Skip to main content
Log in

Distributed-observer-based fault diagnosis and fault-tolerant control for time-varying discrete interconnected systems

  • Original Research
  • Published:
Journal of Ambient Intelligence and Humanized Computing Aims and scope Submit manuscript

Abstract

In this paper, we study distributed-observer-based fault diagnosis and propose a fault-tolerant control approach for a class of discrete interconnected systems. The distributed fault observers are designed to estimate faults based on the improved fast adaptive fault estimation (FAFE) algorithm. As a result of the improved FAFE algorithm, the constraints which are necessary to the general FAFE algorithm can be reduced while the fault estimation accuracy can be maintained. Based on the online fault estimates, the distributed output feedback controllers are developed to accommodate faults. To solve the observers and controllers, the corresponding algorithms are proposed. Finally, various fault situations are considered in detail in a simulation, and the results verify the accuracy of the theory and method.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9

Similar content being viewed by others

References

  • Alwi H, Edwards C (2014) Robust fault reconstruction for linear parameter varying systems using sliding mode observers. Int J Robust Nonlinear Control 24(14):1947–1968

    Article  MathSciNet  Google Scholar 

  • Choi YH, Yoo SJ (2018) Decentralized adaptive output-feedback control of interconnected nonlinear time-delay systems using minimal neural networks. J Frankl Inst 355(1):81–105

    Article  MathSciNet  Google Scholar 

  • Deutscher J (2016) Fault detection for linear distributed-parameter systems using finite-dimensional functional observers. Int J Control 89(3):550–563

    Article  MathSciNet  Google Scholar 

  • Du M, Mhaskar P (2014) Isolation and handling of sensor faults in nonlinear systems. Automatica 50(4):1066–1074

    Article  MathSciNet  Google Scholar 

  • Duong VA, Tsai YW (2017) A novel interpretation of the invariance condition of mismatched uncertain large-scale systems. Int J Control Autom Syst 15(4):1577–1591

    Article  Google Scholar 

  • Gao Z (2015) Fault estimation and fault-tolerant control for discrete-time dynamic systems. IEEE Trans Ind Electron 62(6):3874–3884

    Google Scholar 

  • Hrizi O, Boussaid B et al (2013) Fast adaptive fault estimation algorithm: application to unicycle robot. In: Conference on control and fault-tolerant systems (SysTol), IEEE, pp 714–719

  • Lam HK, Wu L, Lam J (2015) Two-step stability analysis for general polynomial-fuzzy-model-based control systems. IEEE Trans Fuzzy Syst 23(3):511–524

    Article  Google Scholar 

  • Li L, Yang Y et al (2015) On fault-tolerant control configurations for a class of nonlinear systems. J Frankl Inst 352(4):1397–1416

    Article  MathSciNet  Google Scholar 

  • Li H, Pan Y et al (2016) Switched fuzzy output feedback control and its application to a mass–spring–damping system. IEEE Trans Fuzzy Syst 24(6):1259–1269

    Article  Google Scholar 

  • Liu D (2015) A fault-tolerant architecture for ROIA in cloud. J Ambient Intell Humanz Comput 6(5):587–595

    Article  Google Scholar 

  • Liu X, Gao X, Han J (2016) Observer-based fault detection for high-order nonlinear multi-agent systems. J Frankl Inst 353(1):72–94

    Article  MathSciNet  Google Scholar 

  • Majdzik P, Akielaszek-Witczak A et al (2016) A fault-tolerant approach to the control of a battery assembly system. Control Eng Pract 55:139–148

    Article  Google Scholar 

  • Mokryani G, Pierluigi S, Antonio P (2013) Fault ride-through enhancement of wind turbines in distribution networks. J Ambient Intell Humaniz Comput 4(6):605–611

    Article  Google Scholar 

  • Rigatos G, Siano P, Zervos N (2013) An approach to fault diagnosis of nonlinear systems using neural networks with invariance to Fourier transform. J Ambient Intell Humaniz Comput 4(6):621–639

    Article  Google Scholar 

  • Tabatabaeipour SM, Bak T (2014) Robust observer-based fault estimation and accommodation of discrete-time piecewise linear systems. J Frankl Inst 351(1):277–295

    Article  MathSciNet  Google Scholar 

  • Tong SC, Li YM, Zhang HG (2011) Adaptive neural network decentralized backstepping output-feedback control for nonlinear large-scale systems with time delays. IEEE Trans Neural Netw 22(7):1073–1086

    Article  Google Scholar 

  • Wang T, Tong S (2017) Observer-based output-feedback asynchronous control for switched fuzzy systems. IEEE Trans Cybern 47(9):2579–2591

    Article  Google Scholar 

  • Yao L, Qin J, Wang A, Wang H (2013) Fault diagnosis and fault-tolerant control for non-Gaussian non-linear stochastic systems using a rational square-root approximation model. IET Control Theory Appl 7(1):116–124

    Article  MathSciNet  Google Scholar 

  • You F, Li H, Wang F, Guan S (2015) Robust fast adaptive fault estimation for systems with time-varying interval delay. J Frankl Inst 352(12):5486–5513

    Article  MathSciNet  Google Scholar 

  • Zhou Z, Tan Y, Shi P (2016) Fault detection of a sandwich system with dead-zone based on robust observer. Syst Control Lett 96:132–140

    Article  MathSciNet  Google Scholar 

Download references

Acknowledgements

This work was supported by the National Science Foundation of China (Grant no. 61273190). The authors would like to thank the editor and reviewers for the valuable comments and constructive suggestions to improve the paper.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Jianfei Zhao.

Ethics declarations

Conflict of interest

The authors declare no conflict of interest regarding the publication of this paper.

Data availability statement

The simulation data used to support the findings of this study are included within the article.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendix

Appendix

The Appendix provides rigorous proofs of Theorem 1 and Theorem 3, which are fundamental to this paper.

1.1 A1 Proof of Theorem 1

Select the following Lyapunov function of systems (6):

$$V(k)=\sum\limits_{{i=1}}^{N} {\left( \begin{gathered} \widetilde {x}_{i}^{T}(k){P_{i1}}{\widetilde {x}_i}(k)+\widetilde {f}_{i}^{T}(k){P_{i2}}{\widetilde {f}_i}(k) \hfill \\ +\,\sum\limits_{{j=1}}^{N} {\sum\limits_{{l=1}}^{{{h_{ij}}}} {\widetilde {x}_{j}^{T}(k - l)\overline {G} _{{ij}}^{T}{R_{ij}}{{\overline {G} }_{ij}}{{\widetilde {x}}_j}(k - l} )} +\sum\limits_{{l=1}}^{{{h_{ii}}}} {\widetilde {x}_{i}^{T}(k - l){S_i}{{\widetilde {x}}_i}(k - l)} \hfill \\ \end{gathered} \right)} .$$
(26)

Take the first order difference of V(k). It follows that

$$\begin{aligned} & \Delta V(k)=V(k+1) - V(k) \\ & \quad =\sum\limits_{{i=1}}^{N} {\left( {\widetilde {x}_{i}^{T}(k+1){P_{i1}}{{\widetilde {x}}_i}(k+1) - \widetilde {x}_{i}^{T}(k){P_{i1}}{{\widetilde {x}}_i}(k)} \right.} \\ & \quad +\,\widetilde {f}_{i}^{T}(k+1){P_{i2}}{\widetilde {f}_i}(k+1) - \widetilde {f}_{i}^{T}(k){P_{i2}}{\widetilde {f}_i}(k)+\sum\limits_{{j=1}}^{N} {\widetilde {x}_{j}^{T}(k)\overline {G} _{{ij}}^{T}{R_{ij}}{{\overline {G} }_{ij}}{{\widetilde {x}}_j}(k)} \\ & \quad - \,\sum\limits_{{j=1}}^{N} {\widetilde {x}_{j}^{T}(k - {h_{ij}})\overline {G} _{{ij}}^{T}{R_{ij}}{{\overline {G} }_{ij}}{{\widetilde {x}}_j}(k - {h_{ij}})} \\ & \quad \left. {+\,\widetilde {x}_{i}^{T}(k){S_i}{{\widetilde {x}}_i}(k) - \widetilde {x}_{i}^{T}(k - {h_{ii}}){S_i}{{\widetilde {x}}_i}(k - {h_{ii}})} \right). \\ \end{aligned}$$
(27)

Due to \(||{\widetilde {g}_{ij}}||=||{g_{ij}}({x_j}(k - {h_{ij}})) - {g_{ij}}({\widehat {x}_j}(k - {h_{ij}}))|| \le ||{Z_{ij}}({\widetilde {x}_j}(k - {h_{ij}}))||\) and \(\sum\limits_{{i=1}}^{N} {\sum\limits_{{j=1}}^{N} {\widetilde {x}_{j}^{T}(k)\overline {G} _{{ij}}^{T}{R_{ij}}{{\overline {G} }_{ij}}{{\widetilde {x}}_j}(k)} } =\sum\limits_{{i=1}}^{N} {\widetilde {x}_{i}^{T}(k)\left( {\sum\limits_{{j=1}}^{N} {\overline {G} _{{ji}}^{T}{R_{ji}}{{\overline {G} }_{ji}}} } \right){{\widetilde {x}}_i}(k)}\), we can obtain

$$\begin{gathered} \Delta V(k) \le \sum\limits_{{i=1}}^{N} {\left( {\widetilde {x}_{i}^{T}(k+1){P_{i1}}{{\widetilde {x}}_i}(k+1)} \right.} +\widetilde {x}_{i}^{T}(k)\left( {\sum\limits_{{j=1}}^{N} {\overline {G} _{{ji}}^{T}{R_{ji}}{{\overline {G} }_{ji}}} - {P_{i1}}+{S_i}} \right){\widetilde {x}_i}(k) \\ +\,\widetilde {f}_{i}^{T}(k+1){P_{i2}}{\widetilde {f}_i}(k+1) - \widetilde {f}_{i}^{T}(k){P_{i2}}{\widetilde {f}_i}(k) - \widetilde {x}_{i}^{T}(k - {h_{ii}}){S_i}{\widetilde {x}_i}(k - {h_{ii}}) \\ +\,\left. {\sum\limits_{{j=1}}^{N} {\widetilde {x}_{j}^{T}(k - {h_{ij}})\left( {Z_{{ij}}^{T}{Z_{ij}} - \overline {G} _{{ij}}^{T}{R_{ij}}{{\overline {G} }_{ij}}} \right){{\widetilde {x}}_j}(k - {h_{ij}})} - \widetilde {g}_{i}^{T}{{\widetilde {g}}_i}} \right), \\ \end{gathered}$$
(28)

where, \({\widetilde {g}_i}={\left( {\begin{array}{*{20}{c}} {\widetilde {g}_{{i1}}^{T}}&{\widetilde {g}_{{i2}}^{T}}&{.....}&{\widetilde {g}_{{iN}}^{T}} \end{array}} \right)^T}\), in (28), there are

$$\begin{gathered} \widetilde {x}_{i}^{T}(k+1){P_{i1}}{\widetilde {x}_i}(k+1) \hfill \\ \quad =\,\eta _{i}^{T}(k)\overline {\varphi } _{{i1}}^{T}{P_{i1}}{\overline {\varphi } _{i1}}{\eta _i}(k) - 2\eta _{i}^{T}(k)\overline {\varphi } _{{i1}}^{T}{P_{i1}}{H_i}{C_i}{\widetilde {x}_i}(k - {h_{ii}}) \hfill \\ \quad +\,\widetilde {x}_{t}^{T}(k - {h_{ii}})C_{t}^{T}H_{i}^{T}{P_{i1}}{H_i}{C_i}{\widetilde {x}_i}(k - {h_{ii}}), \hfill \\ \end{gathered}$$
(29)
$$\begin{gathered} \widetilde {f}_{i}^{T}(k+1){P_{i2}}{\widetilde {f}_i}(k+1) \hfill \\ \quad =\,\eta _{i}^{T}(k)\overline {\varphi } _{{i2}}^{T}{P_{i2}}{\overline {\varphi } _{i2}}{\eta _i}(k)+2\eta _{i}^{T}(k)\overline {\varphi } _{{i2}}^{T}{P_{i2}}{\Gamma _i}{\Lambda _{i1}}{C_i}{H_i}{C_i}{\widetilde {x}_i}(k - {h_{ii}}) \hfill \\ \quad \quad +\,\widetilde {x}_{i}^{T}(k - {h_{ii}})C_{i}^{T}H_{i}^{T}C_{i}^{T}\Lambda _{{i1}}^{T}{\Gamma _i}{P_{i2}}{\Gamma _i}{\Lambda _{i1}}{C_i}{H_i}{C_i}{\widetilde {x}_i}(k - {h_{ii}}), \hfill \\ \end{gathered}$$
(30)

where, \({\eta _i}(k)={\left( {\begin{array}{*{20}{c}} {\widetilde {x}_{i}^{T}(k)}&{\widetilde {f}_{i}^{T}(k)}&{\widetilde {g}_{i}^{T}}&{w_{i}^{T}(k)}&{\Delta f_{i}^{T}(k)} \end{array}} \right)^T}.\)

Substitute (29), (30) into (28). Then, according to (7), and considering the zero initial condition, it follows that

$$\begin{aligned} J & =\sum\limits_{{k=0}}^{\infty } {\sum\limits_{{i=1}}^{N} {\left( {\frac{1}{{{\gamma _i}}}\widetilde {y}_{i}^{T}(k){{\widetilde {y}}_i}(k) - {\gamma _i}\mu _{i}^{T}(k){\mu _i}(k)} \right)} } \\ & \le \,\sum\limits_{{k=0}}^{\infty } {\left( {\sum\limits_{{i=1}}^{N} {\left( {\frac{1}{{{\gamma _i}}}\widetilde {y}_{i}^{T}(k){{\widetilde {y}}_i}(k) - {\gamma _i}\mu _{i}^{T}(k){\mu _i}(k)} \right)+\Delta V(k)} } \right)} \\ & =\,\sum\limits_{{k=0}}^{\infty } {\sum\limits_{{i=1}}^{N} {\eta _{i}^{T}(k)\left( {\begin{array}{*{20}{c}} {{\Phi _{i1}}}&{{\Phi _{i2}}} \\ {\Phi _{{i2}}^{T}}&{{\Phi _{i3}}} \end{array}} \right)} } {\eta _i}(k). \\ \end{aligned}$$
(31)

In (31), there are

$${\Phi _{i1}}=\overline {\varphi } _{{i1}}^{T}{P_{i1}}{\overline {\varphi } _{i1}}+\overline {\varphi } _{{i2}}^{T}{P_{i2}}{\overline {\varphi } _{i2}}+{\Phi _{i10}}+\sum\limits_{{j=1}}^{N} {\overline {G} _{{ji0}}^{T}{R_{ji}}{{\overline {G} }_{ji0}}} ,$$
(32)
$${\Phi _{i2}}=\overline {\varphi } _{{i2}}^{T}{P_{i2}}{\Gamma _i}{\Lambda _{i1}}{C_i}{H_i}{\widetilde {C}_i} - \overline {\varphi } _{{i1}}^{T}{P_{i1}}{H_i}{\widetilde {C}_i},$$
(33)
$${\Phi _{i3}}=\widetilde {Z}_{i}^{T}{\widetilde {Z}_i} - \widetilde {{\overline {G} }}_{i}^{T}{\widetilde {R}_i}{\widetilde {{\overline {G} }}_i} - {\widetilde {S}_{i0}}+\widetilde {C}_{i}^{T}H_{i}^{T}{P_{i1}}{H_i}{\widetilde {C}_i}+\widetilde {C}_{i}^{T}H_{i}^{T}C_{i}^{T}\Lambda _{{i1}}^{T}{\Gamma _i}{P_{i2}}{\Gamma _i}{\Lambda _{i1}}{C_i}{H_i}{\widetilde {C}_i},$$
(34)

where

$${\Phi _{i10}}=diag\left( {\begin{array}{*{20}{c}} { - {P_{i1}}+{S_i}+\frac{1}{{{\gamma _i}}}{C_i}^{T}{C_i}}&{ - {P_{i2}}}&{ - I}&{ - {\gamma _i}I}&{ - {\gamma _i}I} \end{array}} \right),$$
$${\overline {\Phi } _{i10}}=diag\left( {\begin{array}{*{20}{c}} { - {P_{i1}}+{S_i}+\sum\limits_{{j=1}}^{N} {\overline {G} _{{ji}}^{T}{R_{ji}}{{\overline {G} }_{ji}}} +\frac{1}{{{\gamma _i}}}{C_i}^{T}{C_i}}&{ - {P_{i2}}}&{ - I}&{ - {\gamma _i}I}&{ - {\gamma _i}I} \end{array}} \right),$$
$${\widetilde {C}_i}=\left( {\begin{array}{*{20}{c}} {0 \ldots 0}&{{C_i}}&{0 \ldots 0} \end{array}} \right).$$

For \(- \widetilde {{\overline {G} }}_{i}^{T}{\widetilde {R}_i}{\widetilde {{\overline {G} }}_i}\) in \({\Phi _{i3}}\),due to \({\widetilde {{\overline {G} }}_i}={\widetilde {G}_i}+{\widetilde {W}_i}{\widetilde {F}_i}{\widetilde {N}_i}\) and Lemma 1,we can deduce:

$$- \,\widetilde {{\overline {G} }}_{i}^{T}{\widetilde {R}_i}{\widetilde {{\overline {G} }}_i} \le - \widetilde {G}_{i}^{T}{\widetilde {R}_i}{\widetilde {G}_i}+\widetilde {G}_{i}^{T}{\widetilde {R}_i}{\widetilde {W}_i}\Gamma _{{\alpha i}}^{{ - 1}}\widetilde {W}_{i}^{T}{\widetilde {R}_i}{\widetilde {G}_i}+\widetilde {N}_{i}^{T}{\Gamma _{\beta i}}{\widetilde {N}_i}.$$
(35)

By (3235), it follows that

$$\begin{aligned} \left( {\begin{array}{*{20}{c}} {{\Phi _{i1}}}&{{\Phi _{i2}}} \\ {\Phi _{{i2}}^{T}}&{{\Phi _{i3}}} \end{array}} \right) \le & \,\left( {\begin{array}{*{20}{c}} {{\Phi _{i10}}}&0 \\ 0&{\widetilde {Z}_{i}^{T}{{\widetilde {Z}}_i} - {{\widetilde {S}}_{i0}} - \widetilde {G}_{i}^{T}{{\widetilde {R}}_i}{{\widetilde {G}}_i}} \end{array}} \right) \\ & +\,\left( {\begin{array}{*{20}{c}} {\overline {\varphi } _{{i1}}^{T}} \\ { - \widetilde {C}_{i}^{T}H_{i}^{T}} \end{array}} \right){P_{i1}}\left( {\begin{array}{*{20}{c}} {{{\overline {\varphi } }_{i1}}}&{ - {H_i}{{\widetilde {C}}_i}} \end{array}} \right) \\ & +\,\,\left( {\begin{array}{*{20}{c}} {\overline {\varphi } _{{i2}}^{T}} \\ {\widetilde {C}_{i}^{T}H_{i}^{T}C_{i}^{T}\Lambda _{{i1}}^{T}} \end{array}} \right){P_{i2}}\left( {\begin{array}{*{20}{c}} {{{\overline {\varphi } }_{i2}}}&{{\Lambda _{i1}}{C_i}{H_i}{{\widetilde {C}}_i}} \end{array}} \right)+\sum\limits_{{j=1}}^{N} {\left( \begin{gathered} \overline {G} _{{ji0}}^{T} \hfill \\ 0 \hfill \\ \end{gathered} \right){R_{ji}}(\begin{array}{*{20}{c}} {{{\overline {G} }_{ji0}}}&0 \end{array})} \\ & +\,\left( \begin{gathered} 0 \hfill \\ \widetilde {G}_{i}^{T}{\widetilde {R}_i}{\widetilde {W}_i} \hfill \\ \end{gathered} \right)\Gamma _{{\alpha i}}^{{ - 1}}\left( {\begin{array}{*{20}{c}} 0&{\widetilde {W}_{i}^{T}{{\widetilde {R}}_i}{{\widetilde {G}}_i}} \end{array}} \right)+\left( \begin{gathered} 0 \hfill \\ \widetilde {N}_{i}^{T} \hfill \\ \end{gathered} \right){\Gamma _{\beta i}}\left( {\begin{array}{*{20}{c}} 0&{{{\widetilde {N}}_i}} \end{array}} \right). \\ \end{aligned}$$
(36)

According to the Schur complements, we can determine that

$$\left( {\begin{array}{*{20}{c}} {{\Phi _{i1}}}&{{\Phi _{i2}}} \\ {\Phi _{{i2}}^{T}}&{{\Phi _{i3}}} \end{array}} \right)<0,$$

is equivalent to

$$\begin{gathered} \left( {\begin{array}{*{20}{c}} {{\Phi _{i10}}}&0&{\overline {\varphi } _{{i1}}^{T}} \\ *&{\widetilde {Z}_{i}^{T}{{\widetilde {Z}}_i} - {{\widetilde {S}}_{i0}} - \widetilde {G}_{i}^{T}{{\widetilde {R}}_i}{{\widetilde {G}}_i}}&{ - \widetilde {C}_{i}^{T}H_{i}^{T}} \\ *&*&{ - P_{{i1}}^{{ - 1}}} \\ *&*&* \\ *&*&* \\ *&*&* \\ *&*&* \end{array}} \right.\left. {\begin{array}{*{20}{c}} {\overline {\varphi } _{{i2}}^{T}}&0&0&{\overline {G} _{{i0}}^{{'T}}} \\ 0&{\widetilde {G}_{i}^{T}{{\widetilde {R}}_i}{{\widetilde {W}}_i}}&{\widetilde {N}_{i}^{T}}&0 \\ 0&0&0&0 \\ { - P_{{i2}}^{{ - 1}}}&0&0&0 \\ *&{ - {\Gamma _{\alpha i}}}&0&0 \\ *&*&{ - \Gamma _{{\beta i}}^{{ - 1}}}&0 \\ *&*&*&{ - \widetilde {R}_{i}^{{' - 1}}} \end{array}} \right) \hfill \\ \quad +\,\,\left( \begin{gathered} 0 \hfill \\ \widetilde {C}_{i}^{T}H_{i}^{T} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right){\left( \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ {\Lambda _{i1}}{C_i} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)^T}+\left( \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ {\Lambda _{i1}}{C_i} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right){\left( \begin{gathered} 0 \hfill \\ \widetilde {C}_{i}^{T}H_{i}^{T} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)^T}<0. \hfill \\ \end{gathered}$$
(37)

Owing to:

$$\left( \begin{gathered} 0 \hfill \\ \widetilde {C}_{i}^{T}H_{i}^{T} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right){\left( \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ {\Lambda _{i1}}{C_i} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)^T}+\left( \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ {\Lambda _{i1}}{C_i} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right){\left( \begin{gathered} 0 \hfill \\ \widetilde {C}_{i}^{T}H_{i}^{T} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)^T} \le \left( \begin{gathered} 0 \hfill \\ \widetilde {C}_{i}^{T}H_{i}^{T} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)\Pi _{{i1}}^{{ - 1}}{\left( \begin{gathered} 0 \hfill \\ \widetilde {C}_{i}^{T}H_{i}^{T} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)^T}+\left( \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ {\Lambda _{i1}}{C_i} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right){\Pi _{i1}}{\left( \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ {\Lambda _{i1}}{C_i} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)^T},$$
(38)

by the Schur complements and robust stability theory, we can know that, if the condition in Theorem 1 holds,the matrix inequality (37) will hold. Therefore, the errors of state \({\widetilde {x}_i}(t)\) and fault estimation \({\widetilde {f}_i}(t)\), \(i=1 \ldots N\) are robust and stable, and the error systems (6) satisfy the H performance \({\left\| {{y_i}(k)} \right\|_2}<{\gamma _i}{\left\| {{\mu _i}(k)} \right\|_2}\). Thus, Theorem 1 is proved.

1.2 A2 Proof of Theorem 3

We define the Lyapunov function of system (12) as follows:

$${V_c}(k)=\sum\limits_{{i=1}}^{N} {\left( {\eta _{{ci}}^{T}(k){P_{ci}}{\eta _{ci}}(k)+\sum\limits_{{j=1}}^{N} {\sum\limits_{{l=1}}^{{{h_{ij}}}} {x_{j}^{T}(k - {h_{ij}})\widetilde {{\overline {G} }}_{{ij}}^{T}{R_{cij}}{{\widetilde {{\overline {G} }}}_{ij}}{x_j}(k - {h_{ij}})} } } \right)} .$$
(39)

Taking the difference of \({V_c}(k)\) along the systems (12), it follows that

$$\begin{aligned} \Delta {V_c}(k) & ={V_c}(k+1) - {V_c}(k) \\ & \le \sum\limits_{{i=1}}^{N} {\left( {\eta _{{ci}}^{T}(k)\widetilde {{\overline {A} }}_{i}^{T}{P_{ci}}{{\widetilde {{\overline {A} }}}_i}{\eta _{ci}}(k)+g_{i}^{T}\widetilde {{\overline {G} }}_{{ci}}^{T}{P_{ci}}{{\widetilde {{\overline {G} }}}_{ci}}{g_i}} \right.} +\mu _{{ci}}^{T}(k)\widetilde {{\overline {D} }}_{i}^{T}{P_{ci}}{\widetilde {{\overline {D} }}_i}{\mu _{ci}}(k)+2\eta _{{ci}}^{T}(k)\widetilde {{\overline {A} }}_{i}^{T}{P_{ci}}{\widetilde {{\overline {G} }}_{ci}}{g_i} \\ & \quad +\,2\eta _{{ci}}^{T}(k)\widetilde {{\overline {A} }}_{i}^{T}{P_{ci}}{\widetilde {{\overline {D} }}_i}{\mu _{ci}}(k)+2g_{i}^{T}\widetilde {{\overline {G} }}_{{ci}}^{T}{P_{ci}}{\widetilde {{\overline {D} }}_i}{\mu _{ci}}(k) \\ & \quad +\,x_{i}^{T}(k)\left( {\sum\limits_{{j=1}}^{N} {\widetilde {{\overline {G} }}_{{ji}}^{T}{R_{cji}}{{\widetilde {{\overline {G} }}}_{ji}}} } \right){x_i}(k) - \sum\limits_{{j=1}}^{N} {x_{j}^{T}(k - {h_{ij}})\widetilde {{\overline {G} }}_{{ij}}^{T}{R_{cij}}{{\widetilde {{\overline {G} }}}_{ij}}{x_j}(k - {h_{ij}})} \\ & \quad +\,\sum\limits_{{j=1}}^{N} {x_{j}^{T}(k - {h_{ij}})Z_{{cij}}^{T}{Z_{cij}}{x_j}(k - {h_{ij}})} \left. { - \sum\limits_{{j=1}}^{N} {g_{{ij}}^{T}({x_j}(k - {h_{ij}})){g_{ij}}({x_j}(k - {h_{ij}}))} } \right), \\ \end{aligned}$$
(40)

where, \({g_i}={\left( {\begin{array}{*{20}{c}} {{g_{i1}}^{T}({x_1}(k - {h_{i1}}))}& \ldots &{{g_{iN}}^{T}({x_N}(k - {h_{iN}}))} \end{array}} \right)^T}.\)

According to (13), and take into account the zero initial condition, we have

$$\begin{aligned} {J_c} & \le \sum\limits_{{k=0}}^{\infty } {\left( {\sum\limits_{{i=1}}^{N} {\left( {\frac{1}{{{\gamma _{ci}}}}y_{i}^{T}(k){y_i}(k) - {\gamma _{ci}}\mu _{i}^{T}(k){\mu _i}(k)} \right)} +\Delta V(k)} \right)} \\ & =\,{\sum\limits_{{k=0}}^{\infty } {\sum\limits_{{i=1}}^{N} {\left( \begin{gathered} {\eta _{ci}}(k) \hfill \\ {g_i} \hfill \\ {\mu _i}(k) \hfill \\ {x_1}(k - {h_{i1}}) \hfill \\ . \hfill \\ . \hfill \\ {x_N}(k - {h_{iN}}) \hfill \\ \end{gathered} \right)} } ^T}\left( {\begin{array}{*{20}{c}} {{\Phi _{ci1}}}&0 \\ 0&{{\Phi _{ci2}}} \end{array}} \right)\left( \begin{gathered} {\eta _{ci}}(k) \hfill \\ {g_i} \hfill \\ {\mu _i}(k) \hfill \\ {x_1}(k - {h_{i1}}) \hfill \\ . \hfill \\ . \hfill \\ {x_N}(k - {h_{iN}}) \hfill \\ \end{gathered} \right), \\ \end{aligned}$$
(41)

where

$${\Phi _{ci1}}=\left( {\begin{array}{*{20}{c}} {\widetilde {{\overline {A} }}_{i}^{T}{P_{ci}}{{\widetilde {{\overline {A} }}}_i}+\sum\limits_{{j=1}}^{N} {\widetilde {{\overline {G} }}_{{ji0}}^{T}{R_{cji}}{{\widetilde {{\overline {G} }}}_{ji0}}} - {P_{ci}}+\frac{1}{{{\gamma _{ci}}}}C_{i}^{{'T}}C_{i}^{'}}&{\widetilde {{\overline {A} }}_{i}^{T}{P_{ci}}{{\widetilde {{\overline {G} }}}_{ci}}}&{\widetilde {{\overline {A} }}_{i}^{T}{P_{ci}}{{\widetilde {{\overline {D} }}}_i}} \\ *&{\widetilde {{\overline {G} }}_{{ci}}^{T}{P_{ci}}{{\widetilde {{\overline {G} }}}_{ci}} - I}&{\widetilde {{\overline {G} }}_{{ci}}^{T}{P_{ci}}{{\widetilde {{\overline {D} }}}_i}} \\ *&*&{\widetilde {{\overline {D} }}_{i}^{T}{P_{ci}}{{\widetilde {{\overline {D} }}}_i} - {\gamma _{ci}}I} \end{array}} \right),$$
$$\begin{aligned} {\Phi _{ci2}} & =\widetilde {Z}_{{ci}}^{T}{\widetilde {Z}_{ci}} - \widetilde {{\widetilde {{\overline {G} }}}}_{i}^{T}{\widetilde {R}_{ci}}{\widetilde {{\widetilde {{\overline {G} }}}}_i} \\ & =\,diag\left\{ {Z_{{ci1}}^{T}{Z_{ci1}} - \widetilde {{\overline {G} }}_{{i1}}^{T}{R_{ci1}}{{\widetilde {{\overline {G} }}}_{i1}}, \ldots ,Z_{{ciN}}^{T}{Z_{ciN}} - \widetilde {{\overline {G} }}_{{iN}}^{T}{R_{ciN}}{{\widetilde {{\overline {G} }}}_{iN}}} \right\}. \\ \end{aligned}$$

Due to \({\widetilde {{\widetilde {{\overline {G} }}}}_i}={\widetilde {{\widetilde {G}}}_i}+{\widetilde {{\widetilde {W}}}_i}{\widetilde {F}_i}{\widetilde {N}_i}\) and Lemma 1, we can deduce:

$${\Phi _{ci2}} \le \widetilde {Z}_{{ci}}^{T}{\widetilde {Z}_{ci}} - \widetilde {{\widetilde {G}}}_{i}^{T}{\widetilde {R}_{ci}}{\widetilde {{\widetilde {G}}}_i}+\widetilde {{\widetilde {G}}}_{i}^{T}{\widetilde {R}_{ci}}{\widetilde {{\widetilde {W}}}_i}\Gamma _{{c\alpha i}}^{{ - 1}}\widetilde {{\widetilde {W}}}_{i}^{T}{\widetilde {R}_{ci}}{\widetilde {{\widetilde {G}}}_i}+\widetilde {N}_{i}^{T}{\Gamma _{c\beta i}}{\widetilde {N}_i}.$$
(42)

So, according to the Schur complements, \({\Phi _{ci2}}<0\) will hold when (15) holds.

Note that

$$\begin{aligned} {\Phi _{ci1}}= & \left( \begin{gathered} \widetilde {{\overline {A} }}_{i}^{T} \hfill \\ \widetilde {{\overline {G} }}_{{ci}}^{T} \hfill \\ \widetilde {{\overline {D} }}_{i}^{T} \hfill \\ \end{gathered} \right){P_{ci}}\left( {\begin{array}{*{20}{c}} {{{\widetilde {{\overline {A} }}}_i}}&{{{\widetilde {{\overline {G} }}}_{ci}}}&{{{\widetilde {{\overline {D} }}}_i}} \end{array}} \right)+\left( {\begin{array}{*{20}{c}} { - {P_{ci}}+\frac{1}{{{\gamma _{ci}}}}C_{i}^{{'T}}C_{i}^{'}}&0&0 \\ 0&{ - I}&0 \\ 0&0&{ - {\gamma _{ci}}I} \end{array}} \right) \\ & +\sum\limits_{{j=1}}^{N} {\left( \begin{gathered} \widetilde {{\overline {G} }}_{{ji0}}^{T} \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)} {R_{cji}}\left( {\begin{array}{*{20}{c}} {{{\widetilde {{\overline {G} }}}_{ji0}}}&0&0 \end{array}} \right). \\ \end{aligned}$$
(43)

Thus, by the Schur complements, \({\Phi _{ci1}}<0\) will hold when (44) holds

$$\left( {\begin{array}{*{20}{c}} { - {P_{ci}}+\frac{1}{{{\gamma _{ci}}}}C_{i}^{{'T}}C_{i}^{'}}&0&0&{\widetilde {{\overline {A} }}_{i}^{T}}&{\widetilde {{\overline {G} }}_{{i0}}^{{'T}}} \\ *&{ - I}&0&{\widetilde {{\overline {G} }}_{{ci}}^{T}}&0 \\ *&*&{ - {\gamma _{ci}}I}&{\widetilde {{\overline {D} }}_{i}^{T}}&0 \\ *&*&*&{ - P_{{ci}}^{{ - 1}}}&0 \\ *&*&*&*&{ - \widetilde {R}_{{ci}}^{{' - 1}}} \end{array}} \right)<0.$$
(44)

The matrix inequality (44) can be expressed as:

$$\begin{gathered} \left( {\begin{array}{*{20}{c}} { - {P_{ci}}+\frac{1}{{{\gamma _{ci}}}}\widetilde {C}_{i}^{T}{{\widetilde {C}}_i}}&0&0&{\widetilde {A}_{i}^{T}}&{\widetilde {G}_{{i0}}^{{'T}}} \\ *&{ - I}&0&{\widetilde {G}_{i}^{T}}&0 \\ *&*&{ - {\gamma _{ci}}I}&{\widetilde {D}_{i}^{T}}&0 \\ *&*&*&{ - P_{{ci}}^{{ - 1}}}&0 \\ *&*&*&*&{ - \widetilde {R}_{{ci}}^{{' - 1}}} \end{array}} \right)+\left( \begin{gathered} \widetilde {E}_{{1ci}}^{T} \hfill \\ 0 \hfill \\ \widetilde {E}_{{1ci}}^{T} \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)F_{i}^{T}{\left( \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ {\widetilde {M}_{1i}} \hfill \\ 0 \hfill \\ \end{gathered} \right)^T}+\left( \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ {\widetilde {M}_{1i}} \hfill \\ 0 \hfill \\ \end{gathered} \right){F_i}{\left( \begin{gathered} \widetilde {E}_{{1ci}}^{T} \hfill \\ 0 \hfill \\ \widetilde {E}_{{1ci}}^{T} \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)^T} \hfill \\ \quad +\,\left( \begin{gathered} 0 \hfill \\ \widetilde {N}_{i}^{T} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)\widetilde {F}_{i}^{T}{\left( \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ {\widetilde {W}_{ci}} \hfill \\ 0 \hfill \\ \end{gathered} \right)^T}+\left( \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ {\widetilde {W}_{ci}} \hfill \\ 0 \hfill \\ \end{gathered} \right){\widetilde {F}_i}{\left( \begin{gathered} 0 \hfill \\ \widetilde {N}_{i}^{T} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)^T}+\left( \begin{gathered} \widetilde {N}_{i}^{{'T}} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)\widetilde {F}_{i}^{{'T}}{\left( \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ {\widetilde {{\widetilde {W}}}_i} \hfill \\ \end{gathered} \right)^T}+\left( \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ {\widetilde {{\widetilde {W}}}_i} \hfill \\ \end{gathered} \right)\widetilde {F}_{i}^{'}{\left( \begin{gathered} \widetilde {N}_{i}^{{'T}} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)^T}<0. \hfill \\ \end{gathered}$$
(45)

By Lemma 1, we can note that (45) will hold when (46) holds:

$$\begin{gathered} \left( {\begin{array}{*{20}{c}} { - {P_{ci}}+\frac{1}{{{\gamma _{ci}}}}C_{i}^{{'T}}C_{i}^{'}}&0&0&{\widetilde {A}_{i}^{T}}&{\widetilde {G}_{{i0}}^{{'T}}} \\ *&{ - I}&0&{\widetilde {G}_{{ci}}^{T}}&0 \\ *&*&{ - {\gamma _{ci}}I}&{\widetilde {D}_{i}^{T}}&0 \\ *&*&*&{ - P_{{ci}}^{{ - 1}}}&0 \\ *&*&*&*&{ - \widetilde {R}_{{ci}}^{{' - 1}}} \end{array}} \right)+{\varepsilon _{c2i}}{\left( \begin{gathered} \widetilde {E}_{{1ci}}^{T} \hfill \\ 0 \hfill \\ \widetilde {E}_{{1ci}}^{T} \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)_i}{\left( \begin{gathered} \widetilde {E}_{{1ci}}^{T} \hfill \\ 0 \hfill \\ \widetilde {E}_{{1ci}}^{T} \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)^T}+\varepsilon _{{c2i}}^{{ - 1}}\left( \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ {\widetilde {M}_{1i}} \hfill \\ 0 \hfill \\ \end{gathered} \right){\left( \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ {\widetilde {M}_{1i}} \hfill \\ 0 \hfill \\ \end{gathered} \right)^T} \hfill \\ \quad +\,\left( \begin{gathered} 0 \hfill \\ \widetilde {N}_{i}^{T} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)\Gamma _{{cNi}}^{{ - 1}}{\left( \begin{gathered} 0 \hfill \\ \widetilde {N}_{i}^{T} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)^T}+\left( \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ {\widetilde {W}_{ci}} \hfill \\ 0 \hfill \\ \end{gathered} \right){\Gamma _{cWi}}{\left( \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ {\widetilde {W}_{ci}} \hfill \\ 0 \hfill \\ \end{gathered} \right)^T}+\left( \begin{gathered} \widetilde {N}_{i}^{{'T}} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)\Gamma _{{cN'i}}^{{ - 1}}{\left( \begin{gathered} \widetilde {N}_{i}^{{'T}} \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} \right)^T}+\left( \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ {\widetilde {{\widetilde {W}}}_i} \hfill \\ \end{gathered} \right){\Gamma _{cW'i}}{\left( \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ {\widetilde {{\widetilde {W}}}_i} \hfill \\ \end{gathered} \right)^T}<0. \hfill \\ \end{gathered}$$
(46)

And according to the Schur complements (46), is equivalent to (14) in Theorem 3.

Thus, \({\Phi _{ci1}}<0\), \({\Phi _{ci2}}<0\) will hold when (14) and (15) hold. Then, Theorem 3 is proved.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Xi, X., Zhao, J., Liu, T. et al. Distributed-observer-based fault diagnosis and fault-tolerant control for time-varying discrete interconnected systems. J Ambient Intell Human Comput 11, 459–482 (2020). https://doi.org/10.1007/s12652-018-1130-7

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s12652-018-1130-7

Keywords

Navigation