基于模糊模式识别的漏钢预报
PREDICTION OF STEEL LEAKAGE BASED ON FUZZY PATTERN RECOGNITION
-
摘要: 提出了一种基于模糊模式识别和采用学习方法的模糊决策网络的漏钢预报方法,同时对热电偶测温系统做了改进.实验表明该模糊漏钢预报系统不但可以准确对温度模式进行分类和预测拉漏事故,而且对于连铸生产过程故障诊断和产品质量评价也有一定的指导意义.Abstract: A decision-making network based on pattern clustering method with learning mechanism is presented and the traditional thermocouple system is rearranged for steel leakage alarm. This method is shown by numerous experiments to be very effective in clusterring temperature patterns, detecting possible leakages of liquid steel and their positions and reasons.