|
 WH78 |
| WR82 |
| WHX124 |
| WHX132 |
| WHX150 |
| WHX155 |
| WHX157 |
| WHX159 |
| SS-2184 |
| SS-800 |
| 81X,81XH |
| M315 |
| M450 |
| M630 |
| M900 |
| 9835M |
| 110M |
| 111M |
| 150M |
| 5907D |
|
 |
| BACK |
|
|
High-Voltage Power Transmission |
|
Inspection robot system adopts a mobile robot as carrier. Themobile robot is capable of carrying a variety of inspection instruments (e.g. visible light camera, infrared! ultravioletcameras, defect detector, and crossing distance meter, scanpan, etc.) and moving along power transmission to fulfil lineinspection tasks. Using mobile robot to inspect overhead highvoltagetransmission line(hereinafter referred to as inspection robot) has a lot of advantages. The robot can reach whereverthe line extends, inspect with high precision in close range,without inspection blind-area, be much cheaper than the helicopter inspection. And it is especially fit for the inspectiontask when crossing high mountains, virgin forest ,great riversor other geographic circumstances unreachable for manualwork. The study of the inspection robot has become a hottopic in both application and research field of special robot athome and abroad. Since the transmission line is hanging in thewild with high voltage and high current, vibrating or dancingbecause of the wind inspiration, and also the geographical environment the line experienced, the inspection robotpractical in project should move autonomously, inspectautonomously, and locate with the robot and inspectionpictures data information relative to the line.
So mobilenavigation, location and non-collision obstacles overcoming are the key technological problems in the autonomous are the key technological problems in the autonomous inspection robot study The robot named LineScout[I-2] developed by Canada Hydro Quebec Research Institute in 2006, and the robot "Expliner" [3] created by Japan Kansai Electric Power Company (KEPCO) and the Japanese Electric Power Systems, Inc. (JPS) in 2008 are operated by remote control, without autonomously move function. Peters and his colleagues[4] use multi-distance sensors to measure the same obstacle, and classify obstacles with rough neurocomputer approach. Shenyang Institute of Automation, Chinese Academy of Sciences also developed a local autonomous robot along overhead ground line. Switch-type laser sensor is installed on the bottom of the mechanism arm to detect whether the arm has aimed at the ground line. And there is a camera on the mechanism arm to serve the vision control when the arm is grasping the line. They also use machine vision method to the robot's local autonomous control, which means that it uses the line's geometry constrains to do the obstacles' visual detectionand use two dimensional image to calculate the gripperposition relative to the line in order to serve the obstacleovercomingvision control, The research above is the avoidedcollision problem on local autonomous robots, Paper[9]-[ 101,which also introduce inspection robots, didn't presentnavigation, location and non-collision problem.
A transmission chain consists of tension towers,suspension towers, live-lines and hard wares , including the transmission line's passage way. No matter how long the lineis, we can describe it into two principle kinds of passageway(Fig. I (a)and (b) and adjacent tower sections.
1 If the robot's path is a live-line(A phase), its movingenvironment can be abstracted as Fig . I (c) . It makes infraredand visible images of the three phase lines(A/B/C) and theirhard wares, ground lines(G I/G2) and their hard wares, towers,passage ways during the robot is moving on the A phase line .
The inspection robot needs to know the inspectionenvironment and the moving environment for its autonomousinspection, autonomous movement. The inspection environment(Fig. I (ajandrb) means the scan pan(PTZ),focusing parameters the camera required, including thefollows: (l)three phase lines, two ground lines spatialposition relative to the robot;(2)towers' height;(3)spatialstructure dimension of the passage way . As the picturesshown, there are two classic section structures. The movingenvironment means the robot's navigation, location and noncollisionenvironment parameters, including (I )obstacles typeand their location;(2)non-collision obstacles approaching andovercoming;(3)the section the robot is in, etc . The aboveenvironment information can be classified as globalenvironment and local environment considering the robot. Theglobal environment is the mechanism and the electricalequipment parameters given by workers that the robot canrequire by the tower number or section number. The localenvironment that consists of the robot navigation, location andthe local non-collision environment, which is unknown, needssensors to detect. What is more, strong electromagneticenvironment is the unique characteristic of the inspectionrobot, because it is always running on the live-line.
The inspection robot's basic architecture is shown in Fig.
2 (a), In it, (I )Inspection module: the global environment, thelocal environment of the moving path and the robot conditionare needed to be detected. GPS, TLDDB which is embeddedin the robot software are for the global environment detection,Machine vision and multi-electromagnetic sensors are for thelocal environment detection and the robot attitudeidentification during the obstacle-overcoming process.
Inclination sensors and photoelectric encoders are used for theskip identification of the driving wheels. Voltage sensors andcurrent sensors are used for the power capacity detection.
(2)Identification module: to fulfill the identification algorithmof the global environment, the local environment and the robotpose and condition. The global environment identificationuses GPS to inquire TLDDB following the search algorithm,which can outputs the local section number, the rough distanceto the tower ahead, and the tower's structureinformation(including the tower is suspension tower or thetension tower, how many dampers beside the tower, etc) .Thelocal environment rough identification module works in thisway, use visual image to count the rough distance between therobot current position and the adjacent obstacle ahead, senddecelerate signals to the fast moving robot before it comes tothe obstacle. The local environment exact identificationmodule uses multi-electromagnetic sensors' identificationalgorithm to detect the obstacle type and to exactly locate therobot's position relative to the obstacle. And the poseidentification algorithm is to identify the robot pose relative tothe live-line. Skip identification of the driving wheels and thepower capacity identification are fulfilled by the conditionidentification module. (3)Decision planning module: toaccomplish the inspection and moving decisions, motionplanning. The planning is for the obstacle types and the robotconditions. (4)Control module: to produce the control codesaccording to the decision and planning results, to dominate the control parts. (5)Execution module: to supply the moving,inspection, and power transmission.
|
|
|
