摘要
高压直流输电线路的运行数据直接用于其高频率稳态控制中,会导致控制结果出现偏差,为此,研究了一种基于正态云神经网络的高压直流输电线路高频率稳态控制方法,构建正态云神经网络模型;分析高压直流输电线路结构及接线方式,依据分析结果选取输电线路的逆变侧与整流侧频率差信号及其变化率作为控制信号;以所选取控制信号为输入,通过所构建模型的正态云部分处理输入信号,降低其噪声与不确定性;处理后信号输入所构建模型的广义动态模糊神经网络部分,获得稳态控制信号输出,实现对高压直流输电线路的高频率稳态控制。实验结果显示:该方法可实现高频率下,正常运行与加入功率扰动时,高压直流输电线路的稳态控制,能够快速有效抑制输电线路的振荡情况,保持其发电机功率与直流输送功率快速恢复稳定,进行实时监测所得参数数据产生的噪声影响很小。该方法具有稳定性和可靠性。
Directly using the operation data of high-voltage direct current(HVDC)transmission lines in their high-frequency steady-state control can lead to deviations in the control results.To address this problem,a high-frequency steady-state control method for HVDC transmission lines based on a normal cloud neural network is explored.A normal cloud neural network model is constructed.The structure and wiring mode of HVDC transmission lines are analyzed,and the frequency difference signals between the inverter side and the rectifier side of the lines and their rates of change are selected as the control signals according to the analysis results.The control signals are used as inputs,and the input sig-nals are processed by the normal cloud part of the model to reduce noise and uncertainty.The processed signals are then input into the generalized dynamic fuzzy neural network part of the model to obtain the steady-state control signal output and realize the high-frequency steady-state control of the high-voltage direct current transmission line.Experimental re-sults show that the method can achieve steady-state control of high-voltage DC transmission lines at high frequency.When normal operation and power disturbances are added,it can quickly and effectively suppress the oscillation of the transmission line,maintain the generator power and DC transmission power to quickly restore stability,and the noise ef-fect generated by real-time monitoring of the obtained parameter data is very small.This method ensures the stability and reliability of the DC transmission system.
作者
叶友泉
孙杰杉
YE Youquan;SUN Jieshan(Quanzhou Electric Power Skills Research Institute of Fujian Electric Power Co.,Ltd.,Quanzhou 362000,Fujian,China)
出处
《西昌学院学报（自然科学版）》
2024年第3期52-59,共8页
Journal of Xichang University(Natural Science Edition)
基金
福建省教育厅福建省中青年教师教育科研项目(JAT231235)。
关键词
正态云
神经网络
高压直流
高频率稳
广义动态模糊
normal cloud
neural network
high-voltage direct current
high-frequency steady-state
generalized dy-namic fuzzy
