摘要
由于传统系统无法对多种谐波进行预警处理,导致预警误差较大,为了解决该问题,提出了基于神经元网络的电网谐波联动预警系统设计。根据神经元网络工作原理设计系统总体结构,采用2205型号采集卡对数据进行采集,使用4046型号单片集成锁相环,在一定范围内跟踪电压信号变化,保证输入与输出信号频率一致,设计数字处理器与总线连接电路,使不同神经元网络节点都能接收到相同数据。在软件功能内设置电网参数,实时检测电流变化,结合神经元模型设置输入矢量,获取神经元输出值,采用最小学习算法,调节连接权值,通过分析超过额定值电流和时间关系,实现联动预警系统设计。由实验结果可知,该系统最大预警误差为0.0063,不会对系统预警造成任何影响。
Because the traditional system can not deal with many kinds of harmonics early warning,leading to a large error in early warning,in order to solve this problem,the design of a network harmonic linkage early warning system based on neural network is proposed.According to the working principle of neuron network,the overall structure of the system is designed.2205 acquisition card is used to collect data.4046 monolithic integrated phase-locked loop is used to track the change of voltage signal in a certain range to ensure that the frequency of input and output signals is the same.The connection circuit between digital processor and bus is designed to make different neurons.Network nodes can receive the same data.In the software function,power network parameters are set,current changes are detected in real time,input vectors are set with neuron model,output values of neurons are obtained,connection weights are adjusted by minimum learning algorithm,and the design of linkage early warning system is realized by analyzing the relationship between current exceeding rated value and time.The experimental results show that the maximum early warning error of the system is 0.0063,which will not have any impact on the early warning of the system.
作者
杜俊杰
梁俊伟
和立辉
杜洋
DU Jun-jie;LIANG Jun-wei;HE Li-hui;DU Yang(State Grid Xingtai Power Supply Company,Xingtai 054000,China;Beijing Zhongke RuideTechnology Development,Beijing 100000,China)
出处
《电子设计工程》
2019年第11期5-8,14,共5页
Electronic Design Engineering
基金
河北省电力公司科技资金(kj2018-033)
关键词
神经元网
电网谐波
联动预警
采集卡
锁相环
超过额定值
neuron network
grid harmonics
linkage early warning
acquisition card
phase-locked loop
exceeding rated value