Research on Automatic Sampling Technology of Transformer Oil Using Robots

Transformer oil is an important insulation medium, and its performance will decrease after long-term operation. Therefore, in daily operation and maintenance work, transformers need to regularly take oil for oilization tests. However, there are the following problems with manual oil sample collection: firstly, it is difficult to ensure personal safety during the oil sample collection process; secondly, the quality of oil sampling is greatly affected by the operational level of personnel; thirdly, the efficiency of oil sampling is low, and the operation of traditional mechanical valves is relatively cumbersome. In response to this issue, this article designs an automatic transformer oil sampling robot, which uses a mechanical arm to drive a gripper to grab and rotate the oil valve, and uses a screw motor to drive a syringe to sample oil. In order to achieve reliable grasping of oil valves by robot grippers, a method for estimating the pose of oil valves based on semantic segmentation was proposed. The experiment was conducted in the on-site environment of a 220kV substation, and the results showed that the proposed algorithm achieved an actual success rate of 83.3% for grasping oil valves, which can meet the requirements of automatic sampling of transformer oil.