摘要
以2座±500 kV直流换流站为例,分析了高压直流换流站一次设备的常见故障类型、原因及跳闸情况,提出了一种基于决策树的故障预测模型.通过数据采集与处理、模型构建、预警决策等环节实现故障预测、风险评估与预警.在2座换流站进行了为期1年的试点应用,结果表明,该方法能够准确预测大多数故障.
Abstract
Taking two±500 kV DC converter stations as examples,this paper analyzes the common fault types,causes and tripping situations of primary equipment in HVDC converter stations,and puts forward a fault prediction model based on decision tree,which realizes the fault prediction,risk assessment and early warning through the links of data acquisition and processing,model construction,and early warning decision. It has been applied in two converter stations for one year,and the results show that the method can accurately predict most of the faults.
维普
万方数据