一种基于Adaboost和BP神经网络的装备故障预测方法研究
Research on Equipment Fault Prediction Methods Based on Adaboost and BP Neural Networks
马志刚 1董鹏 2刘晓亮3
作者信息
- 1. 海军工程大学管理工程与装备经济系 武汉 430034;海军装备部 北京 100841
- 2. 海军工程大学管理工程与装备经济系 武汉 430034
- 3. 92555部队 上海 200439
- 折叠
摘要
为了提升装备的故障预测准确性,论文提出了基于Adaboost和BP神经网络的装备故障预测方法.主要是设立多组BP神经网络作为相互独立的故障弱预测器对故障进行预测,预测结果输入Adaboost故障弱预测器,经过多组神经网络的权重设置形成故障强预测器,进而准确预测故障.研究表明,论文所提方法具有较高的预测精度和稳定性,能够广泛应用于装备故障预测领域,可为装备故障预测提供有效的解决方案.
Abstract
In order to improve the accuracy of equipment fault prediction,the paper proposes the equipment fault prediction method based on Adaboost and BP neural network.Multiple groups of BP neural network are set up as an independent fault weak pre-dictor to predict faults,and the prediction results are input into Adaboost fault weak predictor,and the fault strong predictor is formed through the weight setting of multiple groups of neural networks,so as to accurately predict faults.The study shows that the proposed method has high prediction accuracy and stability,and can be widely used in the field of equipment fault prediction,and can provide an effective solution for equipment fault prediction.
关键词
故障预测/神经网络/AdaboostKey words
fault prediction/BP neural network/Adaboost引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据