Identification of Static and Dynamic Distribution System Aggregate Load Using Probabilistic Neural Networks

使用概率神经网络识别静态和动态配电系统总负荷

• 批准号: 9210878
• 负责人: James Patton
• 金额: $ 9.86万
• 依托单位: University of Maine
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1992
• 资助国家: 美国
• 起止时间: 1992-09-15 至 1996-02-29
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9210878&HistoricalAwards=false
• 关键词: Identification; Static; Dynamic; Distribution; System

The objectives of the proposed research are: 1) develop a neural net-based real-time load model for both dynamic and distribution system analyses; and 2) design a "smart" monitor for controlling distribution system capacitors, voltage regulators, and load tap changing transformers. The research will determine dynamic and distribution system models that represent load as a function of voltage and frequency. Tests will be made on Central Maine Power Company feeders to obtain data with which to train a probabilistic neural network. The neural network learning algorithm will be implemented with a digital signal processor in a "smart" monitor that will be able to map a vector of measured and other relevant variables to load parameters will be used for dynamic system studies and will allow real-time volt-var control on a distribution feeder. An important goal of the proposed research is to produce load models or modeling techniques that can be used in standard ways by other researchers applying well-known analysis methods.

这项研究的目标是：1)开发基于神经网络的实时负荷模型，用于动态和配电系统分析；以及2)设计用于控制配电系统电容器、电压调节器和有载调压变压器的“智能”监控器。这项研究将确定将负荷表示为电压和频率函数的动态和配电系统模型。将在缅因州中部电力公司的馈线上进行测试，以获得用于训练概率神经网络的数据。神经网络学习算法将用数字信号处理器在智能监视器中实现，该监视器将能够将测量和其他相关变量的矢量映射到负载参数将用于动态系统研究并将允许对配电馈线进行实时无功控制。拟议研究的一个重要目标是产生可供其他应用知名分析方法的研究人员以标准方式使用的负荷模型或建模技术。