基于模糊聚类的绿色工艺评价样本分类方法

针对绿色工艺评价样本具有不确定性、多维性以及量纲差异大的特点,为实现样本的合理分类,提出一种基于核的模糊可能性聚类新算法.该方法将核模糊聚类算法、可能性聚类算法和减法聚类算法相结合,以提高聚类的准确率;使用聚类有效性指标作为分类条件,自适应确定最佳分类数.仿真实验结果表明,该算法具有较好的有效性和鲁棒性,并将该算法运用在绿色工艺评价样本分类中,得到了较好的分类效果,验证了算法的实用性.

Sample Classification Method for Green Process Evaluation Based on Fuzzy Clustering

Due to the uncertainty， multidimensionality and significant difference of the evaluation samples of green process， a novel algorithm of kernel-based fuzzy possibilistic clustering was proposed in order to achieve reasonable sample classification. Kernel fuzzy clustering， possibilistic clustering and subtraction clustering were combined to improve the accuracy of clustering and cluster validity index was used as the classification condition to obtain the optimal classification number. The simulation results showed that this algorithm has good validity and robustness. When the algorithm is applied to classify the evaluation samples of green process， good classification effects are gained， which verifies its practicability.