BazTech - Yadda

1

Research on electric vehicle charging load prediction and charging mode optimization

Zhang Zhiyan, Shi Hang, Zhu Ruihong, Zhao Hongfei, Zhu Yingjie

Archives of Electrical Engineering

|

2021

|

Vol. 70, nr 2

399--414

EN

To reduce the influence of the disorderly charging of electric vehicles (EVs) on the grid load, the EV charging load and charging mode are studied in this paper. First, the distribution of EV charging capacity and state of charge (SOC) feature quantity are analyzed, and their probability density function is solved. It is verified that both EV charging capacity and SOC obey the skew-normal distribution. Second, considering the space-time distribution characteristics of the EV charging load, a method for charging load prediction based on a wavelet neural network is proposed, and compared with the traditional BP neural network, the prediction results show that the error of the wavelet neural network is smaller, and the effectiveness of the wavelet neural network prediction is verified. The optimization objective function with the lowest user costs is established, and the constraint conditions are determined, so the orderly charging behavior is simulated by the Monte Carlo method. Finally, the influence of charging mode optimization on power grid operation is analyzed, and the result shows that the effectiveness of the charging optimization model is verified.

2

Soft computing-based technique as a predictive tool to estimate blast-induced ground vibration

Arthur Clement Kweku, Temeng Victor Amoako, Ziggah Yao Yevenyo

Journal of Sustainable Mining

|

2019

|

Vol. 18, iss. 4

287--296

EN

The safety of workers, the environment and the communities surrounding a mine are primary concerns for the mining industry. Therefore, implementing a blast-induced ground vibration monitoring system to monitor the vibrations emitted due to blasting operations is a logical approach that addresses these concerns. Empirical and soft computing models have been proposed to estimate blast-induced ground vibrations. This paper tests the efficiency of the Wavelet Neural Network (WNN). The motive is to ascertain whether the WNN can be used as an alternative to other widely used techniques. For the purpose of comparison, four empirical techniques (the Indian Standard, the United State Bureau of Mines, Ambrasey-Hendron, and Langefors and Kilhstrom) and four standard artificial neural networks of backpropagation (BPNN), radial basis (RBFNN), generalised regression (GRNN) and the group method of data handling (GMDH) were employed. According to the results obtained from the testing dataset, the WNN with a single hidden layer and three wavelons produced highly satisfactory and comparable results to the benchmark methods of BPNN and RBFNN. This was revealed in the statistical results where the tested WNN had minor deviations of approximately 0.0024 mm/s, 0.0035 mm/s, 0.0043 mm/s, 0.0099 and 0.0168 from the best performing model of BPNN when statistical indicators of Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Relative Root Mean Square Error (RRMSE), Correlation Coefficient (R) and Coefficient of determination (R2) were considered.

3

Epileptic seizure detection based on improved wavelet neural networks in long-term intracranial EEG

Geng D., Zhou W., Zhang Y., Geng S.

Biocybernetics and Biomedical Engineering

|

2016

|

Vol. 36, no. 2

375--384

EN

Automatic seizure detection is of great importance for speeding up the inspection process and relieving the workload of medical staff in the analysis of EEG recordings. In this study, a method based on an improved wavelet neural network (WNN) is proposed for automatic seizure detection in long-term intracranial EEG. WNN combines the traditional back propagation neural network (BPNN) with wavelet transform. Compared with classic WNN architectures, a modified point symmetry-based fuzzy c-means (MSFCM) algorithm is applied to the initialization of wavelet transform's translations, which has been successful in multiclass cancer classification. In addition, Fast-decaying Morlet wavelet is chosen as the activation function to make the WNN learn faster. Relative amplitude and relative fluctuation index are extracted as a feature vector to describe the variation of EEG signals, and the feature vector is then fed into WNN for classification. At last, post-processing including smoothing, channel fusion and collar technique is adopted to achieve more accurate and stable results. This system performs efficiently with the average sensitivity of 96.72%, specificity of 98.91% and false-detection rate of 0.27 h_1. The proposed approach achieves high sensitivity and low false detection rate, which demonstrates its potential for clinical usage.

4

The soft rock socketed monopile with creep effects – a reliability approach based on wavelet neural networks

Kozubal J., Tomanovic Z., Zivaljevic S.

Archives of Mining Sciences

|

2016

|

Vol. 61, no. 3

571--585

EN

In the present study the numerical model of the pile embedded in marl described by a time dependent model, based on laboratory tests, is proposed. The solutions complement the state of knowledge of the monopile loaded by horizontal force in its head with respect to its random variability values in time function. The investigated reliability problem is defined by the union of failure events defined by the excessive horizontal maximal displacement of the pile head in each periods of loads. Abaqus has been used for modeling of the presented task with a two layered viscoplastic model for marl. The mechanical parameters for both parts of model: plastic and rheological were calibrated based on the creep laboratory test results. The important aspect of the problem is reliability analysis of a monopile in complex environment under random sequences of loads which help understanding the role of viscosity in nature of rock basis constructions. Due to the lack of analytical solutions the computations were done by the method of response surface in conjunction with wavelet neural network as a method recommended for time sequences process and description of nonlinear phenomenon.

PL

W niniejszym studium zaprezentowany jest problem pojedynczego pala osadzonego w miękkiej skale, zastosowano wiskoplastyczny model materiału bazujący na wynikach badań laboratoryjnych zespołu z Uniwersytetu Montenegro. Rozwiązanie uzupełnia stan wiedzy dla pali obciążonych poziomą siłą w głowicy zmienną w sposób losowy w czasie. Badany problem niezawodności został określony przez sumę zdarzeń – awarii – zdefiniowanych jako przekroczenie maksymalnie dopuszczalnego poziomego przemieszczenia głowicy pala niezależnie w wszystkich stanach obciążenia. Zastosowano program metody elementów skończonych, ABAQUS, do budowy trójwymiarowego modelu z dwuwarstwowym wiskoplastycznym modelem dla margla. Parametry mechaniczne modelu zarówno w części plastycznej i reologicznej zostały skalibrowane na podstawie wyników badań laboratoryjnych wykonanych na przestrzeni ostatnich czterech lat na próbkach z jednorodnego złoża margla w Montenegro. Ważnym aspektem problemu jest analiza niezawodności pojedynczego pala dla złożonego mechanicznie środowiska w ramach sekwencji losowych obciążeń. Przedstawione zadanie pozwala dostrzec istotę lepkiej części modelu. Ze względu na brak rozwiązań analitycznych oraz długotrwałość procesu obliczeniowego obliczenia niezawodnościowe przeprowadzono metodą powierzchni odpowiedzi bazując na sieciach falkowo-neuronowych. Sieć poprzez nadanie jej struktury rejestru została dostosowana do opisu procesu o nieliniowym charakterze zjawiska i dla obciążeń zmiennych w czasie.

5

Navigation of autonomous mobile robot using different activation functions of wavelet neural network

Panigrahi P. K., Ghosh S., Parhi D. R.

Archives of Control Sciences

|

2015

|

Vol. 25, no. 1

21--34

EN

An autonomous mobile robot is a robot which can move and act autonomously without the help of human assistance. Navigation problem of mobile robot in unknown environment is an interesting research area. This is a problem of deducing a path for the robot from its initial position to a given goal position without collision with the obstacles. Different methods such as fuzzy logic, neural networks etc. are used to find collision free path for mobile robot. This paper examines behavior of path planning of mobile robot using three activation functions of wavelet neural network i.e. Mexican Hat, Gaussian and Morlet wavelet functions by MATLAB. The simulation result shows that WNN has faster learning speed with respect to traditional artificial neural network.

6

A fuzzy wavelet neural network stabilizer design using genetic algorithm for multi-machine systems

Ganjefar S., Tofighi M.

Przegląd Elektrotechniczny

|

2013

|

R. 89, nr 5

19-25

EN

This paper presents a new method to design power system stabilizer (PSS) using fuzzy wavelet neural network (FWNN) for stability enhancement of a multi-machine power system. In the proposed approach, Wavelet Neural Network (WNN) is used to construct a well localized in both time and frequency domains consequent part for each fuzzy rule of a Takagi-Sugeno-Kang (TSK) fuzzy model. In designing the FWNN stabilizer the activation function of hidden layer neurons is substituted with dilated and translated Mexican Hat wavelet function. In the proposed method, an efficient genetic algorithm (GA) approach is used to obtain the optimal values of such parameters as translation, dilation, weights, and membership functions. These parameters are tuned through simulation of non-linear model of power system under chosen disturbance by minimizing a non-explicit based objective function. Results are promising and demonstrate the capabilities of the proposed FWNN stabilizer in damping of overall power oscillations in the system. It is worth noting that the proposed FWNN stabilizer, moreover, significantly improves the dynamic response characteristics, reducing the number of fuzzy rules as well as a fast convergence of network.

PL

W artykule opisano metodę projektowania stabilizatora systemu elektroenergetycznego z wykorzystaniem sieci neuronowej bazującej na rozmytej teorii falkowej. W celu optymalizacji parametrów sieci zastosowano algorytm genetyczny oraz wykonano symulacje uwzględniające odpowiednie zakłócenia w sieci. Wykonane badania wykazały, że proponowany algorytm pozwala na skuteczne tłumienie oscylacji mocy w systemie elektroenergetycznym.

7

Prediction of Oil Temperature for Substation Distribution Transformers Using Wavelet Neural Networks

Lipiński P.

Przegląd Elektrotechniczny

|

2008

|

R. 84, nr 12

202-204

EN

In this paper a new approach for predicting oil temperature for substation distribution transformer is presented. It takes advantage of a wavelet neural network (WNN), which is used to predict the oil temperature in the following 32 minutes. The in this article we choose optimal parameters for WNN and demonstrate that linear WNN performs better in terms of mean square error (MSE) of predicted values than linear and nonlinear neural networks.

PL

W artykule zostało przedstawione nowe podejście do przewidywania temperatury oleju w transformatorze. W podejściu tym zostały użyte neuronowe sieci falkowe w celu przewidzenia temperatury oleju z wyprzedzeniem 32 minut. W artykule przedstawiono sposób wyboru optymalnych parametrów dla sieci neuronowo falkowej. Wykazano także że, neuronowe sieci falkowe są w stanie przewidzieć temperaturę oleju z większą dokładnością niż liniowe sieci falkowe i nieliniowe sieci falkowe. Do porównania dokładności przewidywań jako kryterium porównania przyjęto błąd MSE.

8

Detekcja i identyfikacja zespołu QRS w sygnałach elektrokardiograficznych stymulatorowych

Duraj A., Krawczyk A.

Przegląd Elektrotechniczny

|

2006

|

R. 82, nr 12

38-41

PL

W niniejszym opracowaniu proponujemy połączenie dwóch bardzo szybko rozwijających się dziedzin, a mianowicie sieci neuronowych oraz przekształceń falkowych. Nowa metoda sieci falkowo - neuronowej nie tylko wykrywa zespół QRS ale jednocześnie dokonuje jego klasyfikacji na rytm endogenny. Eksperymenty prowadzone były dla sygnałów elektrokardiograficznych pochodzących od pacjentów z wszczepionym układem stymulującym pracę serca. Testowano 150 sygnałów o częstotliwości 128Hz udostępnionych przez Instytut Kardiologii w Warszawie.

EN

We described QRS detection method in our earlier works. Although results a lot of authors are good, problems detection still arise. These problems mostly arise at important parts of the signals to be classified due to massive and fast signal changes. On the other hand, noisy signals lead to false detection QRS. This article presents a new QRS detected method. We combine the wavelet transform and neural networks to develop a new method for detection QRS and classification of patient rhythm and rhythm descending from pacemaker. We distinguish two categories of methods. In the first one, the wavelet part is essentially decoupled from learning. A signal is decomposed on some wavelet and the wavelet coefficients are furnished to a neural network. In the second category, wavelet theory and neural networks are combined into a single method. The idea of using wavelets in neural networks has been proposed. We limit the scope of this article to the first category.

9

Detekcja zespołu QRS przy zastosowaniu sieci falkowo-neuronowych

Duraj A., Krawczyk A., Kumor M.

Przegląd Elektrotechniczny

|

2005

|

R. 81, nr 12

98-100

PL

W artykule przedstawiono wstępne wyniki badań prowadzone w zakresie wykrywania zespołu ORS w oparciu o sieć falkowo -neuronową. Źródłem sygnałów testowych była ogólnodostępna baza sygnałów elektrokardiograficznych MIT-BIH zawierająca zapisy EKG trwające od 15 minut do 24h z różnymi patologiami chorób serca oraz sygnały elektrokardiograficzne pochodzące od pacjentów z wszczepionym układem stymulujący prace serca uzyskane w Instytucie Kardiologii w Warszawie.

EN

The paper follows the authors' former works which dealt with the methods based on wavelet-neural network methods. The aim of the paper is to describe tools based on wavelet-neural network and to present selected results. Two kinds of ECG signals were tested. The first gmup of ECG signals were taken from the open source of ECG signals (MIT-BIH). This contains the ECG records lasting from 15 min till 24 hrs with some pathological syndromes. The secound group of ECG signals were taken from Institute of Cardiology, Warsaw.