基于AO-ESN的海上风电机组接地电阻预测

PREDICTION OF IMPACT RESISTANCE OF OFFSHORE WIND TURBINE BASED ON AO-ESN

针对目前2种主流求解海上风电机组接地电阻值精度不足的问题,该文提出基于天鹰算法改进的回声状态网络(AOESN),对海上风电机组接地电阻进行预测分析。分别利用BP、ESN、AO-ESN这3种网络进行预测,结果表明:AO-ESN模型的预测精度相较于BP和ESN模型分别提高了18%和14.46%,误差低至0.54%,所搭建的模型可对海上风电机组接地电阻进行精准预测。

In response to the current two mainstream solution of offshore wind turbine with insufficient accuracy of ground resistance methods,in this paper,the improved Echo state network(AO-ESN)based on the Aquila optimizer is proposed to predict and analyze the grounding resistance of offshore wind turbines.The three networks of BP,ESN and AO-ESN are used to make predictions.The results show that the prediction accuracy of the AO-ESN prediction model is 18%and 14.46%higher than the BP and ESN models,and the error is as low as 0.54%.The model built in this paper can be accurately predict the grounding resistance of offshore wind turbines,and provide reference for transient analysis and lightning protection of offshore wind turbines.