Automatic Identification System Of High Voltage Isolating Switch For Substation Intelligent Inspection Robot

This paper focuses on the development trend of smart grid,and the application of intelligent inspection robots in process control of substation is analyzed.Combined with the difficulty of the current robot automatically identifying high-voltage disconnector switching state,this paper proposes an object recognition algorithm based on local feature to realize the state matching of high voltage isolating switch during switching operation.According to the instructions issued by the base station layer through the magnetic track or GPS intelligent navigation,the intelligent inspection robot with a visible light camera moves to the target device,and after adjust the tilt angle to capture the photo of the isolating switch,analyzed and identified its state.In this process,there will be scale changes,viewing angle translation or rotation and other issues of the target,the complex background of the substation equipment and the transformed weather will bring noise and light interference,that's a big challenge to the matching recognition.In this paper,we first propose a new information histogram based on SIFT key point detection,which is different from the traditional histogram,which can achieve a high degree of matching between the information content of pixel gray level bearer and its importance.When the histogram is applied to image enhancement,it can effectively reduce background noise and highlight image detail.In the pre-treatment of the photo of isolate switch,it has obvious improvement on the illumination effect and noise interference.In this paper,an automatic recognition algorithm based on SIFT key point matching is proposed,which can effectively overcome the influence of scale,translation and rotation on the matching of target recognition,and has good adaptability.In order to realize the status automatically,the paper chooses the sub-gate state template and the closing state template diagram for the high-voltage isolating switch.The real-time graph and the template map are key points with setting a matching threshold to realize the switch judgment.In the SIFT key point matching analysis experiment,the selection of template has a great influence on the accuracy and efficiency of matching,the local image template is chosen for the analysis,the experimental results show that the recognition accuracy rate is 99.75% and the recognition efficiency is high.