Research On Simulation Method Of Wind Power Time Series

As large scale wind power is incorporated into power system, new challengesto the reliable operation of power system are occured. The strong randomness andstrong volatility have become important factors restricting wind power to expandingthe scale of wind power development. The analysis of network frame extension,conventional power generation capacity, long-term system reserve capacity, energystorage devices and service requirements, often need to provide the future windpower time series, which are based on the statistical characteristics of historical data.It has become the key technical problem that how to simulate the wind powerscenarios in the future realistically, eliminating or reducing the bad effect to thedistortion of statistical characteristics of historical data caused by the serious windpower abandoning as far as possible. This paper carries an in-depth research on thewind power time series simulation method and evaluation standard, and mendingstrategy of abandoned wind power, provides good technical means to overcome theproblem above.There are two methods on wind power time series simulation of indirect anddirect. Indirect method need to simulate the wind speed time series first, and thenget wind power time series according to the certain wind-electrical conversioncharacteristics. The difficulty of the method is how to get the wind-electricalconversion characteristics exactly. Direct method uses historical data to establishappropriate model of wind farm output, to simulate the scenarios of wind farm windpower in the future. This paper adopts the method of stochastic differentialequations, numerical integral to generate wind power scenarios in the future directly,considering the statistical characteristics of the historical data. At the same time, aevaluation method of wind power time series simulated by different methods isproposed, assessment criterias of the mean and standard deviation of power, powerlevel probability distribution, power level variation probability distribution andpower duration curve are presented, which are used to analysis the similaritybetween the wind power time series and the historical data.The abandon rate of wind power is high in China’s wind power operation now.It is easily leading to the decrease of credibility of simulation results to generatewind power time series based on the statistical characteristics of historical datadirectly. Therefore, a mending strategy of abandoned wind power is proposed toachieve the reduction of the statistical characteristics of historical data, and thengenerate wind power time series. The example results show that the wind power simulation generating method that adopts the mending strategy of abandoned windpower is consistent with the expected effect, and it is helpful to providing moredependable basic wind power data which are appropriately restored to their originalappearances. At last, this paper generates the wind power time series based on thedirect method and mending strategy of abandoned wind power, and the wind powertime series are used to power system reliability assessment. Specifically, this paperdescribes the steps of application of non-sequential monte carlo method ingenerating capacity adequacy assessment, and verifies the role of simulation data inadequacy index calculation.This research work is supported by the National High Technology Research andDevelopment of China (863Program)(NO.2011AA05A105).