Abstract
Electricity spot prices are complex processes characterized by nonlinearity and extreme volatility. Previous work on nonlinear modeling of electricity spot prices has shown encouraging results, and we build on this area by proposing an Expectation Maximization algorithm for maximum likelihood estimation of recurrent neural networks utilizing the Kalman filter and smoother. This involves inference of both parameters and hyper-parameters of the model which takes into account the model uncertainty and noise in the data. The Expectation Maximization algorithm uses a forward filtering and backward smoothing (Expectation) step, followed by a hyper-parameter estimation (Maximization) step. The model is validated across two data sets of different power exchanges. It is found that after learning a posteriori hyper-parameters, the proposed algorithm outperforms the real-time recurrent learning and the extended Kalman Filtering algorithm for recurrent networks, as well as other contemporary models that have been previously applied to the modeling of electricity spot prices.
Similar content being viewed by others
Notes
The data is freely available at http://www.ieso.ca/.
The data was downloaded from http://www.omel.es.
References
Amjady N (2006) Day-ahead price forecasting of electricity markets by a fuzzy neural network. IEEE Trans Power Syst 21(2)
Amjady N, Hemmati M (2008) Day-ahead price forecasting of electricity markets by a hybrid intelligent system. Eur Trans Elec Power
Aggarwal SK, Sani LM, Kumar A (2008) Electricity price forecasting in Ontario electricity market using wavelet transform in artificial neural network based model. Int J Control Auto Syst 6(5):639–650
Bunn D (2004) Modelling prices in competitive electricity markets. Wiley, London
Catalao JPS, Mariano SJPS, Mendes VMF, Ferreira LAFM (2007) Short-term electricity prices forecasting in a competitive market: a neural network approach. Elec Power Syst Res 77(10):1297–1304
Casdagli M (1989) Nonlinear prediction of chaotic time series. Physica D 35:335–356
Cernansky M, Benuskova L (2003) Simple recurrent network trained by RTRL and extended Kalman filter algorithms. Neural Network World 13(3):223–234
Contreras J, Conejo AJ, Espinola R (2002) Forecasting next-day prices by time-series models. IEEE Trans Power Syst 17(2):342–348
Contreras J, Espínola R, Nogales FJ, Conejo AJ (2003) ARIMA mdels to predict next-day electricity prices. IEEE Trans Power Syst 18(3):1014–1020
Conejo AJ, Plazas MA, Espinola R, Molina AB (2005) Day-ahead electricity price forecasting using the wavelet transform and ARIMA models. IEEE Trans Power Syst 20(2):1035–1042
de Freitas JFG, Niranjan M, Gee AH (2000) Dynamic learning with the EM algorithm for neural networks. J VlSI Signal Proc 26:119–131
Dempster AP, Laird NM, Rubi DB (1977) Mazimum likelihood from incomplete data via the EM algorithm. J R Stat Soc Ser B 39:1–38
Francisco NJ, Javier C (2002) Forecasting next-day electricity prices by time series models. IEEE Trans Power Syst 17(2):342–348
Gao F, Guan X, Cao XR, Papalexopoulos A (2000) Forecasting power market clearing price and quantity using a neural network. IEEE PES Summer Meeting Seattle
Georgilakis P (2006) Market clearing price forecasting in deregulated electricity markets using adaptively trained neural networks. SETN, pp 56–66
Ghahramani Z, Hinton GE (1996) Parameter estimation for linear dynamical systems University of Toronto Technical Report CRG-TR-96-2
Gonzalez AM, San Roque AM, Conzales JG (2005) Modeling and forecasting electricity prices with input/output hidden Markov models. IEEE Trans Power Syst 20(1):13–24
Haykin S (2001) Kalman filtering and neural networks. Wiley, New York
Hong YY, Lee CF (2005) A neuro-fuzzy price forecasting approach in deregulated electricity markets. Elec Power Syst Res 73:151–157
Hong YY, Hisao CY (2001) Locational marginal price forecasting in deregulated electric markets using a recurrent neural network. PESWM 2:539–544
Karsaz A, Mashhadi HR, Eshraghnia R (2006) Cooperative co-evolutionary approach to electricity load and price forecasting in deregulated electricity markets. Power India Conference, pp 1–6
Kim B, Velas JP, Lee J, Park J, Shin J, Lee KY (2006) Short-term system marginal price forecasting using system-type neural network architecture. PSCE, pp 1753–1758
Li C, Wang S (2006) Next-day power market clearing price forecasting using artificial fish-swarm based neural network. ISSN, pp 1290–1295
Liu Z, Yang H, Lai M (2005) Electricity price forecasting model based on chaos theory. IPEC, pp 1–5
Mandal P, Senjyu T, Urasaki N, Funabashi T, Srivastava AK (2007) A novel approach to forecast electricity price for PJM using neural network and similar days Method. IEEE Trans Power Syst 22(4):2058–2065
Mandal P, Senjyu T, Funabashi T (2006) Neural netowrks approach to forecast several hour ahead electricity prices and loads in deregulated market. Energ Conve Manag 47:15–16
Mori H, Awata A (2006) A hybrid method of clipping and artificial neural network for electricity pricew zone forecasting. PMAPS, pp 1–6
Niimura T, Ko HS, Ozawa K (2002) A day-ahead electricity price prediction based on a fuzzy-neuro autoregressive model in a deregulated electrictiy market. IJCNN, pp 1362–1366
Nikolaev N, de Menezes L (2008) Sequential Bayesian Kernel Modelling with non-gaussian noise. Neural Networks 21(1): 36–47
Nikolaev N, Iba H (2006) Adaptive learning of polynomial networks: genetic programming, backpropagation and Bayesian methods. Springer, New York
Nogales FJ, Contreras J, Conejo AJ, Espinola R (2002) Forecasting next-day electricity prices by time-series models. IEEE Trans Power Syst 17(2):342–348
Pindoriya NM, Singh SN, Singh SK (2008) An adaptive wavelet neural network-based energy price forecasting in electricity markets. IEEE Trans Power Syst 23(3):1423–1432
Pino R, Parreno J, Gomez A, Priore P (2008) Forecasting next-day price of electricity in the Spanish energy market using artificial neural networks. Eng Appl Artif Intell 21(1):53–62
Puskorius GV, Feldkamp LA (1994) Neurocontrol of nonlinear dynamical systems with Kalman filter trained recurrent networks. IEEE T Neural Networ 5(2):279–297
Rauch HE, Tung F, Striebel CT (1965) Maximum likelihood estimates of linear dynamic models. AIAA J 3(8):1445–1450
Rodriguez CP, Anders GJ (2006) Energy price forecasting in the Ontario competitive power system market. IEEE Trans Power Syst 21(2):887–896
Ruck DW, Rogers SK, Kabrisky M, Maybeck P, Oxle ME (1992) Comparative analysis of backpropgation and the extended Kalman filter for training multilayer perceptrons. IEEE Trans Patt Anal Mach Intell 14(6):686–691
Schafer AM, Zimmerman G (2007) Recurrent neural networks are universal approximators. Int J Neural Syst 17(4):253–263
Schottky B, Saad D (1999) Statistical mechanics of EKF learning in neural networks. J Phys A 32(9):1605–1621
Zhang L, Luh PB, Kasiviswanathan K (2003) Energy clearing price predication and confidence interval estimation with cascaded neural networks. IEEE Trans Power Syst 18(1):99–105
Shumway RH, Stoffer DS (1982) An approach to time series smoothing and forecasting using the EM algorithm. J Time Ser Anal 3(4):253–264
Szkuta BR, Sanabria LA, Dillon TS (1999) Electricity price short-term forecasting using artificial neural networks. IEEE Trans Power Syst 14(3):2116–2120
Wang A, Ramsay B (1997) Prediction of system marginal price in the UK power pool, vol 4, In: International conference on neural networks, pp 2116–2120
Williams RJ, Zipser D (1989) A learning algorithm for continuously running fully connected recurrent neural networks. Neural Comp 1:270–280
Srinivasan D, Yong FC, Liew AC (2007) Electricity price forecasting using evolved neural networks. ISAP, pp 1–7
Xu YY, Hsieh R, Lyu YL, Shen YC, Chuang SC, Pao HT (2004) Forecasting electricity market prices: a neural network based approach. IJCNN 4:2789–2794
Yamin HY, Shahidehpour SM, Li L (2004) Adaptive short-term electricty price forecasting using artificial neural networks in teh restructured power markets. Elec Power Energ Syst 25:571–581
Yang B, Chen Y, Zhao Z, Han Q (2007) Forecasting of market clearing price by using GA based neural network, in proc. ICIC 2:1278–1286
Yang H, Lai M (2005) Chaotic characteristics of electricity price and its forecasting model. Aus J Elec Elec Eng 2(2):117–125
Zareipour H, Canizares CA, Bhattacharya K (2006) Application of public-domain market information to forecast Ontario’s wholesale electricity prices. IEEE Trans Power Syst 21(4):1707–1717
Zhang L, Luh PB, Kasiviswanthan K (2003) Energy clearing price prediction and confidence interval estimation with cascaded neural networks. IEEE Trans Power Syst 18(1):99–105
Zhang L, Luh PB (2005) Neural network-based market clearing price prediction and confidence interval estimation with an improved extended Kalman filter method. IEEE Trans Power Syst 20(1):29–66
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Mirikitani, D., Nikolaev, N. Nonlinear maximum likelihood estimation of electricity spot prices using recurrent neural networks. Neural Comput & Applic 20, 79–89 (2011). https://doi.org/10.1007/s00521-010-0344-1
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00521-010-0344-1