Abstract
In actual industrial systems, numerous processes are characterized as distributed parameter systems (DPSs) that exhibit strong nonlinearity, complex boundary conditions, and unknown dynamics. Achieving accurate control of such processes poses a significant challenge. In light of this, a data-driven spatiotemporal least squares support vector machine (LS-SVM) inverse control method has been developed specifically for complex DPSs. First, a spatiotemporal LS-SVM model is proposed to capture the dynamics of DPSs by leveraging available data. Subsequently, by employing Taylor expansion on this spatiotemporal model, a state model is derived to explicitly establish the relationship between the control input and output variables. Building upon this foundation, an explicit control input is obtained through inversion and spatial fuzzy strategy, allowing for effective tracking of spatiotemporal dynamics. This control approach takes into account the influence of each input on all spatial points, thereby ensuring a favorable control effect for DPSs. Theoretical analysis and stability proofs affirm the stability of the proposed control approach for nonlinear DPSs. Furthermore, the efficacy of this controller is demonstrated through two actual experiments. First, the proposed approach exhibits superior control performance, as evidenced by a nearly threefold improvement in tracking accuracy on several sensors compared to the fuzzy controller. Second, the proposed controller maintains a smaller error during the tracking process, with deviations bounded within 1 °C even in the presence of disturbances.
Similar content being viewed by others
Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
Abbreviations
- DPS:
-
Distributed parameter system
- LS-SVM:
-
Least square support vector machine
- PDE:
-
Partial differential equations
- MIMO:
-
Multi-input multi-output
- RKHS:
-
Reproducing kernel Hilbert space
- RBF:
-
Radial basis function
- RE:
-
Relative error
- ODE:
-
Ordinary differential equation
- LPS:
-
Lumped parameter system
- SBFs:
-
Spatial basis functions
- MPC:
-
Model predictive control
- UUB:
-
Uniformly ultimately bounded
- MAPE:
-
Mean absolute percentage error
- \(x_{i}\) :
-
The ith spatial point \((i = 1,2, \cdots ,N)\)
- \(\varphi (x)\) :
-
A nonlinear mapping function to reflect the spatial distribution from low dimensions to high dimensions
- \(N\) :
-
Spatial sensor number
- \(y(x, \, t)\) :
-
System output at position x
- \(\gamma\) :
-
Regularization factor
- \({{\varvec{\upalpha}}}(t_{k} )\) :
-
Lagrange multiplier coefficient
- \({\mathbf{z}}(t_{k} )\) :
-
The time-series variable, which is a vector related with Lagrange multiplier coefficients and control input.
- \(\eta_{\tau }\)(\(\tau = 1,2, \cdots ,L\)):
-
The Lagrange multiplier for \({{\varvec{\upalpha}}}(t_{\tau } )\)
- \(k_{\alpha } (z(t_{i} ),z(t_{j} ))\) :
-
Kernel function for \({{\varvec{\upalpha}}}(t_{k} )\)
- \(g({\mathbf{u}}(t_{k} ))\) :
-
A function with respect to \({\mathbf{u}}(t_{k} )\) that is represented by the LS-SVM model
- \(r(x,t_{k} )\) :
-
The target temperature for controlling
- \(G(x_{p} )\) :
-
Membership function for pth rule of the fuzzy model
- \(e_{r} (x, \, t_{k} )\) :
-
The tracking error at \(t_{k}\)
- \(\Lambda\) :
-
Diagonal matrix that denotes PDEs \((\partial y/\partial x,\partial^{2} y/\partial x^{2} , \cdots )\) of the system output for spatial location point x
- \(m\) :
-
The number of system input
- \(K(x,x_{i} )\) :
-
Kernel functions with respect to spatial point \(x_{i}\) and satisfying \(K(x,x_{i} ) = \varphi (x)\varphi (x_{i} )\)
- \(w( \cdot )\) :
-
Weight coefficient in the LS-SVM
- \(b(t_{k} )\) :
-
Bias term for the LS-SVM model
- \(\hat{y}(x, \, t)\) :
-
Estimation of system output at position x
- \(u_{j} (t)\) :
-
The jth control input \((j = 1,2, \cdots ,m)\)
- \({{\varvec{\Omega}}} = \left[ {K(x, \, x_{1} ), \cdots ,K(x, \, x_{N} )} \right]\) :
-
A vector of kernel functions
- \({\mathbf{E}}(x)\) :
-
A vector with respect to spatial point x
- \(t_{k}\) :
-
The kth sampling time
- \(\theta_{{\text{s}}}\) :
-
The bias term for the modeling of \({{\varvec{\upalpha}}}(t_{k} )\)
- \(R_{k} [\Delta {\mathbf{u}}(t_{k} )]\) :
-
The higher-order infinitesimal term with
- \(\Delta {\mathbf{u}}(t_{k} )\;\rho_{0}\) :
-
A positive number
- \(\overline{\upsilon }\) :
-
A small positive number
- \(r(x, \, t_{k} )\) :
-
The reference signal
- \(R_{{k{ + }1}}\) :
-
Abbreviation of the high-order term \(R_{k} \left[ {\Delta u\left( {t_{k} } \right)} \right]\)
- \(h_{p}\) :
-
The weight coefficient of the control input at \(x_{p} (p = 1,2, \cdots ,N)\) with values between (0, 1)
References
Zheng, D., Hoo, K.A.: Low-order model identification for implementable control solutions of distributed parameter systems. Comput. Chem. Eng. 26(7–8), 1049–1076 (2002)
Wang, M., Li, H.X., Xin, C., Yun, C.: Deep learning-based model reduction for distributed parameter systems. IEEE Trans. Syst. Man Cybernet. Syst. 46(99), 1664–1674 (2016)
Wang, Z., Li, H.X.: Incremental spatiotemporal learning for online modeling of distributed parameter systems. IEEE Trans. Syst. Man Cybernet. Syst. 99, 1–11 (2018)
Ray, W.H.: Advanced Process Control. McGraw-Hill Companies, New York (1981)
Li, H.X., Qi, C.K.: Modeling of distributed parameter systems for applications - a synthesized review from time-space separation. J. Process Control 20(8), 891–901 (2010)
Xu, B., Lu, X.: Spatiotemporal kernel-local-embedding modeling approach for nonlinear distributed parameter systems. J. Process Control 119, 101–114 (2022)
Khani, F., Haeri, M.: Robust model predictive control of nonlinear processes represented by wiener or Hammerstein models. Chem. Eng. Sci. 129, 223–231 (2015)
Chen, L.Q., Li, H.X.: Tensor decomposition based spatiotemporal modeling for distributed thermal processes. In: 2019 IEEE International Conference on Systems, Man and Cybernetics (SMC). IEEE.
Xu, K.K., Li, H.X., Yang, H.D.: Dual least squares support vector machines based spatiotemporal modeling for nonlinear distributed thermal processes. J. Process Control 54, 81–89 (2017)
Lu, X.J., Zou, W., Huang, M.: A novel spatiotemporal LS-SVM Method for complex distributed parameter systems with applications to curing thermal process. IEEE Trans. Industr. Inf. 12(3), 1156–1165 (2016)
Lu, X.J., Xu, B.W., He, P.Z.: A novel low-order spatiotemporal modeling method for nonlinear distributed parameter systems. J. Process Control 106, 84–93 (2021)
Xu, Y., Yang, H., Jiang, B., Vincent, C.: Backstepping based sliding mode fault-tolerant control for linear interconnected parabolic distributed parameter systems. IET Control Theory Appl. 14, 1928–1936 (2020)
Zhang, Y., Li, Y., Chen, M.: Iterative learning control for linear generalized distributed parameter system. Neural Comput. Appl. 31(9), 1–9 (2019)
Maidi, A., Diaf, M., Coffiou, J.P.: Optimal linear PI fuzzy controller design of a heat exchanger. Chem. Eng. Process. 47(5), 938–945 (2008)
Wu, H.N., Li, H.X.: Adaptive neural control design for nonlinear distributed parameter systems with persistent bounded disturbances. IEEE Trans. Neural Netw. 20(10), 1630–1644 (2009)
Wang, M.L., Paulson, J.A., Yan, H.: An adaptive model predictive control strategy for nonlinear distributed parameter systems using the Type-2 Takagi-Sugeno model. Int. J. Fuzzy Syst. 18(5), 1–14 (2016)
Wang, J.W., Li, H.X., Wu, H.N.: A membership-function-dependent approach to design fuzzy pointwise state feedback controller for nonlinear parabolic distributed parameter systems with spatially discrete actuators. IEEE Trans. Syst. Man Cybernet. Syst. 47(7), 1486–1499 (2017)
Xu, B.W., Lu, X.J.: A data-driven spatiotemporal model predictive control strategy for nonlinear distributed parameter systems. Nonlinear Dyn. 108(2), 1269–1281 (2022)
Aggelogiannaki, E., Sarimveis, H.: Robust nonlinear H∞ control of hyperbolic distributed parameter systems. Control. Eng. Pract. 17(6), 723–732 (2009)
Ai, L., Ye, S.: Model predictive control for nonlinear distributed parameter systems based on LS-SVM. Asian J. Control 15(5), 1407–1416 (2013)
Zhang, X.X., Jiang, Y., Li, H.X.: SVR learning-based spatiotemporal fuzzy logic controller for nonlinear spatially distributed dynamic systems. IEEE Trans. Neural Netw. Learn. Syst. 24, 1635–1647 (2013)
Deng, H., Li, H.X.: A novel neural approximate inverse control for unknown nonlinear discrete dynamical systems. IEEE Trans. Cybernet. 35(1), 115–123 (2005)
Li, H.X., Deng, H.: An approximate internal model-based neural control for unknown nonlinear discrete processes. IEEE Trans. Neural Netw. 17(3), 659–670 (2006)
Suykens, J.A.K., Vandewalle, J.: Least squares support vector machine classifiers. Neural Process. Lett. 9(3), 293–300 (1999)
Xu, B., Lu, X.: A spatiotemporal fuzzy control approach for complex distributed parameter system with application to thermal processes. J. Process Control 127, 102993 (2023)
Funding
This work was partially supported by National Natural Science Foundation of China (52075556), and the Key R&D Program of Hunan Province (2021SK2016).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
Xinjiang Lu is the corresponding author, and all the authors declare that they have no conflict of interests that could have appeared to influence the work reported in connection with the submitted work.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Xu, B., Lu, X. Spatiotemporal LS-SVM inverse control for nonlinear distributed parameter systems with application to heating process. Nonlinear Dyn 111, 17229–17246 (2023). https://doi.org/10.1007/s11071-023-08771-6
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11071-023-08771-6