基于循环Legendre模糊神经网络的DFIG二阶滑模容错控制  

Second-Order Sliding Mode Fault Tolerant Control for DFIG Based on Recurrent Legendre Fuzzy Neural Network

作　　者：徐鹏涛 李东东 赵耀 XU Pengtao;LI Dongdong;ZHAO Yao(School of Electrical Engineering,Shanghai University of Electric Power,Shanghai 200090,China)

机构地区：[1]上海电力大学电气工程学院,上海200090

出　　处：《上海电力大学学报》2024年第5期405-414,420,共11页Journal of Shanghai University of Electric Power

基　　金：国家自然科学基金(51977128);上海市自然科学基金(21ZR1425400);上海市青年科技启明星计划资助项目(21QC1400200)。

摘　　要：针对双馈感应发电机(DFIG)易受外界干扰而对并网产生影响的问题,将循环Legendre模糊神经网络(RLFNN)与二阶滑模控制(SOSMC)应用于DFIG控制中,从而提高了DFIG在传感器故障和不确定条件下的功率跟踪能力。首先,SOSMC采用超螺旋算法进行推导,并使用Lyapunov第二定理证明了控制系统的渐近稳定性。其次,提出了使用RLFNN来估计不确定部分,RLFNN的控制律与参数可在线训练,以进一步确保系统鲁棒性。仿真结果表明,所提出的方法能够使DFIG在发生传感器故障、参数变化以及外部干扰情况下保持正常运行,实现了有效容错控制。Addressing the susceptibility of doubly fed induction generators(DFIG)to external disturbances that adversely affect power grid,recurrent Legendre fuzzy neural network(RLFNN)and second-order sliding mode control(SOSMC)are employed for DFIG control.The aim is to enhance the power tracking capabilities of DFIG under sensor faults and uncertain conditions.Firstly,the control law of SOSMC is derived using the super-twisting algorithm,and the asymtotic stability of the control system is proven using Lyapunov’s second theorem.Then the utilization of RLFNN is proposed to estimate uncertain components,with the control law and parameters of RLFNN to be trained online,enhancing the robustness of the system.Simulation results demonstrate that the proposed method enables DFIGs to maintain sensor faults,parameter variations,and external disturbances normal operation under achieving effective fault-tolerant control.

关 键 词：双馈感应发电机 容错控制 超螺旋算法 二阶滑模控制 循环Legendre模糊神经网络 

分 类 号：TM315[电气工程—电机]