At the past IEEE PES General Meeting in Vancouver, there was an interesting discussion on the noisiness of the energy forecasting community during our tutorial and panel sessions. First my colleague and friend Prof. Pierre Pinson mentioned that the current wind forecasting research community is too "noisy"; there are too many wind forecasting papers being published every year; he (or maybe a student of his) had to read 450 papers to write the literature review for the PhD dissertation.
450 papers, FOUR HUNDRED AND FIFTY!
"You were too lucky," I made a comment, "I had to read 1200 papers on load forecasting to get my PhD!"
I often felt sick and desperate reading those papers, because most of them were produced just for the sake of meeting the publication requirement to get an advanced degree, a tenure or a promotion. Nicely speaking, those papers were "on the theoretical level without much practical value". In reality, they were, are and will be useless to everybody other than the students who got some degrees out of them or the professors who got tenured or promoted.
On the other hand, I felt really excited when reading a high quality paper. Unfortunately, among the 1200+ load forecasting papers I read, there were only a dozen or two high quality papers.
Nowadays, there are hundreds of papers on load forecasting published every year. If you count energy forecasting papers that include forecasting load, price, wind and solar, you are looking at thousands of papers per year. For the ones who are lucky enough to not have to read 1200 papers, they may want to ask a question:
What shall I read in such a noisy world?
With a miserable experience reading and tracking literature, I would love to share with you my recommended reading list. Hopefully this list can offer a shortcut for the energy forecasters.
Books and thesis
As an energy forecaster, you may not need time series analysis, but you must have a solid background in basic mathematics and statistics to appreciate regression analysis.
- Carl D. Meyer (2000) Matrix Analysis and Applied Linear Algebra. SIAM.
- R. Lyman Ott and Michael T. Longnecker (2008) An Introduction to Statistical Methods and Data Analysis. Cengage Learning.
- Michael Kutner, Christopher Nachtsheim, John Neter and William Li (2004) Applied Linear Statistical Models. McGraw-Hill/Irwin.
Then you can move on to time series analysis by reading the three books below in the order being listed. The first book by Wei is a highly readable book in time series analysis. It was recommended by my professor David Dickey to me. And I'm now recommending it to you guys. The second one by Box, Jenkins and Reinsel is the classical time series analysis book and a must read. The third one by Brocklebank and Dickey is a perfect handbook with good mix of theory and practice for those who are using SAS as the tool for time series forecasting. It is primarily devoted to PROC ARIMA of SAS.
- William W. S. Wei (2005) Time Series Analysis: Univariate and Multivariate Methods. Addison Wesley.
- George Box, Gwilym M. Jenkins, and Gregory Reinsel (1994) Time Series Analysis: Forecasting and Control. Prentice Hall.
- John C. Brocklebank, and David A. Dickey (2003) SAS for Forecasting Time Series. SAS Publishing.
When it comes to forecasting, you have to read this online textbook, not because it's online and free, but because it is really good.
The ones below are on energy forecasting. The book by my former boss Willis is super readable and practical. If you are a planner, that book is a must read. My MS thesis on spatial load forecasting is a recent advancement of the methodology in Willis' book. Weron's book was from his dissertation and is a little academic, but it offers a good overview of statistical approaches to forecasting load and price. My dissertation on short term load forecasting is also being sold on Amazon. You don't have to buy it, because the pdf is available through NCSU library. I'm still working on my book. I don't know when I will finish it, maybe next year, but I'm sure it will be a bible for load forecasters.
- H. Lee Willis (2002) Spatial Electric Load Forecasting. CRC Press.
- Tao Hong (2008) Long Term Spatial Load Forecasting. M.S. thesis, NCSU.
- Rafal Weron (2006) Modeling and Forecasting Electricity Loads and Prices: A Statistical Approach. Wiley.
- Tao Hong (2010) Short Term Electric Load Forecasting. Ph.D. dissertation, NCSU.
- Tao Hong and David A. Dickey (forthcoming) Electric Load Forecasting: Fundamentals and Best Practices.
As I mentioned earlier, there are not many high quality papers to read. I list some really good load forecasting papers below.
Very short term load forecasting:
- James W. Taylor, Lilian M. de Menezes, Patrick E. McSharry (2006) A comparison of univariate methods for forecasting electricity demand up to a day ahead. International Journal of Forecasting.
Short term load forecasting (In my PhD dissertation, there is a fairly comprehensive literature review of short term load forecasting work prior to 2010):
- Shu Fan and Rob J. Hyndman (2012) Short-term load forecasting based on a semi-parametric additive model. IEEE Transactions on Power Systems.
- Shu Fan, K. Methaprayoon, and Wei-Jen Lee (2009) Multiregion load forecasting for system with large geographical area. IEEE Transactions on Industry Applications.
- H. S. Hippert, C. E. Pedriera and R. C. Souza (2002) Neural networks for short-term load forecasting: A review and evaluation. IEEE Transactions on Power Systems.
Medium term load forecasting:
- Bo-Juen Chen, Ming-Wei Chang and Chih-Jen Lin (2004) Load forecasting using support vector Machines: a study on EUNITE competition 2001. IEEE Transactions on Power Systems
Long term load forecasting:
- Tao Hong, Jason Wilson and Jingrui Xie (2014) Long term probabilistic load forecasting with hourly information. IEEE Transactions on Smart Grid.
- Rob J. Hyndman and Shu Fan (2010) Density forecasting for long-term peak electricity demand. IEEE Transactions on Power Systems.
I read the above books, thesis and papers word by word and found them really helpful. The first half of this list applicable to energy forecasters in general, while the second half is mostly on load forecasting. If you find other books and papers helpful, please list them in the comments below or let me know via email.