Dissertation > Excellent graduate degree dissertation topics show

The Research on Shanghai Gas Load Forecasting

Author: PianKun
Tutor: XuXiaoZhong
School: Shanghai Normal University
Course: Applied Computer Technology
Keywords: Rough Set SOFM network (SOFM) Regression support vector machine ( SVR ) The CPSO optimization algorithm
CLC: TU996.3
Type: Master's thesis
Year: 2010
Downloads: 96
Quote: 0
Read: Download Dissertation

Abstract


Natural gas as a green energy , the direction of development of city gas , increase the proportion of natural gas in the energy consumption structure , not only help to promote energy conservation , but also be able to maintain sustainable economic and social development . Shanghai as one of the earliest applications of natural gas city is vigorously promote the application of natural gas , especially in the West-East through greatly promote the rapid development of Shanghai Natural Gas . In order to achieve the efficient operation of the natural gas supply system , optimize the scheduling and scientific management , gas load forecasting work as a basis for decisions is particularly important . This paper through analysis of the Shanghai Gas system load regular commitment to the accurate prediction of the gas load a variety of today 's popular smart prediction techniques , such as : data mining ( DM ) , neural network ( NN ) , support vector machine ( SVM ) , PSO optimization algorithms do deep research , and to explore the two combined forecasting model . The first is a rough set and BP neural network combined using rough sets to load data preprocessing, noise , redundant and irrelevant data to the BP neural network training and prediction , and then as the input variables than simply BP neural network has been greatly improved . The second is a comprehensive self-organizing feature map network (SOFM) and regression support vector machine (SVR) forecast a day load forecasting model . In this model , we use the SOFM training data sample clustering , after clustering after each containing similar characteristics of the data , and then the SVR prediction model were established respectively in each group clustered data , the SVR kernel function parameter choice using the CPSO optimization algorithm . We also use data mining methods of historical load data pre-processing, discrete data points to determine amended to better reflect the regularity of the gas load ; selected based on the characteristics of the filter , select load than A big factor as the the SOFM input variables . The experimental results show that the model , whether in the training time or have a greater increase in the prediction accuracy .

Related Dissertations

  1. Fault Diagnosis Method Based on Support Vector Machine,TP18
  2. Research on Clustering Algorithm Based on Genetic Algorithm and Rough Set Theory,TP18
  3. Research and Application on Design Resource Dynamic Scheduling Faced to Design Reuse,TB47
  4. Study on the Effect Factors of Microcystins、 Allelopathic Inhibition and Microcystins Prediction Model,X52
  5. Research of Environmental Contamination Accidents Emergency Control System Based on GIS and Spatial Data Mining,TP311.13
  6. Research on Financial Distress Prediction of Listed Companies Based on Neighborhood Rough Sets-Neural Network,F276.6;F224
  7. The Method of the Attribute Reduction of Rough Set Based on Genetic Algorithm,TP18
  8. The Application of Data Mining Technology in the Supermarket Marketing System,TP311.13
  9. Support Vector Machine and Its Application to Regional Water Resources Sustainable Utilization,S273
  10. Based on full vector spectrum of rotating machinery fault feature extraction,TH165.3
  11. Study on the Land Evaluation Index System Based on the Rough Set Theory,F224
  12. Research on Method to Determine the Optimization Target of Energy Loss Analysis of Thermal Power Units,TM621
  13. The Related Expansion on Soft Set in the Uncertain Theory,O159
  14. Research on Enterprise Talent Assessment System Based on the Competency Model,F224
  15. Study of Condition and Risk Assessment of Electric Power Supply Equipments,F426.61
  16. Supplier Selection with Multiple Criteria,F274
  17. Based on rough set of supply chain enterprise knowledge flows between impact factor Recognition,F224
  18. Based on the Tourism Industry Cluster to Research the Promotion of Competitiveness of Tourism Industry in Hebei Province,F592.7
  19. The Research of Environmental Early-warning System for Foreign Trade of China,F205;F224
  20. Data Reduction Algorithm Based on Information Entropy,TP18
  21. The Research and Application of Fuzzy Control Based on Granular Computing,TP273.4

CLC: > Industrial Technology > Building Science > Municipal Engineering > City gas supply > Gas requirements,consumption norms and load calculations
© 2012 www.DissertationTopic.Net  Mobile