基于改进的单神经元三相并网逆变器控制策略研究

摘要

为改善LCL型三相并网逆变器控制器无法根据外界条件校正参数的缺点，提出了一种基于改进的单神经元三相并网逆变器控制策略。该控制策略基于并网电流、电容电流反馈的双电流环控制策略，将并网电流外环用单神经元PI自适应控制器代替，根据有监督的Hebb学习规则对自身权值系数进行更新，利用辨识器校正PI参数，实现系统动态调节KP、KI值，仿真结果表明，该控制策略增强了系统鲁棒性、稳定性。%In order to improve the disadvantage that the LCL type of three-phase grid connected inverter isn’t able to revise the parameters according to the external conditions, this paper proposed a new control strategy based on an improved single neuron three-phase grid con-nected inverter. The grid connected current outer ring was replaced by the single neuron proportion integration (PI) self-adaptive controller. According to the supervised Hebb learning rule, one's own weight coefficients were updated. The identifier was used to revise the PI param-eters, so as to dynamically adjust theKP andKI values. The simulation results show that this control strategy enhances the robustness and stability of the system.