| In our country,the main wind energy resources are far away from the load center,which leads the wind power need to be transported to the load centers through large capacity and long distances transmission.Series compensation transmission technology is the most economical and technically feasible measure to increase the capacity of large-capacity long-distance transmission.In order to improve the long-distance transmission capacity of large-capacity wind power,many cluster wind farms usually use series compensation transmission technology.However,it is easy to induce subsynchronous oscillation of the power system and endanger the wind farms.For example,since 2012,in Guyuan,northern Hebei Province,China,the doubly-fed wind power cluster has occurred serious 6-8Hz subsynchronous resonance events,which result in a large number of wind turbines being disconnected from the network and damaged.With the gradual increase of the capacity of the wind turbine,the scale of wind farms is also increasing day by day,and the impacts of subsynchronous oscillation caused by wind farms on the power grid are also increasingly serious.In view of the subsynchronous oscillation of wind farms,this paper has mainly carried out the research on the equivalent modeling of wind farms and the risk assessment of random wind speed.The contents are as follows:(1)Each wind farm usually has dozens or even hundreds of wind turbines.The order of the detailed electromagnetic transient model of the entire wind farm is tens or even hundreds of times higher than that of conventional electromagnetic transient models.If the detailed model of each wind turbine is adopted,the study of subsynchronous oscillation of the electromagnetic transient model will lead to low computational efficiency,high computational complexity,and the "dimensional disaster" problem of small disturbances analysis.Therefore,a wind farm equivalent electromagnetic transient model that satisfies the research requirements of subsynchronous oscillations must be established.Although the current researches have proposed many wind farm equivalence modeling methods,the methods of wind farms for subsynchronous oscillations still lacks in-depth research results.Due to the different control modes of wind turbines in different operation regions,the subsynchronous oscillation damping changes inherently non-linearly,and it is difficult to use one unit to equalize the all wind turbines’ subsynchronous oscillation characteristics in different wind speed intervals.Therefore,this paper proposes to use the K-means method to divide the wind turbines into different aggregation groups in different control intervals.Considering that the current amplitude and phase of the outlet of the wind turbines are different,an equivalent method is proposed to improve the effect of power equivalent of the wind farm.Both time domain and frequency domain analysis results show that the proposed equivalent modeling method satisfies the research requirements of the subsynchronous oscillation of the wind farm and can reproduce the subsynchronous oscillation characteristics of the detailed model,and reduce the computational complexity and increase the analysis efficiency.(2)At present,the deterministic method is widely used to analyze the subsynchronous oscillation safety risk of wind farms under a specific operating condition.It is difficult to give the overall security situation of the wind farm under stochastic variation of wind speed,and the risk of subsynchronous oscillation of wind farms cannot be completely reflected.In order to consider the influence of the stochastic wind speed,based on the existing probabilistic collocation method,this paper proposes a piecewise probabilistic collocation method that considers the essential non-linearity of control modes in different operation intervals.Applying the operating wind speed data of a practical wind farm in North China from 2014 to 2015,a risk assessment method of subsynchronous oscillation wind power based on piecewise probabilistic collocation method was proposed.The research results show that the piecewise probabilistic collocation method overcomes the deficiency of the conventional probabilistic collocation method and can deal with the essential nonlinear problems very well.The results of the piecewise probabilistic collocation method are in good agreement with the results of the classical Monte Carlo method,with high accuracy and high computational efficiency... |