摘要
采用MATLAB2008A和TMS320F28335高速浮点DSP为核心器件半实物仿真装置,对特高压输电系统进行了建模仿真和实时控制的研究。控制系统采用RBF神经网络整定PID算法,动态跟踪特高压交流输电线路的过电压,快速实现系统的闭环控制。仿真试验表明,基于RBF神经网络整定PID控制的特高压磁阀可控电抗器,能动态抑制电网工频暂态过电压,在该控制模式下,单段输电线路的等效长度可以进一步延长,提高特高压交流线路的输电能力和经济性。
MATLAB2008A and high speed floating point DSP, TMS320F28335, are used as core devices to form semi-physical simulator to perform simulation and real-time control of UHV transmission system. PID control algorithm based on RBF neural network is used to dynamically track UHV AC transmission system overvohage and realize closed-loop system control. Simulated tests show that the UHV MCR with PID control based on RBF neural networks is capable of dynamically suppress grid transient overvoltage at working frequency, and under such control mode, the equivalent length of single line section can be further extended, improving line capability and economy.
出处
《电力建设》
2009年第4期29-32,共4页
Electric Power Construction
关键词
特高压
可控电抗器
过电压
神经网络
闭环控制
UHV
controlled reactor
overvoltage
neural network
closed-loop control
