摘要
针对充油设备油位与油箱表面温度、环境温度较复杂的关系,提出一种基于深度学习的充油设备油位实时监测方法,实现油位实时准确监测。首先,通过试验收集充油设备内部不同缺油程度下的设备表面温度分布和环境温度数据;然后,建立神经网络模型,将与油位关系密切的充油设备外表面温度分布、环境温度作为模型输入,充油设备内部油位作为模型的输出;最后,模型训练结束后,在测试集上检验模型的泛化性能和预测性能。结果表明:所提方法能够实现充油设备油位的实时监测。
A deep learning based real-time oil level monitoring method of the oil filling equipment was proposed to address the complex relationship between oil level and tank surface temperature,as well as oil level and environmental temperature,achieving accurate and real-time oil level monitoring.Firstly,surface temperature distribution and environmental temperature data of the oil filling equipment were collected under different degrees of oil deficiency through experiments;then,a neural network model was established,which took the external surface temperature distribution and environmental temperature of the oil filling equipment,which were closely related to the oil level,as model inputs,and the internal oil level of the oil filling equipment as model outputs;finally,afer the model training was completed,the generalized and predicted performances of the model were tested on the test set.The results show that the proposed method can achieve real-time oil level monitoring of the oil flling equipment.
作者
安旺成
杜旭东
司军章
许国斌
刘云飞
An Wangcheng;Du Xudong;Si Junzhang;Xu Guobin;Liu Yunfei(Dingxi Power Supply Branch,State Grid Gansu Electric Pouer Co.,Lad.,Dingxi Gansu 743099,China)
出处
《电气自动化》
2025年第3期6-8,共3页
Electrical Automation
基金
国网甘肃省电力公司科技项目(B1271024002B)。
关键词
充油设备
表面温度分布
环境温度
神经网络
油位实时监测
oil flling equipment
surface temperature distribution
environmental temperature
neural network
real-time oil level monitoring
