混合型超级电容器电源失效自动化监测方法  

Automatic Monitoring Method for Hybrid Supercapacitor Power Failure

作　　者：钟振鑫 ZHONG Zhenxin(Huizhou Power Supply Bureau of Guangdong Power Grid Corporation,Huizhou 516001,China)

机构地区：[1]广东电网有限责任公司惠州供电局,惠州516001

出　　处：《自动化与仪表》2024年第6期124-126,136,共4页Automation & Instrumentation

基　　金：南方电网公司科技资助项目(031300KK52220008)。

摘　　要：敏感特征量可以用于监测电容器电源的状态,该文以其为基础研究基于敏感特征量与机器学习的混合型超级电容器电源失效自动化监测方法。构建混合型超级电容器电源系统数学模型后,获得电容器电源在运行过程中的电压、电流等基础数据,使用时域法计算敏感特征量ESR值,再将其作为输入,利用随机森林模型和旋转森林模型组建多联级森林深度网络,输出混合型超级电容器电源失效监测结果。实验表明,该方法可有效获取混合型超级电容器电源运行过程中的电流,并准确计算其敏感特征量ESR值,同时获得混合型超级电容器电源失效自动化监测结果。The sensitive feature quantity can be used to monitor the state of capacitor power supply.Based on the sensitive feature quantity and machine learning,the automatic monitoring method of hybrid supercapacitor power supply failure is studied.After constructing the mathematical model of the hybrid supercapacitor power supply system,the basic data such as voltage and current during the operation of the capacitor power supply are obtained,and the ESR value of the sensitive characteristic quantity is calculated using the time-domain method,and then the ESR value is used as the input,and the multi-cascaded forest deep network of the random forest model and the rotating forest model components is used to output the failure monitoring results of the hybrid supercapacitor power supply.Experimental results show that this method can effectively obtain the current during the operation of the hybrid supercapacitor power supply,accurately calculate its sensitive characteristic ESR value,and obtain the automatic monitoring results of the hybrid supercapacitor power supply failure.

关 键 词：敏感特征量 机器学习 旋转森林 时域法 灵敏度 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]