Abstract: In order to make the transplanting machine have better planting performance, a cam swing rod planting mechanism is proposed, to ensure good planting erectness, uniformity and planting depth pass rate, it is necessary to analyze the length and position of the connecting rod part of the planting mechanism, establish the corresponding mathematical model on the basis of the working principle of the planting mechanism, and obtain the corresponding design variables, constraint functions and the target trajectory function of the planting point through analysis. Using fgoalattain in MATLAB optimization toolbox to solve the complex multiobjective nonlinear programming problem of transplanting machine planting mechanism, a set of optimized data with a relatively complete set of design variable parameter combinations［L1、L2、L3、L4、L5、L6、θ2、θ3、γ］ can be obtained, and the optimal parameter combination is ［128 mm；52 mm；150 mm；120 mm；336 mm；60 mm；256π/180；316π/180；8π/180］, and then can obtain the trajectory curve, pendulum curve, velocity curve and acceleration curve of the planting point, with the plant distance variation coefficient, uprightness and planting depth pass rate as evaluation index, record the corresponding transplanting data, when the speed is 300, 400 and 500 mm/s, the pass rate and excellent rate of transplanting have reached more than 90% and 80% respectively, which can better meet the requirements of transplanting operations than before optimization.

Key words: planting mechanism, mathematical model, MATLAB, multiobjective optimization

摘要： 为使移栽机具备较好的栽植性能，提出一种凸轮摆杆式栽植机构，为保证良好的栽植直立度、均匀性以及栽植深度合格率，需要对栽植机构连杆部分的长度及位置进行分析，在该栽植机构工作原理的基础上建立相应的数学模型，通过分析得到相应的设计变量、约束函数以及栽植点的目标轨迹函数，利用MATLAB优化工具箱中fgoalattain求解移栽机栽植机构复杂的多目标非线性规划问题，可以得到一组较为完善的设计变量参数组合［L1、L2、L3、L4、L5、L6、θ2、θ3、γ］的优化数据，经过迭代计算最优参数组合为［128 mm；52 mm；150 mm；120 mm；336 mm；60 mm；256π/180；316π/180；8π/180］，进而可以得到栽植点的轨迹曲线、摆角曲线、速度曲线以及加速度曲线，以株距变异系数、直立度、栽植深度合格率为评价指标，记录相应的移栽数据，当速度为300、400、500 mm/s时，移栽合格率及优良率分别达到90%、80%以上，相较于优化前可以更好满足移栽作业的要求。

关键词: 栽植机构, 数学模型, MATLAB, 多目标优化