摘要： 针对混杂系统理论和方法尚不能处理不确定性以及意外事件的发生等问题，将已有的智能理论和混杂系统理论进行有机结合，是一项有效的解决途径。在分析自动化装配中连续和离散变量组成混杂系统基础上，提出了集成装配计划和调度的一类智能混杂系统，给出了智能决策接口以及和智能技术的互连，分析了基于模糊神经网络和粗糙集的智能解决策略，使一类混杂系统具有智能决策功能。通过对一自动化装配实例建立极大代数模型和神经网络的仿真实验，验证了系统分析方法的可行性和有效性。

关键词: 智能混杂系统, 装配, 模糊神经网络, 粗糙集, 极大代数

Abstract: Considering hybrid system theory and method not sufficient enough to cope with unexpected events and uncertainties, the combination of intelligent theories with hybrid system theories can produce an effective solution. On the basis of analyzing the continuous and discrete parts in automatic assembly, a class of intelligent hybrid systems that integrate planning with scheduling in assembly is proposed. Intelligent decision interface and interconnection with intelligent techniques are given. Intelligent solution strategies based on fuzzy-neural network and rough set are analyzed, endowing a class of hybrid systems with intelligent decision ability. The feasibility and efficacy of the analyzing method are illustrated through the construction of a max-plus algebra model and simulation of neural network of an automatic assembly case.

Key words: Intelligent hybrid systems, Assembly, Fuzzy-neural network, Rough set, Max-plus algebra

中图分类号: 

• TP13