基于组合模型的接地网腐蚀速率预测算法

及时掌握并准确预测接地网的运行工况,从而对其进行预警和维护,是保证电力系统安全稳定运行的重要举措。针对接地网腐蚀数据的小样本与非线性特征,且目前单一预测模型存在预测精度不足的问题,提出了一种结合改进最小二乘支持向量机与误差校正的组合模型,将其应用于接地网腐蚀速率预测。该法采用遗传算法优化最小二乘支持向量机参数,为提高模型的预测精度,应用误差预测校正模型修正其预测结果,降低了极大误差出现的可能性,提高了预测模型的稳定性。结果表明,采用组合模型对接地网腐蚀速率进行预测,比采用最小二乘支持向量机模型具有更高的预测精度,更适用于接地网腐蚀速率的预测。

Corrosion Rate Prediction of Grounding Grid Based on Combined Model

It is an important measure to ensure the safe and stable operation of the power system by timely grasping and accurately predicting the operating conditions of the grounding grid and thus alerting and maintaining it.Aiming at the small sample and nonlinear characteristics of grounding grid corrosion data,and the current single prediction model has insufficient prediction accuracy,this paper proposes a combined model of improved Least Squares Support Vector Machine(LSSVM) and error correction,which is applied to grounding grid corrosion rate prediction.The method uses Genetic Algorithm(GA) to optimize the parameters of LSSVM.In order to improve the prediction accuracy of the model,the error prediction correction model is used to correct the prediction results,which reduces the possibility of maximal error and improves the stability of the prediction model.The results show that the combined model is more accurate than LSSVM in predicting the corrosion rate of the grounding grid,and more suitable for predicting the corrosion rate of grounding grid.