柔性装配系统的多目标优化设计

The multiple objective optimization design of flexible assembly system

本文在确定对系统起决定作用的参数、变量集及目标约束集的基础上，采用多层前馈型神经网络作为多目标优化计算器，用改进的遗传算法对神经网络进行训练，为复杂背景下的多目标优化设计提供了一种新方法。

On the basis of determining those played a decisive role upon the system such as the parameters, variable ensemble and assemblage of objective constraints, this paper adopted multi-layer forward feeding typed neural net as the multiple objective optimization calculator and carried out a training on the neural net by the use of inproved genetic algorithm and thus provided a kind of new method for the multiple objective design under complicated backgrounds.