论文部分内容阅读
为了使自动化生产系统始终处于高效的工作状态,在工作过程中就必须对其进行实时监控,以便及时地发现系统中的异常现象并进行诊断和处理。本文基于线性联想记忆的神经网络提出了一种自动化生产系统多过程故障发现及诊断的理论方法。此方法可以同时对多个过程进行监控,克服了以往所采用的方法不能同时监控多过程的局限性,增强了系统监控的实时性。经仿真研究表明:这种方法是较为理想和实用的。
In order to make the automated production system always be in an efficient working state, it must be monitored in real time in the course of work in order to timely find abnormalities in the system and carry out diagnosis and treatment. In this paper, a neural network based on linear associative memory is proposed to find out and diagnose the multi-process fault in automated production system. This method can monitor multiple processes at the same time, overcoming the limitation that the methods used in the past can not simultaneously monitor multiple processes and enhancing the real-time performance of system monitoring. The simulation shows that: This method is more ideal and practical.