Research On The Recognition Of Seeding-strip And The Controling Of Path Tracking For Agricultural Robot

Nowadays,the labour costs are rising in agricultural production.Thus,developing agriculture's intelligent to improve efficiency of agricultural production and reduce the labor intensity in agricultural production is in urgent need.Aiming at the problem of agricultural robots' walking autonomously in field operation,the recognition of seeding_strip and path tracking control system based on embedded system was designed in this dissertation.which makes the seeding_strip be identified accurately and the robot's tracking path be planned and tracked in the non-seedling strip.In this Dissertation,the control systems of seeding strip's identification and path tracking of agricultural robot are studied further.Firstly,on the hardware side,based on the embedded Exy4412 core board,part of the relevant circuit interface was redesigned in dissertation,which includes USB interface circuit,serial port interface circuit,mesh interface circuit,LCD interface circuit and DAC8563 module circuit to build the hardware platform of software running.Secondly,On the software side,based on ARM-Linux operating system,the key algorithms was studied in depth in this dissertation,which combines modern image processing technology with the working environment of agricultural machinery to study the algorithm of seedling identification and the acquisition method of path planning and pose information of agricultural robot.This paper designs the PID controller algorithm combined with MATLAB simulation and describes the method of controlling the robot follow the path based on acquired path and location information.applying the modern control theory.Then,the framework of Linux device driver was researched to design the DAC8563 driver program.Thirdly,the arm-linux operating system was transplant to the hardware platform in this dissertation.By the use of modularization thinking in designing software system,the software of the modules has been finished in this dissertation,which include image-collecting model,image processing module,and the DAC8563 driver module.The image-collecting model and image processing module is based on OpenCv and Qt library by using C++ language to achieve program code.The driver module DAC8563 program was writed in C language,compiled by Cross compiler and downloaded to the embedded Exy4412 to realize function of each module.By testing each module,the results show that various modules could work steadily and accurately.Both robustness and real-time performance meet the design requirements.