间歇生产过程配方的模糊聚类方法

针对间歇生产过程的配方缺少定量分析方法，难以用于过程建模和控制策略实施的问题，提出了一种基于类核函数的配方模糊聚类算法．对ISA配方模型进行了简化，建立了配方样本的相似性测度方法；根据配方样本的模式多样性，建立配方类核来代替配方类心，通过最小化所有配方样本到配方类核距离的加权和来对配方进行聚类，从而得到聚类数目及模糊隶属度矩阵．最后对一类间歇蒸发器的配方使用该方法进行模糊聚类，通过和FCM算法的比较验证了该方法的有效性。

Fuzzy Cluster Analysis Method for Batch Process Recipe

This paper presents a new fuzzy cluster analysis method for batch process recipe that has less quantitative analysis ways before and is difficult to be used for modeling or controlling system. Firstly, the ISA recipe model was simplified, and a recipe distance measurement method was built. According to the distribution characteristic of recipe, a recipe class kernel was presented to replace the recipe class center. Through minimizing all the distance of recipe samples to recipe class kernel, recipe samples were clustered, the class number and fuzzy membership degree matrix was also obtained in the end. At last, compared with FCM algorism, this approach shows its availability to cluster the recipe of a kind of batch evaporator.