Design And Experiment Of Self-Propelled Wheat Whole Growth Cycle Phenotype Monitoring Platform

Along with the rapid growth in global population numbers,there is an increasing demand for food production and quality.Plant genomes have developed rapidly in recent years,but the lack of adequate phenotypic information has limited the exploration of quantitative trait genetics.Automated and high-throughput crop phenotype information collection techniques have an essential impact on obtaining useful phenotypic parameters and identifying crop genetic factors that can accelerate breeding for genetic improvement,increase yields and improve drought tolerance.Current phenotype information collection methods are often manual,which is not only time-consuming and labor-intensive but also subjective,inefficient,and of poor quality.With the aim of improving the efficiency and quality of phenotype data collection,this paper designs a self-propelled phenotype monitoring platform for the whole fertility cycle of wheat,and the main research contents and conclusions are as follows:(1)The overall design and main technical parameters of the phenotyping platform were determined to meet the needs of wheat phenotype information collection,and the drive mode of the platform was determined to be four-wheel independent steering and independent drive(4WS-4WD),and the chassis structure was gantry structure,and the key components were selected and designed.The simulation results show that the centre of mass fluctuates the least when the wheelbase is 1050 mm and the wheelbase is1270 mm,and the comprehensive evaluation index is 5.98.The static analysis of the platform based on the finite element method using ANSYS software shows that the maximum stress is 40.645 MPa,the maximum deformation is 0.29 mm,and the structural strength meets the design requirements.The structural strength meets the design requirements.(2)A hardware design and selection for the motion control system of a phenotypic platform chassis were carried out using a modular approach.The STM32 microcontroller was chosen as the main controller.A kinematic model of the phenotypic platform was established based on the Ackermann steering principle.A platform motion control system based on the PID algorithm was designed.The upper computer software for the chassis motion control system was designed using Python and Py Qt5.To address the impact jitter problem that occurs during platform start-up and shutdown,the torque loading method for the walking motor was optimized.Through a torque single-factor experiment,it was demonstrated that when the motor torque was loaded in a non-linear manner during platform start-up,the gimbal acceleration fluctuated minimally,with a maximum acceleration of 0.02 m/s~2.(3)To collect phenotypic information of wheat at different growth stages,the mechanical structure of the gimbal was designed,key components were selected and strength was verified.The motion control system of the gimbal was designed to achieve precise control.Visual monitoring software was designed and optimized based on SDK and Open CV to control multiple sensors.Sensor parameters were optimized using response surface experiments,and the optimal parameter combination was found to be motion speed A of 0.29 m/s,exposure time B of 2.92 ms,and resolution C of 2048×1536pixels.(4)Experimental studies were conducted on the phenotype monitoring platform on hard ground and in the field to verify its operational performance.The test results show that: the maximum deflection rate when driving in a straight line is 1.24%;the maximum deflection value of the centre of rotation when slewing in place is 55mm;the maximum average error of the mobile wheel turning angle control during four-wheel Ackermann steering is 0.10°,and the maximum deviation rate of the speed control is 2%.In the performance tests,the average range was over 7h,with a climbing capacity of 30°.In the phenotypic information acquisition test,the maximum deviation of the head motor speed control was less than 2%,and it took 40 minutes to complete the phenotypic information acquisition of a 660m~2 wheat field,and the image quality met the requirements for subsequent processing.