• Peer Review
  • Non-profit
  • Global Open Access
  • Green Channel for Rising Stars
Volume 12 Issue 1
Jan.  2025
Turn off MathJax
Article Contents

ZHU Yaoming, ZHANG Lanhong, CHEN Lulu. Vector control of doubly fed induction generator based on sliding mode variable structure [J]. Southern energy construction, 2025, 12(1): 160-167 doi:  10.16516/j.ceec.2024-127
Citation: ZHU Yaoming, ZHANG Lanhong, CHEN Lulu. Vector control of doubly fed induction generator based on sliding mode variable structure [J]. Southern energy construction, 2025, 12(1): 160-167 doi:  10.16516/j.ceec.2024-127

Vector Control of Doubly Fed Induction Generator Based on Sliding Mode Variable Structure

doi: 10.16516/j.ceec.2024-127
  • Received Date: 2024-04-26
  • Accepted Date: 2024-04-27
  • Rev Recd Date: 2024-05-16
  • Available Online: 2024-11-04
  • Publish Date: 2025-01-30
  •   Objective  Current loop is an important control link in the grid-connection process of doubly fed induction generator (DFIG). Aiming at the poor dynamic performance of doubly fed wind power generation system during parameter uptake in traditional PI control and the buffeting problem in traditional reaching law sliding mode control, the sliding mode variable structure is studied and designed.   Method  The research method of vector control of DFIG based on sliding mode variable structure mainly focused on utilizing the advantages of sliding mode variable structure control, such as fast response speed, insensitivity to parameter changes and perturbations, etc., and combining with the characteristics of DFIG to realize accurate vector control. First, the sliding mode surface was designed to ensure that the system state slides on the sliding mode surface, and then the sliding mode controller was designed to stabilize the system state and adjustments were made to enable the system to track the desired trajectory through feedback, to achieve efficient and stable control of DFIG. Additionally, the anti-buffeting factor was combined with the power function to design the reaching law function, proposing a kind of improved power function to improve the controller's performance. A sliding mode controller based on improved power function was then constructed.   Result  Simulations show that the control process of the sliding mode controller based on an improved power function is almost free of overshooting under sudden wind speed changes.   Conclusion  Compared with the traditional PI regulator, the sliding mode controller based on an improved power function has excellent dynamic performance and control accuracy, effectively improving the control effect, stability, and anti-interference ability in the doubly fed wind power generation system.
  • [1] 杨博, 束洪春, 邱大林, 等. 变风速下双馈感应发电机非线性鲁棒状态估计反馈控制 [J]. 电力系统自动化, 2019, 43(4): 60-69. DOI:  10.7500/AEPS20180330007.

    YANG B, SHU H C, QIU D L, et al. Nonlinear robust state estimation feedback control of doubly-fed induction generator under variable wind speeds [J]. Automation of electric power systems, 2019, 43(4): 60-69. DOI:  10.7500/AEPS20180330007.
    [2] NIAN H, YI X L. Coordinated control strategy for doubly-fed induction generator with DC connection topology [J]. IET renewable power generation, 2015, 9(7): 747-756. DOI:  10.1049/iet-rpg.2014.0347.
    [3] 黄守道, 谭健, 许琤, 等. 双馈风力发电机有功、无功的解耦控制 [J]. 电气传动, 2008, 38(9): 7-10. DOI:  10.3969/j.issn.1001-2095.2008.09.002.

    HUANG S D, TAN J, XU Z, et al. Decoupled control of active and reactive power in a doubly fed induction generator [J]. Electric drive, 2008, 38(9): 7-10. DOI:  10.3969/j.issn.1001-2095.2008.09.002.
    [4] 杨文焕, 倪凯峰. 基于模糊控制的双馈风力发电最优控制策略 [J]. 电工技术学报, 2013, 28(4): 279-284. DOI:  10.19595/j.cnki.1000-6753.tces.2013.04.038.

    YANG W H, NI K F. Optimal control strategy for the doubly-fed wind power system based on the fuzzy control technique [J]. Transactions of China electrotechnical society, 2013, 28(4): 279-284. DOI:  10.19595/j.cnki.1000-6753.tces.2013.04.038.
    [5] BEN ELGHALI S E, EL HACHEMI BENBOUZID M, AHMED-ALI T, et al. High-order sliding mode control of a marine current turbine driven doubly-fed induction generator [J]. IEEE journal of oceanic engineering, 2010, 35(2): 402-411. DOI:  10.1109/JOE.2010.2040402.
    [6] 李晓明, 牛玉广, 王世林, 等. 双馈风机自适应神经分散协调预测控制 [J]. 控制理论与应用, 2015, 32(7): 902-911. DOI:  10.7641/CTA.2015.40972.

    LI X M, NIU Y G, WANG S L, et al. Adaptive neural decentralized-coordinated predictive control of double fed induction generator [J]. Control theory & applications, 2015, 32(7): 902-911. DOI:  10.7641/CTA.2015.40972.
    [7] 周涛. 基于一种新型趋近律的自适应滑模控制 [J]. 控制与决策, 2016, 31(7): 1335-1338. DOI:  10.13195/j.kzyjc.2015.0518.

    ZHOU T. Adaptive sliding control based on a new reaching law [J]. Control and decision, 2016, 31(7): 1335-1338. DOI:  10.13195/j.kzyjc.2015.0518.
    [8] 叶佳卓, 邓双喜, 周宁博, 等. 基于幂次变速趋近律的逆变器环流滑模控制 [J]. 电力系统及其自动化学报, 2021, 33(8): 56-61. DOI:  10.19635/j.cnki.csu-epsa.000707.

    YE J Z, DENG S X, ZHOU N B, et al. Circulating current sliding-mode control of inverter based on power variable rate reaching law [J]. Proceedings of the CSU-EPSA, 2021, 33(8): 56-61. DOI:  10.19635/j.cnki.csu-epsa.000707.
    [9] XIONG L Y, WANG J, MI X, et al. Fractional order sliding mode based direct power control of grid-connected DFIG [J]. IEEE transactions on power systems, 2018, 33(3): 3087-3096. DOI:  10.1109/TPWRS.2017.2761815.
    [10] 王君瑞, 彭飘飘, 刘凡齐. 基于滑模变结构双馈风力发电机直接功率控制 [J]. 计算机仿真, 2015, 32(12): 85-88, 196. DOI:  10.3969/j.issn.1006-9348.2015.12.018.

    WANG J R, PENG P P, LIU F Q. Direct power control for doubly fed induction generator based on sliding mode variable [J]. Computer simulation, 2015, 32(12): 85-88, 196. DOI:  10.3969/j.issn.1006-9348.2015.12.018.
    [11] 方馨, 王丽梅, 张康. 基于扰动观测器的永磁直线电机高阶非奇异快速终端滑模控制 [J]. 电工技术学报, 2023, 38(2): 409-421. DOI:  10.19595/j.cnki.1000-6753.tces.211779.

    FANG X, WANG L M, ZHANG K. High order nonsingular fast terminal sliding mode control of permanent magnet linear motor based on disturbance observer [J]. Transactions of China electrotechnical society, 2023, 38(2): 409-421. DOI:  10.19595/j.cnki.1000-6753.tces.211779.
    [12] 林增健, 刘文泽, 谭炜豪. 基于变指数滑模的双馈水电机组功率控制研究 [J]. 三峡大学学报(自然科学版), 2021, 43(6): 81-86. DOI:  10.13393/j.cnki.issn.1672-948x.2021.06.013.

    LIN Z J, LIU W Z, TAN W H. Research on power control of hydroelectric generating units with doubly fed induction generators based on variable exponential sliding mode control [J]. Journal of China Three Gorges University (Natural Sciences Edition), 2021, 43(6): 81-86. DOI:  10.13393/j.cnki.issn.1672-948x.2021.06.013.
    [13] 李培强, 邱时严, 李欣然, 等. 双馈风力发电系统的滑模变结构控制技术 [J]. 电力系统及其自动化学报, 2017, 29(8): 29-35. DOI:  10.3969/j.issn.1003-8930.2017.08.005.

    LI P Q, QIU S Y, LI X R, et al. Sliding mode variable structure control technology used in doubly-fed induction generation system [J]. Proceedings of the CSU-EPSA, 2017, 29(8): 29-35. DOI:  10.3969/j.issn.1003-8930.2017.08.005.
    [14] 赵永玮, 董锋斌, 刘昌建. 基于新型幂次趋近律的DFIG滑模直接功率控制 [J]. 电工技术, 2023(8): 24-30. DOI:  10.19768/j.cnki.dgjs.2023.08.007.

    ZHAO Y W, DONG F B, LIU C J. DFIG sliding mode direct power control based on novel power reaching law [J]. Electric engineering, 2023(8): 24-30. DOI:  10.19768/j.cnki.dgjs.2023.08.007.
    [15] 缪仲翠, 张文宾, 余现飞, 等. 基于转速估计的PMSM分数阶积分滑模控制 [J]. 太阳能学报, 2021, 42(3): 28-34. DOI:  10.19912/j.0254-0096.tynxb.2018-1045.

    MIAO Z C, ZHANG W B, YU X F, et al. Fractional order integral sliding mode control for PMSM based on speed estimation [J]. Journal of solar energy, 2021, 42(3): 28-34. DOI:  10.19912/j.0254-0096.tynxb.2018-1045.
    [16] 张国山, 李现磊. 一种滑模控制新型幂次趋近律的设计与分析 [J]. 天津大学学报(自然科学与工程技术版), 2020, 53(11): 1112-1119. DOI:  10.11784/tdxbz201907050.

    ZHANG G S, LI X L. Design and analysis of a new power reaching law for sliding mode control [J]. Journal of Tianjin University (Science and Technology Edition), 2020, 53(11): 1112-1119. DOI:  10.11784/tdxbz201907050.
    [17] 廖瑛, 杨雅君, 王勇. 滑模控制的新型双幂次组合函数趋近律 [J]. 国防科技大学学报, 2017, 39(3): 105-110. DOI:  10.11887/j.cn.201703017.

    LIAO Y, YANG Y J, WANG Y. Novel double power combination function reaching law for sliding mode control [J]. Journal of National University of Defense Technology, 2017, 39(3): 105-110. DOI:  10.11887/j.cn.201703017.
    [18] 董锋斌, 刘昌建, 赵永玮, 等. 基于扩张状态观测器的DFIG网侧变换器滑模控制 [J]. 电力工程技术, 2023, 42(2): 206-214. DOI:  10.12158/j.2096-3203.2023.02.024.

    DONG F B, LIU C J, ZHAO Y W, et al. Sliding mode control for DFIG grid-side converter based on extended state observer [J]. Electric power engineering technology, 2023, 42(2): 206-214. DOI:  10.12158/j.2096-3203.2023.02.024.
    [19] 高日辉, 陈燕, 马春燕, 等. 基于滑模控制的三相PWM整流器结构设计与性能仿真 [J]. 太原理工大学学报, 2018, 49(2): 258-263. DOI:  10.16355/j.cnki.issn1007-9432tyut.2018.02.013.

    GAO R H, CHEN Y, MA C Y, et al. Physical design and performance simulation of three-phase PWM rectifier based on sliding-mode control [J]. Journal of Taiyuan University of Technology, 2018, 49(2): 258-263. DOI:  10.16355/j.cnki.issn1007-9432tyut.2018.02.013.
    [20] 张兰红, 李胜, 张卿杰. 永磁同步风电系统转速环自抗扰控制器设计 [J]. 控制工程, 2022, 29(9): 1645-1651. DOI:  10.14107/j.cnki.kzgc.20200573.

    ZHANG L H, LI S, ZHANG Q J. Design of active disturbance rejection controller in speed loop for permanent magnet synchronous wind power generation system [J]. Control engineering of China, 2022, 29(9): 1645-1651. DOI:  10.14107/j.cnki.kzgc.20200573.
  • 通讯作者: 陈斌, bchen63@163.com
    • 1. 

      沈阳化工大学材料科学与工程学院 沈阳 110142

    1. 本站搜索
    2. 百度学术搜索
    3. 万方数据库搜索
    4. CNKI搜索

Figures(9)  / Tables(1)

Article Metrics

Article views(212) PDF downloads(10) Cited by()

Related

Vector Control of Doubly Fed Induction Generator Based on Sliding Mode Variable Structure

doi: 10.16516/j.ceec.2024-127

Abstract:   Objective  Current loop is an important control link in the grid-connection process of doubly fed induction generator (DFIG). Aiming at the poor dynamic performance of doubly fed wind power generation system during parameter uptake in traditional PI control and the buffeting problem in traditional reaching law sliding mode control, the sliding mode variable structure is studied and designed.   Method  The research method of vector control of DFIG based on sliding mode variable structure mainly focused on utilizing the advantages of sliding mode variable structure control, such as fast response speed, insensitivity to parameter changes and perturbations, etc., and combining with the characteristics of DFIG to realize accurate vector control. First, the sliding mode surface was designed to ensure that the system state slides on the sliding mode surface, and then the sliding mode controller was designed to stabilize the system state and adjustments were made to enable the system to track the desired trajectory through feedback, to achieve efficient and stable control of DFIG. Additionally, the anti-buffeting factor was combined with the power function to design the reaching law function, proposing a kind of improved power function to improve the controller's performance. A sliding mode controller based on improved power function was then constructed.   Result  Simulations show that the control process of the sliding mode controller based on an improved power function is almost free of overshooting under sudden wind speed changes.   Conclusion  Compared with the traditional PI regulator, the sliding mode controller based on an improved power function has excellent dynamic performance and control accuracy, effectively improving the control effect, stability, and anti-interference ability in the doubly fed wind power generation system.

ZHU Yaoming, ZHANG Lanhong, CHEN Lulu. Vector control of doubly fed induction generator based on sliding mode variable structure [J]. Southern energy construction, 2025, 12(1): 160-167 doi:  10.16516/j.ceec.2024-127
Citation: ZHU Yaoming, ZHANG Lanhong, CHEN Lulu. Vector control of doubly fed induction generator based on sliding mode variable structure [J]. Southern energy construction, 2025, 12(1): 160-167 doi:  10.16516/j.ceec.2024-127
    • 为迎接能源危机带来的挑战,风力发电在新能源发电中比例显著提升。因双馈风力发电机 (Doubly−fed Induction Generator,DFIG) 相对于其他类型的发电系统成本较低,能够实现更高的发电效率,提高能量转换效率[1],得到了广泛应用。DFIG可以通过调节转子回路的功率因数来适应不同的运行工况,从而提高系统的灵活性和稳定性[2]。文献[3]中双闭环PI控制在解耦功率方面已经取得了一定的成功,但是对于双馈感应发电机(DFIG)这样一个多变量、高阶、非线性、强耦合的复杂系统而言,在参数摄动等非匹配扰动情况下,仍然存在挑战。传统的PI调节器难以捕捉系统动态变化和耦合效应的细微变化,难以满足系统在这些情况下的控制要求,因此,为了提高双馈风力发电机的稳定性,需要采用更为精准的控制策略,如模糊控制、滑模控制和预测控制等方法,以更好地应对参数摄动和非匹配扰动带来的挑战。模糊控制能够处理系统模型的不确定性和复杂性,滑模控制则通过引入滑模面来抑制系统的不良影响,预测控制则可以利用系统模型对未来的发展趋势进行预测,并相应地调整控制输入,以达到优化控制性能的目的。这些非线性控制理论的引入为DFIG控制系统带来了新的可能性,有望进一步提高其在各种工况下的稳定性和鲁棒性[4-6]

      滑模控制具有快速的响应特性,能够在短时间内对系统的变化做出调整,保持系统的稳定性和快速性,有效地抑制系统的抖振现象。然而,当系统状态变量处于滑动模态时,滑模控制往往会出现明显的抖振现象。这种抖振现象可能会影响系统的稳定性和控制性能,因此需要通过进一步的优化和改进控制策略来解决这个问题。一些常见的方法包括引入滑模面设计中的非线性元素、融入一些新的控制策略,以及优化滑模控制器的参数等手段。文献[7]提出了趋近律的概念,在快速趋近目标的同时降低抖振现象。这一概念给滑模控制器提供了一种有效的方法,改善了系统的抗抖振性。文献[8]则探索了一种幂次变速趋近律,在逆变器的应用方面,该方法提升了系统动态响应速度,有效抑制环流的影响,实现了更精确的控制。然而,该方法中所选用的趋近律中存在符号函数,其不连续特性仍可能导致系统抖振。文献[9]将分数阶微积分应用于DFIG控制系统中,构建了分数阶滑模面,并引入了分数阶趋近律,保证了系统的快速性。但由于该方法计算量大,物理实现比较困难,在工业中的应用仍具有一定挑战性。

      为了解决参数摄动对DFIG的影响,以及降低滑模控制中的抖振现象。文章提出了一种改进型幂次函数滑模控制,并在幂次函数中引入抗抖振因子函数,将该控制方法应用于DFIG系统中,提高了并网发电过程中的稳定性和抗干扰能力。

    • 双馈风力发电系统如图1所示,主要包括风机、齿轮箱、双馈风力发电机、转子侧变流器、网侧变流器、变压器和滤波电感等部分。

      Figure 1.  Diagram of doubly fed wind power generation system

      dq轴旋转坐标下,DFIG的数学模型[10]如式(1)-式(4)所示:

      电压方程为:

      $$ \left\{ \begin{gathered} {u_{{\mathrm{s}}d}} = {R_{\mathrm{s}}}{i_{{\mathrm{s}}d}} + \dfrac{{\mathrm{d}{\psi _{{\mathrm{s}}d}}}}{{\mathrm{d}t}} - {\omega _{\mathrm{s}}}{\psi _{{\mathrm{s}}{{q}}}} \\ {u_{{\mathrm{s}}q}} = {R_{\mathrm{s}}}{i_{{\mathrm{s}}q}} + \dfrac{{\mathrm{d}{\psi _{{\mathrm{s}}q}}}}{{\mathrm{d}t}} + {\omega _{\mathrm{s}}}{\psi _{{\mathrm{s}}d}} \\ \end{gathered} \right. $$ (1)
      $$ \left\{ \begin{gathered} {u_{{\mathrm{r}}d}} = {R_{\mathrm{r}}}{i_{{\mathrm{r}}d}} + \dfrac{{\mathrm{d}{\psi _{{\mathrm{r}}d}}}}{{\mathrm{d}t}} - {\omega _{{\mathrm{s}}{\mathrm{l}}}}{\psi _{{\mathrm{r}}q}} \\ {u_{{\mathrm{r}}q}} = {R_{\mathrm{r}}}{i_{{\mathrm{r}}q}} + \dfrac{{\mathrm{d}{\psi _{{\mathrm{r}}q}}}}{{\mathrm{d}t}} + {\omega _{{\mathrm{s}}{\mathrm{l}}}}{\psi _{{\mathrm{r}}d}} \\ \end{gathered} \right. $$ (2)

      式中:

      $ {u_{{\mathrm{s}}d}} $——定子d轴电压(V);

      $ {u_{{\mathrm{s}}q}} $——定子q轴电压(V);

      $ {u_{{\mathrm{r}}d}} $——转子d轴电压(V);

      $ {u_{{\mathrm{r}}q}} $——转子q轴电压(V);

      $ {i_{{\mathrm{s}}d}} $ ——定子d轴电流(A);

      $ {i_{{\mathrm{s}}q}} $ ——定子q轴电流(A);

      $ {i_{{\mathrm{r}}d}} $ ——转子d轴电流(A);

      $ {i_{{\mathrm{r}}q}} $ ——转子q轴电流(A);

      $ {\psi _{{\mathrm{s}}d}} $——定子q轴磁链(Wb);

      $ {\psi _{{\mathrm{s}}q}} $——定子d轴磁链(Wb);

      $ {\psi _{{\mathrm{r}}d}} $——转子d轴磁链(Wb);

      $ {\psi _{{\mathrm{r}}q}} $——转子q轴磁链(Wb);

      $ {R_{\mathrm{s}}} $ ——定子电阻(Ω);

      $ {R_{\mathrm{r}}} $ ——转子电阻(Ω);

      $ {\omega _1} $——定子磁场的同步电角速度(rad/s);

      $ {\omega _{{\mathrm{s}}{\mathrm{l}}}} $——转差角速度(rad/s)。

      磁链方程为:

      $$ \left\{ {\begin{array}{*{20}{c}} {{\psi _{{\mathrm{s}}d}} = {L_{\mathrm{s}}}{i_{{\mathrm{s}}d}} + {L_{\mathrm{m}}}{i_{{\mathrm{r}}d}}} \\ {{\psi _{{\mathrm{s}}q}} = {L_{\mathrm{s}}}{i_{{\mathrm{s}}q}} + {L_{\mathrm{m}}}{i_{{\mathrm{r}}q}}} \\ {{\psi _{{\mathrm{r}}d}} = {L_{\mathrm{r}}}{i_{{\mathrm{r}}d}} + {L_{\mathrm{m}}}{i_{{\mathrm{s}}d}}} \\ {{\psi _{{\mathrm{r}}q}} = {L_{\mathrm{r}}}{i_{{\mathrm{r}}q}} + {L_{\mathrm{m}}}{i_{{\mathrm{s}}q}}} \end{array}} \right. $$ (3)

      式中:

      $ {L_{\mathrm{m}}} $——旋转坐标系下的等效互感(H)。

      功率方程:

      $$ \left\{ {\begin{array}{*{20}{c}} {{P_{\mathrm{s}}} = {u_{{\mathrm{s}}d}}{i_{{\mathrm{s}}d}} + {u_{{\mathrm{s}}q}}{i_{{\mathrm{s}}q}}} \\ {{Q_{\mathrm{s}}} = {u_{{\mathrm{s}}q}}{i_{{\mathrm{s}}d}} - {u_{{\mathrm{s}}d}}{i_{{\mathrm{s}}q}}} \\ {{P_{\mathrm{r}}} = {u_{{\mathrm{r}}d}}{i_{{\mathrm{r}}d}} + {u_{{\mathrm{r}}q}}{i_{{\mathrm{r}}q}}} \\ {{Q_{\mathrm{r}}} = {u_{{\mathrm{r}}q}}{i_{{\mathrm{r}}d}} - {u_{{\mathrm{r}}d}}{i_{{\mathrm{r}}q}}} \end{array}} \right. $$ (4)

      式中:

      $ {P_{\mathrm{s}}} $——定子有功功率(W);

      $ {Q_{\mathrm{s}}} $——定子无功功率(W);

      $ {P_{\mathrm{r}}} $——转子有功功率(W);

      $ {Q_{\mathrm{r}}} $——转子无功功率(W)。

    • 双馈风力发电机基于磁场定向的矢量控制通过坐标变换,将三相电机模型等效成dq坐标系下的直流电机模型。通过对转子电流的精准控制,实现了对发电机转矩和功率的精准控制,从而提升了电机的性能和运行效率。

      将定子磁链定向在旋转坐标下的d轴,由式(3)得:

      $$ \left\{ \begin{gathered} {\psi _{{\mathrm{s}}d}} = {L_{\mathrm{s}}}{i_{{\mathrm{s}}d}} + {L_{\mathrm{m}}}{i_{{\mathrm{r}}d}} = {\psi _{\mathrm{s}}} \\ {\psi _{{\mathrm{s}}q}} = {L_{\mathrm{s}}}{i_{{\mathrm{s}}q}} + {L_{\mathrm{m}}}{i_{{\mathrm{r}}q}} = 0 \\ \end{gathered} \right. $$ (5)

      式中:

      $ {\psi _{\mathrm{s}}} $——定子磁链矢量幅值(Wb)。

      将式(5)化简得式(6)。

      $$ \left\{ \begin{gathered} {i_{{\mathrm{s}}d}} = \dfrac{1}{{{L_{\mathrm{s}}}}}({\psi _{\mathrm{s}}} - {L_{\mathrm{m}}}{i_{{\mathrm{r}}d}}) \\ {i_{{\mathrm{s}}q}} = - \dfrac{{{L_{\mathrm{m}}}}}{{{L_{\mathrm{s}}}}}{i_{{\mathrm{r}}q}} \\ \end{gathered} \right. $$ (6)

      忽略定子电阻影响,定子电压方程为:

      $$ \left\{ {\begin{split}& {{u_{{\mathrm{s}}d}} = {R_{\mathrm{s}}}{i_{{\mathrm{s}}d}} + \dfrac{{\mathrm{d}}}{{{\mathrm{d}}t}}{\psi _{{\mathrm{s}}d}} - {\omega _{\text{l}}}{\psi _{{\mathrm{s}}q}} = \dfrac{{\mathrm{d}}}{{{\mathrm{d}}t}}{\psi _{{\mathrm{s}}d}} = 0} \\& {{u_{{\mathrm{s}}q}} = {R_{\mathrm{s}}}{i_{{\mathrm{s}}q}} + \dfrac{{\mathrm{d}}}{{{\mathrm{d}}t}}{\psi _{{\mathrm{s}}q}} + {\omega _{\text{l}}}{\psi _{{\mathrm{s}}d}} = {\omega _{\text{l}}}{\psi _{\mathrm{s}}} = {U_{\mathrm{s}}}} \end{split}} \right. $$ (7)

      式中:

      $ {U_{\mathrm{s}}} $——定子电压矢量幅值(V)。

      将式(7)代入式(4)得到定子功率表达式为:

      $$ \left\{ \begin{gathered} {P_{\mathrm{s}}} = {u_{{\mathrm{s}}d}}{i_{{\mathrm{s}}d}} + {u_{{\mathrm{s}}q}}{i_{{\mathrm{s}}q}} = {U_{\mathrm{s}}}{i_{{\mathrm{s}}q}} = - {U_{\mathrm{s}}}\dfrac{{{L_{\mathrm{m}}}}}{{{L_{\mathrm{s}}}}}{i_{{\mathrm{r}}q}} \\ {Q_{\mathrm{s}}} = {u_{{\mathrm{s}}q}}{i_{{\mathrm{s}}d}} - {u_{{\mathrm{s}}d}}{i_{{\mathrm{s}}q}} = {U_{\mathrm{s}}}{i_{{\mathrm{s}}d}} = {U_{\mathrm{s}}}\dfrac{{({\psi _{\mathrm{s}}} - {L_{\mathrm{m}}}{i_{{\mathrm{r}}d}})}}{{{L_{\mathrm{s}}}}} \\ \end{gathered} \right. $$ (8)

      电磁转矩表达式可化简为:

      $$ {T_{\mathrm{e}}} = {n_{\mathrm{p}}}\dfrac{{{L_{\mathrm{m}}}}}{{{L_{\mathrm{s}}}}}({\psi _{{\mathrm{s}}q}}{i_{{\mathrm{r}}d}} - {\psi _{{\mathrm{s}}d}}{i_{{\mathrm{r}}q}}) = - {n_{\mathrm{p}}}\dfrac{{{L_{\mathrm{m}}}}}{{{L_{\mathrm{s}}}}}{\psi _{\mathrm{s}}}{i_{{\mathrm{r}}q}} $$ (9)

      由式(9)可以看出,转子dq轴电流可以对有功和无功功率解耦控制。将式(6)代入到式(5)得转子磁链方程为:

      $$ \left\{ \begin{gathered} {\psi _{{\mathrm{r}}d}} = {L_{\mathrm{r}}}{i_{{\mathrm{r}}d}} + {L_{\mathrm{m}}}{i_{{\mathrm{s}}d}} = \sigma {L_{\mathrm{r}}}{i_{{\mathrm{r}}d}} + \dfrac{{{L_{\mathrm{m}}}{\psi _{\mathrm{s}}}}}{{{L_{\mathrm{s}}}}} \\ {\psi _{{\mathrm{r}}q}} = {L_{\mathrm{r}}}{i_{{\mathrm{r}}q}} + {L_{\mathrm{m}}}{i_{{\mathrm{s}}q}} = \sigma {L_{\mathrm{r}}}{i_{{\mathrm{r}}q}} \\ \end{gathered} \right. $$ (10)

      σ——总漏磁系数,表达式为:

      $$ \sigma = 1 - \dfrac{{L_{\mathrm{m}}^2}}{{{L_{\mathrm{s}}}{L_{\mathrm{r}}}}} $$ (11)

      将式(10)代入式(2)得:

      $$\left\{\begin{split}& u_{{\mathrm{r}}d}=R_{\mathrm{r}} i_{{\mathrm{r}}d}+\sigma L_{\mathrm{r}} \dfrac{{\mathrm{d}}i_{{\mathrm{r}}d}}{{\mathrm{d}}t}+\Delta u_{{\mathrm{r}}d} \\& u_{{\mathrm{r}}q}=R_{\mathrm{r}} i_{{\mathrm{r}}q}+\sigma L_{\mathrm{r}} \dfrac{{\mathrm{d}}i_{{\mathrm{r}}q}}{{\mathrm{d}}t}+\Delta u_{{\mathrm{r}}q} \\& \Delta u_{{\mathrm{r}}d}=-\left(\omega_1-\omega_{\mathrm{r}}\right) \sigma L_{\mathrm{r}} i_{{\mathrm{r}}q} \\& \Delta u_{{\mathrm{r}}q}=\left(\omega_1-\omega_{\mathrm{r}}\right) \dfrac{L_{\mathrm{s}}}{L_{\mathrm{m}}} \psi_{\mathrm{s}}+\left(\omega_1-\omega_{\mathrm{r}}\right) \sigma L_{\mathrm{r}} i_{{\mathrm{r}}d} \end{split}\right. $$ (12)

      式中:

      urd、∆urq ——交叉耦合项,可通过设置前馈补偿项以提升系统的动稳态性能;

      $ {\omega _{\mathrm{r}}} $ ——转子电角速度(rad/s)。

      结合上述推导,文章构建了基于PI调节器的转子侧变流器控制框图,具体结构如图2所示。

      Figure 2.  Block diagram of rotor side converter control based on PI regulator

    • 滑模控制(Sliding Mode Control,SMC)是非线性控制[11],它通过非连续性的控制来使控制对象运行在滑模面上,促使系统实现“滑动模态运动”[12-14]。 然而,其缺点是当控制目标运行到滑模面后,在滑模面来回穿梭,以锯齿状态趋近平衡点 [15-16]。这种运动方式会造成目标抖振,对系统的控制效果造成影响 [17-19]

      因此,降低滑模控制中的抖振性是滑模控制研究的重点。在解决抖振问题方面,应用在自抗扰控制系统中的幂次函数取得了良好的效果,能够使系统实现无抖振、单调的收敛[20]

    • $$ {\mathrm{fal}}(s,a,\delta ) = \left\{ {\begin{split}& {{{\mathrm{sgn}}} (s) · |s{|^a}}&{\quad |s| \geqslant \delta } \\ & {\dfrac{s}{{{\delta ^{(1 - a)}}}}}&{\quad |s| < \delta } \end{split}} \right. $$ (13)

      s ——输入信号;

      sgn(s) ——符号函数,定义为:

      $$ {\text{sgn}}(s){\text{ = }}\left\{ {\begin{array}{*{20}{l}} {\text{1}}&{\quad s > {\text{0}}} \\ {\text{0}}&{\quad s{\text{ = 0}}} \\ {{{ - 1}}}&{\quad s < {\text{0}}} \end{array}} \right. $$ (14)

      符号函数如下图3所示,即s>0时函数值为l ;s<0时函数值为−1;s=0时函数值为0。

      Figure 3.  The graph of the sign function

      由于符号函数呈现阶跃特性,在s<0时取−1,s>0时取1,控制器的控制效果也会呈现阶跃特性,其控制器的性能也会受到影响。因此,引进一种抗抖振因子函数来改善传统的幂次函数。改进型幂次函数如下式(15)所示。

      $$ {\mathrm{Gfal}}(s,a,\delta ) = \left\{ {\begin{split}& {{\mathrm{G}}(s)\cdot|s{|^a}}&{\quad |s| \geqslant \delta } \\ & {\dfrac{s}{{{\delta ^{(1 - a)}}}}}&{\quad |s| < \delta } \end{split}} \right. $$ (15)

      式中:

      G(s)为抗抖振因子函数,且:

      $$ {\mathrm{G}}(s) = \dfrac{s}{{\left| s \right| + v}} $$ (16)

      式中v>0。抗抖振因子函数G(s)如图4所示,其在零点两侧无穷处分别渐近于−1和1。

      Figure 4.  The graph of the G(x) function

      比较图3图4两种函数的图像,Gfal函数在零点两侧呈现渐进特性,因此在不同情况下的Gfal函数的抗抖振性更好。基于这一观察,将非线性函数fal全部替换为Gfal。

    • 转子电流dq轴误差函数定义如下:

      $$ \left[ {\begin{array}{*{20}{c}} {{e_d}(t)} \\ {{e_q}(t)} \end{array}} \right] = \left[ {\begin{array}{*{20}{c}} {i_{d{\mathrm{r}}}^ * (t)} \\ {i_{q{\mathrm{r}}}^ * (t)} \end{array}} \right] - \left[ {\begin{array}{*{20}{c}} {{i_{d{\mathrm{r}}}}(t)} \\ {{i_{q{\mathrm{r}}}}(t)} \end{array}} \right] $$ (17)

      为了防止稳态误差的产生,影响系统的性能,设计了积分滑模控制器,积分滑模面函数如下式(18)所示:

      $$ s_{d_{q}}(t)=e_{d_{q}}(t)+c_{dq} \int_{0}^{t} e_{d_{q}}(t) {\mathrm{d}} t $$ (18)

      式中,$ {s_{dq}}\left( t \right) = {\left[ {{s_d}\left( t \right){s_q}\left( t \right)} \right]^{\mathrm{T}}} $,$ {e_{dq}}\left( t \right) = {\left[ {{e_d}\left( t \right){e_q}\left( t \right)} \right]^{\mathrm{T}}} $,$ {c_{dq}} $为积分常数,t

      改进型幂次滑模控制器趋近律设计为:

      $$ \dfrac{{{\text{d}}{s_{dq}}(t)}}{{{\text{d}}t}} = - \varepsilon {\mathrm{Gfal}}({s_{dq}},\alpha ,\delta ) $$ (19)

      式中:

      ε ——滑模增益。

      联立式(19)和式(18)得:

      $$ \varepsilon {\mathrm{Gfal}}({s_{dq}},\alpha ,\delta ) + {c_{dq}}{e_{dq}}(t) = \dfrac{{{\mathrm{d}}{i_{dq{\mathrm{r}}}}}}{{{\mathrm{d}}t}} $$ (20)

      为了验证改进型幂次滑模控制的稳定性,定义Lyapunov函数:

      $$ V = \dfrac{1}{2}{s_{dq}}^2 $$ (21)

      联立式(21)和式(18)得:

      $$ \dfrac{{{\mathrm{d}}{s_{dq}}}}{{{\mathrm{d}}t}} = i_{dq{\mathrm{r}}}^* + {c_{dq}}{e_{dq}}(t) + \dfrac{{{R_{\mathrm{r}}}}}{\sigma }{i_{dq{\mathrm{r}}}} - \dfrac{{{u_{dq{\mathrm{r}}}}}}{\sigma } + \dfrac{{\Delta {u_{dq{\mathrm{r}}}}}}{\sigma } $$ (22)

      对式(22)求导:

      $$ \begin{split} \mathop V\limits^ \bullet = & {s_{dq}}\left( {i_{dq{\mathrm{r}}}^* + {c_{dq}}{e_{dq}}(t) + \dfrac{{{R_{\mathrm{r}}}}}{\sigma }{i_{dq{\mathrm{r}}}} - \dfrac{{{u_{dq{\mathrm{r}}}}}}{\sigma } + \dfrac{{\Delta {u_{dq{\mathrm{r}}}}}}{\sigma }} \right) = \\& - {s_{dq}}\varepsilon {\mathrm{Gfal}}({s_{dq}},\alpha ,\delta ) \end{split} $$ (23)

      δ>0,α∈(0,1)时,Lyapunov函数V正定,且V≤0,满足滑动模态存在不等式条件:

      $$ \underset{{s}_{dq}\to 0}{\mathrm{lim}}{s}_{dq}{{\mathop {{s}}\limits^ \bullet}_{dq}}\leqslant 0 $$ (24)

      由此可知采用上述滑模控制律时,系统满足Lyapunov稳定性条件。

      系统采用1.2节中的基于定子磁链的矢量控制策略,结合上述基于改进型幂次函数的滑模控制器,引入到传统矢量控制中基于PI控制的电流环,具体结构图如下图5所示。

      Figure 5.  Sliding mode controller based on power integral function

    • 在Matlab软件上搭建上述2种调节器的双馈风力发电系统仿真模型,仿真参数如表1所示。

      参数 数值
      定子侧电阻/Ω 1.91
      转子侧电阻/Ω 0.0621
      定子自感/H 0.0167
      转子自感/H 0.0167
      定转子间互感/H 0.0165
      电网频率/Hz 50
      额定电压/V 380
      极对数 2
      转动惯量/kg·m2 0.2

      Table 1.  Parameters of DFIG simulation

      风机模型的参数如下:空气密度为1.2 m/s,风机直径为66 m,桨距角为0°,最佳叶尖速比为8.1,齿轮箱变速比为108.18。

      图6展示了风速稳定在10 m/s时,在PI控制下(下面简称Ⅰ型系统),和改进滑模控制下(下面简称Ⅱ型系统)的功率曲线图。

      Figure 6.  Power waveform at steady wind speed

      图6为风速稳定情况下的功率曲线图,Ⅱ型系统无功功率稳定在0,有功功率也保持稳定,体现了良好的静态性能,和Ⅰ型系统相比,Ⅱ型系统的无功功率更快到达了稳定,体现了快速性。

      图7图9展示了在风速突变时候的动态响应,第一次突变1 s时风速由8 m/s突变至10 m/s,第二次突变风速由10 m/s突变至12 m/s。有功功率跟随参考值的波形图如图7所示。

      Figure 7.  The waveform graph of active power following the reference value

      相比较而言,Ⅰ型系统风速突变情况下,有功功率的追踪过程波动较大,并且在第二次风速突变时,Ⅰ型系统约在1.3 s时才追踪到参考值。Ⅱ型系统跟踪过程无超调,跟踪速度更快。

      图8展示了风速突变过程中电磁转矩的波形图,在风速两次突变过程中,Ⅱ型系统的过渡过程与Ⅰ型系统相比平滑了许多,体现了Ⅱ型系统更高的控制精度和优良的动态性能。

      Figure 8.  The waveform graph of electromagnetic torque

      图9展示了风速突变时发电机转子转速的波形图,可见在第二次风速突变的时候,Ⅱ型系统几乎无超调,而Ⅰ型系统波动较大。可见Ⅱ型系统具有更好的跟踪效果。

      Figure 9.  The waveform graph of generator rotor speed

    • 积分滑模控制器融合了滑模控制的鲁棒性和积分控制的精确性,改善了系统的抗抖振性能,能够实现对系统的精确跟踪和控制,提高系统的动态和静态性能。使DFIG系统具有较快的响应速度,能够迅速适应风速和负荷变化,保持系统稳定运行。

      对DFIG发电系统中转子侧电流环中的积分滑模控制器进行了研究,提出了一种结合抗抖振因子的幂次函数Gfal函数,降低了积分滑模控制器的抖振现象。通过仿真分析了双馈风力发电系统的功率、转速和转矩。与传统的PI调节器对比基于幂次函数的积分滑模控制器各项参数动态响应过程无超调,跟踪效果更好。仿真证明改进型幂次滑模控制器表现出了更好的动静态性能,有效抑制外界不确定因素对系统的影响,提高系统的稳定性和鲁棒性。

Reference (20)

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return