Fund Project:Project supported by the National Natural Science Foundation of China (Grant No. 61561007) and the Natural Science Foundation of Guangxi Province, China (Grant No. 2017GXNSFAA198168)
Received Date:26 June 2020
Accepted Date:13 August 2020
Available Online:11 January 2021
Published Online:20 January 2021
Abstract:There are complex nonlinear behaviors such as bifurcation and chaos in a single-phase H-bridge photovoltaic inverter under proportional integral control, which will increase the harmonic content of the output current greatly and reduce the stability of system operation and reliability of power supply. There are the problems suffering the complicated modeling and difficulty in determining the control coefficients in existing chaos control methods. The exponential delay feedback control is a further development of the delay feedback control, which has the advantages of requiring no precise mathematical model of the system and simple implementation. However, our research shows that when the exponential delay feedback control is directly applied to the system, the feedback intensity cannot be controlled, which will bring too big a disturbance to the system. Based on it, an improved exponential delayed feedback control method is proposed in this paper. Firstly, a feedback signal is formed by the difference between the output current of the system and its own delay, then the feedback signal is used to obtain the control signal through an exponential link, a subtraction link and an proportion link, and the control signal is applied to the controlled system in the form of a feedback. At the same time, the discrete mapping model of the system is established and its Jacobian matrix expression is determined. Finally, the limiting conditions of the feedback control coefficient of the control signal are derived based on the stability criterion, and the control is applied to the system. In order to verify the control effect of this method, a lot of simulation experiments are conducted. The results show that the problems that the exponential delay feedback cannot control the feedback strength and causes excessive disturbance to the system will be effectively solved by this method. When the bifurcation parameters vary greatly, the chaos behaviors in the system will be suppressed effectively, the stable operating domain of the system will be expanded greatly and the harmonic content of the output current will be reduced. Keywords:proportional integral control/ single-phase H-bridge photovoltaic inverter/ chaos control/ improved exponential delay feedback control
全文HTML
--> --> --> -->
2.1.离散模型
PI调节光伏逆变器原理如图1(a)和图1(b)实线所示. 光伏阵列的输出接至具有最大功率点跟踪功能的Boost升压变换器的输入端, 与升压电路并联的稳压电容C提供H桥逆变器直流侧电压E, 桥臂上2对开关管(S1S3)和(S2S4)(带反并联二极管)采用双极性正弦脉宽方式进行调制, 输出端是电感L和电阻R组成的阻感性负载. 输出电流i经过电流传感器采样后与参考电流iref相比较, 经过PI控制器得到调制信号icon. 图 1 PI调节单相H桥光伏逆变器控制系统 (a)引入EDFC系统原理图; (b) 引入IEDFC系统原理图 Figure1. PI regulating single-phase H-bridge photovoltaic inverter control system: (a) System schematic diagram with EDFC applied; (b) system schematic diagram with IEDFC applied
$\frac{{{\rm{d}}{i_{{\rm{con}}}}(t)}}{{{\rm{d}}t}} = {k_{\rm{p}}}\frac{{{\rm{d}}{i_{\rm{e}}}\left( t \right)}}{{{\rm{d}}t}} + {k_{\rm{i}}}{i_{\rm{e}}}\left( t \right),$
$\frac{{{\rm{d}}{i_{{\rm{con}}}}\left( t \right)}}{{{\rm{d}}t}} = - {k_{\rm{p}}}\frac{{{\rm{d}}i\left( t \right)}}{{{\rm{d}}t}} - {k_{\rm{i}}}i\left( t \right) + u\left( t \right).$
表1电路参数设定值 Table1.Set values of circuit parameters.
图 2 未引入混沌控制时电感电流峰值处分岔图(n = 100+400k, k = 1, 2, 3, ···) (a) kp为分岔参数时分岔图; (b) E为分岔参数时分岔图 Figure2. Bifurcation diagram with inductance current at peak value without chaos control (n = 100+ 400k, k = 1, 2, 3, ···): (a) Bifurcation diagram with kp as bifurcation parameter; (b) bifurcation diagram with E as bifurcation parameter.
结合(3)式和(9)式, 分别以kp, E为分岔参数, 以Ts为采样周期, 得到系统引入EDFC后电感电流峰值处分岔图如图4(a)和图4(b)所示. 引入EDFC后, kp从0.6增大到2.0、E从200 V增大到600 V时, 特征值λ1, λ2, λ3的轨迹图如图5(a)和图5(b)所示. 图 4 引入EDFC后电感电流峰值处分岔图 (n = 100+400k, k = 1, 2, 3, ···) (a) kp为分岔参数时分岔图; (b) E为分岔参数时分岔图 Figure4. Bifurcation diagram with inductance current at peak value with EDFC applied (n = 100+ 400k, k = 1, 2, 3···): (a) Bifurcation diagram with kp as bifurcation parameter; (b) bifurcation diagram with E as bifurcation parameter
图 5 引入EDFC后特征值轨迹图 (a) kp从0.6增大到2.0; (b) E从200 V增大到600 V Figure5. Eigenvalue trajectory with EDFC applied: (a) with kp increasing from 0.6 to 2.0; (b) with E increasing from 200 V to 600 V.
由上述分析可知, 当kp = 1.8或E = 500 V时, 系统进入混沌态, 为使其恢复至稳定态, 引入IEDFC. 根据(21)式, 当kp = 1.8时, 令k1 = k2 = 0.707, 当E = 500 V时, 令k1 = 0.45, k2 = 0.5. 结合(3)式和(14)式, 分别以kp, E为分岔参数, 以Ts为采样周期, 得到电感电流峰值处的分岔图如图6(a)和图6(b)所示. 图 6 引入IEDFC后电感电流峰值处分岔图 (n = 100+400k, k = 1, 2, 3, ···) (a) 以kp为分岔参数时分岔图; (b) 以E为分岔参数时分岔图 Figure6. Bifurcation diagram with inductance current at peak value with IEDFC applied (n = 100+400k, k = 1, 2, 3, ···): (a) Bifurcation diagram with kp as bifurcation parameter; (b) bifurcation diagram with E as bifurcation parameter
在引入IEDFC后, kp从0.6增大至2.0或E从200 V增大至600 V时特征值λ2, λ3的轨迹图如图7所示. 可以看出, 即使kp增大至1.8或E增大至500 V时系统也没有出现分岔与混沌现象, 特征值λ2, λ3均在单位圆内. 这说明当分岔参数变化较大时, IEDFC能有效抑制系统的非线性行为, 使系统保持稳定运行, 有效地解决了系统直接引入EDFC带来过大扰动的问题. 图 7 引入IEDFC后特征值轨迹图 (a) kp从0.6增大至2.0; (b) E从200 V增大至600 V Figure7. Eigenvalue trajectory with IEDFC applied: (a) with kp increasing from 0.6 to 2.0; (b) with E increasing from 200 V to 600 V.
图 12kp = 1.400, k1 = 0.707, k2 = 0.630时, 引入IEDFC后电感电流 (a) 时域波形图; (b) 时域波形局部放大图 Figure12. Inductor current with IEDFC applied for kp = 1.400, k1 = 0.707, k2 = 0.630: (a) Time domain waveform; (b) local magnification diagram of time-domain waveform.
图13(a)—图13(c)分别示出了当kp = 1.4, t = 0.06 s时对系统施加时间延迟反馈法[21]、扩展时间延迟反馈法[22]、基于滤波器的混沌控制法[23], 且在t = 0.14 s时对系统直流侧电压E施加ΔE = 50 V扰动后电感电流的仿真结果. 对比图12(a)与图13(a)—图13(c)可以看出, 引入IEDFC与这三种方法后系统都能由混沌态恢复至稳定态. 但当对系统直流侧电压施加扰动后, 这三种混沌控制方法都无法控制系统继续稳定运行, 而本文提出的IEDFC法仍能有效地控制系统, 使系统继续保持稳定运行. 这表明IEDFC比这三种混沌控制方法鲁棒性更强. 图 13kp = 1.4, t = 0.06 s时, 引入其他混沌控制后电感电流 (a) 引入时间延迟反馈控制后; (b) 引入扩展时间延迟反馈控制后; (c) 引入基于滤波器的混沌控制后 Figure13. Inductor current with other chaos control applied for kp = 1.4, t = 0.06 s: (a) With time-delay feedback control applied; (b) with extended time-delay feedback control applied; (c) with chaos control based on filter applied.