论文部分内容阅读
针对电力传输路由线路故障类参数多识别率不高等问题,利用提取故障特征向量和支持向量机结合的算法识别线路故障类型.提取故障线路特征向量,采集变化量的有效值,计算突变量所占三相突变量有效值总和的比例系数,将比例系数与零序电流判别系数结合构造故障特征向量.训练支持向量机以测试集特征向量作为输入,利用训练好的支持向量机判别分类,实现故障类型识别.实验表明,提出算法可克服多重困难针对输电线路十种类型故障进行学习并识别,并保证精度和效率.
Aiming at the problem of low recognition rate of fault parameters of power transmission routing lines, the fault types are identified by the combination of extracted fault vectors and support vector machines, and the eigenvectors of fault lines are extracted and the effective values of variation are collected to calculate the amount of mutation The proportional coefficient of the sum of the effective values of the three-phase sudden change, and the combination of the proportional coefficient and the zero-sequence current discriminant coefficient to construct the fault eigenvector. The training support vector machine takes the test set eigenvector as the input and uses the trained support vector machine to classify and implement the fault Type identification.Experiments show that the proposed algorithm can overcome multiple difficulties to learn and identify ten types of transmission line faults and ensure the accuracy and efficiency.