Abstract: Aiming at the problem of robust continuous parameter design in the target being best, in which the output value can be got by theoretical calculation, an optimization method based on genetic evolution is proposed. The researched problem in this paper is described. The technical idea of the optimization method based on genetic evolution for the researched problem is presented. In the proposed method, the discretization is replaced by dense sampling, the experiment scheme is replaced by individual, the fixed internal table is replaced by transformable population, and the optimal design scheme is got through genetic evolution process. The genetic algorithm for robust continuous parameter design in the target being best is presented and designed. The calculation flow, the individual coding, the individual fitness, the population initialization, decoding operation and genetic operation are described. The effectiveness of the proposed method is validated by case study.

Key words: robust parameter design, target being best, continuous parameter, optimization method, genetic algorithm

摘要： 针对一类可通过理论计算得到输出特性值的望目特性连续型参数稳健设计问题，提出了一种遗传进化方法。描述了研究的问题；提出了望目特性连续型参数稳健设计遗传进化方法的技术思路：以密集抽样取代离散化处理，以个体取代试验方案，以变化的种群取代固定的内表，通过遗传进化得到最优设计方案。提出并设计了一种望目特性连续型参数稳健设计遗传算法，阐述了算法的计算流程、个体编码、适应度、种群初始化、解码操作及遗传操作。通过案例分析验证了所提方法的有效性。

关键词: 参数稳健设计, 望目特性, 连续型参数, 优化方法, 遗传算法