电网电力系统三相逆变换器优化控制仿真研究

Grid Power System Three-Phase Inverter Optimization Control Simulation Research

电网电力系统中三相逆变换器的性能包括输出波形稳态精度与动态响应速度,针对用普通PID控制算法在负载突变和带非线性负载时三相逆变器动态响应速度慢、输出波形THD高的问题,在双闭环控制算法的基础上,提出用径向基函数神经网络(RBF)算法整定外环(电压环)PI参数,同时用遗传算法(GA)优化径向基函数神经网络(RBF)算法的学习速率与动量因子以提高网络的收敛速度,缩短网络整定PI的时间,提高三相逆变器的性能,并用matlab仿真工具进行仿真。仿真结果表明,改进控制方案能实现三相逆变器高动态特性的同时得到高质量的稳态波形。

In order to overcome the slow dynamic response and high THD of three-phase inverter of grid power system with the sudden change load and nonlinear load, based on the double closed loop control algorithm, this paper proposes an improved scheme that using radial basis function neural network （RBF） algorithm to adjust the outer loop （ voltage loop） PI parameters and using the genetic algorithm （GA） to optimize the learning rate and momentum fac- tor of the RBF to improve the convergence speed and shorten the tuning time of PI parameters. The performance of three phase inverter is improved. The simulation results show that the proposed control scheme can achieve high dy- namic characteristics of three-phase inverter and get high quality steady waveform.