基于数量化一类分析的IGA算法及应用

发布时间：2018-06-02 11:52

本文选题：交互式遗传算法(IGA) + 数量化一类分析　；参考：《浙江大学学报(工学版)》2013年10期

【摘要】：针对交互式遗传算法(IGA)中用户易疲劳的问题,引入数量化一类分析方法,从提高收敛速度和模拟人工评价两方面入手,改善IGA性能.采用均匀设计法确定遗传算法的初始种群,使得各基因单元类型在初始种群中均匀分布,从而提高遗传算法全局最优解的收敛性能;利用数量化一类分析,求取各基因单元类型对适应度评价的贡献权值,指导GA选择、交叉、变异等操作,以加速算法收敛;在GA操作后期,利用各基因单元类型对适应度评价的权值,模拟个体适应度的自动评价,降低用户疲劳度和提高效率.将该方法应用于动漫玩具形态造型中.结果表明,采用该方法可以有效地提高收敛速度和降低用户疲劳度.
[Abstract]:Aiming at the problem of user fatigue in interactive genetic algorithm (IGA), a kind of quantitative analysis method is introduced to improve the performance of IGA from two aspects: improving convergence speed and simulating manual evaluation. The uniform design method is used to determine the initial population of the genetic algorithm, so that the gene unit types are uniformly distributed in the initial population, so as to improve the convergence performance of the global optimal solution of the genetic algorithm. The contribution weight of each gene unit type to fitness evaluation is obtained to guide GA selection, crossover, mutation and other operations in order to accelerate the convergence of the algorithm; in the later stage of GA operation, the weight of each gene unit type to fitness evaluation is used. Automatic evaluation of individual fitness can reduce user fatigue and improve efficiency. The method is applied to the animation toy shape modeling. The results show that the proposed method can effectively improve the convergence speed and reduce user fatigue.
【作者单位】：浙江大学CAD&CG国家重点实验室;
【基金】：国家自然科学基金资助项目(61103106,51275460) 浙江省自然科学基金资助项目(Y1110230) 浙江省公益技术研究工业资助项目(2013C31046) 高等学校博士学科点专项基金资助项目(20110101120005)
【分类号】：TP18

【参考文献】