Research On Autonomous Charging Technology Of Mobile Robot Based On Visual Guidance

In recent years,with the gradual deepening of artificial intelligence research,mankind has entered the era of intelligence.But the problem of robot power supply has always troubled us.So,how to solve the robot's autonomous identification and recognition of ordinary indoor wall outlets for self charging has become the focus of current research.Therefore,the autonomous charging of robots by visual guidance is of great practical significance.Firstly,the method of mobile robot charging recognition socket and contour template matching theory are introduced.The charging socket template and the socket to be identified are processed by image morphology and their characteristics are extracted.At the same time,the image matching between the template and the target socket is realized in the vc2013 environment,and the feasibility of the method is verified.The theoretical knowledge of visual camera system calibration is expounded,and the calibration method is used to calibrate the visual camera,and the calibration results of the visual camera system are given.From this,the distance and angle information of the mobile robot in different locations are measured.After completing the visual recognition and measuring distance angle information,the mobile robot needs to move to the charging socket extremely accurately,so we set up the motion control scheme of the mobile robot.The concept of the mobile robot under the nonholonomic constraint is expounded,the kinematic model of the mobile robot under the non complete constraint is set up,and the Ba is used.The attitude controller is designed by the ckstepping idea.The robustness and convergence of the designed controller are verified by simulation in Matlab environment.In the actual test,it is found that the form of template matching recognition is single and inefficient,and often appears unrecognized.Therefore,we add the S-SIFT operator to the recognition of the charging socket.The experiment shows that the addition of the S-SIFT operator has a better recognition efficiency for the different angles,distance and rotation of the charging socket.Do not have higher precision.At the same time,the recognition time is greatly shortened.Finally,the mobile robot experimental platform was built,the hardware structure of the experimental platform and the software library needed for the development were introduced.The autonomous charging strategy of the mobile robot was summarized.The experiment showed that the autonomous charging effect of the mobile robot was obvious,and the function of autonomous recognition and charging was realized.