基于遗传算法的圆度误差评估

将遗传算法应用于圆度误差的评定.首先简介了误差评定背景和遗传算法及其特点.然后根据尺寸和公差的数学定义^1]给出满足最小区域条件的圆度公差评定的数学模型和适应度函数.接着,以最小二乘解作为初始值,对圆度误差的遗传优化过程进行了详细的论述.最后用实例对算法进行验证.优化过程和实验结果显示了遗传算法在解决形状公差的评定这类非线性问题的优越性,通过并行搜索能最大限度地保证解的全局最优,计算精度高、效率高,且易于理解和实现.

Circularity Error Evaluation Using Genetic Algorithm

A genetic algorithm based method for form evaluation of a circle is proposed. Firstly, after an introduction to the background of the form error evaluation method and that of the genetic algorithm, a mathematical model of circularity error together with a fitness (evaluation) function is given in terms of the Mathematical Definition of Dimensional and Tolerancing Principles , which corresponds to the minimum zone definition and is easy for interpretation and computer implementation. Then, with the least square solution as the initial value, the whole implementation process of the genetic algorithm in the circularity evaluation is dwelt on. Finally, the simulation results of a quoted example are analyzed and a brief conclusion is given. The example simulation indicates that genetic optimization algorithm has such distinctive advantages as solving nonlinear and large space problems, converging to global optimum solution and convenience of implementation, etc., and it is quite suitable for the form error assessment.