基于SVM算法的PQ型控制逆变器出口故障分类

PQ-based Control of Inverter Exit Fault Classification Based on SVM Algorithm

随着全球清洁能源的广泛应用,含有分布式能源的电网将会是未来配电网的主要供电形式。分布式能源常常经PQ控制型逆变器件接入配电网,由于控制模式的应用,对于分布式电源并网条件下,故障下暂态过程将会发生新的变化。现有研究主要集中于馈线保护和DG保护,而针对PQ控制型微源出口故障选相的研究则非常有限。文中首先分析PQ控制型逆变器的数学模型,故障电流特性及控制器特征状态变量故障特性,在此基础之上,形成PQ控制型逆变器故障特征矩阵,然后,进一步分析粒子群算法改进的多分类支持向量机算法流程,最后,在Digsilent软件进行算例仿真计算,验证了所提方法的可行性和有效性。

With the wide application of clean energy in the world,the power grid containing distributed generator will be the main form of power supply in the future distribution network.Distributed generator is often connected to the distribution network via PQ controlled inverter.Due to the application of control mode,new changes will occur in the transient process under the condition of distributed generator connected to the grid.The existing research is mainly focused on feeder protection and DG protection,but the research on PQ controlled micro source outlet fault phase selection is very limited.In this paper,the mathematical model of the PQ controlled inverter,the fault current characteristics and the characteristic state variable fault characteristics of the controller were analyzed.On this basis,the fault feature matrix of the PQ controlled inverter was formed,and then the multi classification support vector machine algorithm improved by particle swarm optimization was further analyzed.Finally,a simulation example was carried out in Digsilent software to verify the feasibility and effectiveness of the proposed method.