Loading [MathJax]/jax/element/mml/optable/GeneralPunctuation.js
A journal of IEEE and CAA , publishes high-quality papers in English on original theoretical/experimental research and development in all areas of automation
Volume 11 Issue 6
Jun.  2024

IEEE/CAA Journal of Automatica Sinica

  • JCR Impact Factor: 15.3, Top 1 (SCI Q1)
    CiteScore: 23.5, Top 2% (Q1)
    Google Scholar h5-index: 77, TOP 5
Turn off MathJax
Article Contents
M. Yang, “New second-level-discrete zeroing neural network for solving dynamic linear system,” IEEE/CAA J. Autom. Sinica, vol. 11, no. 6, pp. 1521–1523, Jun. 2024. doi: 10.1109/JAS.2023.123384
Citation: M. Yang, “New second-level-discrete zeroing neural network for solving dynamic linear system,” IEEE/CAA J. Autom. Sinica, vol. 11, no. 6, pp. 1521–1523, Jun. 2024. doi: 10.1109/JAS.2023.123384

New Second-Level-Discrete Zeroing Neural Network for Solving Dynamic Linear System

doi: 10.1109/JAS.2023.123384
  • Dear Editor,

    This letter deals with a new second-level-discretization method with higher precision than the traditional first-level-discretization method. Specifically, the traditional discretization method utilizes the first-order time derivative information, and it is termed first-level-discretization method. By contrast, the new discretization method not only utilizes the first-order time derivative information, but also makes use of the second-order derivative information. By combining the new second-level-discretization method with zeroing neural network (ZNN), the second-level-discrete ZNN (SLDZNN) model is proposed to solve dynamic (i.e., time-variant or time-dependent) linear system. Numerical experiments and application to angle-of-arrival (AoA) localization show the effectiveness and superiority of the SLDZNN model.

    Because of the real-time requirement, solving dynamic problems becomes more and more important [1]–[3], which has been applied in many fields such as robot control [4]–[6] and AoA localization [7]. ZNN [8]–[10] is a special type of recurrent neural network (RNN) [11]. In recent years, ZNN have shown its effectiveness for dynamic problems solving, such as for dynamic nonlinear inequalities solving, dynamic linear system solving, and dynamic quadratic minimization. For exmaple, a ZNN (CZNN) model with definable convergence time was developed for solving time-variant linear matrix equations in [2]. By performing the following straightforward steps: 1) Defining the error function; 2) Applying the design formula; and 3) Simplifying the result, one could eventually develop the continuous CZNN model for dynamic problems solving.

    Furthermore, for convenient program implementation, the traditional first-level-discretization method [10], [12] is usually applied to CZNN discretization, and the first-level-discrete ZNN (FLDZNN) model could be developed. For instance, in time point tk=kς with ς being the sampling period, the FLDZNN model computes the dynamic solution s(tk+1) on the basis of s(tk) and ˙s(tk). That is, the solution in time point tk+1 is computed by utilizing the information in time point tk, and the precision is O(ς2).

    Generally, to improve the precision, the traditional way is to utilize more information at more time points. In recent years, many high-precision discrete methods utilizing information at several time points have been developed, and many high-precision discrete ZNN (DZNN) models have been obtained. For instance, a four-points first-level-discretization method is designed in [10]. A six-points first-level-discretization method is developed in [3]. However, for rapidly dynamic systems, the information before several time points may be valueless. Therefore, utilizing more information at a few time points to realize higher precision is meaningful. Note that the second-order derivative containing the rapidly changing information, which is very valuable. Upon careful study, we find there is another way to improve the precision. The traditional discrete method utilizes the first-order time derivative. If we can utilize the second-order time derivative, the model precision may be further improved. As a result, this letter proposes a new second-level-discretization method and further develops the SLDZNN model. In time point tk, the SLDZNN model computes the dynamic solution s(tk+1) on the basis of s(tk), ˙s(tk), and ¨s(tk). As a result, the precision is improved as O(ς3).

    The main contributions of this letter are presented as below.

    1) To our knowledge, the second-level-discretization method is proposed for CZNN discretization for the first time, and it is more effective than the traditional first-level-discretization method.

    2) By utilizing the ZNN design formula twice and combining the second-level-discretization method, the SLDZNN model is proposed to solve dynamic linear system, and detailed theoretical analyses are presented.

    3) Numerical experiments considering a specific example and application to AoA localization are displayed to illustrate the efficiency and superiority of the second-level-discretization method and SLDZNN model.

    Problem statement: Consider dynamic linear system during time duration [0,T]

    Y(t)x(t)=z(t) (1)

    in which coefficient matrix Y(t)Rn×n and coefficient vector z(t)Rn. The vector x(t)Rn is the dynamic solution to be attained. Besides, the matrix Y(t) is always nonsingular to guarantee that (1) is solvable.

    CZNN model: To solve dynamic linear system (1), the CZNN model [9] is developed in the following theorem.

    Theorem 1: Let γ>0 be a ZNN design parameter. For nonsingular matrix A(t), the CZNN model for solving dynamic linear system (1) is developed as

    ˙x(t)=Y1(t)(˙z(t)˙Y(t)x(t)γ(Y(t)x(t)z(t))) (2)

    and the synthesized x(t) converges to the theoretical continuous solution x(t) as t>>0.

    Proof: To make (1) Hold true, perform straightforward steps: 1) defining e(t)=Y(t)x(t)z(t); 2) Applying ˙e(t)=γe(t); and 3) Simplifying result. We get

    Y(t)˙x(t)+˙Y(t)x(t)˙z(t)=γ(Y(t)x(t)z(t)). (3)

    By rearranging the above result and doing an inverse operation, (2) is thus obtained. Evidently, CZNN model (2) is originated from ZNN design formula ˙e(t)=γe(t). Choose Lyapunov functional candidate as L(t)=1/2eT(t)e(t). Computing the derivative of L(t) yields ˙L(t)=eT(t)˙e(t)=γeT(t)e(t). Evidently, we have L(t)0 and ˙L(t)0. If ˙L(t)=0, e(t)=0. Therefore, when t>>0, we have e(t)0. That is, x(t) synthesized by (2) converges to x(t) as t>>0.

    Traditional FLDZNN model: To generate discrete models, the traditional first-level-discretization method [12] and the corresponding FLDZNN model are developed.

    Lemma 1: Let ς>0 be the sampling period. The traditional first-level-discretization method [12] is presented as

    s(tk+1)=s(tk)+ς˙s(tk)+O(ς2). (4)

    For simplicity, we use xk to denote x(tk), and other symbols have similar meanings. In addition, we use O(ς2) to denote the vector form of O(ς2). By using (4) for model discretization, the following theorem develops the FLDZNN model.

    Theorem 2: Let ς>0 be the sampling period and denote the computational assignment operator. The FLDZNN model is developed as

    ˙xk=Y1k(˙zk˙Ykxkγ(Ykxkzk))
    xk+1xk+ς˙xk

    and the synthesized xk+1 converges to the theoretical discrete solution xk+1 with O(ς2) precision.

    Proof: Evidently, (5a) is attained by sampling (2) in time point tk. Meanwhile, (5b) is the vector form of (4). On the basis of Theorem 1 and Lemma 1, xk+1 synthesized by FLDZNN model (5) converges to xk+1 with O(ς2) precision. This completes the proof.

    New SLDZNN model: On the basis of the previous analysis, the deeper-level information in time point tk may be utilized to get the higher precision. The new second-level-discretization method is thus proposed in the following theorem.

    Taylor expansion [12] is presented as

    s(tk+1)=s(tk)+ς˙s(tk)+12ς2¨s(tk)+16ς3s(tk)+O(ς4).

    With O(ς3) absorbing the remainder terms, the second-level-discretization method (6) is obtained as follows.

    Lemma 2: Let ς>0 be the sampling period. The new second-level-discretization method is developed as

    s(tk+1)=s(tk)+ς˙s(tk)+12ς2¨s(tk)+O(ς3). (6)

    By utilizing ZNN method twice and applying the second-level-discretization method (6), the SLDZNN model is proposed in the following theorem.

    Theorem 3: Let ς>0 be the sampling period and denote the computational assignment operator. The SLDZNN model is developed as

    ˙xk=Y1k(˙zk˙Ykxkγ(Ykxkzk))
    ¨xk=Y1k(¨zk2˙Yk˙xk¨Ykxk+γ2(Ykxkzk))
    xk+1xk+ς˙xk+12ς2¨xk

    and the synthesized xk+1 converges to the theoretical discrete solution xk+1 with O(ς3) precision.

    Proof: According to the previous analyse, (3) is obtained by utilizing ZNN method [9]. Then, (7a) can be directly developed by sampling (3) in time point tk. To get the second-order derivative, the another error function is defined according to (3)

    ϵ(t)=Y(t)˙x(t)+˙Y(t)x(t)˙z(t)+γ(Y(t)x(t)z(t)).

    Utilizing ZNN design formula ˙ϵ(t)=γϵ(t) yields

    Y(t)¨x(t)+˙Y(t)˙x(t)+˙Y(t)˙x(t)+¨Y(t)x(t)¨z(t)+γ(˙Y(t)x(t)+Y(t)˙x(t)˙z(t))=γ(˙Y(t)x(t)+Y(t)˙x(t)˙z(t)+γ(Y(t)x(t)z(t))).

    Simplifying the above result in time point tk yields (7b). Furthermore, by applying new second-level-discretization method (6) to discretize CZNN model (2), (7c) is directly obtained. On the basis of ZNN knowledge [9] and Lemma 2, xk+1 synthesized by SLDZNN model (7) converges to xk+1 with O(ς3) precision.

    Experiments verification: This section first provides a numerical example to illustrate the effectiveness of the proposed SLDZNN model (7). Then, an application to AoA localization is also conducted for illustrating the practicability of SLDZNN model (7).

    Example 1: Consider dynamic linear system (1) during time duration [0,20] s with Y(t)=

    [cos(0.4t)+4sin(0.4t)sin(0.4t)sin(0.4t)cos(0.2t)cos(0.2t)+4sin(0.2t)sin(0.2t)sin(0.4t)cos(0.4t)cos(0.4t)+4sin(0.4t)sin(0.2t)cos(0.2t)cos(0.2t)sin(0.2t)+4]

    and z(t)=[sin(0.8t)cos(0.8t)sin(0.8t)cos(0.8t)]T.

    On the basis of the previous experience [1], [3], [10], h=γς is usually set as a constant, e.g., h=0.5 in this letter. For comparison, FLDZNN model (5) and SLDZNN model (7) are both applied, and the corresponding numerical results are presented in Fig. 1. By setting the sampling period ς=0.1 s, the trajectories of the dynamic solution synthesized by SLDZNN model (7) are shown in Fig. 1(a). Meanwhile, Fig. 1(b) illustrates that the residual error synthesized by SLDZNN model (7) is between 103 and 104. By contrast, Fig. 1(b) also shows that the residual error synthesized by FLDZNN model (5) is just about 102. Meanwhile, when setting sampling period ς= 0.05 s and 0.01 s, the residual errors attained from two DZNN models are displayed in Figs. 1(c) and 1(d), respectively. For instance, with ς=0.01 s, the residual error generated by SLDZNN model (7) is between 106 and 107. That is, the precision will be increased by 103 times if the sampling period decreases by 10 times. In summary, the precision of SLDZNN model (7) is indeed O(ς3) while the precision of FLDZNN model (5) is O(ς2), which illustrates the superiority of SLDZNN model (7).

    Figure  1.  Synthesized numerical results from FLDZNN model (5) and SLDZNN model (7) for dynamic linear system solving in Example 1. (a) Solution trajectories generated by SLDZNN model (7) with ς=0.1 s; (b) Residual errors with ς=0.1 s; (c) Residual errors with ς=0.05 s; (d) Residual errors with ς=0.01 s.

    Example 2: The AoA localization system consists of a moving object and two moving stations (simply termed MS1 and MS2). Note that (u1(t),v1(t)) and (u2(t),v2(t)) represent the positions of MS1 and MS2, respectively, which are both known. AoA system uses array antenna to detect direction of arrival in real time. Specifically, θ1(t) and θ2(t) mean the arrival angles of the object to two stations, respectively. The target of AoA localization is to compute (estimate) the real-time position on the basis of two arrival angles. According to the previous work [7], the problem of AoA localization system can be formulated as dynamic linear system (1) with

    Y(t)=[tan(θ1(t))1tan(θ2(t))1],x(t)=[u3(t)v3(t)]
    z(t)=[u1(t)tan(θ1(t))v1(t)u2(t)tan(θ2(t))v2(t)]

    where (u3(t),v3(t)) represents the position of the moving object. Consider the AoA localization during time duration [0,30] s.

    By setting h=0.5 and ς=0.1 s, Fig. 2(a) through Fig. 2(d) display some important results generated by SLDZNN model (7). The AoA localization is successfully completed. Specifically, Fig. 2(a) presents the profiles of two arrival angles. Fig. 2(b) illustrates the real-time positions of two moving stations and the predicted localization of the moving object. We can conclude that the predicted trajectory of the moving object basically coincides with the actual trajectory. For details, Fig. 2(c) shows the predicted position of the moving object, which converges to the actual position quickly. To further analyze the localization precision, the residual error

    Figure  2.  Synthesized numerical results from FLDZNN model (5) and SLDZNN model (7) for AoA localization in Example 2. (a) Two arrival angles; (b) Predicted localization of moving object generated by SLDZNN model (7) with ς=0.1 s; (c) Predicted trajectory of moving object computed by SLDZNN model (7) when ς=0.1 s; (d) Residual errors when ς=0.1 s; (e) Residual errors when ς=0.05 s; (f) Residual errors when ς=0.01 s.
    er=

    is used as the index. With \varsigma = 0.1 s, 0.05 s, and 0.01 s, Figs. 2(d)−2(f) respectively display the corresponding residual errors generated by two discrete models. SLDZNN model (7) indeed has {\bf{O}}(\varsigma^3) precision for dynamic linear system solving, which is one-order higher than FLDZNN model (5).

    Conclusion: This letter has investigated the problem of dynamic linear system (1). Considering that the traditional discretization methods only utilize the first-order time derivative, new second-level-discretization method (6) with higher precision has been developed by utilizing the second-order derivative. Aided with second-level-discretization method (6), SLDZNN model (7) has been further proposed. Theoretical analyses and simulations have been conducted to validate the efficacy and superiority of second-level-discretization method (6) and SLDZNN model (7). Developing more second-level-discretization methods with higher precision is a future research direction. Moreover, Combining second-level-discretization methods with zeroing neural network for solving more kinds of dynamic problems is also a significant future research direction.

    Acknowledgments: This work was supported in part by the National Natural Science Foundation of China (62303174), the Fundamental Research Funds for the Central Universities (531118010815), and the Changsha Municipal Natural Science Foundation (kq2208043).

  • [1]
    L. Xiao and Y. Zhang, “Solving time-varying nonlinear inequalities using continuous and discrete-time Zhang dynamics,” Int. J. Comput. Math., vol. 90, no. 5, pp. 1114–1127, Jun. 2013. doi: 10.1080/00207160.2012.750305
    [2]
    W. Li, “A recurrent neural network with explicitly definable convergence time for solving time-variant linear matrix equations,” IEEE Trans. Ind. Inform., vol. 14, no. 12, pp. 5289–5298, Dec. 2018. doi: 10.1109/TII.2018.2817203
    [3]
    D. Guo, F. Xu, Z. Li, Z. Nie, and H. Shao, “Design, verification and application of new discrete-time recurrent neural network for dynamic nonlinear equations solving,” IEEE Trans. Ind. Inform., vol. 14, no. 9, pp. 3936–3945, Sept. 2018. doi: 10.1109/TII.2017.2787729
    [4]
    Z. Li, S. Li, and X. Luo, “An overview of calibration technology of industrial robots,” IEEE/CAA J. Autom. Sinica, vol. 8, no. 1, pp. 23–36, Jan. 2021. doi: 10.1109/JAS.2020.1003381
    [5]
    L. Huang, M. Zhou, K. Hao, and E. Hou, “A survey of multi-robot regular and adversarial patrolling,” IEEE/CAA J. Autom. Sinica, vol. 6, no. 4, pp. 894–903, Jul. 2019. doi: 10.1109/JAS.2019.1911537
    [6]
    N. Tan, P. Yu, Z. Zhong, and F. Ni, “A new noise-tolerant dual-neural-network scheme for robust kinematic control of robotic arms with unknown models,” IEEE/CAA J. Autom. Sinica, vol. 9, no. 10, pp. 1778–1791, Oct. 2022. doi: 10.1109/JAS.2022.105869
    [7]
    M. Yang, Y. Zhang, N. Tan, and H. Hu, “Explicit linear left-and-right 5-step formulas with zeroing neural network for time-varying applications,” IEEE Trans. Cybern., vol. 53, no. 2, pp. 1133–1143, 2023. doi: 10.1109/TCYB.2021.3104138
    [8]
    L. Jin, S. Li, B. Liao, and Z. Zhang, “Zeroing neural networks: A survey,” Neurocomputing, vol. 267, pp. 597–604, Dec. 2017. doi: 10.1016/j.neucom.2017.06.030
    [9]
    Y. Zhang and C. Yi, Zhang Neural Networks and Neural-Dynamic Method. Commack, USA: Nova Science Publishers, Inc., 2011.
    [10]
    Y. Zhang, L. Jin, D. Guo, Y. Yin, and Y. Chou, “Taylor-type 1-step-ahead numerical differentiation rule for first-order derivative approximation and ZNN discretization,” J. Comput. Appl. Math., vol. 273, pp. 29–40, Jan. 2015. doi: 10.1016/j.cam.2014.05.027
    [11]
    L. Xiao, S. Li, J. Yang, and Z. Zhang, “A new recurrent neural network with noise-tolerance and finite-time convergence for dynamic quadratic minimization,” Neurocomputing, vol. 285, pp. 125–132, Apr. 2018. doi: 10.1016/j.neucom.2018.01.033
    [12]
    J. H. Mathews and K. D. Fink, Numerical Methods Using MATLAB, 4th ed. Englewood Cliffs, USA: Prentice Hall, 2004.
  • Related Articles

    [1]Jinyi Xu, Zuowei Zhang, Ze Lin, Yixiang Chen, Weiping Ding. Multi-View Dynamic Kernelized Evidential Clustering[J]. IEEE/CAA Journal of Automatica Sinica, 2024, 11(12): 2435-2450. doi: 10.1109/JAS.2024.124608
    [2]Shengao Lu, Tong Wu, Lixian Zhang, Jianan Yang, Ye Liang. Interpolated Bumpless Transfer Control for Asynchronously Switched Linear Systems[J]. IEEE/CAA Journal of Automatica Sinica, 2024, 11(7): 1579-1590. doi: 10.1109/JAS.2023.124155
    [3]Sheng Cao, Zhiwei Luo, Changqin Quan. Sequential Inverse Optimal Control of Discrete-Time Systems[J]. IEEE/CAA Journal of Automatica Sinica, 2024, 11(3): 608-621. doi: 10.1109/JAS.2023.123762
    [4]Yisha Li, Ya Zhang, Xinde Li, Changyin Sun. Regional Multi-Agent Cooperative Reinforcement Learning for City-Level Traffic Grid Signal Control[J]. IEEE/CAA Journal of Automatica Sinica, 2024, 11(9): 1987-1998. doi: 10.1109/JAS.2024.124365
    [5]Xinli Shi, Xiangping Xu, Guanghui Wen, Jinde Cao. Fixed-Time Gradient Flows for Solving Constrained Optimization: A Unified Approach[J]. IEEE/CAA Journal of Automatica Sinica, 2024, 11(8): 1849-1864. doi: 10.1109/JAS.2023.124089
    [6]Chi Ma, Dianbiao Dong. Finite-time Prescribed Performance Time-Varying Formation Control for Second-Order Multi-Agent Systems With Non-Strict Feedback Based on a Neural Network Observer[J]. IEEE/CAA Journal of Automatica Sinica, 2024, 11(4): 1039-1050. doi: 10.1109/JAS.2023.123615
    [7]Dingxin He, HaoPing Wang, Yang Tian, Yida Guo. A Fractional-Order Ultra-Local Model-Based Adaptive Neural Network Sliding Mode Control of n-DOF Upper-Limb Exoskeleton With Input Deadzone[J]. IEEE/CAA Journal of Automatica Sinica, 2024, 11(3): 760-781. doi: 10.1109/JAS.2023.123882
    [8]Wei Ren, Zhuo-Rui Pan, Weiguo Xia, Xi-Ming Sun. Hierarchical Controller Synthesis Under Linear Temporal Logic Specifications Using Dynamic Quantization[J]. IEEE/CAA Journal of Automatica Sinica, 2024, 11(10): 2082-2098. doi: 10.1109/JAS.2024.124473
    [9]Kailong Liu, Qiao Peng, Yuhang Liu, Naxin Cui, Chenghui Zhang. Explainable Neural Network for Sensitivity Analysis of Lithium-ion Battery Smart Production[J]. IEEE/CAA Journal of Automatica Sinica, 2024, 11(9): 1944-1953. doi: 10.1109/JAS.2024.124539
    [10]Hui Liu, Ciyun Lin, Bowen Gong, Dayong Wu. Automatic Lane-Level Intersection Map Generation using Low-Channel Roadside LiDAR[J]. IEEE/CAA Journal of Automatica Sinica, 2023, 10(5): 1209-1222. doi: 10.1109/JAS.2023.123183
    [11]Zhijia Zhao, Jian Zhang, Shouyan Chen, Wei He, Keum-Shik Hong. Neural-Network-Based Adaptive Finite-Time Control for a Two-Degree-of-Freedom Helicopter System With an Event-Triggering Mechanism[J]. IEEE/CAA Journal of Automatica Sinica, 2023, 10(8): 1754-1765. doi: 10.1109/JAS.2023.123453
    [12]Xueli Wang, Derui Ding, Hongli Dong, Xian-Ming Zhang. Neural-Network-Based Control for Discrete-Time Nonlinear Systems with Input Saturation Under Stochastic Communication Protocol[J]. IEEE/CAA Journal of Automatica Sinica, 2021, 8(4): 766-778. doi: 10.1109/JAS.2021.1003922
    [13]Yanhong Luo, Shengnan Zhao, Dongsheng Yang, Huaguang Zhang. A New Robust Adaptive Neural Network Backstepping Control for Single Machine Infinite Power System With TCSC[J]. IEEE/CAA Journal of Automatica Sinica, 2020, 7(1): 48-56. doi: 10.1109/JAS.2019.1911798
    [14]Fei-Yue Wang, Jie Zhang, Qinglai Wei, Xinhu Zheng, Li Li. PDP: Parallel Dynamic Programming[J]. IEEE/CAA Journal of Automatica Sinica, 2017, 4(1): 1-5.
    [15]Xiaowei Feng, Xiangyu Kong, Hongguang Ma. Coupled Cross-correlation Neural Network Algorithm for Principal Singular Triplet Extraction of a Cross-covariance Matrix[J]. IEEE/CAA Journal of Automatica Sinica, 2016, 3(2): 147-156.
    [16]Jingbei Yang, Shuang Cong, Feng Shuang, Herschel Rabitz. Manipulations Between Eigenstates of 2-Level Quantum System Based on Optimal Measurements[J]. IEEE/CAA Journal of Automatica Sinica, 2016, 3(1): 35-41.
    [17]Yan Song, Haifeng Lou, Shuai Liu. Distributed Model Predictive Control with Actuator Saturation for Markovian Jump Linear System[J]. IEEE/CAA Journal of Automatica Sinica, 2015, 2(4): 374-381.
    [18]Huiyang Liu, Long Cheng, Min Tan, Zengguang Hou. Containment Control of General Linear Multi-agent Systems with Multiple Dynamic Leaders: a Fast Sliding Mode Based Approach[J]. IEEE/CAA Journal of Automatica Sinica, 2014, 1(2): 134-140.
    [19]Hongjing Liang, Huaguang Zhang, Zhanshan Wang, Junyi Wang. Consensus Robust Output Regulation of Discrete-time Linear Multi-agent Systems[J]. IEEE/CAA Journal of Automatica Sinica, 2014, 1(2): 204-209.
    [20]Qiming Zhao, Hao Xu, Sarangapani Jagannathan. Near Optimal Output Feedback Control of Nonlinear Discrete-time Systems Based on Reinforcement Neural Network Learning[J]. IEEE/CAA Journal of Automatica Sinica, 2014, 1(4): 372-384.

Catalog

    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(2)

    Article Metrics

    Article views (142) PDF downloads(17) Cited by()

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return