Study And Test Of Detecting Robot's Obstacle Avoidance Technology Based On The Circle Structured Light In Reservoir Culvert

As a upstream and downstream water conveyance channel,reservoir culvert plays an important role in the irrigation of downstream farmland.They are generally located at the bottom of the dam and buried deeper.Large diameter reservoir culverts can be observed to detect the internal situation of pipes.However,small and medium-sized reservoir culverts are generally smaller in diameter which is about 500 mm,and people can't enter internal detection which can only be judged by experience.Therefore the inspectors are often helpless in abnormal water delivery situation.Detecting the interior of culverts by pipeline robots will obviously improve the rescue capability of inspectors,which is of great significance to further reduce potential safety hazards.At present,there are many types of pipeline detection robots,which play an important role in certain pipeline detection.However,these pipeline robots detect the subtle abnormality of the inner wall of the pipeline,such as pipeline crack,which has no effect on the detection robots working inside the pipeline.Therefore,the detection robot isn't required to avoid obstacles in the pipeline.For the special type of culvert pipes,there are many hidden dangers in the pipeline,not only the narrow cracks on the pipe wall,but also the obstacles inside the pipe,which mainly are the large pits on the inner wall and the blockage on the wall.All of these hinder the continuous operation of the robot in the reservoir culvert.Therefore,the pipeline robot must not only detect the obstacle information inside the pipeline,but also carry out subsequent obstacle avoidance actions to achieve continuous operation of the detection robot inside the reservoir culvert.Therefore,in this paper,the technology of robot obstacle avoidance was tested and studied.The main contents and results are as follows.1)Contacting a visual obstacle avoidance system for reservoir culvert detection robot,which includes detection of mechanical structure,obstacle location and obstacle avoidance research.The mechanical structure of the detection robot is designed,including the obstacle detection system and its hardware components.The obstacle location mainly includes the extraction of the two-dimensional feature information of the obstacle,and calculating the three-dimensional coordinates of the obstacle feature information based on the two-dimensional coordinates of the feature information.The research of obstacle avoidance method is to analyze the obstacle avoidance strategy and build its obstacle avoidance kinematics model2)The method of the circular structure cursor is studied,and a simple circular structure light calibration method was proposed.Ensuring that the calibration method is synchronized with the camera calibration.The oval light stripe is extracted and the contour fitting is adopted by the RANSAC algorithm.Then Obtaining the ellipse equation corresponding to the fitted contour and take a series of angularly-equidistant image points from the ellipse equation,and use the above-mentioned circular structure light calibration method to obtain the three-dimensional coordinate of the space corresponding to the series of image points.Finally,the series of three-dimensional points are used to fit the structured light surface equation and the fitting error analysis is performed.3)An image recognition of the obstacles in reservoir culvert is presented and the characteristic information of the obstacle is separated.The image coordinates of the feature points are obtained,then the coordinates of the coordinates of the world coordinate system are obtained by using the formal calibrated circular curved surface equation.Thus,the location information of the obstacle in the pipeline is obtained,and the type of obstacle including objects and pits are discriminated.4)The control system of visual obstacle avoidance of robot is designed,and the PC monitor program is written,the control program of the lower bit computer and the communication protocol between them are prepared.5)The experimental research on robot visual obstacle avoidance was carried out.The obstacle avoidance test was carried out on the inner wall pits and the wall barrier respectively and the relevant obstacle avoidance data were obtained.Then comparing the test value and the actual valu e and analyzing the obstacle detection size boundary conditions.Obstacle avoidance function can be achieved when the circumferential conditions of width greater than 13 mm and radial thickness of 30 mm to 200 mm are satisfied.By analyzing the obstacle and laser optical center distance and the comparison between the experimental value and the actual value of the deflection angle of the walking wheel,the influence of the contact pressure of the inner wall and the walking wheel on the obstacle avoidance test was obtained.