Ograniczanie wyników
Czasopisma help
Autorzy help
Lata help
Preferencje help
Widoczny [Schowaj] Abstrakt
Liczba wyników

Znaleziono wyników: 536

Liczba wyników na stronie
first rewind previous Strona / 27 next fast forward last
Wyniki wyszukiwania
Wyszukiwano:
w słowach kluczowych:  sieć neuronowa
help Sortuj według:

help Ogranicz wyniki do:
first rewind previous Strona / 27 next fast forward last
EN
The process of garment production has always been a black box. The production time of different clothing is different and has great changes, thus managers cannot make a production plan accurately. With the world entering the era of industry 4.0 and the accumulation of big data, machine learning can provide services for the garment manufacturing industry. The production cycle time is the key to control the production process. In order to predict the production cycle time more accurately and master the production process in the garment manufacturing process, a neural network model of production cycle time prediction is established in this paper. Using a trained neural network to predict the production cycle time, the overall error of 6 groups is within 5%, and that of 3 groups is between 5% and 10%. Therefore, this neural network can be used to predict the future production cycle time and predict the overall production time of clothing.
PL
Czas produkcji różnych ubrań jest inny i podlega dużym zmianom, dlatego menedżerowie nie mogą dokładnie zaplanować produkcji. Wraz z wkroczeniem świata w erę przemysłu 4.0 i gromadzeniem dużych zbiorów danych dobrym rozwiązaniem dla przemysłu odzieżowego jest zastosowanie maszyn uczących się. Czas cyklu produkcyjnego jest kluczem do kontroli procesu produkcyjnego. W celu dokładniejszego przewidywania czasu cyklu produkcyjnego i opanowania procesu produkcyjnego w procesie produkcji odzieży, w artykule opracowano model sieci neuronowej do przewidywania czasu cyklu produkcyjnego. Do przewidywania czasu cyklu produkcyjnego użyto sieci neuronowej, ogólny błąd 6 grup mieścił się w granicach 5%, a 3 grup – między 5% a 10%. W związku z tym zaprezentowana sieć neuronowa może znaleźć zastosowanie w przewidywaniu czasu cyklu produkcyjnego i całkowitego czasu produkcji odzieży.
EN
This work presents a new epileptic seizures epoch classification scheme. Variational mode decomposition (VMD), has been explored for non-recursively decomposing the electroencephalogram (EEG) signals into fourteen band limited intrinsic mode functions (IMFs). Data augmentation (DA), has been used for handling unbalanced classification problem. Normalized energy, fractal dimension, number of peaks, and prominence parameters were computed from the band-limited IMFs for the discrimination of seizure and non-seizure epochs. Bayesian regularized shallow neural network (BR-SNNs) and six other well-known classifiers were tested. Sensitivity, specificity, and accuracy have been used as performance metrics. This study includes two different epoch lengths of 1-second and 2-seconds. A total of 32 test cases for both, class balanced and unbalanced classification problems have been taken for the performance evaluation. The best performance obtained is 100% for all the three metrics from the test cases of database-2 and 3. For database-1, average sensitivity, specificity, and accuracy of 99.71, 99.75, and 99.73% have been achieved, respectively for the 1-second epoch. The presented work shows better performance results compared to many previously reported works.
4
Content available remote AI-based method of vortex core tracking as an alternative for Lambda2
EN
The paper presents a new method of vortex core detection developed for use in CFD simulation result analysis. Apart from the conventional approach involving vector algebra, mainly the Lambda2 method, it focuses on the identification of certain features in a graphic representation of the velocity field. It is done by generating a series of slices of the said field in the postprocessing software and training a Convolutional Neural Network (AI) to recognize vortex cores. The neural network can be integrated into a simple python program and used to quickly identify vortex cores on a large number of images and translate their locations to coordinates of a CFD model for visualisation.
EN
In this paper, we compare the following machine learning methods as classifiers for sentiment analysis: k – nearest neighbours (kNN), artificial neural network (ANN), support vector machine (SVM), random forest. We used a dataset containing 5,000 movie reviews in which 2,500 were marked as positive and 2,500 as negative. We chose 5,189 words which have an influence on sentence sentiment. The dataset was prepared using a term document matrix (TDM) and classical multidimensional scaling (MDS). This is the first time that TDM and MDS have been used to choose the characteristics of text in sentiment analysis. In this case, we decided to examine different indicators of the specific classifier, such as kernel type for SVM and neighbour count in kNN. All calculations were performed in the R language, in the program R Studio v 3.5.2. Our work can be reproduced because all of our data sets and source code are public.
6
Content available General concept of the EMG controlled bionic hand
EN
The article presents a general concept of a bionic hand control system using multichannel EMG signal, being under development at present. The method of acquisition and processing of multi-channel EMG signal and feature extraction for machine learning were described. Moreover, the design of the control system implementation in the real-time embedded system was discussed.
PL
Na podstawie badań opisanych w artykule opracowano zespół wielowarstwowych jednokierunkowych sztucznych sieci neuronowych, odwzorowujących zależności zachodzące między parametrami wejściowymi procesu azotowania niskociśnieniowego a własnościami końcowymi obrabianej stali, ze szczególnym uwzględnieniem twardości i charakterystyki wytworzonej warstwy. Na tej bazie zbudowano model fizyczny wielosegmentowego azotowania niskociśnieniowego dla stali narzędziowych. Końcowym efektem prac jest aplikacja wspierająca projektowanie, symulację oraz optymalizację tych procesów w rzeczywistości przemysłowej.
EN
Based on the research presented in the paper, a set of multilayer unidirectional artificial neural networks was developed, mapping the relationships between the input parameters of the low-pressure nitriding process and the final properties of the processed steel, with particular emphasis on the hardness and characteristics of the obtained layer. On this basis, a physical model of multi-segment low-pressure nitriding for tool steels was developed. The final effect of the presented work is an application supporting the design, simulation and optimization of these processes in industrial reality.
EN
The paper deals with development neural network controller to ensure safe and reliable operation of damaged induction motor. It was chosen field-oriented control algorithm as a basis for neural controller aimed to provide reliable control signals both for healthy and damaged motor. It was investigated different compositions of neural network model. It was received simulation results for modified field-oriented control algorithm with neural network regulator, which showed similar results comparing to original model for healthy motor, which confirms correctness of developed model.
PL
Opisano zastosowanie sterownika wykorzystującego sieci neuronowe do stabilnej i niezawodnej pracy zdrowego lub uszkodzonego silnika indukcyjnego.
EN
The work presented in this paper is a contribution in the theme of monitoring and diagnosing of faults in the three-phase squirrel cage induction machine. The proposed approach is based on the pattern recognition methods and the artificial intelligence techniques. For so doing, measurements of the stator currents are carried out on a machine subject to various faults such as: short-circuit in the stator windings, bar breakage, bearing failure and eccentricity fault. These acquisitions are classified in databases in order to process them and calculate their Power Spectral Density (PSD). Then, another database is formed of the digital data of the PSD images of the currents associated with the type of fault. After that, a process of learning and classification by artificial neural networks was developed. The test results show the efficiency, robustness and correctness of the proposed approach for the discrimination of faults of electrical or mechanical origin affecting the machine.
PL
Praca przedstawiona w tym artykule stanowi wkład w temat monitorowania i diagnozowania uszkodzeń w trójfazowej maszynie indukcyjnej klatkowej. Proponowane podejście opiera się na metodach rozpoznawania wzorców i technikach sztucznej inteligencji. W tym celu pomiary prądów stojana są przeprowadzane na maszynie podlegającej różnym usterkom, takim jak: zwarcie w uzwojeniach stojana, pęknięcie pręta, uszkodzenie łożyska i błąd mimośrodowości. Przejęcia te są klasyfikowane w bazach danych w celu ich przetworzenia i obliczenia ich gęstości widmowej mocy (PSD). Następnie tworzona jest kolejna baza danych z cyfrowymi danymi obrazów PSD prądów związanych z rodzajem uszkodzenia. Następnie opracowano proces uczenia się i klasyfikacji przez sztuczne sieci neuronowe. Wyniki testu pokazują skuteczność, niezawodność i poprawność proponowanego podejścia do rozróżnienia wad pochodzenia elektrycznego lub mechanicznego mających wpływ na maszynę.
EN
In order to solve the problem for temperature electrical resistance furnace. Characterized by their large inertia, nonlinear, long time delay and time-varying property it is rather difficult to obtain satisfactory control results with Performances of conventional PI control cannot achieve good control effect. In this paper a neural network-based adaptive control approach (ACNN) for electrical furnace is developed .using RBF NN to estimate the unknown functions by neural networks and from good choice of the law of adaptation. Based on the resolution of the lyapunov equation. Taking account of all possible parameter variations the adaptive control is designed so that it has the ability to improve the performance of the closed loop system, producing the control signal by using the information from the system. In this case we use a coping mechanism that observes the signal to control and adjust the synaptic weights of neural networks when system parameters change over time. Result shows that the proposed algorithm (ACNN) performs very well when furnace parameter varies the latter allow the neural model to be identified online and, if necessary its parameters to be stabilized and it is very easy to program it online.
PL
Piec elektryczny charakteryzuje się nieliniowością, dużym czasem opóźnienia co utrudnia sterowanie nime. W pracy zaproponowano system sterowania piecem z wykorzystaniem sieci neuronowej System jest zaprojektowany tak, że uwzględnia zmiany parametrów.
EN
The article presents a method enabling estimation of the selected power quality indicators at a given point of a power network, on the basis of the power quality indicators recorded at the nearest vicinity points. For needs of the estimations, artificial neural network algorithms were applied. The result is a neural model that defines the relationship between the power quality indicators of the same type, at adjacent points. The paper presents results of analyses and tests under real operating conditions of the distribution system.
PL
W artykule przedstawiono metodę umożliwiającą estymację wybranych wskaźników jakości energii elektrycznej w zadanym punkcie sieci elektroenergetycznej na podstawie wskaźników jakości energii elektrycznej zarejestrowanych w punktach leżących w najbliższym otoczeniu. Do estymacji wykorzystano algorytmy sztucznych sieci neuronowych. W rezultacie uzyskano neuronowy model określający relację pomiędzy wskaźnikami jakości energii elektrycznej tego samego typu w sąsiadujących ze sobą punktach. W artkule przedstawiono wyniki analiz i testów dla rzeczywistych warunków pracy sieci dystrybucyjnej.
12
Content available remote A modified neural network for antennas optimization
EN
In mobile communications, devices are generally more compact, nevertheless allow data traffic at high speeds. To meet such demand, embedded hardware must present limited dimensions and at the same time be robust enough to ensure high communication speeds. In this work, a modified Hopfield neural network was applied in the optimization of planar antennas. The role of the algorithm presented here, is to find the ideal antenna dimensions to meet the future 5G mobile technology. With this, a significant improvement in resonance, gain and directivity was expected, which are some of the important parameters in antenna analysis. In the literature, no reference was found based on the modified Hopfield neural network applied to the optimization of planar antennas, which further enhances this research, providing an important and unprecedented contribution. The analysis of the results shows the efficiency, robustness, precision and reliability of this approach, encouraging further research in this area.
PL
W pracy przedstawiono zmodyfikowaną sieć neuronową Hopfielda wykorzystaną do optymalizacji planarnej anteny. Rolą algorytmu jest znalezienie wymiarów anteny tak aby można było projektować anteny 5G. Doatkowo można antenę optymalizować pod kątem wzmocniania w rezonansie i kierunkowości.
EN
The work presents the structure and principle of operation of the artificial neuron network constructed for identification of a polymer on the basis of its flammability. The characteristic properties of burning of a polymer are saved in a special form in a database. The network creates a binary standard for each polymer from the database, coding data by means of the signals of the values 1, 0, -1. The network memorizes data related to each polymer detecting the similarities and differences between them and determines the weights which reflect the importance of particular features of its burning process.
PL
Przedstawiono strukturę i zasadę działania sztucznej sieci neuronowej skonstruowanej do identyfikacji polimeru na podstawie jego palności. Charakterystyczne cechy palności każdego polimeru zostały zapisane we wzorcowej bazie danych. Dla każdego polimeru z tej bazy sieć tworzy wzorzec binarny, w którym charakterystyczne cechy palności są kodowane jako sygnały o wartościach -1, 0 i 1. Sieć zapamiętuje palność każdego polimeru, wykrywając podobieństwa i różnice między nimi oraz określając wagi odzwierciedlające znaczenie poszczególnych cech palności.
EN
In this paper, the problem of feedback stabilization for a class of impulsive state-dependent neural networks (ISDNNs) with nonlinear disturbance inputs via quantized input signals is discussed. By constructing quasi-invariant sets and attracting sets for ISDNNs, we design a quantized controller with adjustable parameters. In combination with a suitable ISS-Lyapunov functional and a hybrid quantized control strategy, we propose novel criteria on input-to-state stability and global asymptotical stability for ISDNNs. Our results complement the existing ones. Numerical simulations are reported to substantiate the theoretical results and effectiveness of the proposed strategy.
EN
The linear programming (LP) approach to solve the Bellman equation in dynamic programming is a well-known option for finite state and input spaces to obtain an exact solution. However, with function approximation or continuous state spaces, refinements are necessary. This paper presents a methodology to make approximate dynamic programming via LP work in practical control applications with continuous state and input spaces. There are some guidelines on data and regressor choices needed to obtain meaningful and well-conditioned value function estimates. The work discusses the introduction of terminal ingredients and computation of lower and upper bounds of the value function. An experimental inverted-pendulum application will be used to illustrate the proposal and carry out a suitable comparative analysis with alternative options in the literature.
EN
In this paper, feature weighting is used to develop an effective computer-aided diagnosis system for breast cancer. Feature weighting is employed because it boosts the classification performance more as compared to feature subset selection. Specifically, a wrapper method utilizing the Ant Lion Optimization algorithm is presented that searches for best feature weights and parametric values of Multilayer Neural Network simultaneously. The selection of hidden neurons and backpropagation training algorithms are used as parameters of neural networks. The performance of the proposed approach is evaluated on three breast cancer datasets. The data is initially normalized using tanh method to remove the effects of dominant features and outliers. The results show that the proposed wrapper method has a better ability to attain higher accuracy as compared to the existing techniques. The obtained high classification performance validates the work which has the potential for becoming an alternative to the other well-known techniques.
EN
Autonomous rehabilitation training for assisted patients with injured upper-limbs promotes the regenerative communication between muscle signals and brain consciousness. Surface electromyographic (sEMG) is a type of electrical signals of neuromuscular activity recorded by electrodes on the surface of the human body, which is widely applied for detecting gestures and stimuli reactions. Experimental results proved the importance of the sEMG signals for extracting such reactions, in which, the segmentation and classification of the sEMG are vital tasks. The objective of the present work is to segment and classify the sEMG signals of patients to assist the design of clinical rehabilitation devices based on the classification of sEMG signals. In the pre-processing stage, a dual-tone multi-frequency signaling is designed for signal coding; subsequently, the pre-processed sEMG signal is transformed by the Fast Fourier Transfer. Afterward, a time-series frequency analysisis performed by applyingHiddenMarkov Models.A basic traditional longshort- term memory (LSTM) model is addressed for waveform-based classification to be compared to the proposed improved deep BP (back-propagation)–LSTM for sEMG signal classification. Seventeen performance features are selected for evaluating the proposed multi-classification, deep learning model for classifying six actions, namely moving gesture of grip, slowly moving, flexor, straight lift, stretch, and up-high lift; which were proposed by rehabilitation physician. The experiment results indicated the superiority of the proposed method compared to other well-known classifiers, such as the neural network, support vector machine, decision trees, Bayes inference, and recurrent neural network. The proposed deep BP–LSTM network achieved 92% accuracy, 89% specificity, 91% precision, and 96% F1-score, in the multi-classification of the sEMG signals.
EN
Load distribution analysis on foot surface allows knowing human mechanical behavior and aids the doctor in the detection of gait disorders like, the risk of foot ulcerations, leg discrepancy, and footprint alterations. Plantar pressure data combined with techniques that use integral reasoning produce easy understanding medical tools for assisting in treatment, early detection, and the development of preventive strategies. The present research compares the classification of human plantar foot alterations using Fuzzy Cognitive Maps (FCM) trained by Genetic Algorithm (GA) against a Multi-Layer Perceptron Neural Network (MLPNN). One hundred and fifty-one subject volunteers (aged 7–77) were classified previously with the flat foot (n = 70) and cavus foot (n = 81) by specialized physicians of the Piédica diagnostic center. The trial walking was conducted using plantar pressure platforms FreeMed®. The foot surface was divided into 14 areas that included toe 1 st to 5th, metatarsal joint 1st to 5th, lateral midfoot, medial midfoot, lateral heel, and medial heel. Pressure data were normalized for each area. Better performance in the classification using small amounts of data were found by using Fuzzy rather than non-Fuzzy approach.
EN
This study presents an artificial intelligence technique based on ensemble of artificial neural networks for the purposes of analysis and prediction of labour productivity. The study focuses on the development of model that combines several artificial neural networks on the basis of real-life data collected on a construction site for steel reinforcement works. The data includes conditions, characteristics, features of steel reinforcement works and related efficiencies of workers assigned to particular tasks recorded on site. The proposed ensemble based model combines five supervised learning models - five different multilayer perceptron networks, which contribution in the prediction is weighted due to the application of generalised averaging approach. Testing results show that the proposed ensemble based model achieves the satisfactory evaluation criteria for coefficient of correlation (0.989), root-mean-squared error (2.548), mean absolute percentage error (4.65%) and maximum absolute percentage error (8.98%).
PL
Wydajność pracy ma kluczowy wpływ na czas realizacji i koszty przedsięwzięć budowlanych. W publikacji przedstawiono wyniki prac badawczych nad wykorzystaniem zespołów sztucznych sieci neuronowych w analizie i predykcji wydajności pracy na przykładzie robot zbrojarskich. Analiza została przeprowadzona w oparciu dane zbierane przez wykonawcę w trakcie realizacji robót. Celem pracy badawczej była ocena przydatności danych zebranych przez wykonawcę robot oraz proponowanego narzędzia matematycznego do analizy i predykcji wydajności pracy.
20
Content available remote A machine learning approach to predict explosive spalling of heated concrete
EN
Explosive spalling is an unfavorable phenomenon observed in concrete when exposed to heating load. It is a great potential threat to safety of concrete structures subjected to accidental thermal loads. Therefore, assessing explosive spalling risk of concrete is important for fire safety design of concrete structures. This paper proposed a popular machine learning approach, i.e., artificial neural network (ANN), to assess explosive spalling risk of concrete. Besides, the decision tree method was also used to execute the same mission for a comparison purpose. Twenty-eight groups of heating tests were conducted to validate the proposed ANN model. The ANN model behaved well in assessing explosive spalling of concrete, with a prediction accuracy of 82.1%. This study shows that ANN is a promising method for adequate classification of concrete as material resistant or not resistant to thermal explosive spalling.
first rewind previous Strona / 27 next fast forward last
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.