Aiming at the problem of support vector machine in system harmonic impedance estimation due to improper parameter setting which leads to degradation of estimation accuracy

a system harmonic impedance estimation method based on wavelet support vector machine optimized by improved sparrow search algorithm is proposed. A wavelet support vector machine model with multi-scale feature extraction capability is constructed by replacing the traditional Gaussian kernel function with the wavelet kernel function. The sparrow search algorithm is improved by using iterative chaotic map with infinite collapses

adaptive t-distribution and inverse learning strategy. Parameters of wavelet support vector machine are optimized by the improved algorithm and the optimization model is established based on the optimal parameters which improves the accuracy of system harmonic impedance estimation. Finally

substitute the harmonic voltage and harmonic current data collected at the point of common coupling into the optimization model to obtain the system harmonic impedance estimation results. Simulation analysis and field test results show that the optimized model has higher harmonic impedance estimation accuracy and robustness in the presence of background harmonic fluctuations and outliers in the sample data.