The rapid increase in transmission voltages to fulfil higher demand of the
transmitted powers yields to put the switching surges as the governing factor in the
insulation design process for EHV and UHV systems. The modern methods of insulation
coordination make use of probabilistic concepts and statistical procedures especially for
very high voltage systems. In this paper, a sophisticated and accurate approach to
calculate the switching surge flashover rate (SSFOR) using Adaptive Neuro Fuzzy
Interface System (ANFIS) is developed. The variation of switching overvoltages
distribution along the transmission line under different system conditions is investigated.
The statistical distributions of energization overvoltages for a typical Egyptian 500kV
double-line from High–Dam to Samalut are derived from the results of 100 energization
cases with statistical (random) switching using alternate transient program (ATP). The
statistical distributions and its key values, such as mean value, standard deviation and
2% statistical overvoltages values have been recorded. The effects of various parameters
on the energization overvoltages distribution such as line length, source strength and the
use of shunt reactors to control the switching surge overvoltages are considered. The
developed ANFIS system estimated the SSFOR for different input parameters for both
the switching overvoltages and the insulation strength distributions with high accuracy. |
