Modeling and forecasting electricity consumption

In the Israel, electric power is becoming the main energy form trusted in all financial sectors from the country. As time goes by, while different establishments and properties had been built and developed, the necessity for home-based electricity intake within the region accelerates. Energy consumption is a crucial index from the economic development of a country. Quick changes in sector and the economic system strongly impact energy ingestion. According to the Intercontinental Energy Gross annual (IEA) in the year 2004, the Israel had total installed electrical power generating capability of 12-15.

one particular gigawatts (GW). The country made 53. one particular billion kilowatt-hours (Bkwh) of electricity in 2004, while consuming forty-nine. 4 Bkwh. Conventional thermal sources make up the largest discuss of Filipino electricity supply, comprising much more than 65 percent of the total in 2005.

However , the Philippines is likewise the world’s second-largest producer of geothermal energy. Inspite of several electrical power sources, there are still bunch of conditions that occur including electricity scarcity and high price somewhat because of increase of demand and company bills.

The Philippines is not just the sole region that encounters these particular dilemmas however the other countries in Asia like Lebanon and Arab saudi, and the world too. This kind of pushes analysts and authorities to study the consumption motion from the earlier years to make certain that they learn the behavior and suggest a means to help make the power corporations and to stop uncertainties that might happen in the near future. Through the years, there are plenty of ways and methods developed by the experts and one of them is definitely modeling and forecasting.

Modelling electric energy usage is useful in planning and distribution simply by power utilities. Modeling can be described as process of creating abstract, conceptual, graphical and mathematical models. Models are usually used in the next either difficult or not practical to create experimental conditions in which scientists can easily directly evaluate outcomes. In the field of energy use direct to electricity, modeling is a very important factor in predicting the next pair of electricity consumption.

There are plenty of tactics and mathematical methods that happen to be already utilized and proven effective in identifying the energy ingestion such as Multivariate regression “analysis, neural networks, autoregressive, and many other. Nowadays, time-series analysis was also utilized in the electricty consumption modeling and foretelling of. In figures, signal processing, and numerical finance, a time series is known as a sequence of data points, scored typically at successive time instants spread out at homogeneous time times. Based on Investopedia (2012) it gives you another building approach which in turn requires only data within the modeled varying, thus saving the user the problem of determining influential variables and suggesting a form intended for the relation between them.

For example, measuring the importance of retail sales each month in the year would comprise an occasion series. This is due to sales earnings is well defined, and consistently tested at evenly spaced time periods. Data gathered irregularly or only once aren’t time series. Also, in accordance to Austrilian Bureau of Statistics (2005) an seen time series can be decomposed into 3 components: fashionable (long term direction), the seasonal (systematic, calendar related movements) as well as the irregular (unsystematic, short term fluctuations). Models intended for time series data can easily have many varieties and symbolize different stochastic processes. Various other applications of time-series analysis are in macroeconomics and financial. As of now, modelling and predicting is of the highest optimum of achievement and significance in the modern society particularly in aiding selected dilemmas in electric intake.

Objectives

1 ) To formulate a statistical model intended for the electrical power consumption in the Philippines. installment payments on your To forecast the Philippines’ monthly electrical power consumption for the next three years. three or more. To evaluate the formulated style in forecasting the electric power consumption.

Relevance of the Research

The goal of this kind of study shall be able to outlook the electrical power consumption of the Philippines for the next three years through formulating a model acquired from your data by making use of time-series. This information can bring about much towards the power supply businesses of the Korea in order for them to identify the group of electricity ingestion for the approaching years. The forecasted consequence will help all of them plan and prepare for whatever might happen in the future years specifically in addressing the electricity shortage.

Scope and Limitation

The research focuses generally in modeling the electricity use of the Philippines by using the time-series evaluation. The study can be narrowed towards the forecasting in the monthly electric power consumption for the next three years from the entire Thailand. The data used in modelling will be based upon the 1999-2011 record.

Overview of Related Books

Modelling and forecasting electricity consumption of Malaysian large steel generators This analyze attempts to model and forecast the daily optimum demand of Malaysian significant steel generators and the twelve-monthly maximum require contributed simply by these metallic mills. This attempts to combine both the top-down and bottom-up approaches to prediction the daily and gross annual maximum demand of the stainlesss steel mills. The top-down way uses regression analysis to forecast the annual volume of electric power consumption from the steel generators. The bottom-up approach uses the Unit for Evaluation of Electric Require Electric Weight (MAED_EL) to convert the annual stainlesss steel mills electrical energy consumption (which was previous obtained from the regression model) into by the hour load with the steel mills. The suggested method displays good predicting accuracy, with weekly Mean Absolute Percentage Error (MAPE) of 2. 3%.

This study propose mixture of the top-down and bottom-up methods to forecast the daily maximum demand of Malaysian large steel mills and the annual optimum demand offered by these steel generators. The top-down approach uses regression examination to outlook the twelve-monthly electricity intake of these significant steel generators, based on their relationship with annual metal production and GDP. The projected total annual electricity consumption from regression analysis was then incorporated into the bottom-up model applying MAED_EL to set up the per hour load figure. From the on an hourly basis load curves, the daily and total annual maximum requirements of the stainlesss steel mills will be determined. The[desktop] has the ability to prediction accurately the daily more the large steel mills, with MAPE of less than 3%.

The recommended method nevertheless , is simply based on the assumption the future trend of daily consumption follows the base year. Although this is a slight disadvantage, nevertheless the proposed method has provided the utility which has a better ways to forecast metallic mills’ weight, despite the unavailability of daily production info which is essential in foretelling of. The outcome with this study can benefit the utility in ensuring trusted and economical operation with the national grid, and is likewise useful for evaluation pertaining to advancement future maximum generation and transmission expansion plans. Studies of this research also provide a valuable contribution to the utility in deciding load supervision strategies and designing of tariff constructions.

A possible way of improve the prediction performance is by combining the model using a time series method such as ARIMA. This will enable the model to consider the most recent behavior of stainlesss steel mills fill, and thus boost the accuracy with the forecast. The very best approach however , would be the one that will be able to take into account the daily production data of the metal mills. While using availability of this specific data, many other complex plus more effective methods can be looked into such as Manufactured Neural Network (ANN) and fuzzy geradlinig regression. These kinds of methods can to capture the factors that highly affect steel generators daily fill such as daily production prepare and repair schedule, and so improve the precision of the outlook. (S. Aman et. al, 2011 )

Long term strength consumption predicting using innate programming Taking care of electrical energy source is a sophisticated task. The most important part of electric powered utility useful resource planning is forecasting of the future load demand in the local or countrywide service area. This is usually achieved by constructing versions on relative information, just like climate and former load require data. With this paper, a genetic encoding approach is proposed to forecast permanent electrical power consumption in the place covered by a computer program situated in the southeast of Turkey. The empirical results demonstrate good load forecast with a low error charge.

In this paper, a hereditary programming way on the forecasting of long term electrical power usage of a modest city in Turkey was presented. It uses the genetic programming strategy to forecast future usage through symbolic regression using total annual data with the previous years.

In conventional regression, one has to decide on the approximation function (can become an n-degree polynomial, non-polynomial, or a mix of both) and try to find the coefficients on this selected function. Constructing an approximation function can be a trial. There is an additional form of regression called “symbolic regression. Inside the symbolic regression problem, the aim is to search a symbolic representation of your model, rather than only looking for coefficients of any predefined style. Genetic coding (GP) technique introduced simply by Koza can be utilized for the symbolic regression problem. GP searches for the model and coefficients from the model at the same time. In this research, power ingestion data is usually processed with both conventional analysis and innate programming methods.

Long term electricity consumption predicting can provide information for electricity distribution centers. Power usage in this metropolis is growing; therefore exact forecasts will help authorities for making reliable programs. In this work, a genetic programming centered forecasting technique is presented. Two other curve fitting methods are also shown for assessment with this method. Data found in all three designs are not preprocessed. Genetic coding technique is utilized to form a model and assess the parameters for the model. The goodness of the match produced by the genetic development method is evaluated using sum of square-shaped errors (SSE) method, which is better than the other two methods of regression. It was demonstrated that the hereditary programming can be utilised for electric powered utility resource planning and forecasting of the future load demand in the local or national service region effectively. (K. Karabulot ain. al, 2008)

Electricity intake forecasting in Italy using linear regression models The influence of economic and demographic factors on the total annual electricity consumption in Italia has been researched with the intention to produce a long-term consumption forecasting style. The time period regarded for the historical info is from 1970 to 2007. Several regression designs were created, using historical electricity intake, gross home product (GDP), gross home-based product every capita (GDP per capita) and human population. A first portion of the paper considers the evaluation of GROSS DOMESTIC PRODUCT, price and GDP per capita elasticities of home-based and nondomestic electricity consumption. The home-based and non-domestic short run selling price elasticities are found to be both approximately comparable to 0. 06, while long haul elasticities will be equal to 0. 24 and 0. 2009, respectively. However, the elasticities of GROSS DOMESTIC PRODUCT and GDP per household present larger values. Inside the second part of the paper, different regression models, based on co-integrated or immobile data, happen to be presented. Several statistical testing are employed to check the quality of the proposed models.

A comparison with nationwide forecasts, based on complex econometric models, just like Markal-Time, was performed, showing that the designed regressions will be congruent with the official predictions, with deviations of 1% for the best case and 11% for the worst. These kinds of deviations need to be considered appropriate in relation to time span taken into account. This paper aims to calculate GDP, price and GROSS DOMESTIC PRODUCT per household elasticities of domestic and non-domestic electrical power consumption in Italy. Likewise this paper wants to prediction the future growth of these consumptions using diverse regression versions and compare our results with other offered projections. The elasticity examination showed the price suppleness of domestic and non-domestic consumption is fairly limited, credit reporting some outcomes presented in previous studies. Through the studies, conclusions have been completely acquired.

First, there is no need to consider electricity price as explaining changing in forecasting models for Italian electrical power consumption. Second, pricing policies cannot be accustomed to promote the efficient make use of electricity in Italy. The estimation of GDP and GDP every capita elasticities showed larger values regarding price elasticities, demonstrating that the consumption response to GDP and GDP per capita alterations is relevant. Therefore , there is the have to assure an appropriate level of electricity supply to sustain the economic development in Italia. According to the second target of the paper, different long-term predicting models had been developed and in addition they substantially lead to similar results.

Therefore , in the next years, an increase in the whole electricity consumption, driven by both home-based and nondomestic consumptions, should be expected in Italia with a normal rate corresponding to about 2% per year. Assuming that the data reported represent the reference benchmark, it can guarantee the most accurate projections for total, home-based and non-domestic electricity consumptions respectively, mainly because they fit the info. It is thought that the elasticities, forecasts and comments offered in this newspaper would be useful to energy organizers and coverage makers to develop future cases about the Italian electrical energy consumption. (V. Bianco ou. al., 2009)

Forecasting electric power consumption in New Zealand using economical and market variables The influence of selected financial and market variables within the annual electrical power consumption in New Zealand has been researched. The study uses gross home-based product, common price of electricity and population of recent Zealand through the period 1965″1999. Models will be developed applying multiple geradlinig regression research. It was identified that the electrical power consumption related effectively using variables. Forecasts made applying these types were compared with some readily available national forecasts. The forecasts are also compared to the forecasts of the previously developed Logistic model. Electricity consumption foretelling of models based on economic elements for Home-based and NonDomestic sectors and Total consumption for New Zealand using multiple linear regression have been suggested.

The versions performed efficiently in the record tests conducted, implying their particular significance in forecasting electrical power consumption using the explaining parameters considered. Reviews of these types have been constructed with the national forecasts obtainable in New Zealand. The comparison revealed that the forecasts created by the regression models are extremely comparable while using national predictions. The precision of the forecasts made by these kinds of models depends strongly on the accuracy of forecasts created for the explaining variables. From this paper, straightforward regression have been used to version these parameters. (Z. Mohamed & Pat Bodger, 2003)

Modeling and Forecasting Electricity Demand in the Philippines The Philippine federal government has deregulated electricity generation markets to encourage private investors and actively courted independent electricity producers (IPPs). This has been completed promote efficiency and reduce govt financial debt responsibilities. Until the mid-1980s, the power sector in the Philippines was typically state-owned through the National Electric power Corporation (NPC). After the debt crisis in the early 1980s, the Philippines’ government tightened fiscal coverage, and capital expenditure for extra electricity potential was drastically reduced. This led to a slowdown inside the electricity generating facilities. Concurrently, electricity demand continued to boost. This resulted in tight electric power supply and demand circumstances by the middle of the 1980s. Presently there continue to be durations of generating ability constraints.

This might be the result of problems in forecasting electricity usage. The residential and commercial electricity require in the Korea is patterned. The examination follows Johansen’s vector mistake correction way of estimate the cost and income elasticity in both lengthy and short run. The outcomes indicate a lengthy run cointegrating relationship is located among residential electricity consumption, income, plus the stock of electrical appliances. Inside the industrial sector there definitely seems to be a long-run relationship retains for industrial electricity usage and GDP. The lack of significant price answers appears to be a result of government creation policies. The estimated models are used in forecasting total electricity ingestion suggest that the government’s established forecast for electricity demand would be for the upper destined of the forecast range. (K. Ishi & F. Joutz, 2009) Methodology

1 . Accumulate data with the Philippines’ month-to-month electricity intake from the yr 1999 to 2011 in National Figures Coordination Table to be intended for constructing a time-series model for the electricity consumption. 2 . Making use of the formulated unit, forecast the Philippines’ electrical power consumption for three years. 3. Through the obtained forecasted intake, evaluate the functionality of the model.

References

Clough, L. (2008). Strength profile of Philippines. The encyclopedia of earth. Tenang, S., Titled ping, H., & Mubin, M (2011). Modeling and predicting electricity intake of Malaysian large steel mills. Clinical Research

and Works Vol. 6th (8), pp. 1817-1830. ISSN 1992-2248. Karabulut, K., Alkan, A., & Yilmaz, A. (2008). Permanent energy consumption forecasting employing genetic programming. Mathematical and Computational Applications, Vol. 13, No . two, pp. 71-80. Bianco, V., Manca, U., & Nardini, S. (2009). Electricity ingestion forecasting in Italy applying linear regression models. Elsevier Ltd. Strength 34 (2009) 1413-1421. Mohamed, Z. & Bodger, S. (2003). Forecasting electricity consumption in New Zealand applying economic and demographic parameters. Elsevier Ltd. Energy 50 (2004) 1833-1843.

one particular

Tweet

Related essay