Estimation of Switching Overvoltages on Transmission Lines Using Neuro-Fuzzy Method

Insulation failure caused by switching overvoltages (SOVs) is one of the main sources of transmission lines’ outage، specially، on voltage levels of 345 kV and above. Therefore، the estimation of SOVs is vital in order to control and/or to reduce the switching–related outages. Due to the stochastic behavior of some of the parameters affecting on SOVs، the study of this phenomenon should be carried out based on a statistical study of the switching. Also، in the case of surge arrester installation on the transmission lines، depending on the location of arrester، voltage profile on line is changed and all the simulation should be performed for each new location of arresters، separately. One can conclude that this procedure is complex and time consuming. In this paper، a fuzzy based meta-model is presented which is be able to estimate the switching surge flashover rate (SSFOR)، the maximum value of SOVs on the network and the location where the maximum overvoltage takes place. In the proposed meta model، the effect of altitude on SSFOR and the magnitude of SOVs is considered. This meta-model can be used، directly، for planning the insulation level of transmission lines in order to meet a certain number of outages and locating arresters on the region/nodes of the network of weak operation against SOVs. It is also possible to utilize the proposed meta model، indirectly، for assigning the optimal location of any specified set of arresters on the network without simulating of real network by a transient software، e. g. EMTP/ATP draw. The presented meta model can also be used in the operating stage to decide on the sequence of energizing and re-energizing of different transmission lines connected to the substations with the aim of reducing of maximum SOVs.