摘要
紧急直流功率支援能在交直流系统受到大扰动后快速调制HVDC的直流功率,利用HVDC短时过载能力提高电网的暂态稳定性,因此如何利用广域信号实现紧急直流功率支援的在线预测意义重大。文中构造了一种基于多分辨率分析理论的新结构正交小波神经网络,并将其应用于云广特高压直流的紧急直流功率支援在线预测。该多分辨率正交小波神经网络采用正交尺度函数作为激励函数,能保证逼近函数表达式的唯一,收敛迅速。实验结果表明:该多分辨率分析正交小波神经网络能依据主成分降维后的输入数据准确并可靠地给出紧急直流功率支援控制量。
Emergency DC Power Support （EDCPS） function can quickly modulate available DC power when heavy fault occurs in parallel AC lines or other HVDC systems, and finally enhances transient stability with short-time DC overload capacity. Therefore, how to implement online EDCPS prediction by wide area measurement system （WAMS） signals is of great significance and deserves to be researched. An online EDCPS prediction model based on WAMS signals is introduced, and a newly structured orthogonal wavelet neural network based on multi-resolution analysis is constructed and applied to the ECDPS prediction of HVDC power transmission project from Yunnan to Guangdong. Adopting orthogonal scaling function as activation function, this multi-resolution analysis orthogonal wavelet neural network （MAOWNN） can converge fast and ensure the uniqueness of approximating function expression. Simulation results show that based on dimension-reduced principal components the MAOWNN can accurately give the controlled quantity of EDCPS.
出处
《电网技术》
EI
CSCD
北大核心
2010年第8期12-17,共6页
Power System Technology
基金
"十一五"国家科技支撑计划重大项目(2006BAA02A17)~~
关键词
正交尺度函数
多分辨率分析
正交小波神经网络
紧急直流功率支援
特高压直流
主成分分析
orthogonal scaling function
multi-resolution analysis
orthogonal wavelet neural network （OWNN）
Emergency DC Power Support （EDCPS）
UHVDC
principalcomponent analysis （PCA）
