The safety of substations has an important impact on people’s production and life,which need to conduct regular inspections.However,the distribution area of substations in China is scattered,and the existing manual inspection mode has disadvantages of high labor intensity,low inspection quality and severe interference from bad weather.So replacing manual inspection with robot inspection will greatly reduce the labor burden,improve the quality of inspection,and reduce the risk of inspection.The substation inspection has high requirements on the stability of the robot and the positioning and navigation accuracy.At present,although the robot technology has achieved certain research results,but it is not practical enough;there are certain applications,but its applicability is not strong,there are problems such as large differences in performance indicators,low positioning and navigation accuracy,and vulnerability to interference.This paper mainly focuses on the above problems,based on the existing research,the omnidirectional mobile Mecanum is used to build a patrol robot platform to realize remote movement of the robot by remote control;the robot recognizes the automatic navigation and intersection recognition by the path line of visual recognition;acquiring the encoder position information,RFID electronic tag,attitude angle sensor and other information to achieve precise positioning of the robot;the robot’s motion parameter debugging and trajectory map generation are realized through the man-machine interface.The main contents of this paper are as follows:1)The overall structure and scheme of the robot platform are determined by analyzing the research status of the robot and its positioning and navigation technology at home and abroad.The motion model of the robot is established to realize the remote motion control of the robot.2)The robot master controller and remote controller are designed with STM32F429VGT6 as the control core.At the same time,robot motion control,sensor data acquisition and data interaction are realized.The positioning system of the robot is built by a motor encoder,an RFID tag and an attitude angle sensor.3)The embedded image processing platform was built with the Raspberry Pi,which realized the visual navigation of the road path and the identification of the road intersection.Robot path navigation is implemented using PID control.4)The QT software is used to write the upper computer interface,which realizes the functions of parameter setting,positioning navigation,data analysis and real-time trajectory map generation.5)After several experiments and tests,it shows that the patrol robot designed in this paper runs stably,has good navigation effect and high positioning accuracy,and can achieve the expected effect.Through the operation of the actual road section,the reliability of the online trajectory map of the inspection robot was tested. |