Research On Wind Power Grid-Connected Resonances Suppression Strategy Based On Artificial Neural Network

In recent years,with the widespread application of wind resources,the proportion of wind power generation in the power grid has become larger and larger.As a result of,large amount of wind power plants has gradually been integrated into the power grid.As the wind farms are susceptible to uncertain factors such as wind speed,the integration of a large amount of wind power plants into the power grid will cause various harmonic oscillations,so called "wideband resonances".This problem has gradually become a bottleneck problem that hinders the integration of large wind farms into the power system.Therefore,when a wind farm is connected to the AC power grid,the potential borad resonance frequencies in the system should be accurately detected,and then the corresponding control measures can be taken to suppress the resonance of the wind power system.This article conducts research on this wideband resonance issue and provides a solution for solving the broad frequency resonance problems in the wind power integrated systems.Firstly,this paper analyzes the commonly used system resonance frequency detection methods,and then combines the complex and changeable characteristics of the wind power grid-connected system,which has the characteristics of wide-band frequency resonance.Based on the artificial neural network algorithm,an artificial neural network based wind power is investigated for detecting resonance frequency of grid-connected system,detecting resonance frequency of grid-connected system of wind power.In addition,an active filter control method working in conjunction with the artificial neural network resonance frequency detection module is proposed to perform the corresponding resonance detection and control.At the last,the correctness and effectiveness of the proposed control strategy are verified by in simulation model of wind power grid-connected system in MATLAB/Simulink simulation platform.