【摘要】 为了保证沥青烟电除尘器的除尘效率、降低除尘设备的火灾风险,采用了K-means聚类的方法分析ESP放电信号。首先将ESP伏安特性曲线的二维空间进行了分割,确定了不同运行状态与聚类中心的关系,然后使用K-means聚类算法计算其聚类中心,最后根据当前ESP输入参数与各个聚类中心欧氏距离的关系,从而判断出ESP是否处于火花放电状态。仿真结果表明该方法可以准确地判断出所有火花放电信号。更多还原

【Abstract】 In order to guarantee the cleaning efficiency of asphalt fume electrostatic precipitator and reduce fire risks of the equipment,the K- means cluster algorithm is introduced to analyze the discharge signal of ESP. The 2 dimension space of ESP voltage- ampere characteristics is divided into four parts and each part has a cluster center. The center is mapped with the ESP operation status and it can be calculated by K- means algorithm. The ESP status can be determined according to the Euclidean distances between input parameters and various clustering centers. The simulation result indicates that the method can accurately estimate ESP discharge signals.更多还原