基于神经网络的瞬动校验电流控制技术的研究  

Research on the control technology of instantaneous calibration current based on neural network

作　　者：杜太行[1] 贾兴建[1] 韩春贤[1] 张勇[1] 

机构地区：[1]河北工业大学电气与自动化学院,天津300130

出　　处：《电力系统保护与控制》2009年第10期44-47,58,共5页Power System Protection and Control

摘　　要：在瞬动校验电流控制方法中,提出用补偿系数提高瞬动电流的精度。在分析瞬动电流影响因素并建立等效模型的基础上,采用神经网络控制方法进行控制。以LabVIEW为语言环境,通过Matlab Script节点调用训练好的神经网络,计算出不同等级电流所需的补偿系数,最终由电流产生机构产生瞬动调试电流。为了提高系统的适应性,提出通过变更训练数据样本,实现神经网络的在线调整,进一步提高了瞬动调试电流的控制精度。This paper improves the accuracy of current by referring compensation coefficient in the way of instantaneous calibration current control. The system is controlled by neural network after analyzing the influencing factors of instantaneous calibration current and establishing equivalent model.In LabV1EW,calling neural network which has been trained by Matlab Script node to compute the compensation coefficient.This method can compute the compensation coefficient of different current.Finally, the current producing institution creates instantaneous current.In order to enhance system＇s adaptability,the neural network online adjustment is achieved by changing the training samples to improve the accuracy of instantaneous current furthely.

关 键 词：LABVIEW 神经网络 断路器检测 MATLAB 在线调整 

分 类 号：TM561[电气工程—电器]