# Short-Term Electricity Demand Forecasting: Impact Analysis of Temperature for Thailand

^{1}

^{2}

^{3}

^{*}

## Abstract

**:**

## 1. Introduction

#### 1.1. Motivation and Background

#### 1.2. Review of Related Works

#### 1.2.1. Temperature Effect

#### 1.2.2. Weekend, Holiday/Special Day Effects

#### 1.2.3. Grouping of Dataset

#### 1.3. Model Selection

#### 1.4. Contributions

- The marginal impact of temperature that leads to raising the demand for day hours and night hours is explored for Thailand which is quite useful for tropical countries.
- The quantitative analysis among the variables such as the impact of holidays, working days, working days after a holiday/long holiday, AR effect, special days/events such as Bangkok flood for the demand is discussed in detail.
- The unexpected Bangkok flood and lockdown situation were quite similar to the current Covid-19 in terms of electricity demand. Therefore, the researcher can extend this methodology to analyze the impact on electricity due to Covid-19.
- Construction of four different scenarios based on similar characteristics of demand which leads to achieving the best prediction capability among the existing literature of the Thai dataset.
- The strategy for the selection of variables, determination of the training length of a dataset, hidden layers and nodes are also major contributions for the improvement of the accuracy are also major contributions of this study.

## 2. Methods

- Scenario 1: only demand for working days.
- Scenario 2: only demand for weekends.
- Scenario 3: only holiday demand and highly fluctuated demand from December 24 to New Years eve.
- Scenario 4: all the demand dataset.

#### 2.1. Model Design

#### 2.1.1. MLR Model Description

- OLS: where errors $\sum =\mathit{I}$
- GLS: for orbitrary covariance ∑
- GLSAR: where AR(p) $\sum =\sum \rho $

#### 2.1.2. OLS and GLSAR Estimation

#### 2.1.3. Performance Measurements

#### 2.2. Artificial Neural Network Approach

#### 2.2.1. Structure of ANNs

#### 2.2.2. Activation Function

#### 2.2.3. Resolving Overfitting

## 3. Results and Discussion

#### 3.1. Selection of Training Length

- Case I: Training period: 911 days, test period: 239 days in the year 2013.
- Case II: Training period: 717 days, test period: 239 days in the year 2013.
- Case III: Training period: 475 days, test period: 239 days in the year 2013.
- Case IV: Training period: 236 days, test period: 239 days in the year 2013.

#### 3.2. MLR Approach: Simple OLS Method

#### 3.2.1. Model Selection

#### 3.2.2. Temperature Effect

#### 3.2.3. Special Day and AR Effect

#### 3.3. ANN Approach: FF-ANN: A Simple DNN

#### 3.4. Computation Time

#### 3.5. Pro and Cons of the Methods

## 4. Conclusions

## Author Contributions

## Funding

## Acknowledgments

## Conflicts of Interest

## Abbreviations

AC | Air Condition |

AR | Auto-regression |

ARMAX | Auto-regressive Moving Average with Exogenous Variable |

CDD | Cooling Degree Day |

Covid-19 | Corona Virus Disease 2019 |

DW | Durbin–Watson |

EGAT | Electricity Generating Authority of Thailand |

FF-ANN | Feed Forward Artificial Neural Network |

GA | Genetic Algorithm |

GLSAR | Generalized Least Square Auto Regression |

HWT | Holt Winters Triple |

MAPE | Mean Absolute Percentage Error |

MEA | Metropolitan Electricity Authority |

MLR | Multiple linear regression |

MW | Megawatt |

OLS | Ordinary Least Square |

PSO | Particle Swarm Optimization |

STDF | Short-term Demand Forecasting |

RegSARIMA | Regression Seasonal ARIMA |

ReLU | Rectified Linear Unit |

SVM | Support Vector Machine |

## Appendix A. Data Pre-Processing

#### Appendix A.1. Monthly and Seasonal Pattern

**Figure A2.**(

**a**) Hourly demand profile with seasonality and trend. (

**b**) Monthly variation of demand. (

**c**) Monthly variation of demand: 2010–2013.

#### Appendix A.2. Weekly, Daily and Holiday Patterns

**Figure A3.**(

**a**) Intra-day variation of demand for working and special days. (

**b**) Intra-day variation of demand on the working day (next day of holiday).

**Figure A4.**(

**a**) Impact of long holiday on electricity demand. (

**b**) Impact of New Year on electricity demand. (

**c**) Impact of Sonkran festival on electricity demand (

**d**) Impact of massive Bangkok flood on electricity demand.

#### Appendix A.3. Temperature

**Figure A5.**(

**a**) Effect of temperature during peak hour. (

**b**) Effect of temperature at two different hours.

#### Appendix A.4. Variable Identification

Types | Variables | Description |
---|---|---|

Deterministic | WD | Week dummy [Mon <Tue ...<Sat<Sun] |

MD | Month dummy [Feb <Mar <...<Nov <Dec] | |

DayAfterHoliday | Binary 0 or 1 | |

DayAfterLongHoliday | Binary 0 or 1 | |

DayAfterSongkran | Binary 0 or 1 | |

DayAfterNewyear | Binary 0 or 1 | |

Temperature | Temp | Forecasted temperature |

MaxTemp | Maximum forecasted temperature | |

Square temperature | Square of the forecasted temperature | |

MA2pmTemp | Moving avearage of temperature at 2pm | |

Lagged | load1d_cut2pm | 1 day ahead untill 2pm and 2 day ahead after 2pm load |

load2d_cut2pm | 2 days ahead untill 2pm and 3 day ahead after 2pm load | |

load3d_cut2pmR | 3 days ahead untill 2pm and 4 days ahead after 2pm load | |

load4d_cut2pmR | 4 days ahead untill 2pm and 5 days ahead after 2pm load | |

Interaction | WD:Temp | Interaction of week day dummy to temperature |

MD:Temp | Interaction of month dummy to temperature | |

WD:load1d_cut2pm | Interaction of week day dummy to load1d_cut2pm | |

WD:load2d_cut2pm | Interaction of week day dummy to load2d_cut2pm |

## Appendix B. Figures and Tables

**Figure A11.**(

**a**) Forecasting during Sonkran festival (Scenario 4). (

**b**) Forecasting during last week of December (Scenario 4).

## References

- Murakoshi, C.; Namagami, H.; Xuan, J.; Takayama, A.; Takayama, H. State of residential energy consumption in Southest Asia: Need to promote smart appliances because urban household consumption is higher than some develped countries. ECEEE Summer Study Proc.
**2017**, 1489–1499. Available online: https://www.eceee.org/library/conference_proceedings/eceee_Summer_Studies/2017/7-appliances-products-lighting-and-ict/state-of-residential-energy-consumption-in-southeast-asia-need-to-promote-smart-appliances-because-urban-household-consumption-is-higher-than-some-developed-countries/ (accessed on 4 May 2020). - Panklib, K.; Prakasvudhisarn, C.; Khummongkol, D. Electricity Consumption Forecasting in Thailand Using an Artificial Neural Network and Multiple Linear Regression. Energy Sources Part B Econ. Plan. Policy
**2015**, 10, 427–434. [Google Scholar] [CrossRef] - ADB. Key Indicators for Asia and the Pacific 2017. Online
**2017**. Available online: https://www.adb.org/sites/default/files/publication/357006/06-rt-energy-electricity.pdf (accessed on 1 May 2020). - Kaur, N.; Sood, S.K. An Energy-Efficient Architecture for the Internet of Things (IoT). IEEE Syst. J.
**2017**, 11, 796–805. [Google Scholar] [CrossRef] - Samson, O. Electricity and the Fourth Industrial Revolution. 2018. Available online: https://www.researchgate.net/publication/324876698_ELECTRICITY_AND_THE_FOURTH_INDUSTRIAL_REVOLUTION (accessed on 4 May 2020).
- Clarke, L.; Jiang, K.; Akimoto, K.; Babiker, M.; Blanford, G.; Fisher-Vanden, K.; Hourcade, J.C.; Krey, V.; Kriegler, E.; Löschel, A.; et al. Chapter 6—Assessing transformation pathways. In Climate Change 2014: Mitigation of Climate Change. IPCC Working Group III Contribution to AR5; Cambridge University Press: Cambridge, UK, 2014. [Google Scholar]
- Momami, M. Factors Affecting Electricity Demand in Jordan. Energy Power Eng.
**2013**, 5, 50–58. [Google Scholar] [CrossRef][Green Version] - Parkpoom, S.; Harrison, G.P. Analyzing the Impact of Climate Change on Future Electricity Demand in Thailand. IEEE Trans. Power Syst.
**2008**, 23, 1441–1448. [Google Scholar] [CrossRef][Green Version] - Mideksa, T.; Kallbekken, S. The impact of climate change on the electricity market: A review. Energy Policy
**2010**, 38, 3579–3585. [Google Scholar] [CrossRef] - McCulloch, J.; Ignatieva, K. Forecasting High Frequency Intra-Day Electricity Demand Using Temperature. SSRN Electr. J.
**2017**. [Google Scholar] [CrossRef] - Julian, M.C.; Julian, P. Temperature effects on firms’ electricity demand: An analysis of sectorial differences in Spain. Appl. Energy
**2015**. [Google Scholar] [CrossRef][Green Version] - Hossein Motlagh, N.; Mohammadrezaei, M.; Hunt, J.; Zakeri, B. Internet of Things (IoT) and the Energy Sector. Energies
**2020**, 13. [Google Scholar] [CrossRef][Green Version] - Csoknyai, T.; Legardeur, J.; Akle, A.A.; Horvath, M. Analysis of energy consumption profiles in residential buildings and impact assessment of a serious game on occupants behavior. Energy Build.
**2019**, 196, 1–20. [Google Scholar] [CrossRef] - Serralles, R.J. Electric energy restructuring in the European Union: Integration, subsidiarity and the challenge of harmonization. Energy Policy
**2006**, 34, 2542–2551. [Google Scholar] [CrossRef] - Chapagain, K.; Kittipiyakul, S. Short-Term Electricity Demand Forecasting with Seasonal and Interactions of Variables for Thailand. In Proceedings of the 2018 Int Electrl Eng Congress (iEECON), Krabi, Thailand, 7–9 March 2018; pp. 1–4. [Google Scholar] [CrossRef]
- Taylor, J.W.; de Menezes, L.M.; McSharry, P.E. A comparison of univariate methods for forecasting electricity demand up to a day ahead. Int. J. Forecast.
**2006**, 22, 1–16. [Google Scholar] [CrossRef] - Ramanathan, R.; Engle, R.; Granger, C.W.; Vahid-Araghi, F.; Brace, C. Short-run forecasts of electricity loads and peaks. Int. J. Forecast.
**1997**, 13, 161–174. [Google Scholar] [CrossRef] - Dordonnat, V.; Koopman, S.J.; Ooms, M. Dynamic factors in periodic time-varying regressions with an application to hourly electricity load modelling. Comput. Stat. Data Anal.
**2012**, 56, 3134–3152. [Google Scholar] [CrossRef] - Chapagain, K.; Kittipiyakul, S. Performance Analysis of Short-Term Electricity Demand with Atmospheric Variables. Energies
**2018**, 11, 818. [Google Scholar] [CrossRef][Green Version] - Chapagain, K.; Kittipiyakul, S. Short-term Electricity Load Forecasting Model and Bayesian Estimation for Thailand Data. In Proceedings of the 2016 Asia Conf on Power and Electl Engg (ACPEE 2016), Bankok, Thailand, 20–22 March 2016; Volume 55, p. 06003. [Google Scholar]
- Clements, A.E.; Hurn, A.S.; Li, Z. Forecasting day-ahead electricity load using a multiple equation time series approach. Eur. J. Oper. Res.
**2016**, 251, 522–530. [Google Scholar] [CrossRef][Green Version] - Phyo, P.; Jeenanunta, C. Electricity Load Forecasting using a Deep Neural Network. Eng. Appl. Sci. Res.
**2019**, 46, 10–17. [Google Scholar] - Su, W.H.; Jeenanunta, C. Short-term Electricity Load Forecasting in Thailand: An Analysis on Different Input Variables. In IOP Conference Series: Earth and Environmental Science; IOP Publishing: Bristol, UK, 2018. [Google Scholar] [CrossRef]
- Darshana, A.K.; Chawalit, J. Hybrid Particle Swarm Optimization with Genetic Algorithm to Train Artificial Neural Networks for Short-term Load Forecasting. Int. J. Swarm Intell. Res.
**2019**, 10, 1–14. [Google Scholar] - Chapagain, K.; Sato, T.; Kittipiyakul, S. Performance analysis of short-term electricity demand with meteorological parameters. In Proceedings of the 2017 14th Int Conf on Electl Eng/Elx, Computer, Telecom and IT (ECTI-CON), Phuket, Thailand, 27–30 June 2017; pp. 330–333. [Google Scholar]
- Ministry, E. Energy data visualaization for Thailand. Energy Policy Plan. Off.
**2020**. Available online: http://www.eppo.go.th/index.php/en/ (accessed on 2 May 2020). - Wangpattarapong, K.; Maneewan, S.; Ketjoy, N.; Rakwichian, W. The impacts of climatic and economic factors on residential electricity consumption of Bangkok Metropolis. Energy Build.
**2008**, 40, 1419–1425. [Google Scholar] [CrossRef] - Cian, D.E.; Lanzi, E.; Roberto, R. The Impacts of Temperature Change on Energy Demand: A Dynamic Panel Analysis. SSRN Electr. J.
**2007**. [Google Scholar] [CrossRef] - Mirasgedis, S.; Sarafidis, Y.; Georgopoulou, E.; Kotroni, V.; Lagouvardos, K.; Lalas, D. Modeling framework for estimating impacts of climate change on electricity demand at regional level: Case of Greece. Energy Convers. Manag.
**2007**, 48, 1737–1750. [Google Scholar] [CrossRef] - Masood, N.A.; Sadi, M.Z.; Deeba, S.R.; Siddique, R.H. Temperature Sensitivity Forecasting of Electrical Load. In Proceedings of the 2010 4th International Power Engineering and Optimization Conference (PEOCO), Shah Alam, Malaysia, 23–24 June 2010. [Google Scholar]
- Istiaque, A.; Khan, S.I. Impact of Ambient Temperature on Electricity Demand of Dhaka City of Bangladesh. Sci. Res. Publ.
**2018**, 10, 319–331. [Google Scholar] [CrossRef][Green Version] - Zhang, C.; Liao, H.; Mi, Z. Climate impacts: Temperature and electricity consumption. Nat. Hazards
**2019**, 99, 1259–1275. [Google Scholar] [CrossRef] - Zhang, M.; Zhang, K.; Hu, W.; Zhu, B.; Wang, P.; Wei, Y.M. Exploring the climatic impacts on residential electricity consumption in Jiangsu, China. Energy Policy
**2020**, 140, 111398. [Google Scholar] [CrossRef] - Li, B.; Lu, M.; Zhang, Y.; Huang, J. A Weekend Load Forecasting Model Based on Semi-Parametric Regression Analysis Considering Weather and Load Interaction. Energies
**2019**, 12, 3820. [Google Scholar] [CrossRef][Green Version] - Bessec, M.; Fouquau, J. The non-linear link between electricity consumption and temperature in Europe: A threshold panel approach. Energy Econ.
**2008**, 30, 2705–2721. [Google Scholar] [CrossRef][Green Version] - Asadoorian, M.O.; Eckaus, R.S.; Schlosser, C.A. Modeling climate feedbacks to electricity demand: The case of China. Energy Econ.
**2008**, 30, 1577–1602. [Google Scholar] [CrossRef] - Chang, Y.; Kim, C.S.; Miller, J.I.; Park, J.Y.; Park, S. A new approach to modeling the effects of temperature fluctuations on monthly electricity demand. Energy Econ.
**2016**, 60, 206–216. [Google Scholar] [CrossRef][Green Version] - Ang, B.; Wang, H.; Ma, X. Climatic influence on electricity consumption: The case of Singapore and Hong Kong. Energy
**2017**, 127, 534–543. [Google Scholar] [CrossRef] - Fung, W.; Lam, K.; Hung, W.; Pang, S.; Lee, Y. Impact of urban temperature on energy consumption of Hong Kong. Energy
**2006**, 31, 2623–2637. [Google Scholar] [CrossRef] - Maheshwari, A.; Murari, K.K.; Jayaraman, T. Peak Electricity Demand and Global Warming in the Industrial and Residential areas of Pune: An Extreme Value Approach. arXiv
**2019**, arXiv:1908.08570. [Google Scholar] - Li, Y.; Pizer, W.A.; Wu, L. Climate change and residential electricity consumption in the Yangtze River Delta, China. Proc. Natl. Acad. Sci. USA
**2019**, 116, 472–477. [Google Scholar] [CrossRef] [PubMed][Green Version] - Apadula, F.; Bassini, A.; Elli, A.; Scapin, S. Relationships between meteorological variables and monthly electricity demand. Appl. Energy
**2012**, 98, 346–356. [Google Scholar] [CrossRef] - Sailor, D.; Pavlova, A. Air conditioning market saturation and long-term response of residential cooling energy demand to climate change. Energy
**2003**, 28, 941–951. [Google Scholar] [CrossRef] - Lopez, M.; Sans, C.; Valero, S.; Senabre, C. Classification of Special Days in Short-Term Load Forecasting: The Spanish Case Study. Energies
**2019**, 12, 1253. [Google Scholar] [CrossRef][Green Version] - Darbellay, G.A.; Slama, M. Forecasting the short-term demand for electricity: Do neural networks stand a better chance? Int. J. Forecast.
**2000**, 16, 71–83. [Google Scholar] [CrossRef] - Hippert, H.; Bunn, D.; Souza, R. Large neural networks for electricity load forecasting: Are they overfitted? Int. J. Forecast.
**2005**, 21, 425–434. [Google Scholar] [CrossRef] - Chapagain, K.; Kittipiyakul, S. Short-term Electricity Load Forecasting for Thailand. In Proceedings of the 2018 15th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON), Chiang Rai, Thailand, 18–21 July 2018; pp. 521–524. [Google Scholar] [CrossRef]
- Srinivasan, D.; Chang, C.S.; Liew, A.C. Demand forecasting using fuzzy neural computation, with special emphasis on weekend and public holiday forecasting. IEEE Trans. Power Syst.
**1995**, 10, 1897–1903. [Google Scholar] [CrossRef] - Ziel, F. Modeling Public Holidays in Load Forecasting: A German Case Study. J. Mod. Power Syst. Clean Energy
**2018**, 6, 191–207. [Google Scholar] [CrossRef][Green Version] - Darshana, A.K.; Chawalit, J. Combine Particle Swarm Optimization with Artificial Neural Networks for Short-Term Load Forecasting. Int. Sci. J. Eng. Technol
**2017**, 1, 25–30. [Google Scholar] - Cottet, R.; Smith, M. Bayesian Modeling and Forecasting of Intraday Electricity Load. J. Am. Stat. Assoc.
**2003**, 98, 839–849. [Google Scholar] [CrossRef] - Soares, L.J.; Medeiros, M.C. Modeling and forecasting short-term electricity load: A comparison of methods with an application to Brazilian data. Int. J. Forecast.
**2008**, 24, 630–644. [Google Scholar] [CrossRef] - Papalexopoulos, A.D.; Hesterberg, T.C. A regression-based approach to short-term system load forecasting. IEEE Trans. Power Syst.
**1990**, 5, 1535–1547. [Google Scholar] [CrossRef] - Ismail, Z.; Jamaluddin, F.; Jamaludin, F. Time Series Regression Model for Forecasting Malaysian Electricity Load Demand. Asian J. Math. Stat.
**2008**, 1, 139–149. [Google Scholar] [CrossRef][Green Version] - Dilhani, M.H.M.R.S.; Jeenanunta, C. Daily electric load forecasting: Case of Thailand. In Proceedings of the 2016 7th Int Conf of Inf and Comm Tech for Embedded Sys (IC-ICTES), Bangkok, Thailand, 20–22 March 2016; pp. 25–29. [Google Scholar] [CrossRef]
- Hippert, H.S.; Pedreira, C.E.; Souza, R.C. Neural networks for short-term load forecasting: A review and evaluation. IEEE Trans. Power Syst.
**2001**, 16, 44–55. [Google Scholar] [CrossRef] - Hong, T.; Wang, P. Fuzzy interaction regression for short term load forecasting. Fuzzy Opt. Decis. Mak.
**2014**, 13, 91–103. [Google Scholar] [CrossRef] - Shi, H.; Xu, M.; Li, R. Deep Learning for Household Load Forecasting Novel Pooling Deep RNN. IEEE Trans. Smart Grid
**2018**, 9, 5271–5280. [Google Scholar] [CrossRef] - Asaleye, D.A.; Breen, M.; Murphy, M.D. A Decision Support Tool for Building Integrated Renewable Energy Microgrids Connected to a Smart Grid. Energies
**2017**, 10, 1765. [Google Scholar] [CrossRef][Green Version] - Sigauke, C.; Chikobvu, D. Peak electricity demand forecasting using time series regression models: An application to South African data. J. Stat. Manag. Syst.
**2016**, 19, 567–586. [Google Scholar] [CrossRef] - Sigauke, C. Regression-SARIMA modelling of daily peak electricity demand in South Africa. J. Energy South. Afr.
**2012**, 23, 23–30. [Google Scholar] - Haben, S.; Giasemidis, G.; Ziel, F.; Arora, S. Short term load forecasting and the effect of temperature at the low voltage level. Int. J. Forecast.
**2019**, 35, 1469–1484. [Google Scholar] [CrossRef][Green Version] - Soares, L.J.; Souza, L.R. Forecasting electricity demand using generalized long memory. Int. J. Forecast.
**2006**, 22, 17–28. [Google Scholar] [CrossRef][Green Version] - Sigauke, C.; Nemukula, M.M.; Maposa, D. Probabilistic Hourly Load Forecasting Using Additive Quantile Regression Models. Energies
**2018**, 11, 2208. [Google Scholar] [CrossRef][Green Version] - Botoc, C.; Anton, S.G. Is Profitability Driven by Working Capital Management? Evidence for High-growth Firms from Emerging Europe. J. Bus. Econ. Manag.
**2017**, 18, 1135–1155. [Google Scholar] [CrossRef][Green Version] - Lusis, P.; Khalilpour, K.; Andrew, L.; Liebman, A. Short-term residential load forecasting: Impact of calendar effects and forecast granularity. Appl. Energy
**2017**, 205, 654–669. [Google Scholar] [CrossRef] - Goodwin, R. Convert Irregular Time Sereis to Hourly Data in Python and Have Normal Distribution. 2018. Available online: https://stackoverflow.com/questions/51392012/convert-irregular-time-series-to-hourly-data-in-python (accessed on 10 April 2020).
- Hor, C.; Watson, S.J.; Majithia, S. Analyzing the impact of weather variables on monthly electricity demand. IEEE Trans. Power Syst.
**2005**, 20, 2078–2085. [Google Scholar] [CrossRef] - Hong, T.; Wang, P.; Willis, H.L. Short Term Electric Load Forecasting. Int. J. Forecast.
**2010**, 74, 1–6. [Google Scholar] [CrossRef] - Wang, Y.; Niu, D.; Ji, L. Short-term power load forecasting based on IVL-BP neural network technology. Syst. Eng. Procedia
**2012**, 4, 168–174. [Google Scholar] [CrossRef][Green Version]

**Table 1.**Mean Absolute Percentage Error (MAPE) measures for different training lengths using the Scenario 1 dataset.

Simple OLS | FF-ANN | |||||
---|---|---|---|---|---|---|

Training Length (Days) | Testing Days | MAPE (%) | Exe. Time (Sec) | MAPE (%) | Exe. Time (Sec) | |

Case I | 911 | 239 | 1.97 | 109.2 | 2.96 | 565.97 |

Case II | 717 | 239 | 2.33 | 87.38 | 3.41 | 516.46 |

Case III | 475 | 239 | 2.04 | 114.26 | 6.00 | 435.32 |

Case IV | 236 | 239 | 2.44 | 78.82 | 18.00 | 309.81 |

Model | Deterministics and Interaction | Temp | Temp Square | Holiday | Holiday 2 | MAPE (%) |
---|---|---|---|---|---|---|

OLS-(Scenario 1) | Yes | Yes | Yes | Yes | Yes | 1.97 |

Yes | Yes | No | No | No | 2.00 | |

Yes | No | No | No | No | 3.04 | |

OLS-(Scenario 2) | Yes | Yes | Yes | Yes | Yes | 1.78 |

OLS-(Scenario 3) | Yes | Yes | Yes | Yes | Yes | 16.00 |

OLS-(Scenario 4) | Yes | Yes | Yes | Yes | Yes | 2.94 |

HH | Rsq | Adj-Rsq | DW | HH | Rsq | Adj-Rsq | DW | ||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|

OLS | GLSAR-7 | OLS | GLSAR-7 | OLS | GLSAR-7 | OLS | GLSAR-7 | OLS | GLSAR-7 | OLS | GLSAR-7 | ||

0 | 0.97 | 0.96 | 0.97 | 0.95 | 1.80 | 2.00 | 24 | 0.86 | 0.62 | 0.86 | 0.59 | 1.73 | 1.99 |

1 | 0.97 | 0.96 | 0.97 | 0.95 | 1.86 | 2.01 | 25 | 0.88 | 0.68 | 0.88 | 0.66 | 1.79 | 1.99 |

2 | 0.97 | 0.96 | 0.97 | 0.95 | 1.85 | 2.00 | 26 | 0.87 | 0.67 | 0.87 | 0.65 | 1.77 | 1.98 |

3 | 0.97 | 0.95 | 0.97 | 0.95 | 1.85 | 2.00 | 27 | 0.86 | 0.63 | 0.86 | 0.61 | 1.74 | 1.98 |

4 | 0.97 | 0.95 | 0.97 | 0.94 | 1.85 | 2.00 | 28 | 0.85 | 0.62 | 0.85 | 0.60 | 1.78 | 1.97 |

5 | 0.97 | 0.95 | 0.97 | 0.94 | 1.87 | 2.00 | 29 | 0.86 | 0.63 | 0.86 | 0.61 | 1.67 | 1.98 |

6 | 0.97 | 0.94 | 0.96 | 0.94 | 1.85 | 2.00 | 30 | 0.87 | 0.64 | 0.87 | 0.61 | 1.64 | 1.98 |

7 | 0.96 | 0.94 | 0.96 | 0.94 | 1.86 | 2.01 | 31 | 0.87 | 0.64 | 0.87 | 0.61 | 1.59 | 1.98 |

8 | 0.96 | 0.94 | 0.96 | 0.93 | 1.87 | 2.01 | 32 | 0.86 | 0.64 | 0.86 | 0.60 | 1.57 | 1.98 |

9 | 0.96 | 0.36 | 0.96 | 0.93 | 1.88 | 2.00 | 33 | 0.86 | 0.61 | 0.86 | 0.58 | 1.51 | 1.98 |

10 | 0.96 | 0.93 | 0.96 | 0.93 | 1.88 | 2.01 | 34 | 0.85 | 0.59 | 0.85 | 0.57 | 1.44 | 1.99 |

11 | 0.95 | 0.92 | 0.95 | 0.92 | 1.84 | 1.99 | 35 | 0.84 | 0.56 | 0.84 | 0.53 | 1.40 | 1.99 |

12 | 0.94 | 0.92 | 0.94 | 0.91 | 1.78 | 2.00 | 36 | 0.80 | 0.52 | 0.80 | 0.49 | 1.38 | 2.00 |

13 | 0.93 | 0.90 | 0.93 | 0.89 | 1.80 | 2.00 | 37 | 0.80 | 0.52 | 0.80 | 0.49 | 1.35 | 2.00 |

14 | 0.93 | 0.88 | 0.92 | 0.97 | 1.78 | 1.99 | 38 | 0.84 | 0.60 | 0.84 | 0.58 | 1.33 | 2.00 |

15 | 0.92 | 0.85 | 0.92 | 0.84 | 1.65 | 1.99 | 39 | 0.86 | 0.65 | 0.86 | 0.63 | 1.33 | 2.00 |

16 | 0.90 | 0.78 | 0.89 | 0.76 | 1.67 | 1.98 | 40 | 0.87 | 0.68 | 0.87 | 0.66 | 1.33 | 2.00 |

17 | 0.88 | 0.73 | 0.88 | 0.71 | 1.69 | 1.99 | 41 | 0.88 | 0.72 | 0.88 | 0.70 | 1.17 | 1.99 |

18 | 0.87 | 0.69 | 0.86 | 0.67 | 1.73 | 1.99 | 42 | 0.90 | 0.74 | 0.90 | 0.73 | 1.43 | 1.99 |

19 | 0.86 | 0.65 | 0.85 | 0.63 | 1.74 | 2.00 | 43 | 0.91 | 0.77 | 0.91 | 0.75 | 1.45 | 1.99 |

20 | 0.86 | 0.64 | 0.85 | 0.61 | 1.71 | 2.00 | 44 | 0.91 | 0.77 | 0.91 | 0.76 | 1.45 | 2.00 |

21 | 0.85 | 0.61 | 0.84 | 0.59 | 1.69 | 2.00 | 45 | 0.91 | 0.78 | 0.91 | 0.77 | 1.48 | 2.00 |

22 | 0.84 | 0.59 | 0.83 | 0.56 | 1.70 | 2.00 | 46 | 0.92 | 0.79 | 0.92 | 0.78 | 1.46 | 1.99 |

23 | 0.84 | 0.60 | 0.83 | 0.57 | 1.74 | 1.99 | 47 | 0.92 | 0.80 | 0.92 | 0.79 | 1.46 | 2.00 |

Methods with AR(p) | MAPE(%) | Time Elapse (Sec) |
---|---|---|

OLS | 1.97 | 47.00 |

GLSAR-1 | 1.92 | 223.00 |

GLSAR-2 | 1.94 | 247.00 |

GLSAR-3 | 1.92 | 223.00 |

GLSAR-4 | 1.90 | 223.00 |

GLSAR-5 | 1.90 | 230.00 |

GLSAR-6 | 1.90 | 232.00 |

GLSAR-7 | 1.88 | 241.00 |

Nos of Hidden Layers | Nos of Neurons | MAPE (%) | Epochs |
---|---|---|---|

1 | $128\times 1$ | 2.85 | 8000 |

$64\times 1$ | 2.82 | 15,000 | |

$32\times 1$ | 2.92 | 22,000 | |

2 | $16\times 1$ | 2.78 | 45,000 |

$8\times 1$ | 3.18 | 55,000 | |

$4\times 1$ | 3.06 | 75,000 | |

$2\times 1$ | 38.79 | 100,000 | |

$128\times 128$ | 2.78 | 3000 | |

$128\times 64$ | 2.72 | 4000 | |

$\mathbf{64}\times \mathbf{64}$ | 2.72 | 2500 | |

3 | $64\times 64\times 64$ | 2.75 | 3000 |

4 | $64\times 64\times 64\times 64$ | 2.92 | 2500 |

Pre_Intervals | MAPE% | Exe Time (Sec.) | ||
---|---|---|---|---|

Simple OLS | FF-ANN | Simple OLS | FF-ANN | |

1 | 2.03 | 2.89 | 2.96 | 273 |

2 | 2.03 | 2.69 | 3.22 | 539 |

3 | 2.03 | 2.70 | 3.04 | 800 |

4 | 2.05 | 2.65 | 3.20 | 1062 |

5 | 2.05 | 2.56 | 3.40 | 1585 |

239 | 1.97 | NA | 33.4 | NA |

Simple OLS (MAPE%) | GLSAR (MAPE%) | |||||||
---|---|---|---|---|---|---|---|---|

Parameters | Scenario 1 | Scenario 2 | Scenario 3 | Scenario 4 | Scenario 1 | Scenario 2 | Scenario 3 | Scenario 4 |

Mon | 1.99 | NaN | 19.21 | 4.37 | 2.10 | NaN | 18.47 | 4.70 |

Tue | 1.72 | NaN | 15.49 | 2.54 | 1.84 | NaN | 22.96 | 2.89 |

Wed | 1.96 | NaN | 12.49 | 2.44 | 1.99 | NaN | 25.85 | 2.96 |

Thu | 1.78 | NaN | 26.01 | 2.32 | 1.9 | NaN | 30.44 | 2.86 |

Fri | 1.59 | NaN | 11.28 | 2.18 | 1.55 | NaN | 23.97 | 2.87 |

Sat | NaN | 1.76 | NaN | 3.17 | NaN | 1.77 | NaN | 2.69 |

Sun | NaN | 1.77 | NaN | 3.62 | NaN | 1.69 | NaN | 2.54 |

Holiday | NaN | NaN | 14.73 | 9.62 | NaN | NaN | 21.00 | 10.22 |

Holiday 2 | NaN | NaN | 16.62 | 10.92 | NaN | NaN | 22.58 | 9.68 |

Songkran | NaN | NaN | 13.54 | 11.42 | NaN | NaN | 18.48 | 14.76 |

Newyear | NaN | NaN | 26.35 | 17.85 | NaN | NaN | 30.48 | 17.96 |

DayAfterHoliday | 5.13 | NaN | NaN | 5.22 | 4.97 | NaN | NaN | 7.06 |

DayAfterLongHoliday | 7.38 | NaN | NaN | 2.73 | 6.41 | NaN | NaN | 6.59 |

DayAfterSongkran | 3.10 | NaN | NaN | 2.03 | 1.82 | NaN | NaN | 6.09 |

DayAfterNewyear | 15.77 | NaN | NaN | 12.18 | 13.82 | NaN | NaN | 6.46 |

DayAfterLabor | 3.39 | NaN | NaN | 5.14 | 4.85 | NaN | NaN | 9.26 |

DayAfterReligion | 5.54 | NaN | NaN | 5.54 | 5.18 | NaN | NaN | 6.04 |

Overall MAPE | 1.81 | 1.77 | 16.63 | 2.95 | 1.88 | 1.74 | 22.59 | 3.08 |

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Chapagain, K.; Kittipiyakul, S.; Kulthanavit, P. Short-Term Electricity Demand Forecasting: Impact Analysis of Temperature for Thailand. *Energies* **2020**, *13*, 2498.
https://doi.org/10.3390/en13102498

**AMA Style**

Chapagain K, Kittipiyakul S, Kulthanavit P. Short-Term Electricity Demand Forecasting: Impact Analysis of Temperature for Thailand. *Energies*. 2020; 13(10):2498.
https://doi.org/10.3390/en13102498

**Chicago/Turabian Style**

Chapagain, Kamal, Somsak Kittipiyakul, and Pisut Kulthanavit. 2020. "Short-Term Electricity Demand Forecasting: Impact Analysis of Temperature for Thailand" *Energies* 13, no. 10: 2498.
https://doi.org/10.3390/en13102498