课程文献中心 更多>>

关键词： MOSFET   IGBT   zero current switching   IGBT driver   resonant circuits   overcurrent sensor  

摘要： The control of IGBT (insulated gate bipolar transistor) and MOSFET (metal oxide semiconductor field effect transistor) is of great interest nowadays as they are widely used in electric vehicles, photovoltaic applications, and a multitude of systems. The field of power electronics and their correct activation ensures that the transistors are operated without being destroyed. In this work, a double resonant transformer was built and used to produce very high currents. These currents are switched by a full bridge of resonant IGBT transistors to demonstrate the feasibility of exceeding the maximum permissible transistor currents in a resonant system. The system is controlled by the feedback from two current sensors. In this case the currents exceed in a 170% the peak current of the transistor without problems. In this way, resonant circuits with IGBT transistors can be designed with currents lower than the maximum currents of the resonant circuit, therefore reducing the cost of the circuit and reducing the switching losses to nearly zero.

关键词： FIR filters  

摘要： This paper presents a distributed arithmetic (DA) architecture that can efficiently implement finite impulse response (FIR) filters for biomedical signal processor applications. FIR filter design is more efficient when it uses a look-up table (LUT)-based technique rather than a serial one. The design’s performance and efficiency can be improved by using segmented memory banks as well as memory lookup for multiply operation. Verilog HDL is used to model the proposed design, and Synopsys Design Compiler tool is used for synthesis. The FIR filter architecture utilizing DA results in a 24.82% reduction in total power compared with the serial FIR structure. © Shanghai Jiao Tong University 2022.

关键词： design automation   Artificial intelligence (AI)   image processing   power electronics   power electronics   image processing   power electronics  

摘要： Power electronics design automation, implementing artificial intelligence (AI) to optimize the design of power converters, has emerged as a novel research topic given the complexity of power converter design, whose key challenges include power loss modeling across the enormous number of available components. This article proposes a novel end-to-end AI-based tool for extracting nonlinear dynamic properties from semiconductor datasheets, which can enhance the power loss estimation model and accelerate the optimal design of power converters. First, thousands of images from power transistor datasheets are collected and annotated to construct a training database. Then, CenterNet, a neural network for image object detection, is trained for figure segmentation from datasheets and key element detection from figures. Optical character recognition (OCR) and morphological image processing techniques are utilized to extract the specific dynamic data. The results illustrate that the customized tool for power transistor device datasheets in this article can accurately extract the data, significantly reducing the time consumption for transistor data collection and its characteristic modeling work, promising pathways to streamline and optimize power electronics design. The tool has been published online and is actively being updated and improved via http://***.

关键词： MOSFET   Semiconductor device modeling   Integrated circuit modeling   Analytical models   Silicon carbide   Mathematical models   Switches   Analytical switching model   multichip sic power modules   MOSFET   Semiconductor device modeling   unbalanced dynamic currents   unbalanced dynamic currents   Analytical models   Silicon carbide   Mathematical models   Switches   Analytical switching model   multichip sic power modules   paralleled sic metal-oxide-semiconductor field-effect transistors   unbalanced dynamic currents  

摘要： Uneven dynamic currents between paralleled silicon carbide (SiC) metal-oxide-semiconductor field-effect transistors (mosfets) can cause unbalanced switching losses, challenging the circuit reliability. Therefore, it is essential to quantitatively evaluate unbalanced dynamic currents during circuit design and application. However, the existing calculation methods face challenges in modeling or are time-consuming in circuits with paralleled mosfets. To address these issues, this article presents an analytical model to calculate the unbalanced switching currents and switching energy, which is applicable to circuits with any parallel number (n) and easier to use than Spice simulation method. To address the challenge of the high circuit order, the power and driving circuits are decoupled and modeled, reducing the equation order to 3n - 1. To handle the numerous parasitic mutual inductances, the inductance matrices are used for modeling. Moreover, the nonlinear die parameters and their temperature dependence are considered to guarantee accuracy. The accuracy of the presented model is verified by experiments. The model-calculated waveforms fit the tested results well. Under various switching speeds, load currents, and die temperatures, the model can accurately predict the dynamic current differences and unbalanced switching energy, with calculation errors lower than 8% for the current differences and 11% for the switching energy.

关键词： 电力变压器   状态评估   图卷积网络   图池化   数据不平衡  

摘要： 电力变压器运行过程中故障数据远少于正常数据，数据不平衡问题较为严重，并且监测变量间耦合关系复杂，导致状态评估任务的建模难度大、评估精度低。针对相关问题，提出基于双输入残差图卷积网络的电力变压器健康状态评估方法。首先，采用SMOTE Tomek混合采样算法对训练数据进行不平衡数据预处理，解决了故障数据过少、训练数据严重不平衡的问题；然后，考虑到单一度量无法准确描述变量间相关性的问题，提出多度量融合构图方法，通过多个度量方法共同学习变量间的相关性，并构造图结构数据；最后，提出基于切比雪夫图卷积的双输入残差图卷积网络，对所构造的图结构数据进行特征提取，并通过自注意力机制进行特征融合，得到变压器的状态评估结果。在真实电力变压器运行的油中溶解气体及油化试验数据集上进行了对比实验，实验结果表明，所提出方法的状态评估准确率达到94.7%，F1分数达到0.942，高于其他深度学习方法，具有最佳的评估性能。

关键词： 直流变压器   故障   自适应保护   差动电流   直流分量   交流分量  

摘要： 直流变压器(DCT)的故障特性受内部储能元件及其拓扑的影响，传统基于端口电流差动的保护原理难以有效识别区内外故障。针对非隔离型模块化多电平直流变压器（DCMT），提出了一种基于附加桥臂电流测量的改进差动保护方法。分析DCMT不同故障位置的故障电流特性，获得DCMT在区内外故障下的差动电流特性，构建基于附加桥臂电流测量的DCMT差动保护复合判据。考虑到故障严重程度对故障识别的影响，结合比率制动原理，确定了DCMT差动保护阈值自适应整定策略。仿真实验结果表明，与传统方法相比，所提的保护方法可准确识别DCMT内部匝间短路、端口母线故障和外部故障，在不同故障位置和故障电阻条件下有更高的灵敏性。

关键词： 氧化锌避雷器   在线监测   谐波分析法   Nuttall窗   采样频率   自适应调整  

摘要： 针对基于FFT的传统谐波分析法测氧化锌避雷器(MOA)阻性电流时会由于电网频率波动造成频谱泄漏和栅栏效应的问题，对Nuttall窗三谱线插值算法进行改进，利用硬件检测基波频率，推导相关计算公式，大幅减少了原算法计算量，并将改进后算法与采样频率自适应调整相结合，提出一种改进谐波分析法，能够有效解决上述问题，提升阻性电流测量精度。通过仿真验证改进谐波算法的精度与稳定性，并基于该算法对MOA在线监测系统进行软硬件设计。测试结果表明，该系统性能稳定，精度较高，满足MOA在线监测的要求。

关键词： 极端天气   风电功率预测   自编码器   迁移学习  

摘要： 针对寒潮、台风、覆冰等极端天气下风电功率预测精度不足的问题，提出基于气象因子的极端天气事件判别方法，以及基于迁移学习和自编码器的极端天气事件自适应短期风电功率预测方法。首先，通过分析气象要素和机组出力间的耦合特性，定义极端天气判别标准，识别未来将要发生的天气事件类型。其次，基于自编码器预测模型的自相关机制，增加长时间序列信息利用率，采用迁移学习的“预训练-微调”策略，先利用正常天气下的充足样本对预测模型预训练，再针对极端天气下有限样本数据进行微调，根据判别得到的天气事件，自适应地采用该类天气事件下的预测模型进行短期风电功率预测。选取12个风电场的数据集进行分析，通过分析模型在极端天气和所有天气条件下的预测表现，验证了所提方法的有效性。实验结果表明，所提方法可准确预知未来是否会发生极端天气事件，并大幅度提升极端天气下的短期风电功率预测精度，3种极端天气下所提方法的平均预测准确率为87.52%，相较于常规方法平均提升2.47%。

关键词： 变压器室   变压器火灾   数值模拟   机械排烟   排烟量   温度分布  

摘要： 为了研究不同机械排烟量下的室内变压器火灾动力学过程，探索适合于室内不同规模变压器火灾的最佳排烟量，基于FDS数值模拟技术，采用羽流法计算排烟量，研究6,12,18 MW这3种火源功率下室内变压器火灾在机械排烟条件下的燃烧特性。结果表明：无机械排烟时，不同火源功率下的燃烧形式不同；机械排烟量与火源功率呈正相关；不同规模变压器火灾均采用计算的2倍机械排烟量时降温幅度、能见度提升幅度最大，故采用2倍机械排烟量的降温控烟效果最佳。模拟6 MW火源功率的最佳排烟量为23.8 m3/s, 12 MW火源功率的最佳排烟量为41.7 m3/s, 18 MW火源功率的最佳排烟量为58.6 m3/s;合理增大机械排烟量，能减小室内变压器火灾蔓延形成窗口溢流火的危险性，有效抑制火焰的发展；同一火源功率下，随机械排烟量的增大，主变室内温度显著降低，能见度提高。研究结果对户内式变电站主变压器室火灾的安全防护提供了参考依据。

关键词： 无线充电   感测线圈   输出功率   非接触测量  

摘要： 为了降低无线充电系统耦合线圈上高频高压直接测量的技术复杂度和经济成本，该文提出一种多感测线圈协同的无线充电系统发射线圈输出功率非接触测量方法。相比于传统的直接测量方式，该方式仅需若干组独立的感测线圈，无需专用的大功率高频电流、电压传感器，同时可以很好地适应于不同功率等级、汽车水平与垂直地面方向的位置偏移以及屏蔽材料等实际场景。该文利用电磁感应定律，通过在线圈表面设置多组感测线圈进行测量，建立了考虑水平、垂直偏移及功率变化的各感测线圈电压相量与发射线圈输出功率之间的耦合矩阵模型，然后给出了多感测线圈的参数设计依据，并进行了仿真与实验验证，接着探讨了感测线圈形状、数目与采样点数量对测量精度的影响，最后进一步探讨了实际场景将多感测线圈水平放置，可以避免垂直叠放高度影响收发线圈短距离测量的问题。实验结果表明，在不同功率等级下，无线充电系统在水平与垂直地面方向同时发生偏移时，所提方法的最大计量误差在1.5%以内，具有较高的测量准确性和实用价值。