Forcast On The Transmission Line Trip Probability Based On Lightning Location System

Lightning is one of the ten worst natural disasters published by" International Decade of Diminishing Disasters", thunderstrike will harm household appliances, power electronics equipment damage. worse still,it can lead to the building explosion, electrical outage of large area, as well as people injury, having significant negative effects to the economic and social. Statistics show that lightning is the main cause of the power system line trip previous studies mainly focused on the insulation design and lightning protection of transmission line, in research on lightning flashover probability of transmission lines the achievement is much less. The lightning flashover probability prediction can help the running department assess the electricity transmission lines lightning risk timely in order to take the necessary measures to reduce or even avoid losses caused by line trip, which has an important significance.To solve this problem, based on historical data and real-time data of lightning location system, this paper designed a lightning flashover probability forcast algorithem of the transmission lines. The algorithm can be divided into two parts:first, the probability prediction of transmission line lightningstrike, secondly, the probability prediction of transmission line lightning flashover trip.In the transmission line lightningstrike probability prediction algorithm, first, the target area is divided into0.01°x0.01°(approximately1km2) grid cell, meanwhile the grid cell size is calculated; Secondly, the former16min cloud-to-ground thunderstrike measured by LLS with1min interval are divided to12pieces with duration of5min; Next, for each piece of data an clustering algorithm DBSCAN will eliminate noise in the data, and based on the geographical coordinates each piece of data matches with the grid cell to compute grid parameters; Finally, using the polynomial fitting algorithm to forecast the next10min moving path of thundercloud, using inverse distance weighting interpolation algorithm to predict the coverage area of the thundercloud of the corresponding time, and according to the minimum width trigerring transmission line lightningstrike of each grid, the regional topography where the gird is positioned, and the propotion of the hit points in the minimum width trigerring transmission line lingtningstrike and the hit points in the rest area in the grid, we can predict transmission lines lingtningstrike probability of each grid in10min in the future.In the transmission line flashover probability prediction algorithm, first of all, based on the proposed IEEE cumulative probability of lightning current amplitude, using the last time statistics for each forecast thundercloud lightning current amplitude, the cumulative probability of lightning current amplitude is calculated for each section of the time periods; then, based on the standard-recommended procedures the rate of shielding failure and the rate of an arc caused by lightning strike are calculated, based on which the probability of transmission line lightning trip can be obtained, and Depending on this, the probability of the transmission line lightning trip correction based on the number of lightning strikes that may occur is finally caculated. Algorithm testing shows that:the probability of transmission line lightning trip prediction algorithm proposed in this paper can predict the probability of transmission lines lightning trip probability, and gives the quantified value, which is of great engineering value.