| At present,the haze and other climate and environmental issues are becoming increasingly prominent.The problems of environment and energy,such as limited fossil energy reserves,become a hot topic all over the world.All countries and regions are advocating energy transformation and energy structure reform.The "13th Five-Year" energy development planning of China Rene,wable also pointed out that transition which is the high carbon energy model from fossil energy consumption to renewable energy as the main energy consumption low-carbon energy model will been gradually achieved and energy transformation will be promoted.In recent decades,the world is promoting the development and utilization of renewable energy.The world and China's wind power installed capacity increased year by year and China's total installed capacity of wind power is the first in the world since 2010.However,China's wind power accommodation is not in the forefront of the world.The topic is based on the background of high proportion of wind power integration in order to improve the accommodation capacity of wind power as the research target.The research is carried out on the basis of the analysis of wind power operation characteristics with different spatial and temporal scales.The two main research directions of wind power forecasting and wind energy storage are summarized.One is to improve the wind power forecast accuracy from the power system reliability aspects in order to increase the wind power accommodation,the other is energy storage configuration to improve the wind power accommodation from the view of peak load regulation.Analysis and research on the operation characteristics of wind power at different spatial and temporal scales is based on the general idea of big data analysis.The hierarchical structure and data classification framework of wind farm data analysis and the framework of wind power operation characteristics at different spatial and temporal scales are summarized.The month and day data of wind farm and the day and ten-minute data of wind turbine are collected.The seasonality of wind power farm month generation,the volatility and relativity of wind power farm day generation are analyzed.The day average wind speed,the relativity between day generation and average wind speed,the relativity between the day loss generation and average temperature and the relativity of 10-minute wind speed and power of wind turbine are also analyzed.The actual wind power curve is drawn with the median interval estimation method.Under the premise of ensuring the power balance of power system,the power generation research in long time scale for wind power accommodation planning and the output power study in short time scale to real-time dispatch wind power accommodation are established.Through the analysis and research on the wind power data of seasonality,relativity,volatility,distribution characteristics and limitation rate,four specific orientations of improving the accommodation capacity of wind power are distilled.There are the study of wind farm month generation,ultra short term output power prediction for wind turbines,the energy storage capacity of wind farms or its cluster and the integration of energy storage and generating wind turbine system.Based on the research of forecasting volatility and the introduction of the methods of econometrics and non parametric statistics,this paper studies the medium and ultra short term wind power forecasting methods.The month generation data of wind farm is modeling and forecasting with the methods of seasonal variation time series model,Holt-Winter seasonal exponential smoothing model,SARIMA model and combination model based on entropy method.The applicability and rationality of the prediction method is verified by the analysis of the prediction results.The ultra short wind power prediction is studied by the ARIMAX model and quantile regression model and the prediction results extend from the certain point to the uncertain interval.The reliability of prediction is improved and it can provide effective reference for real-time scheduling of wind power accommodation.The wind power storage research is based on the research object of stabilizing wind power fluctuations,improving wind power accommodation and exploring"source-grid-load-storage" operation mode.In this paper,the energy storage methods of the current power system are summarized and analyzed,and three kinds of energy storage arrangement schemes and their structures are put forward,which are based on large-scale centralized energy storage,small capacity distributed energy storage and instantaneous local energy storage.The method which is the combination of wind power accommodation maximization and economic analysis for determining the capacity of vanadium redox flow battery energy storage system with wind farm and its cluster is studied.On the other hand,it is used to stabilize the fluctuation of the wind power system to determine the local energy storage arrangement,that is,from the source to curb the wind power output fluctuations.The two kind model structure of the integration of energy storage and generating wind turbine system model is put forward,and the characteristics of the two kinds of energy storage models are analyzed.Two kinds of the integration of energy storage and generating wind turbine system model are designed respectively.In summary,under the background of high proportion wind power integration,the characteristics of wind power operation is as the basis and the improvement of wind power accommodation capacity is as the target.There are four part contents are studied which are including modeling and forecasting of wind power farm month generation,modeling and forecasting of ultra short term wind turbine power,the method for determining energy storage capacity of wind farms or its cluster based on vanadium redox flow battery and the integration of energy storage and generating wind turbine system model based on local energy storage.The effective technical methods are discussed for improving wind power accommodation. |
PDF Full Text Request