Dynamic Regression Prediction Models for Customer Specific Electricity Consumption
Abstract
:1. Introduction
2. Modelling Time Series with Trend and Seasonality
2.1. Time Series Regression Models
2.1.1. Fourier Series
2.1.2. Loess Smoother
- 1.
- For each i, define the weights depending on the distance of to , and fit a polynomial of low degree d (often ) by solving the weighted least-squares problem
- 2.
- With the just obtained weights define the estimator
- 3.
- Check the residuals , define a robustness weight that relates to the median of the and compute new estimates via the steps 1 and 2, but with the weights
2.2. Time Series Decomposition
- 1.
- A detrended series is computed;
- 2.
- In the second step, the cycle-subseries are formed and smoothed on the detrended series using Loess with and . For example, for a monthly series with a yearly seasonality , the first subseries consists of the January values, the second is the February values, and so on. The collection of smoothed values for the entire cycle-subseries is a temporary seasonal series, ;
- 3.
- A low-pass filter is applied into the smoothed cycle-subseries and consists of the three moving averages followed one by one, where the two first moving averages have a length of , while the last has a length of 3. In the end, a Loess smoothing with and is applied, and the output is defined as ;
- 4.
- The seasonal component from the st loop is ;
- 5.
- A deseasonalized series is computed;
- 6.
- In the last step, the trend component is estimated using the deseasonalized series and smoothing them with and and is given by .
- Unlike SEATS and X11, STL can handle any type of seasonality, not only monthly and quarterly data;
- The seasonal component is allowed to change over time, and the rate of change can be controlled by the user;
- The smoothness of the trend-cycle can also be controlled by the user;
- It can be robust to outliers (i.e., the user can specify a robust decomposition), so that occasional unusual observations will not affect the estimates of the trend-cycle and seasonal components. They will, however, affect the remainder component;
- The implementation of the STL procedure is based purely on numerical methods and does not require any mathematical modelling.
2.3. Seasonal ARIMA Models
2.4. Dynamic Regression Models
2.5. Average Method
3. Predicting Electricity Consumption: Conceptual Issues
- Are dynamic regression models capable of modelling electricity consumption data and generating acceptable forecasts?
- When is it hard to beat the average-method-variant?
- When does at least the best method perform well?
- When does no method perform well?
3.1. Data Exploration and Analysis
3.2. Research Design
3.2.1. Dynamic Harmonic Regression
3.2.2. Dynamic Regression Model with STL Decomposition
4. Numerical Results
4.1. Dynamic Harmonic Regression
4.2. Dynamic Regression with STL
- Residuals: Min 1Q Median 3Q Max
- −2649.28 −341.56 3.91 347.19 2328.30
- Coefficients:
- Estimate Std. Error t value Pr(>|t|)
- (Intercept) −1.297428 7.999 × 10 −5.071 3.96 × 10−07 ***
- STL_t 1.088 × 10 1.747 × 10−02 62.286 < 2 × 10−16 ***
- STL_p1 1.077 × 10 5.120 × 10−03 210.315 < 2 × 10−16 ***
- STL_p2 1.028 × 10 2.901 × 10−03 354.392 < 2 × 10−16 ***
- STL_p3 1.124 × 10 6.259 × 10−03 179.633 < 2 × 10−16 ***
- STL_p4 1.010 × 10 2.268 × 10−03 445.197 < 2 × 10−16 ***
- ---
- Signif. codes: 0 *** 0.001 ** 0.01 * 0.05 . 0.1 1
- Residual standard error: 516.5 on 102233 degrees of freedom
- Multiple R-squared: 0.8041, Adjusted R-squared: 0.8041
- F-statistic: 8.393 × 104 on 5 and 102233 DF, p-value: < 2.2 × 10−16
4.3. Comparison of the Model Performances
5. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Acknowledgments
Conflicts of Interest
Appendix A
Data | Model | RMSE | MAPE | MAE |
---|---|---|---|---|
“” | Average Method | 759.20 | 25.74 | 535.56 |
Dynamic Regression with STL | 712.53 | 20.19 | 465.83 | |
“” | Average Method | 394.81 | 8.52 | 314.64 |
Dynamic Regression with STL | 296.42 | 6.33 | 236.57 | |
“” | Average Method | 295.23 | 11.92 | 223.94 |
Dynamic Regression with STL | 283.24 | 12.15 | 246.71 | |
“” | Average Method | 869.86 | 13.53 | 724.98 |
Dynamic Regression with STL | 803.29 | 12.23 | 649.67 | |
“” | Average Method | 1312.93 | 27.15 | 1100.95 |
Dynamic Regression with STL | 1276.12 | 26.48 | 1081.26 | |
“” | Average Method | 145.23 | 5.93 | 105.05 |
Dynamic Regression with STL | 131.7767 | 5.86 | 107.68 | |
“” | Average Method | 215.02 | 9.63 | 171.34 |
Dynamic Regression with STL | 191.2 | 9.02 | 156.37 | |
“” | Average Method | 802.68 | 84.03 | 640.13 |
Dynamic Regression with STL | 610.43 | 68.01 | 492.64 |
References
- Ozturk, S.; Ozturk, F. Forecasting energy consumption of Turkey by Arima model. J. Asian Sci. Res. 2018, 8, 52–60. [Google Scholar] [CrossRef] [Green Version]
- Fan, S.; Hyndman, J.R. Short-term load forecasting based on a semi-parametric additive model. IEEE Trans. Power Syst. 2011, 27, 134–141. [Google Scholar] [CrossRef] [Green Version]
- Huang, S.-J.; Shih, K.-R. Short-term load forecasting via ARMA model identification including non-Gaussian process considerations. IEEE Trans. Power Syst. 2003, 18, 673–679. [Google Scholar] [CrossRef] [Green Version]
- Taylor, J.W. Short-term electricity demand forecasting using double seasonal exponential smoothing. J. Oper. Res. Soc. 2003, 54, 799–805. [Google Scholar] [CrossRef]
- Weron, R. Modeling and Forecasting Electricity Loads and Prices: A Statistical Approach; John Wiley & Sons: Hoboken, NJ, USA, 2007; p. 403. [Google Scholar]
- Amjady, N. Day-ahead price forecasting of electricity markets by a new fuzzy neural network. IEEE Trans. Power Syst. 2009, 21, 887–896. [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]
- Zhang, Y.; Zhou, Q.; Sun, C.; Lei, S.; Liu, Y.; Song, Y. RBF neural network and ANFIS-based short-term load forecasting approach in real-time price environment. IEEE Trans. Power Syst. 2008, 23, 853–858. [Google Scholar] [CrossRef]
- Lalmuanawma, S.; Hussain, J.; Chhakchhuak, L. Applications of machine learning and artificial intelligence for COVID-19 (SARS-CoV-2) pandemic: A review. Chaos Solitons Fractals 2020, 139, 110059. [Google Scholar] [CrossRef] [PubMed]
- Pugliese, R.; Regondi, S.; Marini, R. Machine learning-based approach: Global trends, research directions, and regulatory standpoints. Data Sci. Manag. 2021, 4, 19–29. [Google Scholar] [CrossRef]
- Zhao, E.; Sun, S.; Wang, S. New developments in wind energy forecasting with artificial intelligence and big data: A scientometric insight. Data Sci. Manag. 2022, 5, 84–95. [Google Scholar] [CrossRef]
- Tarsitano, A.; Amerise, I. Short-term load forecasting using a two-stage sarimax model. Energy 2017, 133, 108–114. [Google Scholar] [CrossRef]
- Badri, M.; Al-Mutawa, A.; Davis, D.; Davis, D. EDSSF: A decision support system (DSS) for electricity peak-load forecasting. Energy 1997, 22, 579–589. [Google Scholar] [CrossRef]
- Wang, E.; Galjanic, T.; Johnson, R. Short-term electric load forecasting at Southern California Edison. In Proceedings of the 2012 IEEE Power And Energy Society General Meeting, San Diego, CA, USA, 22–26 July 2012; pp. 1–3. [Google Scholar]
- Mishra, S.; Shaik, A.G. Performance Evaluation of Prophet and STL-ETS methods for Load Forecasting. In Proceedings of the 2022 IEEE India Council International Subsections Conference (INDISCON), Bhubaneswar, India, 15–17 July 2022; pp. 1–6. [Google Scholar]
- Tian, Y.; Zhou, S.; Wen, M.; Li, J. A short-term electricity forecasting scheme based on combined GRU model with STL decomposition. In IOP Conference Series: Earth and Environmental Science; IOP Publishing: Bristol, UK, 2021; Volume 701, p. 012008. [Google Scholar]
- Lu, Q.; Cai, Q.; Liu, S.; Yang, Y.; Yan, B.; Wang, Y.; Zhou, X. Short-term load forecasting based on load decomposition and numerical weather forecast. In Proceedings of the 2017 IEEE Conference on Energy Internet and Energy System Integration (EI2), Beijing, China, 26–28 November 2017; pp. 1–5. [Google Scholar]
- Permata, R.; Prastyo, D. Others Hybrid dynamic harmonic regression with calendar variation for Turkey short-term electricity load forecasting. Procedia Comput. Sci. 2022, 197, 25–33. [Google Scholar] [CrossRef]
- Royston, J.P. An extension of Shapiro and Wilks W test for normality to large samples. J. R. Stat. Soc. Ser. C Appl. Stat. 1982, 31, 115–124. [Google Scholar]
- Cleveland, W.S. Robust locally weighted regression and smoothing scatterplots. J. Am. Stat. Assoc. 1979, 74, 829–836. [Google Scholar] [CrossRef]
- Cleveland, R.B.; Cleveland, W.S.; McRae, J.E.; Terpenning, I. STL: A seasonal-trend decomposition procedure based on loess. J. Off. Stat. 1990, 6, 3–73. [Google Scholar]
- Persons, W. Indices of Business Conditions: An Index of General Business Conditions; Harvard University Press: Harvard, MA, USA, 1919; pp. 5–107. [Google Scholar]
- Shiskin, J.; Young, A.H.; Musgrave, J.C. The X-11 Variant of the Census Method II consumption Seasonal Adjustment Program; Technical Report #15; Bureau of Economic Analysis: Suitland, MD, USA, 1967. [Google Scholar]
- Brian, M. The X-13A-S seasonal adjustment program. In Proceedings of the 2007 Federal Committee On Statistical Methodology Research Conference, Washington, DC, USA, 5–7 November 2007. [Google Scholar]
- Theodosiou, M. Forecasting monthly and quarterly time series using STL decomposition. Int. J. Forecast. 2011, 27, 1178–1195. [Google Scholar] [CrossRef]
- Hyndman, R.J.; Athanasopoulos, G. Forecasting: Principles and Practice; OTexts: Melbourne, Australia, 2018. [Google Scholar]
- Geurts, M. Time Series Analysis: Forecasting and Control San Francisco; Prentice Hall: Hoboken, NJ, USA, 1976. [Google Scholar]
- Deutscher Wetterdienst. Available online: https://www.dwd.de (accessed on 3 March 2021).
Data | Model | RMSE | MAPE | MAE |
---|---|---|---|---|
“” | Average Method | 526.37 | 52.99 | 395.51 |
Dynamic Harmonic Regression | 658.13 | 50.65 | 536.50 | |
Dynamic Regression with STL | 502.12 | 40.52 | 365.18 | |
“” | Average Method | 1301.08 | 89.18 | 811.46 |
Dynamic Harmonic Regression | 1099.71 | 78.34 | 725.62 | |
Dynamic Regression with STL | 953.37 | 62.12 | 721.95 | |
“” | Average Method | 1138.69 | 366.40 | 908.28 |
Dynamic Harmonic Regression | 813.07 | 280.83 | 753.46 | |
Dynamic Regression with STL | 403.32 | 74.34 | 302.41 | |
“” | Average Method | 891.80 | 35.19 | 675.77 |
Dynamic Harmonic Regression | 1455.22 | 70.84 | 1263.73 | |
Dynamic Regression with STL | 1095.55 | 23.65 | 769.45 | |
“” | Average Method | 1374.82 | 93.59 | 966.60 |
Dynamic Harmonic Regression | 1290.3 | 98.48 | 1017.46 | |
Dynamic Regression with STL | 1250 | 48.33 | 865.28 | |
“” | Average Method | 244.05 | 13.55 | 176.53 |
Dynamic Harmonic Regression | 253.86 | 14.49 | 203.80 | |
Dynamic Regression with STL | 200.88 | 12.03 | 161.64 | |
“” | Average Method | 672.27 | 173.25 | 548.95 |
Dynamic Harmonic Regression | 402.88 | 96.73 | 324.12 | |
Dynamic Regression with STL | 354.39 | 48.57 | 270.09 | |
“” | Average Method | 746.45 | Inf | 561.90 |
Dynamic Harmonic Regression | 838.06 | Inf | 705.59 | |
Dynamic Regression with STL | 723.57 | Inf | 580.34 |
Data | Model | RMSE | MAPE | MAE |
---|---|---|---|---|
“” | Average Method | 810.40 | 154.53 | 715.87 |
Dynamic Harmonic Regression | 467.06 | 86.48 | 370.42 | |
Dynamic Regression with STL | 365.18 | 49.90 | 256.27 | |
“” | Average Method | 2592.55 | 357.94 | 2512.66 |
Dynamic Harmonic Regression | 2130.54 | 303.97 | 1914.59 | |
Dynamic Regression with STL | 2044.07 | 280.74 | 1913.46 | |
“” | Average Method | 1396.29 | 694.94 | 1266.59 |
Dynamic Harmonic Regression | 884.68 | 478.95 | 867.47 | |
Dynamic Regression with STL | 116.17 | 58.51 | 104.95 | |
“” | Average Method | 527.80 | 53.55 | 450.46 |
Dynamic Harmonic Regression | 1136.01 | 103.28 | 1000.73 | |
Dynamic Regression with STL | 225.77 | 13.86 | 163.78 | |
“” | Average Method | 2179.19 | 310.30 | 1645.88 |
Dynamic Harmonic Regression | 1742.89 | 283.96 | 1507.42 | |
Dynamic Regression with STL | 158.46 | 22.99 | 124.11 | |
“” | Average Method | 379.64 | 28.96 | 323.32 |
Dynamic Harmonic Regression | 277.45 | 21.48 | 228.75 | |
Dynamic Regression with STL | 171.56 | 13.97 | 152.21 | |
“” | Average Method | 647.05 | 287.58 | 591.61 |
Dynamic Harmonic Regression | 486.13 | 183.64 | 385.78 | |
Dynamic Regression with STL | 41.55 | 16.80 | 34.62 | |
“” | Average Method | 869.15 | 201.53 | 713.78 |
Dynamic Harmonic Regression | 828.33 | 340.48 | 731.67 | |
Dynamic Regression with STL | 700.69 | 76.51 | 452.43 |
Data | Model | RMSE | MAPE | MAE |
---|---|---|---|---|
“” | Average Method | 475.24 | 132.80 | 448.74 |
Dynamic Harmonic Regression | 686.95 | 141.99 | 510.26 | |
Dynamic Regression with STL | 149.91 | 36.69 | 126.49 | |
“” | Average Method | 2678.93 | 478.91 | 2622.51 |
Dynamic Harmonic Regression | 1475.34 | 253.16 | 1364.24 | |
Dynamic Regression with STL | 537.83 | 88.02 | 481.78 | |
“” | Average Method | 1365.10 | 699.81 | 1241.89 |
Dynamic Harmonic Regression | 104.87 | 50.17 | 89.20 | |
Dynamic Regression with STL | 91.96 | 46.28 | 81.54 | |
“” | Average Method | 230.65 | 16.49 | 213.28 |
Dynamic Harmonic Regression | 1149.15 | 80.77 | 1028.17 | |
Dynamic Regression with STL | 247.50 | 15.38 | 200.62 | |
“” | Average Method | 873.30 | 149.50 | 796.78 |
Dynamic Harmonic Regression | 1339.25 | 177.18 | 938.68 | |
Dynamic Regression with STL | 114.59 | 17.67 | 93.82 | |
“” | Average Method | 220.30 | 20.57 | 210.58 |
Dynamic Harmonic Regression | 366.04 | 31.05 | 303.20 | |
Dynamic Regression with STL | 107.33 | 8.47 | 85.51 | |
“” | Average Method | 395.63 | 168.81 | 339.67 |
Dynamic Harmonic Regression | 428.61 | 173.86 | 349.50 | |
Dynamic Regression with STL | 49.28 | 23.71 | 47.93 | |
“” | Average Method | 727.85 | 37.68 | 665.43 |
Dynamic Harmonic Regression | 1780.46 | 91.34 | 1721.95 | |
Dynamic Regression with STL | 669.09 | 37.08 | 645.56 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2023 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 (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Shaqiri, F.; Korn, R.; Truong, H.-P. Dynamic Regression Prediction Models for Customer Specific Electricity Consumption. Electricity 2023, 4, 185-215. https://doi.org/10.3390/electricity4020012
Shaqiri F, Korn R, Truong H-P. Dynamic Regression Prediction Models for Customer Specific Electricity Consumption. Electricity. 2023; 4(2):185-215. https://doi.org/10.3390/electricity4020012
Chicago/Turabian StyleShaqiri, Fatlinda, Ralf Korn, and Hong-Phuc Truong. 2023. "Dynamic Regression Prediction Models for Customer Specific Electricity Consumption" Electricity 4, no. 2: 185-215. https://doi.org/10.3390/electricity4020012
APA StyleShaqiri, F., Korn, R., & Truong, H. -P. (2023). Dynamic Regression Prediction Models for Customer Specific Electricity Consumption. Electricity, 4(2), 185-215. https://doi.org/10.3390/electricity4020012