集成环境中基于多Agent的时间定额确定方法

Multi-agent Based Approach for Determination of Time-quota in Integrated Environment

对CAPP和生产调度集成环境中的时间定额问题进行了研究 ,提出了一种基于Agent的时间定额确定方法 ;采用智能Agent的方法 ,建立了集成环境中确定时间定额的结构模型。将影响时间定额的因素映射成与工序状态有关的零件Agent和加工方法Agent ,以神经网络Agent为计算工具 ,借助于Agent的判断推理能力 ,将各因素处理成神经网络可以识别和接受的数据模式 ,利用黑板结构处理各Agent之间的信息交互和协同控制 ,以快速解决工序的时间定额问题。实验表明 ,只要选择适当的样本 ,并在系统中不断进行自学习与自组织 ,就能快速获得准确的工序时间定额。

Based on research of time-quota about CAPP and production scheduling in integration environment, it puts forward the Agent-based time-quota method, and sets up the structure model by means of intelligent agent. The factors influencing time-quota can be mapped into part agent and machine method agent. With the help of rule-based reasoning, the agents can transform these factors into data mode. As a tool, artificial neural network agent can calculate time-quota quickly. The experiments show that precise process time-quota can be obtained rapidly with proper samples selected, continuous self-study and self-organization in system.