network of agents that interact to solve problems that are beyond the individual capacities or knowledge of each problem solver. However, the network of MAS still requires a main system to govern or oversees the operation of the agents in order to achieve a unified goal. We had developed a fleet management system (FMS) in order to manage the fleet of agents, plan route for the agents, perform real-time data processing and analysis, and issue sets of general and specific instructions to the agents. This FMS should be able to perform real-time data processing, communicate with the autonomous surface vehicle (ASV) agents and generate bathymetric map according to the data received from each ASV unit. The first algorithm is developed to communicate with the ASV via radio communication using standard National Marine Electronics Association (NMEA) protocol sentences. Next, the second algorithm will take care of the path planning, formation and pattern generation is tested using various sample data. Lastly, the bathymetry map generation algorithm will make use of data collected by the agents to create bathymetry map in real-time. The outcome of this research is expected can be applied on various other multi-agent systems.

Keywords: autonomous surface vehicle, fleet management system, multi agent system, bathymetry

Procedia PDF Downloads 262

Authors: Yi Huang, Clemens Guehmann

Abstract:

In order to monitor the thermal behavior of an asynchronous machine with squirrel cage rotor, a 9th-order extended Kalman filter (EKF) algorithm is implemented to estimate the temperatures of the stator windings, the rotor cage and the stator core. The state-space equations of EKF are established based on the electrical, mechanical and the simplified thermal models of an asynchronous machine. The asynchronous machine with simplified thermal model in Dymola is compiled as DymolaBlock, a physical model in MATLAB/Simulink. The coolant air temperature, three-phase voltages and currents are exported from the physical model and are processed by EKF estimator as inputs. Compared to the temperatures exported from the physical model of the machine, three parts of temperatures can be estimated quite accurately by the EKF estimator. The online EKF estimator is independent from the machine control algorithm and can work under any speed and load condition if the stator current is nonzero current system.

Keywords: asynchronous machine, extended Kalman filter, resistance, simulation, temperature estimation, thermal model

Procedia PDF Downloads 276

Authors: Yu-Jia Jian, Emily Chia-Yu Su, Hui-Ling Hsu, Jian-Jhih Chen

Abstract:

Coronary artery is the major supplier of myocardial blood flow. When fat and cholesterol are deposit in the coronary arterial wall, narrowing and stenosis of the artery occurs, which may lead to myocardial ischemia and eventually infarction. According to the World Health Organization (WHO), estimated 740 million people have died of coronary heart disease in 2015. According to Statistics from Ministry of Health and Welfare in Taiwan, heart disease (except for hypertensive diseases) ranked the second among the top 10 causes of death from 2013 to 2016, and it still shows a growing trend. According to American Heart Association (AHA), the risk factors for coronary heart disease including: age (> 65 years), sex (men to women with 2:1 ratio), obesity, diabetes, hypertension, hyperlipidemia, smoking, family history, lack of exercise and more. We have collected a dataset of 421 patients from a hospital located in northern Taiwan who received coronary computed tomography (CT) angiography. There were 300 males (71.26%) and 121 females (28.74%), with age ranging from 24 to 92 years, and a mean age of 56.3 years. Prior to coronary CT angiography, basic data of the patients, including age, gender, obesity index (BMI), diastolic blood pressure, systolic blood pressure, diabetes, hypertension, hyperlipidemia, smoking, family history of coronary heart disease and exercise habits, were collected and used as input variables. The output variable of the prediction module is the degree of coronary artery stenosis. The output variable of the prediction module is the narrow constriction of the coronary artery. In this study, the dataset was randomly divided into 80% as training set and 20% as test set. Four machine learning algorithms, including logistic regression, stepwise regression, neural network and decision tree, were incorporated to generate prediction results. We used area under curve (AUC) / accuracy (Acc.) to compare the four models, the best model is neural network, followed by stepwise logistic regression, decision tree, and logistic regression, with 0.68 / 79 %, 0.68 / 74%, 0.65 / 78%, and 0.65 / 74%, respectively. Sensitivity of neural network was 27.3%, specificity was 90.8%, stepwise Logistic regression sensitivity was 18.2%, specificity was 92.3%, decision tree sensitivity was 13.6%, specificity was 100%, logistic regression sensitivity was 27.3%, specificity 89.2%. From the result of this study, we hope to improve the accuracy by improving the module parameters or other methods in the future and we hope to solve the problem of low sensitivity by adjusting the imbalanced proportion of positive and negative data.

Keywords: decision support, computed tomography, coronary artery, machine learning

Procedia PDF Downloads 220

Authors: Ananya Ananya, Karthik Rao

Abstract:

Accurate segmentation of knee cartilage in 3-D magnetic resonance (MR) images for quantitative assessment of volume is crucial for studying and diagnosing osteoarthritis (OA) of the knee, one of the major causes of disability in elderly people. Radiologists generally perform this task in slice-by-slice manner taking 15-20 minutes per 3D image, and lead to high inter and intra observer variability. Hence automatic methods for knee cartilage segmentation are desirable and are an active field of research. This paper presents design and experimental evaluation of 2D convolutional neural networks based fully automated methods for knee cartilage segmentation in 3D MRI. The architectures are validated based on 40 test images and 60 training images from SKI10 dataset. The proposed methods segment 2D slices one by one, which are then combined to give segmentation for whole 3D images. Proposed methods are modified versions of U-net and dilated convolutions, consisting of a single step that segments the given image to 5 labels: background, femoral cartilage, tibia cartilage, femoral bone and tibia bone; cartilages being the primary components of interest. U-net consists of a contracting path and an expanding path, to capture context and localization respectively. Dilated convolutions lead to an exponential expansion of receptive field with only a linear increase in a number of parameters. A combination of modified U-net and dilated convolutions has also been explored. These architectures segment one 3D image in 8 – 10 seconds giving average volumetric Dice Score Coefficients (DSC) of 0.950 - 0.962 for femoral cartilage and 0.951 - 0.966 for tibia cartilage, reference being the manual segmentation.

Keywords: convolutional neural networks, dilated convolutions, 3 dimensional, fully automated, knee cartilage, MRI, segmentation, U-net

Procedia PDF Downloads 248

Authors: Fatemeh Sadeghi, Peyman Sheikhzadeh

Abstract:

Introduction: Block Sequential Regularized Expectation Maximization (BSREM) reconstruction algorithm was recently developed to suppress excessive noise by applying a relative difference penalty. The aim of this study was to investigate the effect of various strengths of noise penalization factor in the BSREM algorithm under different acquisition duration and lesion sizes in order to determine an optimum penalty factor by considering both quantitative and qualitative image evaluation parameters in clinical uses. Materials and Methods: The NEMA IQ phantom and 15 clinical whole-body patients with prostate cancer were evaluated. Phantom and patients were injected withGallium-68 Prostate-Specific Membrane Antigen(68 Ga-PSMA)and scanned on a non-time-of-flight Discovery IQ Positron Emission Tomography/Computed Tomography(PET/CT) scanner with BGO crystals. The data were reconstructed using BSREM with a β-value of 100-500 at an interval of 100. These reconstructions were compared to OSEM as a widely used reconstruction algorithm. Following the standard NEMA measurement procedure, background variability (BV), recovery coefficient (RC), contrast recovery (CR) and residual lung error (LE) from phantom data and signal-to-noise ratio (SNR), signal-to-background ratio (SBR) and tumor SUV from clinical data were measured. Qualitative features of clinical images visually were ranked by one nuclear medicine expert. Results: The β-value acts as a noise suppression factor, so BSREM showed a decreasing image noise with an increasing β-value. BSREM, with a β-value of 400 at a decreased acquisition duration (2 min/ bp), made an approximately equal noise level with OSEM at an increased acquisition duration (5 min/ bp). For the β-value of 400 at 2 min/bp duration, SNR increased by 43.7%, and LE decreased by 62%, compared with OSEM at a 5 min/bp duration. In both phantom and clinical data, an increase in the β-value is translated into a decrease in SUV. The lowest level of SUV and noise were reached with the highest β-value (β=500), resulting in the highest SNR and lowest SBR due to the greater noise reduction than SUV reduction at the highest β-value. In compression of BSREM with different β-values, the relative difference in the quantitative parameters was generally larger for smaller lesions. As the β-value decreased from 500 to 100, the increase in CR was 160.2% for the smallest sphere (10mm) and 12.6% for the largest sphere (37mm), and the trend was similar for SNR (-58.4% and -20.5%, respectively). BSREM visually was ranked more than OSEM in all Qualitative features. Conclusions: The BSREM algorithm using more iteration numbers leads to more quantitative accuracy without excessive noise, which translates into higher overall image quality and lesion detectability. This improvement can be used to shorter acquisition time.

Keywords: BSREM reconstruction, PET/CT imaging, noise penalization, quantification accuracy

Procedia PDF Downloads 85

Authors: Hina Iqbal, Mustansar Ali Ghazanfar, Sandor Szedmak

Abstract:

Recommender systems uses artificial intelligence practices for filtering obscure information and can predict if a user likes a specified item. Kernel mapping Recommender systems have been proposed which are accurate and state-of-the-art algorithms and resolve recommender system’s design objectives such as; long tail, cold-start, and sparsity. The aim of research is to propose hybrid framework that can efficiently integrate different versions— namely item-based and user-based KMR— of KMR algorithm. We have proposed various heuristic algorithms that integrate different versions of KMR (into a unified framework) resulting in improved accuracy and elimination of problems associated with conventional recommender system. We have tested our system on publically available movies dataset and benchmark with KMR. The results (in terms of accuracy, precision, recall, F1 measure and ROC metrics) reveal that the proposed algorithm is quite accurate especially under cold-start and sparse scenarios.

Keywords: Kernel Mapping Recommender Systems, hybrid recommender systems, cold start, sparsity, long tail

Procedia PDF Downloads 326

Authors: Ksenia Volkova, Artur Petrosyan, Ignatii Dubyshkin, Alexei Ossadtchi

Abstract:

Brain-computer interfaces are a growing research field producing many implementations that find use in different fields and are used for research and practical purposes. Despite the popularity of the implementations using non-invasive neuroimaging methods, radical improvement of the state channel bandwidth and, thus, decoding accuracy is only possible by using invasive techniques. Electrocorticography (ECoG) is a minimally invasive neuroimaging method that provides highly informative brain activity signals, effective analysis of which requires the use of machine learning methods that are able to learn representations of complex patterns. Deep learning is a family of machine learning algorithms that allow learning representations of data with multiple levels of abstraction. This study explores the potential of deep learning approaches for ECoG processing, decoding movement intentions and the perception of proprioceptive information. To obtain synchronous recording of kinematic movement characteristics and corresponding electrical brain activity, a series of experiments were carried out, during which subjects performed finger movements at their own pace. Finger movements were recorded with a three-axis accelerometer, while ECoG was synchronously registered from the electrode strips that were implanted over the contralateral sensorimotor cortex. Then, multichannel ECoG signals were used to track finger movement trajectory characterized by accelerometer signal. This process was carried out both causally and non-causally, using different position of the ECoG data segment with respect to the accelerometer data stream. The recorded data was split into training and testing sets, containing continuous non-overlapping fragments of the multichannel ECoG. A deep convolutional neural network was implemented and trained, using 1-second segments of ECoG data from the training dataset as input. To assess the decoding accuracy, correlation coefficient r between the output of the model and the accelerometer readings was computed. After optimization of hyperparameters and training, the deep learning model allowed reasonably accurate causal decoding of finger movement with correlation coefficient r = 0.8. In contrast, the classical Wiener-filter like approach was able to achieve only 0.56 in the causal decoding mode. In the noncausal case, the traditional approach reached the accuracy of r = 0.69, which may be due to the presence of additional proprioceptive information. This result demonstrates that the deep neural network was able to effectively find a representation of the complex top-down information related to the actual movement rather than proprioception. The sensitivity analysis shows physiologically plausible pictures of the extent to which individual features (channel, wavelet subband) are utilized during the decoding procedure. In conclusion, the results of this study have demonstrated that a combination of a minimally invasive neuroimaging technique such as ECoG and advanced machine learning approaches allows decoding motion with high accuracy. Such setup provides means for control of devices with a large number of degrees of freedom as well as exploratory studies of the complex neural processes underlying movement execution.

Keywords: brain-computer interface, deep learning, ECoG, movement decoding, sensorimotor cortex

Procedia PDF Downloads 162

Authors: Wei-Chia Huang, Jane Wang

Abstract:

Damage to nerves is considered one of the most irreversible injuries. The regeneration of nerves has always been an important topic in regenerative medicine. In general, damage to human tissue will naturally repair overtime. However, when the nerves are damaged, healed flesh wound cannot guarantee full restoration to its original function, as truncated nerves are often irreversible. Therefore, the development of treatment methods to successfully guide and accelerate the regeneration of nerves has been highly sought after. In order to induce nerve tissue growth, nerve conduits are commonly used to help reconnect broken nerve bundles to provide protection to the location of the fracture while guiding the growth of the nerve bundles. To prevent the protected tissue from becoming necrotic and to ensure the growth rate, the conduits used are often modified with microstructures or blended with neuron growth factors that may facilitate nerve regeneration. Electrical stimulation is another attempted treatment for medical rehabilitation. With appropriate range of voltages and stimulation frequencies, it has been demonstrated to promote cell proliferation and migration. Biodegradability are critical for medical devices like nerve conduits, while conductive polymers pose great potential toward the differentiation and growth of nerve cells. In this work, biodegradability and conductivity were combined into a novel biodegradable, photocurable, conductive polymer composite materials by embedding conductive nanoparticles in poly(glycerol sebacate) acrylate (PGSA) and 3D-printed into nerve conduits. Rat pheochromocytoma cells and rat neuronal Schwann cells were chosen for the in vitro tests of the conduits and had demonstrate selective growth upon culture in the conductive conduits with built-in microchannels and electrical stimulation.

Keywords: biodegradable polymer, 3d printing, neural regeneration, electrical stimulation

Procedia PDF Downloads 95

Authors: Shraddha Jain Sharma, Ratnalata Gupta

Abstract:

Predicting stock movements in the financial market is both difficult and rewarding. Analysts and academics are increasingly using advanced approaches such as machine learning techniques to anticipate stock price patterns, thanks to the expanding capacity of computing and the recent advent of graphics processing units and tensor processing units. Stock market prediction is a type of time series prediction that is incredibly difficult to do since stock prices are influenced by a variety of financial, socioeconomic, and political factors. Furthermore, even minor mistakes in stock market price forecasts can result in significant losses for companies that employ the findings of stock market price prediction for financial analysis and investment. Soft computing techniques are increasingly being employed for stock market prediction due to their better accuracy than traditional statistical methodologies. The proposed research looks at the need for soft computing techniques in stock market prediction, the numerous soft computing approaches that are important to the field, past work in the area with their prominent features, and the significant problems or issue domain that the area involves. For constructing a predictive model, the major focus is on neural networks and fuzzy logic. The stock market is extremely unpredictable, and it is unquestionably tough to correctly predict based on certain characteristics. This study provides a complete overview of the numerous strategies investigated for high accuracy prediction, with a focus on the most important characteristics.

Keywords: stock market prediction, artificial intelligence, artificial neural networks, fuzzy logic, accuracy, deep learning, machine learning, stock price, trading volume

Procedia PDF Downloads 79