基于加权模糊C均值聚类的空预器积灰监测研究

电站空预器积灰会严重影响机组运行经济性。提出加权模糊C均值聚类算法对空预器积灰程度进行监测,该方法计算多维样本中每一维数据的标准差,将其作为权重,计算样本与类心之间的加权欧式距离,降低模糊C均值聚类算法对离群点的敏感度。利用人工数据对该方法进行验证,结果表明,相比于传统模糊C均值聚类算法,提出的方法对离群点识别更加准确,分类结果更加合理。进一步将此方法用于空预器实际运行数据中,结果表明,此方法能够有效反映出空预器积灰程度随运行时间的变化。

Air pre-heater fouling monitoring based on weighted fuzzy c-means clustering algorithm

The deposition in air pre-heater will seriously affect the operation economy of the unit.A method for air pre-heater fouling monitoring is proposed based on weighted fuzzy C-means clustering algorithm.This method calculates the standard deviation of each one-dimensional data in multi-dimensional samples,and takes it as the weight to calculate the weighted Euclidean distance between the sample and the cluster center,so as to reduce the sensitivity of fuzzy c-means clustering algorithm to outliers.The artificial data is used to verify the method.The results show that,compared with the traditional fuzzy c-means clustering algorithm,the proposed method is more accurate for outlier recognition and the classification result is more reasonable.Furthermore,this method is applied to the actual operation data of air preheater,and the results show that this method can effectively reflect the change of ash deposition degree of air preheater with operation time.