基于神经网络的SiC功率管导通电阻估测技术研究  

Research on On-State Resistance Estimation of SiC Power Transistor Based on Neural Network

作　　者：崔江[1] 陈一凡 范士颖 林华[1] 沈勇 王友仁[1] 陈则王[1] Cui Jiang;Chen Yifan;Fan Shiying;Lin Hua;Shen Yong;Wang Youren;Chen Zewang(Nanjing University of Aeronautics and Astronautics,Nanjing 211106,China;Aeronautical Science and Technology Key Laboratory of Fault Diagnosis and Health Management Technology,AVIC Shanghai Aero Measurement Controlling Research Institute,Shanghai 201601,China)

机构地区：[1]南京航空航天大学,江苏南京211106 [2]航空工业上海航空测控技术研究所故障诊断与健康管理技术航空科技重点实验室,上海201601

出　　处：《航空科学技术》2021年第9期81-86,共6页Aeronautical Science & Technology

基　　金：航空科学基金(201933052001,20183352030);中央高校基本科研业务费专项资金资助项目(NS2021021)。

摘　　要：SiC功率管器件广泛应用在航空领域的电能变换、配电等场合,其健康状况十分重要。在SiC器件的健康监测应用中,导通电阻的检测是一项十分重要的技术。为了能够简单准确地得到碳化硅(SiC)MOSFET功率器件的导通电阻,本文提出了一种基于神经网络的SiC MOSFET器件导通电阻估测方法。本文搭建SiC MOSFET导通电阻测试电路仿真和物理试验平台,并使用BP神经网络(BP neural networks,BPNN)对不同温度、不同栅极电压以及不同漏极电流下SiC MOSFET器件的导通电阻数据进行详细描述。最后,对基于BPNN的SiC MOSFET导通电阻估测方法进行效果验证。结果表明,该方法具有精度高和泛化能力强的优点,能够实现SiC MOSFET器件导通电阻的有效估测。SiC power devices are applied widely in the applications of aeronautics,mainly in terms of electrical energy transform and distribution.The health surveillance technology has been a focus in the past time for these devices,whose on-state resistance detection technology is crucial.In order to obtain on-state resistance of silicon carbide(SiC)MOSFET power devices simply and accurately,this paper proposes a neural network based on-state resistance estimation method for SiC MOSFET devices.This paper builds a simulation and physical experiment platform for SiC MOSFET on-resistance test circuit,and uses BP neural network to describe in detail on-stateresistance data of SiC MOSFET devices under different temperatures,different gate voltages and different drain currents.Finally,the effect of SiC MOSFET on-state resistance estimation method based on BP neural network is verified.The results show that this method has the advantages of high accuracy and strong generalization ability,and can realize effective estimation of on-state resistance of SiC MOSFET devices.

关 键 词：碳化硅功率管 导通电阻 神经网络 误差分析 温度特性 

分 类 号：V240.2[航空宇航科学与技术—飞行器设计]