摘要
针对加热炉模糊控制器模糊规则库难以建立的问题,提出了一种基于历史数据的加热炉温度模糊控制规则提取方法。此方法包含3个环节,首先选择输入输出变量和典型工况的历史数据并对数据进行预处理,再对历史数据使用模糊C-均值聚类算法以确定规则数目和输入变量的隶属度函数,最后对每条规则使用支持向量回归机算法确定规则的后件参数。应用此方法能够有效地提取加热炉模糊控制器的模糊规则,实验结果验证了该方法的有效性。
Considering the difficulty in establishing fuzzy rule set of heating furnace controller,a method for extracting temperature fuzzy control rules of heating furnace based on historical data was proposed,which consists of three steps,i. e. having input/output variables and the historical data of typical working conditions selected and processed; making use of C-means clustering algorithm to determine the number of rules and the input variables' membership function; and having support vector regression algorithm used to determine parameters of each fuzzy rule's consequent. This method can effectively extract furnace fuzzy controller's fuzzy rules and the experimental results demonstrate the effectiveness of this method.
出处
《化工自动化及仪表》
CAS
2016年第9期940-944,共5页
Control and Instruments in Chemical Industry
关键词
模糊控制
规则
加热炉
历史数据
fuzzy control
rule
heating furnace
historical data
