风电场阻尼电力系统振荡的机理及时滞影响

Mechanism of Wind Farm Damping Power System Oscillation and Time-delay Influence

为改善风电场对电力系统的阻尼影响,提出了附加有功控制策略并分析了通信时滞对附加控制效果的影响。首先,分析了风电场阻尼电力系统的机理,推导了风电场提供正阻尼的解析条件。基于该条件,设计了附加有功控制策略,将风电场公共母线的频率增量作为附加控制器的输入,附加控制器的输出经通信网络作用于各风电机组,以改善风电场整体动态行为,达到对电力系统产生正阻尼的作用。然后,考虑了通信时滞对控制效果的影响,提出并证明了求解风电场提供零阻尼的临界时滞以及计算系统发生Hopf分岔的临界时滞的2个命题,以作为附加有功控制的约束条件。最后,通过时域仿真表明,附加有功控制可显著增强电力系统阻尼;在引入时滞因素后发现,随着时滞的增大,系统阻尼相应变化。根轨迹分析与频率特性分析验证了风电场提供正阻尼条件及时滞影响相关命题的正确性。

In order to improve the damping effect of the wind farm on the power system,this paper proposes a novel active power control strategy and analyzes the impact of communication time-delay on the control strategy.Firstly,the mechanism of the wind farm restraining oscillation of the power system is studied and an analytical condition for enabling the dynamic behaviors of the wind farm to be improved is derived.According to the condition,the additional control strategy is designed for the purpose of producing positive damping to the power system.The frequency increment is collected from the common bus of the wind farm as the input signal of the controller.Meanwhile,the output signals from the controller are individually transmitted to the active control loops of each wind turbogenerator.Secondly,considering the impact of commutation timedelay on the system performance,two propositions,as the constraining conditions for the design of the active power controller,are presented and proved,the proposition of solving the critical time-delay of the wind farm in supplying zero-damping and the critical time-delay resulting in Hopf bifurcation with the calculating system due to the wind farm supply to the power system with negative damping.Finally,as shown by simulation results,the additional active power control can markedly enhance the damping of the power system.Furthermore,it is found after introducing the time-delay factor that,with the time-delay increase,system damping varies as a result.The correctness of the analytical condition and the two propositions are verified by the root locus analysis and frequency domain analysis.