基于小样本数据的社区电梯故障预测

Community Elevator Fault Prediction Based on Small Sample Data

设计一种基于融合聚类算法的电梯故障预测模型,该模型应用粒子群算法、聚类算法、集成学习等人工智能技术完成电梯小样本数据故障预测。首先利用社区电梯系统历史维保数据进行数据聚类。结合聚类结果,利用聚类后具有可类比性的历史电梯故障数据集,分别在每个数据集进行集成学习回归故障预测。

In this paper,an elevator fault prediction model based on fusion clustering algorithm is designed.The model uses artificial intelligence technology such as particle swarm optimization algorithm,clustering algorithm and ensemble learning to complete the elevator fault prediction with small sample data.Firstly,the historical maintenance data of community elevator system is used for data clustering.Combined with the clustering results,the historical elevator fault data sets with comparability after clustering are used to carry out ensemble learning regression fault prediction in each data set.