基于Apriori和C5.0算法的智能电表故障预测

Fault Prediction of Smart Electricity Meter Based on Apriori and C5.0 Algorithm

针对智能电表故障状态的预测问题,提出了一种基于Apriori算法和C5.0算法建立智能电表故障识别模型,实现智能电表故障的预测。首先,对智能电表历史故障数据库进行数据挖掘预处理,并采用Apriori算法进行强关联因素深度挖掘。然后,将强关联因素组成的数据集合分为训练数据集和测试数据集两部分,采用C5.0算法对训练数据集进行数据挖掘,生成智能电表故障初步预测规则。接着,根据测试集的数据对初步预测规则的正确性进行评估:如果准确度满足要求,确定预测规则;如果不满足,则返回训练集。最后,根据获得的预测规则建立智能电表故障状态预测模型进行智能电表故障预测。算例分析结果证明,智能电表故障状态预测模型具有较高的精度,可获得极为准确的故障状态预测结果。

Aiming at the problem of fault state prediction of smart electricity meter,a smart electricity meter fault identification model based on Apriori algorithm and C5.0 algorithm is proposed to realize the fault prediction of smart electricity meter.Firstly,the historical fault database of smart electricity meter is preprocessed by data mining,and the Apriori algorithm is used for deep mining of strong correlation factors.Secondly,the data set composed of strong correlation factors is divided into training data set and test data set.C5.0 algorithm is used to conduct data mining on training data set and generate preliminary prediction rules for smart meter faults.Then,the correctness of the preliminary prediction rule is evaluated according to the data of the test set.If the accuracy meets the requirements,the prediction rule is determined.If not,the training set is returned.Finally,a smart meter fault state prediction model is established according to the obtained prediction rules to predict smart meter faults.The result of example analysis proves that the smart meter fault state prediction model has high precision and can obtain extremely accurate fault state prediction results.