-
摘要: 针对一类难以建立精确模型的单输入单输出(Single-input single-output, SISO) 非线性离散动态系统, 提出了一种数据驱动模型的自适应控制方法. 所提方法首先设计具有直链与增强结构的随机配置网络(Stochastic configuration network, SCN), 建立了一种可同时表征非线性系统低阶线性部分与高阶非线性项(未建模动态)的数据驱动模型, 并采用增量学习方法与监督机制, 对模型结构与模型参数进行同步更新优化, 保证了数据驱动模型的无限逼近能力, 解决了传统自适应控制采用交替辨识算法存在的建模精度低、模型收敛性无法保证的问题. 进而利用直链部分与增强部分, 分别设计了线性控制器及虚拟未建模动态补偿器, 建立了基于SCN 数据驱动模型的自适应控制新方法, 分析了其稳定性与收敛性, 通过数值仿真实验和采用交替辨识算法的传统自适应控制方法进行对比, 实验结果表明了所提方法的有效性.Abstract: For a class of single-input single-output (SISO) nonlinear discrete dynamical systems which are difficult to establish an accurate model, a novel adaptive control method is proposed based on data-driven model. In the proposed approach, stochastic configuration network (SCN) is first employed to build the data-driven nonlinear system model, which adopts direct link and enhancement nodes to approximate the low-order linear and the high-order nonlinear parts (unmodeled dynamics) of system, respectively. Besides, this paper employed an incremental learning algorithm and supervision mechanism to optimize the model structure and model parameters synchronously, which guarantee the universal approximation property of the data-driven model, solving the problems of low modeling accuracy and unguaranteed model convergence existing in traditional adaptive control with alternate identification algorithm. Then, the direct link and enhancement nodes are used to design the linear controller and virtual unmodeled dynamics compensator respectively. A new adaptive control approach based on SCN data-driven model is established, and the stability and convergence of the proposed control method are proved. Finally, simulation comparisons between our proposed method and the classic adaptive control method with alternate identification algorithm are made, showing the effectiveness of the proposed method.
-
图 2 基于SCN数据驱动模型的自适应控制方法结构图
Fig. 2 Structure diagram of adaptive control method with SCN-based data-driven model
图 8 基于SCN数据模型的灰分含量跟踪控制输出
Fig. 8 Output of ash content tracking control based on SCN data-driven model
图 9 基于SCN数据模型的重介质选煤灰分含量估计误差曲线
Fig. 9 Estimation error curve of ash content in dense medium separation process based on SCN data model
表 1 模型性能对比
Table 1 Performance comparison of models
模型性能指标 增强节点个数 离线建模
时间 (s)模型在线平均
绝对误差传统RVFLNN模型 17 0.257 19 0.004 6 SCN模型 9 0.245 82 0.001 3 
下载: 导出CSV
表 2 控制系统模型估计性能对比
Table 2 Comparison of performance of model estimates for control systems
基于不同模型的自适应控制系统 ${\rm MAE}$ 基于线性模型的自适应控制 0.009 2 基于BP交替辨识模型的自适应控制 0.007 0 基于ANFIS交替辨识模型的自适应控制 0.005 1 基于SCN数据模型的自适应控制 0.001 3
下载: 导出CSV
-
[1] Gao C H, Jian L, Liu X Y, Chen J M, Sun Y X. Data-driven modeling based on volterra series for multidimensional blast furnace system. IEEE Transactions on Neural Networks, 2011, 22(12): 2272−2283 doi: 10.1109/TNN.2011.2175945 [2] Diaz S, Luque J, Romero M C, Escudero J I. Power systems monitoring and control using telecom network management standards. IEEE Transactions on Power Delivery, 2005, 20(2): 1349−1356 doi: 10.1109/TPWRD.2004.833918 [3] Wu Z W, Wu Y J, Chai T Y, Sun J. Data-driven abnormal condition identification and self-healing control system for fused magnesium furnace. IEEE Transactions on Industrial Electronics, 2015, 62(3): 1703−1715 doi: 10.1109/TIE.2014.2349479 [4] Biswal G R, Maheshwari R P, Dewal M L. Modeling, control, and monitoring of S3RS-based hydrogen cooling system in thermal power plant. IEEE Transactions on Industrial Electronics, 2012, 59(1): 562−570 doi: 10.1109/TIE.2011.2134059 [5] 柴天佑. 自动化科学与技术发展方向. 自动化学报, 2018, 44(11): 1923−1930 doi: 10.16383/j.aas.2018.c180252Chai Tian-You. Development direction of automation science and technology. Acta Automatica Sinica, 2018, 44(11): 1923−1930 doi: 10.16383/j.aas.2018.c180252 [6] Landau I D. A survey of model reference adaptive techniques-theory and applications. Automatica, 1974, 10(4): 353−379 doi: 10.1016/0005-1098(74)90064-8 [7] 柴天佑, 岳恒. 自适应控制. 北京: 清华大学出版社, 2016.Chai Tian-You, Yue Heng. Adaptive Control. Beijing: Tsinghua University Press, 2016. [8] Ceperic E, Ceperic V, Baric A. A strategy for short-term load forecasting by support vector regression machines. IEEE Transactions on Power Systems, 2013, 28(4): 4356−4364 doi: 10.1109/TPWRS.2013.2269803 [9] Milanese M, Vicino A. Optimal estimation theory for dynamic systems with set membership uncertainty: An overview. Automatica, 1991, 27(7): 997−1009 [10] Zadeh L A. Fuzzy logic, neural networks, and soft computing. Microprocessing and Microprogramming, 1993, 37(3): 77−84 [11] 徐凤霞, 朱全民, 赵东亚, 李少远. 基于U模型的非线性控制系统设计方法十年发展综述. 控制与决策, 2013, 28(7): 961−971 doi: 10.13195/j.cd.2013.07.4.xufx.023Xu Feng-Xia, Zhu Quan-Ming, Zhao Dong-Ya, Li Shao-Yuan. U-model based design methods for nonlinear control systems a survey of the development in the 1st decade. Control and Decision, 2013, 28(7): 961−971 doi: 10.13195/j.cd.2013.07.4.xufx.023 [12] Chen S, Billings S A. Neural networks for nonlinear dynamic system modelling and identification. International Journal of Control, 1991, 56(2): 319−346 [13] Panjapornpon C, Saksomboon P, Dechakupt T. Real-time application of pH control in a carbon dioxide bubble column reactor by input/output linearizing control coupled with pH target optimizer. Industrial and Engineering Chemistry Research, 2018, 58(2): 771−781 [14] Gao Y, Er M J. NARMAX time series model prediction: Feedforward and recurrent fuzzy neural network approaches. Fuzzy Sets and Systems, 2005, 150(2): 331−350 doi: 10.1016/j.fss.2004.09.015 [15] Chen L, Narendra K S. Nonlinear adaptive control using neural networks and multiple models. Automatica, 2001, 37(8): 1245−1255 doi: 10.1016/S0005-1098(01)00072-3 [16] 王兰豪, 贾瑶, 柴天佑. 再磨过程的泵池液位和给矿压双速率区间控制. 自动化学报, 2017, 43(6): 993−1006Wang Lan-Hao, Jia Yao, Chai Tian-You. Double rate interval control of pump pool liquid level and feed pressure in regrinding process. Acta Automatica Sinica, 2017, 43(6): 993−1006 [17] Zhang Y J, Chai T Y, Wang D H. An alternating identification algorithm for a class of nonlinear dynamical systems. IEEE Transactions on Neural Networks and Learning Systems, 2017, 28(7): 1−12 doi: 10.1109/TNNLS.2017.2705379 [18] Zhang Y J, Chai T Y, Wang H, Wang D H. Nonlinear decoupling control with ANFIS-based unmodeled dynamics compensation for a class of complex industrial processes. IEEE Transactions on Neural Networks and Learning Systems, 2018, 29(6): 2352−2366 doi: 10.1109/TNNLS.2017.2691905 [19] Wang H, Wang A P, Brown M, Harris C J. One-to-one mapping and its application to neural networks based control systems design. International Journal of Systems Science, 1996, 27(2): 161−170 doi: 10.1080/00207729608929200 [20] Ljung L. Convergence analysis of parametric identification methods. IEEE Transactions on Automatic Control, 1978, 23(5): 770−783 doi: 10.1109/TAC.1978.1101840 [21] Wang D, Li M. Stochastic configuration networks: Fundamentals and algorithms. IEEE Transactions on Cybernetics, 2017, 47(10): 3466−3479 doi: 10.1109/TCYB.2017.2734043 [22] Pao Y H, Park G H, Sobajic D J. Learning and generalization characteristics of the random vector functional-link net. Neurocomputing, 1994, 6(2): 163−180 doi: 10.1016/0925-2312(94)90053-1 [23] Dai W, Zhou X Y, Li D P, Zhu S, Wang X S. Hybrid parallel stochastic configuration networks for industrial data analytics. IEEE Transactions on Industrial Informatics, 2021, 18(4): 2331−2341 [24] Dai W, Zhang L Z, Fu J, Chai T Y, Ma X P. Model-data-based switching adaptive control for dense medium separation in coal beneficiation. Control Engineering Practice, 2020, 98: 1−12 [25] Zhang L J, Xia X H, Zhang J F. Medium density control for coal washing dense medium cyclone circuits. IEEE Transactions on Control Systems Technology, 2014, 23(3): 1117−1122 -
图(9) / 表(2)
计量
- 文章访问数: 720
- HTML全文浏览量: 188
- PDF下载量: 112
- 被引次数: 0
