| The automatic packing operation of fruits and vegetables replaces the current manual operation,which fundamentally eliminates the potential safety problems caused by the long-term coexistence of human and large machines in operation,avoids the food safety problems caused by the direct contact between workers and fruits and vegetables,and guarantees the work efficiency and stability.Therefore,in order to realize the security of automatic packing,this paper takes apple as an object to study.The structure design of automatic packing manipulator,the algorithm design of automatic packing path optimization of manipulator and the design of apple position and attitude recognition system are studied.The main research contents and conclusions are as follows.(1)Structural design of automatic packing manipulator.According to the actual working environment,a structure model of horizontal rotating manipulator suitable for apple automatic packing was designed,and the machining material,average thickness and width were used as test factors to determine the specification parameters of manipulator links.The driving torques of the two rotatory joints of the manipulator are used as the basis for selecting the driving motors of the manipulator.The safety and reliability of the manipulator by finite element analysis.The test results show that the specification parameter of the manipulator Link1 and manipulator Link2 are aluminum alloy 6061 with an average thickness of 10 mm and 8 mm,and an average width of115 mm and 90 mm,respectively.The driving torque of Joint1 and Joint2 are34.0680 NM and 3.3978 NM,respectively.According to the finite element analysis,the maximum stress of Link1 and Link2 are 12.69 MPa and 14.34 MPa,and the maximum displacement are 0.3730 mm and 0.3617 mm,respectively.The maximum stress of the ladder axle in Joint1 and Joint2 are 87.15 MPa and 124 MPa,the maximum displacement are 0.01081 mm and 0.01295 mm,respectively.The above four important parts of stress and displacement are affordable,and meet the design requirements.The test results show that the structure design and specification parameters of the manipulator conform to the actual grasping requirements and safety specifications.(2)Design of automatic packing path optimization algorithm for manipulator.The goal of path optimization is to minimize the total packing time.Based on this,this paper proposes an improved genetic algorithm fused with the optimal parameter selection algorithm as the path optimization algorithm,including the improved genetic algorithm and the optimal parameter selection algorithm.The implementation process of improved genetic algorithm is to define the rule of chromosome generation,set fitness function,crossover operation,mutation operation,comparison and replacement operation,etc.The realization process of the optimal parameter selection algorithm is to use SPSS to fit the target path time and program running time of the output of the improved genetic algorithm corresponding to each set of parameters and 365 sets of non-repeating parameter values,which include population size,crossover probability,mutation probability and comparison substitution probability.Then the optimal parameters suitable for the improved genetic algorithm are obtained by means of exhaustive method.In order to reflect the optimization performance of the improved genetic algorithm fused with the optimal parameter selection algorithm,the improved genetic algorithm fused with the optimal parameter selection algorithm and the traditional genetic algorithm which uses the fixed empirical value are tested for comparison.Test results show that in the premises of 4 grasping points and 14 placing points of the cross experiment,the total time of improved genetic algorithm with optimal parameter selection algorithm is 23.86 s,which is less than the total time required by the other three experimental algorithms.The comparison shows that the proposed algorithm is advanced.Among the experimental results of optimization performance comparison of different grasping points and placing points,it is also shown that the improved genetic algorithm,which integrates the optimal parameter selection algorithm,has better optimization effect and universality.(3)Design of apple position and attitude recognition system.The design of apple position and attitude recognition system includes design of hardware structure and software algorithm.Among them,the hardware structure mainly includes image acquisition device and apple position and attitude adjustment mechanism,and the software algorithm mainly includes apple position recognition and apple attitude recognition.Image acquisition device is mainly composed of camera,lens,lighting source,detection source,etc.The apple position and attitude adjustment mechanism is composed of two synchronous wheels and a drive motor.By using image segmentation,image noise reduction and target positioning,the position of apples in the image are recognized among multiple apple targets,and the final apple position is determined by mapping with the rectangular coordinate system.Then,the attitude of the apple is identified by two experimental schemes according to the performance characteristics of the detection light source,and finally determined to be able to be grasped.The experimental results show that the recognition accuracy of apple position recognition algorithm is 100%,indicating that the position of multiple apples is in good condition.The accuracy of the two apple attitude recognition methods are higher than 96.60% and98.79%.The time of each time in apple position and attitude recognition reach 1.64 s and 1.31 s respectively.The above research results provide a safe and effective packing method for fruits and vegetables,a new design method for the automatic packing manipulator,and a certain method and technical basis for the path optimization of the manipulator and the attitude recognition of apple,which is helpful to realize the safety and automaticity of apple packing. |
PDF Full Text Request