The current level flowing through the circuit in the opening and closing states is not the same. The traditional methods can not respond to the change of current level in time

resulting in delayed or inaccurate identification results. Therefore

in order to effectively identify the switch status

a HOG based method for identifying the three-phase opening and closing status of the isolation switch at the outlet of hydropower station units is proposed.Normalize the HOG feature vectors

and based on this

combine the sampled edge nodes of the isolation switch image to calculate their gradient size and direction

thus defining the Hough Transform principle to achieve edge transformation of the isolation switch image at the outlet of hydropower station units based on HOG. Using the recognition threshold

determine the opening and closing behavior status of the isolation switch at the outlet of the hydropower station unit

and complete the design of the recognition algorithm. The experimental results show that based on the above method

the closing and opening states of the isolation switch at the outlet of the hydropower station unit can be accurately identified according to the different values of the circulating current.