旋转式分插机构运动学多目标非劣解群自动寻求  

Automatic Search of Pareto Solutions of Multi-objective for Rotary Transplanting Mechanism Kinematics

作　　者：俞高红[1] 何琰[1] 陈建能[1] 赵匀[1] 

机构地区：[1]浙江理工大学机械与自动控制学院,杭州310018

出　　处：《农业机械学报》2009年第6期47-52,共6页Transactions of the Chinese Society for Agricultural Machinery

基　　金：国家自然科学基金资助项目(50605057);"十一五"国家科技支撑计划资助项目(2006BAD28B01);浙江省自然科学基金资助项目(Y107361)

摘　　要：针对水稻插秧机分插机构优化过程中的多目标、非线性、交互性、模糊性以及优化结果不唯一等问题,建立了运动学参数的计算模型,通过对插秧性能各要求的模糊化得出各要求的隶属度,同时提出了对干涉的判定方法。在以各隶属度最大化为目标,运用优化算法求得一组较优解的基础上对该组解进行了区间扩张,将区间搜索转换为多目标优化问题,再次运用优化算法并通过Visual Basic 6.0软件编程实现了自动搜索非劣解群的目的。结果表明:利用该方法可以迅速搜索得到运动学参数的范围,较人机交互优化方法更为快捷、方便,提高优化效率。Aiming at the problem of multi-objective, nonlinearity, interaction and fuzziness of transplanting mechanism of rice transplanter in optimization process, calculation model of kinematical parameter that could satisfy the performance requirements for transplanting seedlings was established. Firstly, by blurring the performance requirement, the membership was obtained and interference judgment method was put forward. Then the optimization model was built up, which took the maximum of every membership as the object function, and adopted motion non-interference as the restrictive condition. Based on the optimal result, the solution was expanded as an interval and the interval reaching was converted as multi-objective optimization. Then automatic search of the non-inferior solution was realized by coding the optimization program with Visual Basic 6.0. The results showed that by using the method it could quickly search the interval that could meet the kinematical performance requirement and it was easier and more efficient to manipulate than human-computer conversation.

关 键 词：旋转式分插机构 复优化 多目标 数值化 区间搜索 

分 类 号：S223.91[农业科学—农业机械化工程]