An Accurate and Fast Converging Short-Term Load Forecasting Model for Industrial Applications in a Smart Grid

Abstract:

Short-term load forecasting (STLF) models are very important for electric industry in the trade of energy. These models have many applications in the day-to-day operations of electric utility(ies) like energy generation planning, load switching, energy purchasing, infrastructure maintenance and contract evaluation. A large variety of STLF models have been developed which trade-off between forecast accuracy and convergence rate. This paper presents an accurate and fast converging STLF (AFC-STLF) model for industrial applications in a smart grid. In order to improve the forecast accuracy, modifications are devised in two popular techniques: (1) mutual information (MI) based feature selection, and (2) enhanced differential evolution (EDE) algorithm based error minimization. On the other hand, the convergence rate of the overall forecast strategy is enhanced by devising modifications in the heuristic algorithm and in the training process of the artificial neural network (ANN). Simulation results show that accuracy of the newly proposed forecast model is 99.5% with moderate execution time, i.e., we have decreased the average execution of the existing Bi-level forecast strategy by 52.38%.

 

Existing System:

 

In order to optimize the performance of SG, especially its distribution part, a decision making entity is needed. Where, proper decision making leads to reduction in the electricity cost of end user(s) along with minimization of total power losses and alleviation of peak to average ratio [19]. Keeping these objectives in mind, the current research in SGs majorly focuses on the optimization techniques of power scheduling [20], [21], [22]. However, prior to scheduling, an accurate STLF model is needed to properly plan the ongoing grid operations subject to efficient management of resources. High randomness and non linearity in history load curves make the STLF highly challenging. In the literature, many STLF models have been presented [1], [10], [11], [15], [16], however, accuracy of these models is either not satisfactory or their convergence rate is slow. For example, [12] uses a hybrid ANN based approach to increase the forecast accuracy, however, in doing so the complexity of the overall strategy is increased in terms of implementation which causes its convergence rate to decrease. In another ANN based work [10], the convergence rate is improved by paying the cost of increased forecast error.

 

Proposed System:

 

The proposed methodology consists of three modules: feature selector, forecaster, and optimizer. At first, the candidate inputs (load time series) are given to the feature selection module, which uses MI (entropy) based technique to rank the candidate inputs. The ranked candidate inputs are then passed through two filters; one to remove redundancy and the other to remove irrelevancy. The selected inputs are given to the forecast module (note: this module consists of ANN). Where, from the selected inputs, training and validation samples are constructed based on previously observed data. These training and validation samples are given to the ANN which forecasts load of the next day and gives it to the optimization module. At first, the optimization module calculates error signal between target and the forecast ones. Then, using iteration based mEDE algorithm, the optimizer minimizes this error.

 

Conclusion:

 

The proposed AFC-STLF model forecasts next day’s load on the basis of lagged input samples till the current day. For every next day the model needs information till the current day. Thus, the prediction model never fails provided the lagged input samples till the current day. However, the proposed model is not designed to forecast two or more days ahead load. In the feature selection process, downsized input features significantly decreased the ANN training time; information loss must be avoided here. The MI+ANN forecast strategy achieved a faster convergence rate, however, at the cost of accuracy. Similarly, the Bi-level forecast strategy minimized the forecast error, however, at the cost of increased execution time. In order to overcome the indicated trade-off, we have proposed AFC-STLF model for SGs. The newly proposed AFCSTLF model achieved approximately 99.5% accuracy which is better than the existing Bi-level (97.6%) and MI+ANN (95.9%) based forecast strategies, respectively. The AFC-STLF model has decreased the average execution time of the existing Bi-level forecast strategy by 52.38%. We also conclude that these results provide justification of the correctness of our modifications in the selected modules.

 

References:

 

[1] L. Hernandez, C. Baladrn, J. M. Aguiar, B. Carro, A. J. Sanchez- Esguevillas, and J. Lloret, “Short-Term Load Forecasting for Microgrids Based on Artificial Neural Networks”, Energies, vol. 2013, iss. 6, pp. 1385–1408, 2013.

 

[2] L. Gelazanskas, and K. A. Gamage, “Demand side management in smart grid: A review and proposals for future direction”, Sustainable Cities and Society, vol. 11, iss. 2014, pp. 22–30, 2014.

 

[3] A. Kailas, V. Cecchi, and A. Mukherjee, “A survey of communications and networking technologies for energy management in buildings and home automation”, Journal of Computer Networks and Communications, vol. 2012, article ID 932181, pp. 1–12, 2012.

 

[4] M. Chinnici, A. Capozzoli, and G. Serale, “Chapter 10: Measuring energy efficiency in data centers”, In Pervasive Computing, 1st Edition Next Generation Platforms for Intelligent Data Collection, pp. 299–351, 2016.

 

[5] Y. Yan, Y. Qian, H. Sharif, and Tipper, “A Survey on Smart Grid Communication Infrastructures: Motivations, Requirements and Challenges”, IEEE Communications Surveys & Tutorials, vol. 15, no. 1, pp. 5–20, 2013.

 

[6] S. Mohammadi, S. Soleymani, and B. Mozafari, “Scenario-based stochastic operation management of MicroGrid including Wind, Photovoltaic, Micro-Turbine, Fuel Cell and Energy Storage Devices”, Electrical Power and Energy Systems, no. 54, iss. 2014, pp. 525-535, 2014.

 

[7] P. D. Martini, “A Future of Customer Response”, DR 2.0, Newport Consulting, Prepared for the Association for Demand Response and Smart Grid, USA, Department of Energy, 2013.

 

[8] P. Siano, “Demand response and smart gridsA survey”, Renewable and Sustainable Energy Reviews, no. 30, iss. 2014, pp. 461-478, 2014.

 

[9] H-T. Zhang, F-Y. Xu, and L. Zhou, “Artificial neural network for load forecasting in smart grid”, In Machine Learning and Cybernetics (ICMLC), 2010 International Conference on, vol. 6, pp. 3200–3205, IEEE, 2010.

 

[10] N. Amjaday, and F. Keynia, “Day-Ahead Price Forecasting of Electricity Markets by Mutual Information Technique and Cascaded Neuro- Evolutionary Algorithm”, IEEE Transactions on Power Systems, vol. 24, no. 1, pp. 306–318, 2009.