Dissertation > Excellent graduate degree dissertation topics show

Application of Markov Chain in Mid-long Term Load Combination Forecasting

Author: HuangYinHua
Tutor: PengJianChun
School: Hunan University
Course: Electrical Engineering
Keywords: Power Systems Long-term load forecasting Combined Forecast Markov chain Transition probability
CLC: F407.61
Type: Master's thesis
Year: 2010
Downloads: 353
Quote: 5
Read: Download Dissertation


Long-term load forecasting is the basis for urban power grid planning work, one of which is essential for network planning provides basic data, the power system is to ensure reliable power supply and economic operation of the premise. Reasonable and accurate load forecasting power investment will directly affect the network layout and operation is reasonable, has great economic and social benefits. Since long-term load forecast by many uncertain factors, single forecasting model difficult to guarantee in any case be able to obtain satisfactory predictions. How to get simple, practical, good stability prediction model and how to improve the accuracy of load forecasting has become the focus of research scholars. In this paper, learn and digest Markov chains, based on the combination of its introduction into the long term load forecasting, respectively, from the series combination of prediction and forecasting the parallel combination of two aspects of the analysis period. This paper first outlines the power load forecasting current research and the purpose and significance of load forecasting, load forecasting are discussed classification and long term load forecasting method, also introduced the basic theory of Markov for the later development to do the groundwork. On this basis, considering the characteristics of Markov chain and apply it to combination forecasting go. For the series combination of forecasts, combined with Markov chain transition probabilities may reflect the influence of random factors applicable to random fluctuations of the dynamic characteristics of the process will be conducted with the combination of gray prediction model. This method makes up gray prediction model in predicting the outcome in terms of accuracy and trustworthiness demonstrated the inherent defects. Prediction results show that the method is feasible to improve forecasting accuracy; For the parallel combination of forecasts, on the one hand, for a single model has a specific scope and conditions of use of Markov process without aftereffect will be characterized by their application Screening of the load model. Study confirmed the relative use of the model has not been screened forecast screening model predictions with the predicted results are more satisfactory. The other hand, gives the smallest error as a criterion states and state Markov chain probability preliminary estimates, then the probability distribution of Markov chain fitting time-varying state laws. By-step transition probability matrix estimation problem into multiple constraints autoregressive model, and then use one-step transition probability matrix of the initial state probability distribution estimates and dynamically acquire portfolio weights. Examples show that this method small amount of calculation, high accuracy. Finally, combining Markov in the application of long-term load forecasting, analysis and summarizes the prediction error sources and further explore ways to improve the accuracy of load forecasting.

Related Dissertations

  1. Power System Dynamic Voltage Stability Simulation Study Based on Precise Integration Method,TM712
  2. Researches on Urban Rail Transportation Operation Management System Test and Evaluation Method,TP311.52
  3. Detection and Tracking of Moving Object in Complex Background,TP391.41
  4. Traffic Prediction Model Based on Wavelet and Markov,TP393.06
  5. Research of a Medium and Long-term Load Forecasting Method,TM715
  6. Research on the Distribution System of Machinery Factory,TH183
  7. Research on Dps Impedance Calculation Method and Simulation Software,TN86
  8. Design and Simulation of Power System for Deep Space Prober,TN86
  9. Based on the value of content and links to pages Algorithm,TP393.092
  10. The Fruther Research on Ruin Problem Model of Gamblers,F840
  11. Portfolio quality prediction based on probabilistic method,X832
  12. Linearly polarized pulsed laser fields lithium atom Rydberg transitions population,O562
  13. Applications of a Multi-state Risk Factor/Morality Model in Life Insurance,F840.6
  14. The Research and Development of the System of Load Modeling Online Based Component,TM714
  15. Research on Power System Short-term Load Forecasting Approaches,TM715
  16. Study on Establishement of RPS and REC Market in Electrical Power Supply System in China,F426.61;F426.2
  17. Angular power power system research and implementation methods,TM933.3
  18. The Study of Management for Domain Ontology and Its Application in Power System,TM769
  19. Dynamic Behavior of Discrete Mapping,O415.5
  20. Combination Forecasting Method in Shandong Province telecom business revenue Empirical Study,F626

CLC: > Economic > Industrial economy > Industrial economic theory > Industrial sector economy > Electrical and electronics industry > Electricity, motor industry
© 2012 www.DissertationTopic.Net  Mobile