Research On Intelligent Mobile Robot Based On Path Planning And Image Recognition Technology

As the size of grid increases and the level of intelligence increases,substation management and control are gradually developing towards intensive management and intelligence.The inspection machine is used to replace the manual inspection in part or in full,which has become an inevitable trend of equipment inspection.This study intends to build 4 substations near Yi Chun.It plans to equip the intelligent inspection robot.It uses automatic navigation technology,visible light image and thermal infrared image recognition technology,and sound detection technology.The automatic navigation of inspection robot is the basis for performing the inspection task in the substation.This study starts with the trackless navigation method.The point cloud matching algorithm of lidar was selected to establish the navigation map.The vehicle positioning information was obtained by using the navigation technology and multi-sensor fusion technology.Moreover,genetic algorithm was used to plan the task path of robot for the layout characteristics of electrical equipment in different substations.The simulation test show that it can provide a method for the inspection robot to perform the self-inspection task with high precision.For the detection of electrical equipment and the reading of instrument values in the substation.In order to achieve the same scene,both visible light and thermal infrared images were acquired.This study used a binocular vision system for image acquisition.The semi-global matching algorithm was used to smooth the constraints of image.Template matching and LARK were used for feature matching of two images.The result shows that the effect of image enhancement was obvious,the accuracy of image positioning,recognition and assisted navigation of image matching was high,which meet the inspection accuracy requirements of different substations of robots.