基于改进粒子滤波的风电系统并网侧谐波估计

Harmonic Estimation of Wind Power Systems at the Grid-connection Side Based on the Improved Particle Filter

受各类非线性特征影响,风力发电系统并网侧会随机产生大量谐波,对风电安全并网和电网稳定运行造成影响。为了准确检测谐波并确保相关设备稳定工作,提出了一种基于误差模型改进粒子滤波的谐波估计方法。该方法在标准粒子滤波的基础上,利用预测值和观测值间的误差进行粒子补偿,并计算其协方差矩阵;根据粒子权重和协方差进行重采样参数调整,以克服粒子多样性丧失导致的滤波发散的问题。基于实际观测结果,通过比较无迹卡尔曼滤波、传统粒子滤波和改进粒子滤波的谐波估计效果开展了实验,结果表明改进粒子滤波方法可以更准确地估计电网谐波幅值和相位。相比于标准粒子滤波,3次谐波的最大估计误差降低了67.07%,并且收敛速度更快,具有较高的理论研究意义和工程应用价值。

Due to the influence of various nonlinear characteristics,the grid-connection side of wind power systems sporadically generates a substantial amount of harmonics,posing challenges to the safe integration of wind energy and the stable operation of the electric grid.In pursuit of precise harmonic detection and ensuring the stability of associat-ed equipments,a harmonic estimation method based on an error model-enhanced particle filter is introduced.Building upon the standard particle filter framework,the proposed method compensates particles using the discrepancy between predicted and observed values,subsequently computing their covariance matrix,resampling parameters are adjusted based on particle weights and covariance to address the divergence issue arising from the loss of particle diversity.Based on the actual observation results,experiments are carried out by comparing the harmonic estimation perfor-mances of the unscented kalman filter,conventional particle filter,and the improved particle filter,results underscore that the enhanced particle filter offers a more accurate estimation of grid harmonic magnitude and phase.Compared to the standard particle filter,the maximum estimation error for the third harmonic was reduced by 67.07%,with a no-tably faster convergence rate,and this method holds significant theoretical research implications and potential value for engineering applications.