基于多源报警信息贝叶斯网络与关联离散系数的配电网故障台区定位方法OA

Considering the imbalanced development of regional network and the imperfect fault information of regional distribution networks,which leads to inaccurate fault location of regional network,a fault location method of distribution networks based on Bayesian network and associated discrete coefficient is proposed,which uses the multi-source alarm information to improve the accuracy of fault location and power supply reliability.The multi-source alarm information mainly includes sectional switches overcurrent alarm information,distribution transformer fault or alarm information,and telephone complaints information.Firstly,a distribution network structure model in the form of a linked list was established,Bayesian network approach was used to preliminarily identify the faulty section.and the Z value was obtained to rule out distribution transformer faults.Secondly,the correlation discrete coefficient between alarm information and preliminary fault area was constructed,and the correlation discrete coefficient curve was obtained.Then the minimum value point of the correlation discrete coefficient curve was searched to obtain the final location results.Finally,an example of a distribution network topology diagram in a certain urban area was used to verify the correctness and effectiveness of the proposed method.