基于虚拟响应面分析的水稻插秧机分插机构参数优化与试验

针对现有高速分插机构参数优化方法盲目性较大或计算复杂的问题,以椭圆齿轮三插臂分插机构为例,采用虚拟中心组合响应面试验设计方法,以插秧株距140mm时的插秧穴口宽度、推秧角度和轨迹高度3个指标为目标函数,以椭圆齿轮偏心率、齿数、行星架初始安装角度和插植臂安装角度4个参数为设计变量,构建响应面回归模型,采用优化算法求解响应面近似回归模型,直接求得一组满足多个约束条件的最佳参数组合。以求解的最佳参数组合:椭圆齿轮偏心率0.18、齿数23齿、行星架安装角度15°、插植臂安装角度25°设计并试制三插臂分插机构。通过理论分析和台架试验,回归模型计算的目标函数值与理论计算值和台架试验实测值之间的最大相对误差为0.668%。分析结果表明:采用虚拟响应面法优化可直接求得满足多个约束条件的参数组合,计算精度较高,提高了高速分插机构参数选择的针对性。

Parameter optimization and experiment of transplanting mechanism of rice transplanter based on virtual response surface analysis

The existing method to optimizing parameters of transplanting mechanism was usually blindness or computational complex.Taking three separating planting arms of elliptic gear transplanting mechanism as an example,a response surface regression model was built by adopting virtual central composite response surface experiment design,of which planting-hole width at the row spacing of 140 mm,push seedling angle and track height were adopted as objective functions,and centricity ratio of elliptical gear,teeth number,initial angle of planet carrier and initial angle of planting arm were taken as design variables.An optimum parameter combination,which meets multi-constraint conditions,was directly acquired by using optimization algorithm to solve the response surface regression model.The three separating planting arms of elliptic gear transplanting mechanism was designed and trial-produced according to the optimal portfolio parameters:centricity ratio of elliptical gear of 0.18,teeth number of 23,initial angle of planting arm of 15 and initial angle of planting arm of 25.The maximum relative error between the approximate value of regression model with the precise value of the theoretical analysis model and the measured value of test-bed was only 0.668% according to theoretical analysis and bench test.The result indicated that we could directly acquire the optimum parameter combination to meet multi-constraint conditions by adopting virtual response surface optimization method.Its computational accuracy was relatively high and it could improve the pertinence in selecting the parameters of transplanting mechanism.