摘要
传统比例-积分-微分(proportion-integration-differentiation,PID)控制技术在电网产生扰动时无法兼顾快速性和鲁棒性,易造成系统不稳定失衡,向电网注入大量的谐波。对此现象,提出了一种改进反向传播(back propagation,BP)神经网络的分数阶PID控制器来提高电网的鲁棒性和对响应的快速性。该算法采用分数阶PID控制器跟踪电流外环的参考电流,并针对分数阶PID控制器的5个参数采用BP神经网络实时在线整定,消除了人为调参所带来的不确定性。对于BP神经网络在整定参数过程中无法整定得到最优解,引入变化的惯性因子和学习速率,提高了BP神经网络的求解效率。仿真结果验证表明,所提控制算法对并网电流能够实现快速跟踪,鲁棒性好。
The traditional proportion-integration-differentiation(PID)control technology can’t take into account the rapidity and robustness when the power grid is disturbed,which is easy to cause system instability and imbalance,and inject a large number of harmonics into the power grid.A fractional order PID controller based on improved back propagation(BP)neural network was proposed to improve the robustness of power grid and the rapidity of response.The fractional order PID controller was used to track the reference current of the current inner loop,and BP neural network was used to adjust the five parameters of the fractional order PID controller in real time.The uncertainty caused by artificial parameter adjustment was eliminated.For the BP neural network can not get the optimal solution in the process of setting parameters,the variable inertia factor and learning rate were introduced to improve the solution efficiency of BP neural network.Simulation results show that the proposed control algorithm can track the grid connected current quickly and has good robustness.
作者
吴亚雄
杨旭红
方浩旭
张苏捷
WU Ya-xiong;YANG Xu-hong;FANG Hao-xu;ZHANG Su-jie(Automatic Engineering of Shanghai University of Electric Power, Shanghai 200090, China;State Grid Pudong Power Supply Company, SMEPC, Shanghai 200122, China)
出处
《科学技术与工程》
北大核心
2022年第13期5243-5249,共7页
Science Technology and Engineering
基金
国家自然科学基金(51777120)。
关键词
三相LCL并网逆变器
BP神经网络
分数阶PID控制
自适应参数调整
three-phase LCL grid-connected inverter
BP neural network
fractional order PID control
adaptive parameter adjustment
