粗糙数据的Max-Min分级模型

针对复杂制造系统生产线优化调度等众多实际领域中的不确定性处理问题,在分析了现有进化计算机制的特点以及存在的不足基础上,建立了一种数组分级的Max-Min模型,讨论了该模型中两个基本参数(分级参数和分级调节参数)的性质及其作用,证明了在分级参数充分大时,Max-Min分级模型即为某种意义下的比例分级。分析结果表明,Max-Min分级模型是一种可反映一定松弛意识的较比例分级更广泛的分级方法,适应于不确定信息的分类问题,弥补了现行分级方法不适合粗糙数据的不足。最后给出了不确定环境中进化机制的个体复制方案。

Max-Min classification model for rough data

Evolutionary computational methods play an important role in many practical problems such as production line optimization in complex manufacturing systems with uncertain environments. This paper reviews the current development of evolutionary computational methods and the short comings of fuzzy information processing. A Max-Min classification model for rough data was then developed with two basic parameters, a classification parameter and an adjustment parameter for the classification. The Max-Min classification model is shown to be a proportional classification model for sufficiently large classification parameters. Therefore, the Max-Min classification model is a more extensive classification method compared to the common proportional classification method, which makes the Min-Max method better for uncertain information than current classification methods which are not fit for rough data. The Min-Max model includes an evolutionary reproduction procedure for individuals in an under uncertainty environment.