臺灣博碩士論文加值系統

由於傳統電網的集中式配電系統無法應付電廠發生緊急斷電事故時大量用戶所受到的劇烈影響，因此取而代之的即是分散式供電系統的智慧電網架構。然而，傳統分散式電網只考慮當前的電力狀況來進行配電，這不僅造成了多餘發電資源浪費的可能性，也可能導致電力成本無法有效降低的問題。為了解決電力成本的問題，本論文提出利用模型預測優化控制應用於智慧電網之配電管理策略。透過自相關移動平均模型預測未來的電力狀況以決定樂觀與保守的交易態度層級，並根據態度有計畫性地使用電池儲能系統。經由粒子群優化演算法找到最佳交易對象，便能有效降低電力成本的負擔。舉例來說，當未來電量充足且市場買價較當前時刻的買價低時，能將部分電池儲能設備中的電力資源提前進行拍賣以獲得較多的利益。反之，當未來電量短缺且市場賣價較當前時刻的賣價高時，可優先購買較多的電力儲存至電力儲能設備中以降低電力使用成本。因此，我們的實驗顯示預測模型誤差率小於10%。從兩種買賣電力的案例來看，實驗結果也顯示我們提出的方法在拍賣電力時會比傳統不包含預測模型機制的方法多賺177%的電力費用，而在購買電力時會比傳統電網的機制少支付31.08%的電力費用，對整體智慧電網而言，在參考未來時段的電力狀況下也能比傳統電網省下19.38%的電力費用，著實有效地降低電力成本並賺取更多的利益。

There is a fatal problem in the traditional centralized power system. When some accidents occur in the power plants, a large number of users will be affected. Thus, the traditional centralized power system was replaced by the decentralized grid. However, the decentralized grid only focuses on the current electricity situation to process distribution. It not only wastes the remaining power generation but also keeps the electricity cost higher. To solve this cost saving problem, the Thesis proposed a Model Predictive Optimization (MPO) for distribution management system in smart grids. We can predict the future electricity situation by Autoregressive Integrated Moving Average (ARIMA) model and can determine the distribution attitude such as optimistic attitude and pessimistic attitude. Besides, according to the attitude, we use the Energy Storage System (ESS) effectively. After the prediction of bidding price and bidding capacity, we could find an optimal trading pairs through the Particle Swarm Optimization (PSO). For example, when the surplus electricity situation is predicted in the future and the future utility selling price is lower than the current price, we will discharge ESS to trade with other micro-grids to earn more cost. Conversely, while the power shortage is predicted in the future and the future utility buying price is higher than the current price, we can first buy more energy to charge the ESS to reduce the cost savings.

Therefore, our experimental results show that the error rate of prediction model is less than 10\%. Besides, we can reduce the cost savings effectively in smart grids by our proposed MPO method. For one case, the proposed cost of our proposed method in one micro-grid is 177\% of the traditional grid. Another case could show that the negative cost by our proposed method in another micro-grid is smaller than that by the traditional grid by 31.08\%. For smart grids, our proposed MPO method totally saves the much more cost of 19.38\% if four time slots is used for prediction. The overhead of prediction model and the optimization efficiency for our proposed MPO method are 5.65 seconds and 2.42 seconds.

1 Introduction 1
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3 Thesis Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2 Related Work 7
2.1 Smart Grid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Research On Model Predictive Control . . . . . . . . . . . . . . . . . . 11
2.3 Research On Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.4 Research On Optimization . . . . . . . . . . . . . . . . . . . . . . . . 15
3 Preliminaries 17
3.1 Problem Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.2 Assumptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.2.1 Architecture Assumptions . . . . . . . . . . . . . . . . . . . . 18
3.2.2 Electricity Assumptions . . . . . . . . . . . . . . . . . . . . . 18
3.2.3 Transaction Assumptions . . . . . . . . . . . . . . . . . . . . . 19
3.3 Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.4 Parameter Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.4.1 Parameters for Prediction Model . . . . . . . . . . . . . . . . . 21
3.4.2 Parameters for Optimizer . . . . . . . . . . . . . . . . . . . . . 23
4 Distribution Management System Design 25
4.1 Model Predictive Control . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.2 Prediction with ARIMA model . . . . . . . . . . . . . . . . . . . . . . 27
4.3 Particle Swarm Optimization . . . . . . . . . . . . . . . . . . . . . . . 37
5 Experiments 47
5.1 Experimental Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
5.1.1 Experimental Environment . . . . . . . . . . . . . . . . . . . . 47
5.1.2 Demand Load Data, Generation Data, ESS Specification, and Dynamic Utility Electricity Price . . . . . . . . . . . . . . . . . 48
5.2 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
5.2.1 Evaluation of ARIMA Model Prediction . . . . . . . . . . . . . 50
5.2.2 Optimization Efficiency . . . . . . . . . . . . . . . . . . . . . 53
5.2.3 Cost Savings on the MPO Distribution Management . . . . . . 54
6 Conclusions and Future Work 60
Bibliography 62

[1] F. Birol. World Energy Outlook. International Energy Agency, November 2014.
[2] Z. Wang and L. Wang. Adaptive negotiation agent for facilitating bi-directional energy trading between smart building and utility grid. IEEE Transactions on Smart Grid, 4(2):702–710, June 2013.
[3] D. Gielen. Energy Technology Perspectives. International Energy Agency, May 2014.
[4] Department of Energy. The Smart Grid: An Introduction. http://energy.gov/oe/downloads/smart-grid-introduction-0, 2008.
[5] N. Abi-Samra, J. Holt, J.McDaniel, andM. Baustert. Advanced Distribution Management System (ADMS). http://www.dnvkema.com/newsletters/energy-notes/ADMS.aspx, May 2014.
[6] K. Holkar and L. Waghmare. An overview of model predictive control. International Journal of Control and Automation, 3(4):47–63, December 2010.
[7] J. Kennedy and R. Eberhart. Particle swarm optimization. In Proceedings of the IEEE International Conference on Neural Networks, volume 4, pages 1942–1948, November 1995.
[8] I. Koutsopoulos and L. Tassiulas. Optimal control policies for power demand scheduling in the smart grid. IEEE Journal on Selected Areas in Communications, 30(6):1049–1060, July 2012.
[9] E. Lee and H. Bahn. A genetic algorithm based power consumption scheduling in smart grid buildings. In Proceedings of the International Conference on Information Networking (ICOIN), pages 469–474, February 2014.
[10] Taiwan Power Company. Time Price and Seasonal Price. http://www.taipower.com.tw/UpFile/PowerSavFile/main_6_2_2.pdf, April 2013.
[11] Ministry of Economic Affairs Bureau of Energy. The Manual of Residential Energy Conservation. http://ebook.energypark.org.tw/books/admin/
1/pdf/source/14140257687908.pdf, August 2014.
[12] M. R. Patel. Wind and Solar Power Systems: Design, Analysis, and Operation. CRC press, July 2005.
[13] L. M. Fraas and L. D. Partain. Solar Cells and their Applications, volume 236. John Wiley & Sons, August 2010.
[14] J. Larminie, A. Dicks, and M. S. McDonald. Fuel Cell Systems Explained, volume 2. Wiley New York, February 2003.
[15] K. Alanne and A. Saari. Distributed energy generation and sustainable development. Renewable and Sustainable Energy Reviews, 10(6):539–558, December 2006.
[16] S. J. Qin and T. A. Badgwell. An overview of industrial model predictive control technology. In Proceedings of the AIChE Symposium Series, volume 93, pages 232–256, June 1997.
[17] E. F. Camacho and C. B. Alba. Model Predictive Control. Springer Science & Business Media, January 2013.
[18] B. Yu and J. Zhu. Neural network model predictive control with genetic algorithm optimization and its application to turbofan engine starting. In Proceedings of the International Conference on Intelligent Human-Machine Systems and Cybernetics (IHMSC), volume 2, pages 262–265, August 2010.
[19] S. Lek, M. Delacoste, P. Baran, I. Dimopoulos, J. Lauga, and S. Aulagnier. Application of neural networks to modelling nonlinear relationships in ecology. Ecological Modelling, 90(1):39–52, September 1996.
[20] L. Davis. Handbook of Genetic Algorithms, volume 115. Van Nostrand Reinhold Company, January 1991.
[21] D. Molina, C. Lu, V. Sherman, and R. G. Harley. Model predictive and genetic algorithm-based optimization of residential temperature control in the presence of time-varying electricity prices. IEEE Transactions on Industry Applications, 49(3):1137–1145,May-June 2013.
[22] H. White. Maximum likelihood estimation of misspecified models. Econometrica: Journal of the Econometric Society, 50(1):1–25, January 1982.
[23] W. Yang, F. Yang, and J. Chen. Distributed predictive control of grid-connected solar PV generation based on data-driven subspace approach. In Proceedings of the International Electronics and Application Conference and Exposition (PEAC), pages 1087–1092, November 2014.
[24] Y. Jia and X. Liu. Distributed model predictive control of wind and solar generation system. In Proceedings of the 33rd Chinese Control Conference (CCC), pages 7795–7799, July 2014.
[25] F. Oldewurtel, A. Ulbig, A. Parisio, G. Andersson, and M.Morari. Reducing peak electricity demand in building climate control using real-time pricing and model predictive control. In Proceedings of the 49th IEEE Conference on Decision and Control (CDC), pages 1927–1932, December 2010.
[26] L. Xie and M. D. Ilic. Model predictive dispatch in electric energy systems with intermittent resources. In Proceedings of the IEEE International Conference on Systems, Man and Cybernetics (SMC), pages 42–47, October 2008.
[27] Y. Zong, D. Kullmann, A. Thavlov, O. Gehrke, and H.W. Bindner. Application of model predictive control for active load management in a distributed power system with high wind penetration. IEEE Transactions On Smart Grid, 3(2):1055–1062, June 2012.
[28] E. Mayhorn, K. Kalsi, M. Elizondo, W. Zhang, S. Lu, N. Samaan, and K. ButlerPurry. Optimal control of distributed energy resources using model predictive control. In Proceedings of the IEEE Power and Energy Society General Meeting, pages 1–8, July 2012.
[29] I. Ojo. Autoregressive integrated moving average. Asian Journal of Mathematics and Statistics, 3(4):225–236, September 2010.
[30] J. L. Mathieu, P. N. Price, S. Kiliccote, and M. A. Piette. Quantifying changes in building electricity use, with application to demand response. IEEE Transactions on Smart Grid, 2(3):507–518, September 2011.
[31] G. E. Box, G. M. Jenkins, and G. C. Reinsel. Time Series Analysis: Forecasting and Control. Holden-Day, June 2008.
[32] S. Hui-Lan. Application of time series analysis-autoregressive model for the enterovirus cases during 1999 to 2008 in Taiwan. Master’s thesis, Asia University, June 2009.
[33] J. Palomares-Salas, J. De la Rosa, J. Ramiro, J.Melgar, A. Aguera, and A.Moreno. ARIMA vs. neural networks for wind speed forecasting. In Proceedings of the IEEE International Conference on Computational Intelligence for Measurement Systems and Applications (CIMSA), pages 129–133, May 2009.
[34] H. Zhang, F. Xu, and L. Zhou. Artificial neural network for load forecasting in smart grid. In Proceedings of the International Conference on Machine Learning and Cybernetics (ICMLC), volume 6, pages 3200–3205, July 2010.
[35] M. Beccali, M. Cellura, V L. Brano, and A. Marvuglia. Forecasting daily urban electric load profiles using artificial neural networks. Energy Conversion and Management, 45(18):2879–2900, November 2004.
[36] A. A. Moghaddam and A. Seifi. Study of forecasting renewable energies in smart grids using linear predictive filters and neural networks. IET renewable power generation, 5(6):470–480, November 2011.
[37] K. E. Reddy and M. Ranjan. Solar resource estimation using artificial neural networks and comparison with other correlation models. Energy Conversion and Management, 44(15):2519–2530, September 2003.
[38] X. Wang, P. Guo, and X. Huang. A review of wind power forecasting models. Energy Procedia, 12:770–778, December 2011.
[39] L. Van, J. Peter, and E. H. Aarts. Simulated Annealing. Springer, September 1987.
[40] F. Glover and M. Laguna. Tabu Search. Springer, January 1999.
[41] M. Gambardella, M. B. A. Martinoli, and R. P. T. St ¨utzle. Ant Colony Optimization and Swarm Intelligence, volume 4150. Springer Science & Business Media, September 2006.
[42] P. Faria, T. Soares, T. Pinto, T. M. Sousa, J. Soares, Z. Vale, and H. Morais. Dispatch of distributed energy resources to provide energy and reserve in smart grids using a particle swarm optimization approach. In Proceedings of the IEEE Symposium on Computational Intelligence Applications in Smart Grid (CIASG), pages 51–58, April 2013.
[43] G. Asadi, M. Gitizadeh, and A. Roosta. Welfare maximization under real-time pricing in smart grid using PSO algorithm. In Proceedings of the 21st Iranian Conference on Electrical Engineering (ICEE), pages 1–7, May 2013.
[44] S. Hu. Akaike information criterion. Center for Research in Scientific Computation, March 2007.
[45] G. Schwarz. Estimating the dimension of a model. The Annals of Statistics, 6(2):461–464, March 1978.
[46] E. Vattekar. Analysis and model of consumption patterns and solar energy potentials for residential area smart grid cells. Master’s thesis, Norwegian University, June 2014.