Wyniki wyszukiwania - BazTech

1

Selection of an algorithm for classifying data quoted on the Day Ahead Market of TGE S.A. in MATLAB and Simulink using Deep Learning Toolbox

Tchórzewski Jerzy, Mielcarz Tomasz

Studia Informatica : systems and information technology

|

2023

|

Vol. 1(28)

83--108

EN

The article contains an analysis leading to the selection of an algorithm for classifying data listed on the Day-Ahead Market of TGE S.A. in MATLAB and Simulink using Deep Learning Toolbox. In this regard, an introduction to deep learning methods, classification methods, and classification algorithms is provided first. Particular attention was paid to the essence of three important deep learning methods in the classification, i.e. the methods called: Stochastic Gradient Descent Momentum, Root Mean Square Prop and Adaptive Moment Estimation. Then, three architectures of artificial neural networks used in deep learning were characterized, i.e.: Deep Belief Network, Convolutional Neural Network and Recurrent Neural Network. Attention was paid to the selection parameters of algorithms for learning deep artificial neural networks that can be used in classification, such as: accuracy, information losses and learning time. Practical aspects of research experiments were also shown, including selected results of research conducted on volume and fixing 1 data quoted on the TGE S.A. Day-Ahead Market. After analyzing the obtained test results for the hourly system, it was noted that the least suitable algorithm for classification purposes was the Stochastic Gradient Descent Momentum algorithm, which in each case had worse results than the other two algorithms, i.e. the Adaptive Moment Estimation algorithm and the Root Mean algorithm Square Prop. However, the best algorithm turned out to be the Adaptive Moment Estimation algorithm, which obtained the highest accuracy, which was at a level comparable to the Root Mean Square Prop algorithm, with the latter algorithm having larger losses.

2

Prediction capabilities of the LSTM and Perceptron models based on the Day-Ahead Market on the Polish Power Exchange S.A.

Ruciński Dariusz

Studia Informatica : systems and information technology

|

2023

|

Vol. 1(28)

69--82

EN

The main purpose of the research was to examine the properties of models for two kinds of neural networks, a deep learning models in which the Long Short-Term Memory was chosen and shallow neural model in which the Perceptron Neural Network was chosen. The subject of the examination was the Day-Ahead Market system of PPE S.A. The article presents the learning results of both networks and the results of the predictive abilities of the models. The research was conducted based on data published on the Polish Stock Exchange for the 2018 year. The MATLAB environment was chosen as a tool for providing the examinations. The determination index (R2) and the mean square error (MSE) was adopted as the network evaluation criterion for the learning ability and for the prediction ability of both networks.

3

Artificial Neural Network based on mathematical models used in quantum computing

Ruciński Dariusz

Studia Informatica : systems and information technology

|

2022

|

Vol. 2(27)

27--48

EN

The article is a proposition of a new approach to building a neural model based on the system of Day-Ahead Market operating at TGE S.A. The reason for the proposed method is an attempt to find a better model for the DAM system. The proposed methodology is based on using mathematical models used in quantum computing. All calculations performed on learning the Artificial Neuron Network are based on operations described in Hilbert space. The main idea of calculations is to replace the data from the decimal system into the quantum state in Hilbert space and perform learning operations for a neural model of the DAM system in a special manner which relay on the teaching model for each position of the quantum register for all data. The obtained results were compared to the “classical” neural model with the use of a comparative model.

4

Analysis of data quoted on the Day-Ahead Market of TGE S.A. using Statistics and Machine Learning Toolbox

Tchórzewski Jerzy, Longota Bartłomiej

Studia Informatica : systems and information technology

|

2022

|

Vol. 2(27)

49--74

EN

The publication contains the results of research in the field of cluster analysis carried out using data quoted on the Day-Ahead Market of TGE S.A. Two methods were used in the analysis, one hierarchical known as the Ward’s method, and the other non-hierarchical - the k-means method. Many interesting research results have been obtained, which are illustrated, among others, in in the form of dendrograms, silhouette graphs and graphs in the form of clusters. Data on the volume and the volumeweighted average price of electricity were examined for various types of quotations: fixing 1, fixing 2 and continuous quotations. The research was carried out in the MATLAB and Simulink environments using a library called Machine and Statistics Learning Toolbox. Selected test results were interpreted.

5

Hourly identification and simulation of the TGE S.A. Day-Ahead Market system

Marlęga Radosław, Tchórzewski Jerzy R.

Control and Cybernetics

|

2022

|

Vol. 51, No. 4

523--555

EN

The paper presents selected research results concerning the identification and simulation of the TGE S.A. Day-Ahead Market (DAM) system of the day for electricity delivered and sold, listed for the following hours: 5:01-6:00, 11:01-12:00, 17:01-18:00 and 23:01-24:00 in 2019, which were obtained in the MATLAB and Simulink environment using the System Identification Toolbox. As a result of identification, four respective discrete parametric arx models were obtained, which were then subject to quality assessment. Then, a simulation model was built in the Simulink environment, which was used for simulation tests and for assessing the sensitivity of the model created using the data from 2019 as the basis and the data from 2020 for verification. The obtained results confirm the correctness of both the performed discrete parametric identification and the possibility of testing the quality of the model and its sensitivity with the use of the DAM system model in the MATLAB and Simulink environment.

6

The Influence of the Artificial Neural Network type on the quality of learning on the Day-Ahead Market model at Polish Power Exchange joint-stock company

Ruciński Dariusz

Studia Informatica : systems and information technology

|

2019

|

Vol. 1-2(23)

77--93

EN

The work contains the results of the Day-Ahead Market modeling research at Polish Power Exchange taking into account the numerical data on the supplied and sold electricity in selected time intervals from the entire period of its operation (from July 2002 to June 2019). Market modeling was carried out based on three Artificial Neural Network models, ie: Perceptron Artificial Neural Network, Recursive Artificial Neural Network, and Radial Artificial Neural Network. The examined period of the Day-Ahead Market operation on the Polish Power Exchange was divided into sub-periods of various lengths, from one month, a quarter, a half a year to the entire period of the market's operation. As a result of neural modeling, 1,191 models of the Market system were obtained, which were assessed according to the criterion of the least error MSE and the determination index R2.

7

Krótkookresowe prognozy ceny energii elektrycznej na Rynku Dnia Następnego

Popławski T., Weżgowiec M.

Śląskie Wiadomości Elektryczne

|

2012

|

Nr 5 (104)

10--14

PL

Artykuł porusza problematykę prognozowania krótkookresowego ceny energii elektrycznej na Rynku Dnia Następnego. Badania przeprowadzone zostały na rzeczywistych danych, dla których skonstruowano szereg czasowy. Wykonano analizę statystyczną i przeprowadzono testy pozwalające na wybór odpowiedniej metody i modelu predykcyjnego. Zaproponowano kilka modeli prognostycznych dla przewidywania ceny w horyzoncie dobowym. Dokonano analizy oceny prezentowanych modeli.

EN

Article raises the problem of short-term forecasting of the electricity price on the Day-Ahead-Market. Researches were conducted on real data, for which the time series was constructed. Statistical analysis and tests were performed which allow to choose the proper method and predictive model. It was proposed several forecasting models to predict price in the twenty-four hours horizon. It was performed analysis and evaluation used tor presented models.