Research On Broadband Oscillation Risk Assessment Method Considering Time-Varying Characteristics Of Wind Power Grid-Connected System

As the proportion of grid-connected wind power increases year by year,and the series compensation technology and a large number of nonlinear power electronic devices are widely used,the broadband oscillation problem of doubly-fed wind power remote transmission system is becoming increasingly prominent,the risk of oscillation instability increases,and the wind power consumption is restricted.Therefore,based on an in-depth exploration of the oscillation mechanism and characteristics of the gridconnected system of doubly-fed wind power via series-compensated lines,and taking into account the uncertainty factors in the system,the probability assessment of the broadband oscillation risk of the grid-connected system of wind power is important for the grid-connected planning and oscillation suppression of wind power.Based on the above research background,the work carried out in this thesis is as follows.Firstly,the mathematical models of each sub-module of the grid-connected doubly-fed wind power system are established.Based on this,the analytical expression of impedance for the grid-connected doubly-fed wind power system via series compensation is derived,and the generation mechanism of the sub-synchronous oscillation of the grid-connected doubly-fed wind farm system via series compensation is analyzed.At the same time,in order to explore the oscillation characteristics of wind farm grid-connected systems under single-machine equivalent and multi-machine equivalent models,theoretical research and time-domain simulation were conducted to study the oscillation characteristics of single-machine systems and multi-machine systems under various influencing factors.The research results show that the subsynchronous oscillations may occur at higher compensation degrees,lower wind speeds,and unreasonable configuration of converter control parameters.Under different influencing factors,the single turbine equivalent for the whole wind farm generates certain errors,but the results of both equivalence analyses are basically consistent when the operating parameters are the same.Secondly,aiming at the time-varying oscillation risk of the grid-connected wind power system,most existing studies use deterministic methods to analyze the occurrence of oscillation or not under a specific condition,and cannot quantitatively assess the oscillation risk.Based on the consideration of the uncertainty factors in the system,a sub-synchronous oscillation risk assessment method for the grid-connected wind farm system based on Latin hypercube and extended Prony is proposed.First,the time-varying characteristics are analyzed based on impedance models.Then,based on the equivalent model of the actual oscillating doubly-fed wind farm connected to the grid through series compensation,Latin Hypercube Sampling and extended Prony are used to obtain assessment indicators.Combining with risk theory,quantitative evaluation of sub-synchronous oscillation risks during long-distance transmission processes for doubly-fed wind power through series compensation is achieved.Finally,the inherent time-varying and nonlinear characteristics of wind power grid-connected systems,which make it difficult to model quantitative indicators.A probabilistic assessment method for the risk of broadband oscillations in gridconnected doubly-fed wind power systems is proposed based on nonlinear approximation methods and modal identification theory.Firstly,the system oscillation mode parameters are accurately identified from the measurement signals by modal identification.Then,a multi-index quantification model is established using nonlinear approximation methods,which can achieve quantitative evaluation of broadband oscillation risk for doubly-fed wind power grid-connected systems with series compensation.The results show that the established multi-index evaluation model can not only reflect the broadband fluctuation characteristics of wind power grid-connected systems with series compensation but also quickly and accurately quantitatively assess the time-varying broadband oscillation risk due to uncertainty,and then comprehensively characterize the influence of uncertainty factors on the oscillation risk in the system.