Analysis Of Wind Power Characteristics And Study Of Its Time Series Generation

Research and large scale application of wind power has brought renewable powerdevelopment to a new era. It is also a big challenge for traditional power system operationand control. It has become a hot issue as how to cope with problems like unpredictablepower system operating point which are the results of large scale wind power integration. Inorder to control wind power, we should firstly make thorough description and detailedanalysis of it which is, however, scarcely reported up to now.Dynamic modeling experiment is the most direct and effective way for the research ofcharacteristics of power system integrating wind power which, however, is rather difficultto accomplish. Therefore, digital simulation is a prevailing method adopted by most of theresearchers. Though methods for the modeling of wind power time series is currentlyreported extensively. The wind power time series generated with these methods can’t wellcontain the characteristics of the original wind power.In light of the above two problems, this paper mainly defines and analyzes in detail thetime-domain characteristics of wind power which include wind power state duration time,wind power state transition and wind power fluctuation, etc. The feature of eachcharacteristic is summarized and the applicable value in the field of power system planning,operation and control is also analyzed. After a brief introduction of random variablesgeneration methodology, the paper transforms the traditional Markov Chain Monte Carlomethod (MCMC) in two aspects considering the above characteristics. This methodologysolves the disadvantages of traditional MCMC method that are incapability of transformingto other states and distortion of probability density function resulting from the impropervalue obtaining within each state. What’s more, the results of the method well contain theabove-mentioned time-domain characteristics of the original data. At last, the effectivenessand practicality of the proposed method is validated by means of statistical analysis andcomparison of the simulation results with the help of large amount of wind power data.