Visual Servo Control Method Of Dry Ice Cleaning Robot For Substation Pillar Insulators

Because the pillar insulators in the substation are in the high electric field intensity environment and the open-air natural environment for a long time,various industrial pollution and natural pollution are easily deposited on the surface of the insulator,causing the insulator pollution flashover phenomenon,which brings security risks to the safe and stable operation of the entire substation.This thesis designs a substation pillar insulator cleaning robot based on the visual servo control method,which can realize efficient and safe cleaning of insulators in the state of no power failure,solve the impact of power failure cleaning on the quality of substation power supply,and solve the safety of manual cleaning.Hidden dangers and cleaning dead corners.At the same time,in order to improve the cleaning efficiency and not damage the surface of the insulator,dry ice cleaning is specially used as the cleaning method of the insulator.This thesis firstly analyzes the constraints of the insulator cleaning robot in the complex environment of live work in substations,and conducts design verification of the robot as a whole and important components according to geometric constraints and electrical constraints,to ensure the safety of the robot live work.Then,the corresponding research on the robot visual servo control method is carried out,and the target recognition and positioning and contamination detection,as well as the robot motion trajectory control method and the dry ice cleaning control strategy are analyzed and designed respectively.The vision system identifies and locates the pillar insulators using the YOLOv4 detection algorithm to perform real-time detection and two-dimensional positioning of the target insulators to achieve a balance between detection accuracy and recognition speed,and the overall recognition rate reaches over 95%.The relative depth information is used to convert the two-dimensional coordinates of the plane to obtain the real three-dimensional pose of the insulator to complete the positioning,and the overall positioning accuracy is within 1cm.After completing the identification and positioning of the insulator,the insulator’s umbrella plate inclination angle information is also extracted to adjust the cleaning angle of the end to achieve no dead angle cleaning;by detecting the degree of contamination on the surface of the insulator,the cleaning strategy can be adjusted in real time to achieve the best cleaning effect.The research on the control method of the robot motion trajectory is firstly established by the D-H method to establish the coordinate model of the robot connecting rod,and the accuracy of the model is verified by the forward and inverse kinematics analysis of the robot.According to the analysis of the operation process,the motion planning of the robot’s operation trajectory is carried out,and the trajectory planning effects of the cubic polynomial interpolation method and the quintic polynomial interpolation method are compared under the MATLAB simulation software,and the quintic interpolation method with a smoother trajectory curve is selected.Finally,the fuzzy-PID controller is used to control the joints of the manipulator in real time,which improves the tracking accuracy of each joint and realizes precise operation control.In addition,the dry ice cleaning process is controlled by the BP neural network PID controller,and the cleaning strategy is adjusted in real time according to the degree of contamination to achieve optimal cleaning.Finally,the prototype of the robot designed in this paper is developed,the corresponding hardware modules and software system design are analyzed,and the pressure test and on-site cleaning operation test are carried out in the Hunan Provincial Disaster Prevention and Mitigation Center.The reliability of the robot’s visual servo control system for efficient cleaning of different types of insulators in a live working environment was verified,and the data comparison verified that the robot improved the cleaning efficiency by 5%compared with manual operations,and the cleaning time was 8 minutes.