摘要
IGBT模块在各领域的现代设备中都起着至关重要的作用,对IGBT模块的状态评估可以有效避免因模块损坏而造成的严重后果。研究采用随机森林、XGBoost、分类与回归树、k近邻、支持向量回归、Adaboost、极端随机树和梯度提升决策树8种算法,根据IGBT模块的任务剖面对其进行状态评估。该研究根据多种人工智能算法的实验结果分析了IGBT设备的任务剖面,为IGBT状态评估提供了相关的解决思路。
IGBT modules play an important role in modern equipment in various fields.The evaluation of the status of IGBT modules can effectively prevent the serious consequences caused by module damage.The research adopts 8 algorithms of Random Forest, XGBoost, Classification And Regression Tree,k-Nearest Neighbor, Support Vector Regression,Adaboost,Extremely Randomized Trees and Gradient Boosting Decision Tree,and conducts state assessment according to the task profile of the IGBT module.The experimental results of the artificial intelligence algorithm analyze the mission profile of the IGBT device,and provide relevant solutions for the evaluation of the IGBT status.
出处
《工业控制计算机》
2021年第7期142-145,共4页
Industrial Control Computer
关键词
IGBT
状态评估
任务剖面
人工智能
IGBT
state assessment
mission profile
artificial intelligence
