Key Technology Research And Test Of Broccoli Selective Harvesting Equipment

China is a large country of broccoli planting and production,and its planting area and total output rank among the top in the world.The planting area of Zhejiang Province accounts for about 1 / 4 of the country.At present,China ’s broccoli harvesting industry is a labor-intensive industry,which is in the stage of manual harvesting,with low mechanization and automation.With the expansion of planting area and total output,the defects of low efficiency of manual harvesting and high labor cost limit the development of broccoli industry.Based on the above situation,this paper takes the field broccoli as the research object,takes the automatic and mechanized selective harvesting as the goal,and studies the broccoli harvesting equipment.Combined with the experimental analysis of the biological characteristics of broccoli,the broccoli harvesting end-effector was designed and optimized.A field broccoli recognition algorithm based on Laws filter and Gabor filter is proposed.Based on the isochronous linear interpolation,the path and trajectory planning of the harvesting route of the Delta manipulator were carried out.The physical prototype of broccoli harvesting system was developed and the harvesting test was carried out.The main research contents of this paper are as follows :1)Design and optimization of harvesting end-effector.Analyze the agronomic requirements of broccoli harvesting,build a broccoli biological characteristics test bench,and obtain broccoli biological parameters and broccoli stem cutting force characteristics.Based on the test data,the harvesting end-effector is designed.The structure is optimized by dynamic analysis and genetic algorithm,and the physical prototype of the end-effector is built.2)Visual recognition and positioning module design and construction.A field broccoli recognition algorithm based on Laws filtering and Gabor filtering was proposed,and the F1 score was 95.67 %.For different shooting environment image sets,F1 score remained above 94 %.The pixel center and pixel diameter of the broccoli head are obtained by least square circle fitting.Through the analysis of imaging and ranging principle,the real diameter is obtained,and the selective recognition of broccoli head is realized.The depth camera is used to obtain the three-dimensional positioning and depth information of the broccoli head,and the coordinate system transformation is realized by hand-eye calibration.3)Manipulator module analysis and moving platform motion planning.The kinematics model of Delta manipulator is established,and the degree of freedom and inverse kinematics analysis are completed.The fifth-order polynomial motion trajectory and the improved ’ L ’ path are selected as the motion trajectory and path of the moving platform of the manipulator.Through isochronous linear interpolation,the motion law of the moving platform and the active arm is analyzed,which lays a foundation for the driving control of the manipulator.4)The key module prototype construction and harvesting test of broccoli harvesting equipment.The control system of harvesting equipment is designed.The portable computer is used as the upper computer to complete the communication between each module,and the virtual prototype and physical prototype are built by integrating each module.The cutting performance test bench of the end effector was built,and the cutting success rate was 93.3 %,which verified that the cutting performance of the tool was good.The field planting environment was simulated and the harvesting experiment was carried out.The success rate of harvesting was 90 %,the harvesting efficiency was 6 / min,and the degree of broccoli head damage was low.The feasibility of the key technology of selective harvesting of broccoli was verified.