Study on the method of switch state detection based on image recognition in substation sequence control

The key question in substation sequence control becomes whether the devices executed the instructions sent by the control center and reached to the target state. The general criterion commonly adopted is the signal of remote measurement, remote signal and Five-proof Blocking System. In this study a new assistant criterion based on image recognition is discussed. The paper carries on the analysis of the framework of multi-criteria sequence control system, and then, based on Digital Image Processing, a method recognized the target state of switch in substation is proposed. It recognizes switch state by extracts switch character-SIFT, constructs KD-tree, and matches image features based on cluster analysis. Our proposed method is verified through experimental study on conditions of salt and pepper noise, varying Illumination and image rotation, the experimental results reveal the method yields an improvement in enhancing reliability of sequence control. The whole system has deployed in a 110KV substation, and it acquired favorable effect.