RU   UA  

Anastasia Usova

Faculty of computer science and technology (CST)

Automated control systems

Speciality Automated control systems

Computerized system for optimizing electricity consumption in a residential building

Scientific adviser: Ph.D., Assoc. Sergey Hmelevoy

Resume Abstract

Abstract

Content

• Introduction
• 1. Relevance of the problem of optimizing electricity consumption
• 2. Analysis of approaches for optimizing electricity consumption
• 3. The object of optimization in the residential sector
• 4. Applying mathematical methods
• 4.1 Particle swarm optimization
• 4.2 Algorithm for solving linear programming problems using the inner point method>
• 4.3 Greedy algorithm
• 4.4 Genetic algorithm
• 5. Goals and tasks of the research
• 6. The formal statement of the problem
• 7. Applying a genetic algorithm to Demand Response
• Conclusions
• References

Introduction

Currently, energy consumption is an essential condition for human existence. At the same time, population growth and people's desire to improve living standards lead to the need for a rapid increase in energy capacity. That entails the problem of depletion of raw material resources, so people are required to use energy more economically and rationally. The irrational use of fuel and energy resources also contributes to higher electricity prices.

Optimization of electricity consumption becomes necessary to solve the above problems.

Therefore, the issue of energy conservation remains relevant. Energysaving is a process that reduces the need for energy resources and energy carriers per unit of the final useful effect of their application. [1] However, it should be noted that reliable power supply to consumers requires a balance of electricity production and consumption with the necessary amount of redundancy [2].

The task of the master's thesis is to develop a system for optimizing electricity consumption. In this work optimizing electricity consumption will be considered on the example of an apartment building, in the conditions of an OSMD OSMD. OSMD (Association of co-owners of an apartment building) is a legal entity established by owners to facilitate the use of their common property and the management, maintenance and use of the common property. [3]. The object of optimization of electricity consumption in the residential sector will be home electrical appliances and they will be optimized for switching on time points, to minimize energy costs and reduce the peak load on the network.

1. Relevance of the problem of optimizing electricity consumption

With the development of mankind, electricity consumption increases along with the volume of resource extraction to provide the energy industry. Scientists have estimated that modern power plants will one day not be able to meet the demand for electricity, because annual consumption is growing by an average of 15-20% [4]. Let's take a concrete example, see fig. 1.

As can be seen from fig. 1 the bulk of global electricity consumption growth is in Asia (almost 80%, with China accounting for almost 60%). China's electricity demand has accelerated on the back of sustained economic growth and industrial demand. Demand also increased in India, South Korea, Japan and Indonesia. Electricity consumption in the United States, which declined by 1% in 2017, recovered in 2018 (+2.2%). Most of this increase came from the residential sector (+6.2%), mainly due to increased electricity consumption for household appliances (which accounts for about half of electricity consumption) and air conditioning (almost 90% of American homes use centralized or individual air conditioners). Economic growth and industrial demand also contributed to the growth of electricity consumption in Canada, Brazil and Russia. It has also increased in Africa, especially in Egypt, and in the middle East, helped by Iran. Electricity consumption in Europe remained stable in 2018: decreased in France and Germany, stabilized in other major countries (Great Britain, Italy, Spain) and increased in the Netherlands, Poland and Turkey [5].

In order to prevent future problems related to the lack of electrical capacity, many governments are working on programs aimed at encouraging people to reduce their electricity consumption by using energy efficient technologies, tariff systems and energy saving equipment.

For the time being, the management of electricity consumption is relevant and an important tool for managing the balance of supply and demand in energy markets.

2. Analysis of approaches for optimizing electricity consumption

In many countries, the Demand Response program is being introduced to optimize electricity consumption Demand Response. Demand Response it is a change in the consumption of electricity to end consumers relative to their normal load profile in response to changing electricity prices over time or in response to incentive payments provided to reduce consumption during periods of high electricity prices on the wholesale market or when system reliability is jeopardized [6]. Analysis of literature sources shows that there is a whole class of problems related to solving the Demand Response problem [7-9].

The shape of the load power can be adjusted using the following methods [10-12]: peak clipping, valley filling, load shifting, energy efficiency, flexible load shape, see fig. 2.

Load shifting is the most effective and widely used method for load management in residential power supply networks.

3. Optimization objects in the residential sector

Household appliances are the most important optimization target in the residential sector. The analysis of instruments by type and frequency of activation implies some classification.

In particular, [13] considers household electrical appliances with cyclic operation and thermostatic control. In [14] the authors divide electrical appliances with cyclic operation mode and unmanaged operation mode. In [15] they are classified into electrical appliances: working in the presence of a person in the house, in need of human control, working without the presence of a person in the house. In [16] the authors distinguish the following classes of electrical appliances: uncontrolled appliances, controlled electrical appliances with uninterrupted load, controlled electrical appliances with interrupted load. In [17] electrical appliances are divided into: non-optimizing electrical appliances, electrical appliances with thermostatic control, with cyclic operation.

Based on the above information in this paper, it is proposed to divide household appliances into the following categories:

• unoptimized;
• with thermostatic control;
• with cyclic operation mode;
• needing a human presence in the house.

Analysis of literature sources shows that optimization of devices with thermostatic control and devices with cyclic mode of operation is possible and widely used by a number of authors.

4. Applying mathematical methods

In the studied sources, the following methods were used to optimize electricity consumption: the particle swarm optimization (PSO), the algorithm for solving the linear programming problem by the inner point method, the genetic algorithm (GA), and the greedy algorithm.

4.1 Particle swarm optimization

The particle swarm optimization is a numerical optimization method that does not require knowing the exact gradient of the optimized function. The data is presented in binary form [18].

In the particle swarm method, each particle represents a potential solution to the problem. The behavior of a particle in the hyperspace of the search for a solution is always adjusted according to its experience and that of its neighbors. Each particle remembers its best position with the achieved local best value of the fitness function and knows the best position of the particles - their neighbors, where the current global optimum is reached.

This method was used by the authors in [16]. The paper presents a demand management strategy based on the method of planning the load of the power grid for the day ahead. The planning of home devices is shown in such a way as to exclude peak values of electricity consumption and minimize energy bills. The particle is a two-dimensional array of 0 and 1. The rows that represent individual electrical appliancesndashare the hour number columns. 0 – turning off the electrical appliance at a given hour, 1 – turning on. The particle is expanded to a vector: S = [S₁, S₂, S₃, ..., S₂₃, S₂₄].

4.2 Algorithm for solving linear programming problems using the inner point method

Linear programming is a method for solving extreme problems on sets of n-dimensional vector space defined by systems of linear equations and inequalities.

According to the inner point methods, you can only select the source point for searching inside a valid area. In this case, the set of points is divided into acceptable and invalid points, depending on the restrictions. The set of acceptable points is also divided into boundary and internal points, depending on the restrictions. The authors in [17] consider the problem of optimizing household energy consumption, without alternative energy sources, by constructing an optimal schedule for the use of household appliances. According to the classification proposed in [19], this problem reduces to the task at hand integer programming and proposed an algorithm for its solution by the interior point method. The target function of the original task is defined as the total cost of energy consumed. The moments of switching on and off devices are optimized within the framework of dividing into numbers of time intervals implemented in advance.

4.3 Greedy algorithm

The greedy algorithm is a method for solving optimization problems based on the fact that the decision-making process can be divided into elementary steps, at each of which a separate decision is made. The decision at each step should be optimal in the current step should be taken without regard to previous or subsequent decisions [20].

In the greedy algorithm, the decision-making process can be divided into elementary steps, each of which takes a separate decision. The decision made at each step should be optimal only for the current step and should be made without taking into account previous or subsequent decisions. This algorithm divides the problem into smaller subsets of the same problem. The authors in [21] consider extensive demand management strategies using a greedy algorithm. The paper analyzes the dynamics of electricity consumption by household appliances in the home network. The purpose of the proposed planning is to minimize total electricity costs for the consumer and balance the network load.

4.4 Genetic algorithm

Genetic algorithms are adaptive search methods that are used to solve functional optimization problems. They are based on mechanisms and models of evolution, and genetic processes of biological algorithms.

The algorithm takes a group of solutions and searches for the most appropriate ones. Then slightly changes them – gets new solutions, among which the best are again selected and the worst are discarded. Thus, at each step of the work, the algorithm selects the most suitable solutions (conducts selection), assuming that they will give even better solutions at the next step (evolve) [23].

The block–algorithm scheme is shown in fig. 3.

Genetic algorithms are used for solving problems of Demand Response the large number of authors [19], [24-27]. For example, in the work [19] the chromosome represents the switching moment for the i electrical appliance. During the operation of the algorithm, loads on the power grid will be shifted over time, depending on the price of electricity, and the time for switching on the electrical appliance is optimized. The fitness function of the total cost of energy consumption for the planning interval seeks to minimize the total cost of electricity, taking into account the limitations.

Also consider the use of a genetic algorithm to minimize the cash cost of electricity for households [26], [27], for office sectors [27], industrial [26]. GA is used to reduce peak loads on the grid [19], [24-26]. Optimization of the energy consumption can be applied various constraints [24], [25], in particular, take into account the comfort of the user (at the time of activation of electrical appliances) [24], [25]. The obtained results of simulations in the above works show that the GA is effective in this task.

In the table 1, compare the algorithms against predefined criteria.

Of all the methods considered, it is proposed to optimize electricity consumption using a genetic algorithm for the following reasons:

• simple and transparent implementation;
• in a very large data set, it is not guaranteed to provide the optimal solution, but it allows you to quickly find a good one.
• is effective when the search space is very large and a large number of parameters are involved.

5. Goals and tasks of the research

The goals of this work are:

• minimizing energy costs and limiting peak load of electricity consumption;
• dynamic restructuring of the schedule for switching on electrical appliances in case of manual change.

TTo achieve the set goals, it is necessary to solve the following tasks:

• domain analysis;
• research methods for optimizing electricity consumption;
• modification of existing methods if they do not fully meet the existing conditions;
• software implementation;
• an experiment that confirms the effectiveness of the method.

6. The formal statement of the problem

The classical purpose of these works is to minimize energy costs, in addition, in the works [24-26] takes into account the limitations peak load that can be applied one of the above methods (fig.2).

There are M – apartments in the house, and N – electrical appliances in each apartment. For each managed appliance user can t_i – turn on time i appliance, where i = [1, ... , N] and Δt_i – duration of the operation time of the electric appliance.

Hence the finish time of the operation of the appliance: t’_i = t_i + Δt_i.

For each of the electrical devices can get the function of the cardinality estimate for i appliance – P_i.

For i appliance, the evaluation function of power per day will be equal to the sum of the capacities of each of the inclusions, which depend on the time of activation of the appliance and the length of time of operation of the appliance. Therefore, it is possible to obtain the following formula: P_t=∑_tP_i(t_i,Δt )

In order to obtain the total power P needed to find the sum of all P_t: P_t=∑P_t

The cost of electricity consumed is denoted by S. The cost for i appliance in the day will be equal to the sum of the values of power for each of the runs at different times of day: S_t=∑_tS_i(P_i(t_i,Δt),t_i)

The cost of electricity is not a constant and can change at different times of the day.

In order to get the total cost of S electricity consumed, it is necessary to find a sumS_t: S=∑S_t

The goal of this work is to minimize the total cost of electricity, taking into account the day and night tariff: S → min

This is achieved by shifting the switching times of electrical appliances. Thus, in order to maximize the time the inclusion was less than the maximum time limit off of the appliance: max(t_i – t_i+1) < t_max

7. Applying a genetic algorithm to Demand Response

It is assumed that a microcontroller is installed in each apartment in an apartment building, which will control electrical appliances. At the same time, payment for electricity is charged at two rates: night and day.

In turn, users provide information to the administrator about the devices that will be regulated. Name of the electrical appliance and its power consumption. Users also make a schedule for their electrical equipment in the apartment.

The computerized system will distribute the operating time for thermostatically controlled and cyclically operated electrical appliances.

At the output of the energy consumption optimization system, it is expected to obtain an optimal operating schedule for electrical appliances with a decrease in demand, which entails minimizing the total cost of electricity, taking into account the day and night tariffs, as well as reducing the peak load of electricity consumption. Planned output format:

Y={[N_i(Appliance)], [VKL_i,t(Current state of the electrical appliance)],
[t_i(opening Hours)], [P_i(Power)], [EcostSavings(energy saving)]};

Expected results: minimization of electricity costs and reduction of peak electricity consumption in residential buildings.

Conclusions

So we can conclude that in this work we have analyzed the use of Demand Response methods, analyzed the strategies of load generation, considered the classifications of electrical equipment, analyzed the approaches for optimizing the energy consumption. An overview of the mathematical methods and algorithms that can be applied to the original problem was made. The genetic algorithm was selected. The formal problem statement was completed. The application of the genetic algorithm to Demand Response was considered.

References

1. Стукова Т.М. Основной принцип формирования энергосберегающих программ // Научные исследования – 2015 – с. 52–53.
2. Кулешов М. Управление потреблением на рынке электроэнергии // Управление эффективностью и результативностью – 2012 – с. 44–53.
3. Функции и деятельность ОСМД [Электронный ресурс] / Ольга Папинова. – Электрон. текст. – 2016 – Режим доступа: https://ru.wikipedia.org/wiki/Объединение_совладельцев_многоквартирного_дома– Загл. с экрана.
4. Зачем надо экономить электроэнергию [Электронный ресурс] / Ася Малютина. – Электрон. текст. – [Россия, 2015]. – Режим доступа: https://recyclemag.ru/article/zachem-nado-ekonomit-elektro– Загл. с экрана.
5. Мировое потребление электроэнергии [Электронный ресурс] – Режим доступа:https://yearbook.enerdata.ru/electricity/electricity-domestic-consumption-data.html– Загл. с экрана.
6. National Assessment & Action Plan on Demand Response – 2016 – Режим доступа:https://ru.wikipedia.org/wiki/Управление_спросом_на_электроэнергию#cite_note-1– Загл. с экрана.
7. Na Li, Lijun Chen and Steven H. Low Optimal Demand Response Based on Utility Maximization in Power Networks Power and Energy Society General Meeting // IEEE Power and Energy Society General Meeting, 2011. – p. 1–8.
8. Dehnavi E, Abdi H. Optimal pricing in time of use demand response by integrating with dynamic economic dispatch problem // Energy 109, 2016. – p. 1086–1094.
9. C. Eid, E. Koliou, M. Valles, J. Reneses, e R. Hakvoort, Time–based pricing and electricity demand response: Existing barriers and next steps, Util. Policy, vol. 40, pp. 15–25, Jun. 2016.
10. D. Javor, A. Janjic, Application of demand side management techniques in successive optimization procedures // Communications in dependability and Quality Management, An International Journal 4 (19), 2016. – p.40–51.
11. Kishor& C. Muley, Sandeep Bhongade, Load Management Techniques and Pricing Model for Demand Side Management – A Review // International Journal on Emerging Technologies 10(1), 2019. – p.42–46.
12. Mohamed AboGaleela, Magdy El Marsafawy, Mohamed El–Sobki Optimal Scheme with Load Forecasting for Demand Side Management (DSM) in Residential Areas // Energy and Power Engineering, Vol. 5, 2013. – p. 889–896.
13. Lujano Rojas JM, Monteiro C, Dufo Lopez R, Bernal–Agustin JL. Optimum residential load man–agement strategy for real time pricing demand response programs // Energy policy, Vol.45, 2012. – p.671–679.
14. Bradak Z., Kaczmarczyk V., Fiedler P. Optimal scheduling of domestic appliances via MILP // Energies, Vol.8, Iss. 1. 2015. – p.217–232.
15. Mahmood, D.; Javaid, N.; Alrajeh, N.; Khan, Z.A.; Qasim, U.; Ahmed, I.; Ilahi, M. Realistic scheduling mechanism for smart homes // Energies Vol. 9, 2016, – p.202.
16. Kumar, K.S.S., Naik, M.G.: Load Shifting Technique on 24Hour Basis for a Smart–Grid to Reduce Cost and Peak Demand Using Particle Swarm Optimization // IRJET, Vol.04, 2017. – p.1180–1185.
17. Гребенюк Г.Г., Крыгин А.А. Оптимизация энергопотребления домохозяйства на основе прогноза графика максимальной нагрузки бытовых приборов. – Управление большими системами: сборник трудов 69, – 2017.
18. Метод роя частиц [Электронный ресурс] – Режим доступа: https://ru.wikipedia.org/wiki/Метод_роя_частиц – Загл. с экрана.
19. Albadi M.H., El Saadany E.F. A summary of demand response in electricity markets // Electric Power Systems Research, Vol.11, Iss. 78, 2008. – p.1989–1996.
20. Жадные алгоритмы, теоретические основы, применение [Электронный ресурс] – Режим доступа: https://genew.ru/tema--jadnie-algoritmi-teoreticheskie-osnovi-primenenie.html– Загл. с экрана.
21. Cui, Tiansong, Hadi Goudarzi, Safar Hatami, Shahin Nazarian, and Massoud–Pedram. Concurrent optimization of consumer's electrical energy bill and producer's power generation cost under a dynamic pricing model // IEEE PES Innovative Smart Grid Technologies (ISGT), 2012. – p. 1–6.
22. Oladeji, O., Olakanmi, O.O. A genetic algorithm approach to energy consumption scheduling under demand response. In 2014 IEEE 6th International Conference on Adaptive Science & Technology (ICAST) (pp. 1–6). IEEE. – 2014
23. Генетические алгоритмы [Электронный ресурс] – Режим доступа: http://lazysmart.ru/iskusstvenny-j-intellekst/geneticheskie-algoritmy-ili-kak-uchebn/– Загл. с экрана.
24. Javaid, N., Javaid, S., Abdul, W., Ahmed, I.; Almogren, A., Alamri, A., Niaz, I.A. A Hybrid Genetic Wind Driven Heuristic Optimization Algorithm for Demand Side Management in Smart Grid // Energies, Vol.10, N 319, 2017, – p.1–7.
25. R.R. El Razky, A.A. Daoud and . K. El–serafi, Optimization of Residential Load Consumption during Energy Peaks using Smart Metering // Renewable Energy and power quality journal (RE&PQJ), Vol.1, N 15, 2017. – p.227–232.
26. M. Awais, N. Javaid, N. Shaheen, Z. Iqbal, G. Rehman, K. Muhammad, I. Ahmad. An Efficient Genetic Algorithm Based Demand Side Management Scheme for Smart Grid. In Electrical Engineering COMSATS Institute of Information Technology Islamabad – 2015. – Taipei, Taiwan, p.351–356.
27. Nusrat Shaheen, Nadeem& Javaid, Najmun Nisa, Amber Madeeha Zeb, Zahoor Ali Khan, Umar Qasim. Appliance Scheduling for Energy Management with User Preferences // International Conference on Innovative Mobile and Internet Services in Ubiquitous Computing 10th. 2016. – p. 328–334.