The subject of the research is the methods of constructing and training neural networks as a nonlinear modeling apparatus for solving the problem of predicting the energy consumption of metallurgical enterprises. The purpose of this work is to develop a model for forecasting the consumption of the power system of a metallurgical enterprise and its experimental testing on the data available for research of PJSC "Dneprospetsstal". The following tasks have been solved: analysis of the time series of power consumption; building a model with the help of which data on electricity consumption for a historical period is processed; building the most accurate forecast of the actual amount of electricity for the day ahead; assessment of the forecast quality. Methods used: time series analysis, neural network modeling, short-term forecasting of energy consumption in the metallurgical industry. The results obtained: to develop a model for predicting the energy consumption of a metallurgical enterprise based on artificial neural networks, the MATLAB complex with the Neural Network Toolbox was chosen. When conducting experiments, based on the available statistical data of a metallurgical enterprise, a selection of architectures and algorithms for learning neural networks was carried out. The best results were shown by the feedforward and backpropagation network, architecture with nonlinear autoregressive and learning algorithms: Levenberg-Marquard nonlinear optimization, Bayesian Regularization method and conjugate gradient method. Another approach, deep learning, is also considered, namely the neural network with long short-term memory LSTM and the adam learning algorithm. Such a deep neural network allows you to process large amounts of input information in a short time and build dependencies with uninformative input information. The LSTM network turned out to be the most effective among the considered neural networks, for which the indicator of the maximum prediction error had the minimum value. Conclusions: analysis of forecasting results using the developed models showed that the chosen approach with experimentally selected architectures and learning algorithms meets the necessary requirements for forecast accuracy when developing a forecasting model based on artificial neural networks. The use of models will allow automating high-precision operational hourly forecasting of energy consumption in market conditions. Keywords: energy consumption; forecasting; artificial neural network; time series.
Об'єкт даного дослідження визначає управління ризиками централізованої аптечної мережі, що пов'язані з маркетингом відносин між мережею та різними групами клієнтів. Розвиток конкурентного ринку призводить аптечні мережі до необхідності вести діалог з клієнтом, надаючи йому певні переваги, тим самим зменшувати власні ризики. Предметом дослідження є моделювання оптимального портфелю клієнтів централізованої аптечної мережі як одного з інструментів управління ризиками. Спираючись на основи портфельної теорії Марковіца та багатокритеріальної оптимізації, в даній роботі побудовано базову модель оптимального портфелю клієнтів централізованої аптечної мережі, що враховує три групи клієнтів-лояльних, випадкових та інтернет-замовлення. На відміну від класичної двокритеріальної моделі (мінімізація ризику при максимізації доходу) в модель додано критерій максимізації ентропії, що підсилює ефект диверсифікації. Також розглянуто чотири модифікації базової моделі. Перша з них поглиблює аналіз портфелю клієнтів мережі до окремих аптек, що входять до цієї мережі. Три наступні моделюють різні маркетингові стратегії, в яких перевага надається одній з груп клієнтів. Для розв'язування відповідних до побудованих моделей багатокритеріальних задач розроблено програмне забезпечення в пакеті Matlab, за допомогою якого методом послідовних поступок знайдено множини парето-оптимальних портфелів клієнтів. Верифікацію моделей виконано на реальних даних, що були надані однією з аптечних мереж. Моделювання оптимальних портфелів клієнтів централізованої аптечної мережі дозволяє усунути недоліки в управлінні мережею та обрати оптимальну комбінацію розподілу груп лояльних, випадкових та інтернет-клієнтів. Завдяки цьому забезпечується можливість впливати на ці групи споживачів, впроваджуючи відповідні програми лояльності, що в кінцевому рахунку веде до підвищення прибутку. Результати моделювання будуть корисними для автоматизації бізнес-процесів будь-яких торгівельних мереж, управління ризиками, аналізу програм лояльності для підвищення ефективності їхнього функціонування. Ключові слова: аптечна мережа, лояльні клієнти, інтернет-клієнти, модель оптимального портфелю, багатокритеріальна задача.
The subject of the article's research is the formalization of unstructured or semistructured verbal information for a fuzzy production system. The purpose of the work is to develop a method of constructing membership functions for fuzzy sets of terms of a linguistic variable that will allow formalizing unstructured or semistructured verbal information for fuzzy production systems. The following tasks are solved in the article: to develop a method of constructing the membership functions to determine the sequence of stages: the stage of modeling verbal information in the form of digraphs, the stage of constructing the order relation on the elements of this model, the step of determining a linguistic variable based on the created model, and determining the functions of fuzzy sets of linguistic terms variable. Methods are used: graph theory, mathematical induction, fuzzy modeling. Results obtained: a method for constructing the membership function of linguistic variables that formalizes unstructured or semi-structured qualitative information for fuzzy production systems is developed. For this purpose, the process of constructing the membership function has been broken down into stages. The implementation of the first stage requires the creation of a model of unstructured or semistructured verbal information. Three models of information based on oriented trees are considered with increasing complexity. A model based on an acyclic oriented graph is considered as a generalization. Such a model is the basis for processing information that has a structure of greater complexity. The second stage provides a theoretical basis for constructing the order relation for the elements of the models under consideration. For the implementation of the third stage, a method of identifying the order elements on the basis of the positional system is proposed. Based on the ID of each ordered element, functions of fuzzy sets of terms of a linguistic variable are constructed. Appropriate procedures have been developed to implement the steps. Conclusions: application of the method will allow automating the assignment of vectors of input and output information, to automate the formation of fuzzy sets of terms of the corresponding linguistic variables, will allow to build fuzzy products as a knowledge base of a fuzzy production system, and to train such a system.
The risk management of a centralized pharmacy network identifies the research object. The paper proposes a strategy of complex diversification for the pharmacy network. By constructing portfolio models for complex diversification and solving relevant multicriteria problems, multiple pareto-optimal portfolios have been found for successive risk management. Based on the fundamentals of Markowitz portfolio theory and multicriteria optimization, this paper builds four models of the optimal portfolios for centralized pharmacy network. In contrast to the classic two-criteria model (risk minimization while maximizing income), our models have been introduced to maximize entropy, which enhances the diversification effect. Matlab software has been developed for solving multicriteria problems. Model verification was performed on real data provided by one of the pharmacy networks. The modeling results will be useful for automating the business processes of any trading network, managing risk, analysing loyalty programs to improve the effectiveness of their operations.
An up-to-date issue of a modern metallurgical enterprise is the increase of its energy efficiency, which is related, first of all, with energy saving. Therefore, the purpose of this paper is to develop a model for forecasting the metallurgical enterprise power system consumption and its experimental testing based on the PJSC “Electrometallurgical plant “Dniprospetsstal” named after A. M. Kuzmin data. In order to build a forecasting model, a neural network apparatus in the MATLAB system was used and it was done in two stages. At the first stage, as an experiments series result, the optimal architecture and algorithm of neural network training were determined. In the second stage, the dependence of the modeling graphs load error from the influence of daily consumption graphs is identified. The MATLAB software package has been adapted for the needs of “Dniprospetsstal” named after A. M. Kuzmin. Neural networks designed in this way can be used to solve applied issues of electrometallurgy, in particular, the long-term estimation of time series of hourly power for the 24 hours ahead.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.