【摘要】 对监控对象的知识获取是过程监控的一个关键技术 ,粗集神经网络过程监控模型采用了事例知识获取的方法 ,该方法简单易行 ,但存在样本分布空间不能覆盖论域空间的问题。为解决该问题 ,本文采用了决策树增量学习法和神经网络完全学习相结合的方法。经分析表明 ,该方法符合粗集神经网络过程监控模型的特点 ,能较好地从新增样本中获取知识 ,并具有较快的学习速度更多还原

【Abstract】 Acquiring the knowledge of monitored object is a key technology in process minitoring.Case based knowledge acquisition method is used in the monitor model of Rough Set Neural Network,which is simple and available,but there presents the problem that Sample space can not cover domain space.In this paper,a new approach is set forth that integrating both decision tree incremental learning and neural network global learning.Through theory analysis,it’s indicated that this approach is agree with the characteristic of Rough Set Neural Network minitoring model,new knowledge can be rather easily obtained from newly added samples,and the learning speed is rather fast.更多还原