Multistability of Switched Neural Networks With Piecewise Linear Activation Functions Under State-Dependent Switching

This paper is concerned with the multistability of switched neural networks with piecewise linear activation functions under state-dependent switching. Under some reasonable assumptions on the switching threshold and activation functions, by using the state-space decomposition method, contraction mapping theorem, and strictly diagonally dominant matrix theory, we can characterize the number of equilibria as well as analyze the stability/instability of the equilibria. More interesting, we can find that the switching threshold plays an important role for stable equilibria in the unsaturation regions of activation functions, and the number of stable equilibria of an <inline-formula> <tex-math notation="LaTeX">$n$ </tex-math></inline-formula>-neuron switched neural network with state-dependent parameters increases to <inline-formula> <tex-math notation="LaTeX">$3^{n}$ </tex-math></inline-formula> from <inline-formula> <tex-math notation="LaTeX">$2^{n}$ </tex-math></inline-formula> in the conventional one. Furthermore, for two-neuron switched neural networks, the precise attraction basin of each stable equilibrium point can be figured out, and its boundary is composed of the stable manifolds of unstable equilibrium points and the switching lines. Two simulation examples are discussed in detail to substantiate the effectiveness of the theoretical analysis.

[1]  A. Stephen Morse,et al.  Switched nonlinear systems with state-dependent dwell-time , 2003, Syst. Control. Lett..

[2]  Jun Zhao,et al.  Dissipativity Theory for Switched Systems , 2005, CDC 2005.

[3]  Wu-Hua Chen,et al.  Multistability in a class of stochastic delayed Hopfield neural networks , 2015, Neural Networks.

[4]  Wei Xing Zheng,et al.  Exponential stability of multiple equilibria for memristive Cohen-Grossberg neural networks with non-monotonic activation functions , 2015, 2015 5th Australian Control Conference (AUCC).

[5]  Foss,et al.  Multistability and delayed recurrent loops. , 1996, Physical review letters.

[6]  Daniel Liberzon,et al.  Switching in Systems and Control , 2003, Systems & Control: Foundations & Applications.

[7]  CHIH-WEN SHIH,et al.  Multistability in Recurrent Neural Networks , 2006, SIAM J. Appl. Math..

[8]  Wei Xing Zheng,et al.  Complete Stability of Neural Networks With Nonmonotonic Piecewise Linear Activation Functions , 2015, IEEE Transactions on Circuits and Systems II: Express Briefs.

[9]  Tianping Chen,et al.  Multistability and New Attraction Basins of Almost-Periodic Solutions of Delayed Neural Networks , 2009, IEEE Transactions on Neural Networks.

[10]  Linlin Liu,et al.  Multistability of Recurrent Neural Networks With Piecewise-Linear Radial Basis Functions and State-Dependent Switching Parameters , 2020, IEEE Transactions on Systems, Man, and Cybernetics: Systems.

[11]  Zhigang Zeng,et al.  Multistability of Recurrent Neural Networks With Nonmonotonic Activation Functions and Mixed Time Delays , 2016, IEEE Transactions on Systems, Man, and Cybernetics: Systems.

[12]  Jun Wang,et al.  Global exponential stability of a general class of recurrent neural networks with time-varying delays , 2003 .

[13]  D. Chillingworth DYNAMICAL SYSTEMS: STABILITY, SYMBOLIC DYNAMICS AND CHAOS , 1998 .

[14]  Robert Shorten,et al.  On the Stability of Switched Positive Linear Systems , 2007, IEEE Transactions on Automatic Control.

[15]  Jinde Cao,et al.  Multistability of neural networks with discontinuous activation function , 2008 .

[16]  Tianping Chen,et al.  Multistability of Neural Networks With Mexican-Hat-Type Activation Functions , 2012, IEEE Transactions on Neural Networks and Learning Systems.

[17]  Jun Wang,et al.  Multiperiodicity and Exponential Attractivity Evoked by Periodic External Inputs in Delayed Cellular Neural Networks , 2006 .

[18]  Mauro Forti,et al.  Limit Set Dichotomy and Multistability for a Class of Cooperative Neural Networks With Delays , 2012, IEEE Transactions on Neural Networks and Learning Systems.

[19]  Jianbin Qiu,et al.  Robust mixed H/sub 2//H αfiltering design for discrete-time switched polytopic linear systems , 2008 .

[20]  T.-H. Kim,et al.  A new stability analysis of switched systems , 2000, Autom..

[21]  Zhang Yi,et al.  Multistability of discrete-time recurrent neural networks with unsaturating piecewise linear activation functions , 2004, IEEE Transactions on Neural Networks.

[22]  Chang-Yuan Cheng,et al.  Multistability and convergence in delayed neural networks , 2007 .

[23]  Mauro Forti,et al.  Convergence and Multistability of Nonsymmetric Cellular Neural Networks With Memristors , 2017, IEEE Transactions on Cybernetics.

[24]  Jun Wang,et al.  Global uniform asymptotic stability of memristor-based recurrent neural networks with time delays , 2010, The 2010 International Joint Conference on Neural Networks (IJCNN).

[25]  J J Hopfield,et al.  Neurons with graded response have collective computational properties like those of two-state neurons. , 1984, Proceedings of the National Academy of Sciences of the United States of America.

[26]  Aleksej F. Filippov,et al.  Differential Equations with Discontinuous Righthand Sides , 1988, Mathematics and Its Applications.

[27]  Kai Zhang,et al.  Exponential stability for switched Cohen–Grossberg neural networks with average dwell time , 2011, Proceedings of the 29th Chinese Control Conference.

[28]  Zhigang Zeng,et al.  Multistability of periodic delayed recurrent neural network with memristors , 2012, Neural Computing and Applications.

[29]  Jinde Cao,et al.  Coexistence and local μ-stability of multiple equilibrium points for memristive neural networks with nonmonotonic piecewise linear activation functions and unbounded time-varying delays , 2016, Neural Networks.

[30]  Jinde Cao,et al.  Delay-dependent multistability in recurrent neural networks , 2010, Neural Networks.

[31]  Masahiko Morita,et al.  Associative memory with nonmonotone dynamics , 1993, Neural Networks.

[32]  Zhigang Zeng,et al.  Global exponential stability of recurrent neural networks with time-varying delays in the presence of strong external stimuli , 2006, Neural Networks.

[33]  Jinde Cao,et al.  Multistability and multiperiodicity of delayed Cohen–Grossberg neural networks with a general class of activation functions , 2008 .

[34]  Zhigang Zeng,et al.  Complete stability of delayed recurrent neural networks with Gaussian activation functions , 2017, Neural Networks.

[35]  Jinde Cao,et al.  Stability in Cohen–Grossberg-type bidirectional associative memory neural networks with time-varying delays , 2006 .

[36]  Wei Xing Zheng,et al.  A New Method for Complete Stability Analysis of Cellular Neural Networks With Time Delay , 2010, IEEE Transactions on Neural Networks.

[37]  Wei Xing Zheng,et al.  Multistability and Instability of Neural Networks With Discontinuous Nonmonotonic Piecewise Linear Activation Functions , 2015, IEEE Transactions on Neural Networks and Learning Systems.

[38]  Jinde Cao,et al.  Multistability of memristive Cohen-Grossberg neural networks with non-monotonic piecewise linear activation functions and time-varying delays , 2015, Neural Networks.

[39]  Jun Wang,et al.  Passivity of Switched Recurrent Neural Networks With Time-Varying Delays , 2015, IEEE Transactions on Neural Networks and Learning Systems.

[40]  Tianping Chen,et al.  Coexistence and local stability of multiple equilibria in neural networks with piecewise linear nondecreasing activation functions , 2010, Neural Networks.

[41]  Jinde Cao,et al.  Multistability in bidirectional associative memory neural networks , 2008 .

[42]  Zhigang Zeng,et al.  Multistability of Recurrent Neural Networks With Nonmonotonic Activation Functions and Unbounded Time-Varying Delays , 2018, IEEE Transactions on Neural Networks and Learning Systems.

[43]  Lin-Bao Yang,et al.  Cellular neural networks: theory , 1988 .

[44]  Jinde Cao,et al.  Multistability of Second-Order Competitive Neural Networks With Nondecreasing Saturated Activation Functions , 2011, IEEE Transactions on Neural Networks.

[45]  Freeman J. Dyson,et al.  DISTRIBUTION OF EIGENVALUES FOR A CLASS OF REAL SYMMETRIC MATRICES. , 1971 .

[46]  Zhigang Zeng,et al.  Multiple Mittag–Leffler Stability of Fractional-Order Recurrent Neural Networks , 2017, IEEE Transactions on Systems, Man, and Cybernetics: Systems.

[47]  Jinde Cao,et al.  Stability in delayed Cohen–Grossberg neural networks: LMI optimization approach , 2005 .

[48]  Wei Wu,et al.  Global exponential stability of Cohen-Grossberg neural networks with distributed delays , 2008, Math. Comput. Model..

[49]  Fang Liu,et al.  Exponential Stability Analysis for Neural Networks With Time-Varying Delay , 2008, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).

[50]  Jinde Cao,et al.  Multistability of Memristive Neural Networks with Non-monotonic Piecewise Linear Activation Functions , 2015, ISNN.

[51]  Jun Zhao,et al.  Dissipativity based stability of switched systems with state-dependent switchings , 2007, 2007 46th IEEE Conference on Decision and Control.

[52]  Zhigang Zeng,et al.  Stability analysis of delayed cellular neural networks described using cloning templates , 2004, IEEE Trans. Circuits Syst. I Regul. Pap..

[53]  D. Hill,et al.  A notion of passivity for switched systems with state-dependent switching , 2006 .

[54]  Jinde Cao,et al.  Multistability of competitive neural networks with time-varying and distributed delays , 2009 .

[55]  Zhigang Zeng,et al.  Multistability analysis of a general class of recurrent neural networks with non-monotonic activation functions and time-varying delays , 2016, Neural Networks.

[56]  Zhengrong Xiang,et al.  Stability and $${L_{\infty }}$$L∞-Gain Analysis for Positive Switched Systems with Time-Varying Delay Under State-Dependent Switching , 2016, Circuits Syst. Signal Process..

[57]  Lingjie Chen,et al.  Stability and Stabilization of a Class of Multimode Linear Discrete-Time Systems With Polytopic Uncertainties , 2009, IEEE Transactions on Industrial Electronics.

[58]  Zhigang Zeng,et al.  Multistability of Delayed Recurrent Neural Networks with Mexican Hat Activation Functions , 2017, Neural Computation.