| In the process of the modern manufacture,more and more AGVs are applied to aid in carrying merchandise in the fields such as manufacturing industry and play an important role.It can promote the efficiency and competitive edge greatly that designing and developing a functional and suitable-priced AGV with embedded system and make it serve for carrying things in the maching factories,which will have a significant effect in the society and economy development.On the basis of analyzing the recent and important research achievement about AGV at home and abroad,it is designed that the automation mechanism and motion control system solution of AGV in detail in this dissertation,and has developed a muli-functional visual-guided AGV prototype.The main research contents are as described below :For the mechanism design,a solution of Mecanum-wheels is chosen to make the AGV movable in every direction,and then the motion-velocity Jacobbi matix of the AGV is analized and computed.The force analysis between the AGV and the ground is executed to determine the specifications of the servomotor.Besides,it is mainly designed that the automatic mechanism connecting with the docking shelf.Using the function of line tracking on the ground and the RFID positioning solution,the relative position between them is adjusted by using Kinect to recognize the central color bar of the docking shelf to adjust the position precision.The main controller controls the push rod and other components to connect with the docking shelf.In addition,the 3D model of the automatic charging mechanism is designed.Utilizing the the AGV's patrol motion and the sensors it carries,the automatic-charging is executed after the initial positioning and accurate positioning adjustment of the charging position,then the charging pile is docked and begins to charge.For the control system solution of AGV,it is separated into the main part and the subordinate part based on the operating system of Linux and the open-source ROS software framework,of which the main one is in charge of the coputation and decision functions while the subordinate one is in charge of the motion-control in servo-motor.It is the Raspberry pi playing as the main controller that installs Linux operation system,and has communication with the microchip Arduino.It is well designed that the controlling process of the motor with the Arduino,and the code of the controlling program.Besides,the communication protocols between Raspberry pi and Arduino is designed,and the RFID tag's functional-serial-number of the spots in the factory is described.For the obstacle-avoidance of the AGV,it is chosen that the combination of multi-sensors incluing the Kinect camera using image-processing algorithm,the infrared-sensor for human-body and ultrasonic sensor.As for the locationing in the factory,the RFID solution is adopted.In addition,by utilizing the ROS's sensor driver module,it is realized that voice controlling on AGV's basic motion.For the rectifing deviation of the line tracking of AGV,it is proposed that an improved increamental PID algorithm with integral-separated and anti-integral-saturated.By setting a proper threshold value of the integral-separated for the AGV,it's applied to the rectifing deviation of the line-tracking of AGV,which enables AGV to rectify the errors of line-tracking which is caused by the ground's mutual friction,meaning that it can move more relible and stable and will not lose response to the controller.Lastly,after manufacturing and assembling an AGV prototype,and code-and-debug repeatedly,some experiments are done to test the performance of the motion-control of the AGV,the accuracy of the locationing,the feasibility and reliability of the controlling mechanism of the auto-connection mechanism with the docking shelf,as long as the autocharging.And it proves that they all work and are practical,which means that it's qualified to the requirement of the project. |
PDF Full Text Request