Stability analysis of nonlinear digital systems under hardware overflow constraint for dealing with finite word-length effects of digital technologies

Abstract The purpose of this paper is to examine stability and originate stability criteria for nonlinear digital systems under the influence of saturation overflow, both in the absence and presence of external interference. The developed approaches can be employed to analyse overflow oscillation-free implementation of a nonlinear digital system under saturation overflow nonlinearity, caused by the finite word-length limitation of a digital hardware, such as computer processor or micro-controller. Asymptotic stability is examined in the absence of disturbance, whereas in the presence of external interference, the form of stability ensured is uniformly ultimately bounded stability, in which the states trajectories converge to an ellipsoidal region around the origin. In most of the studies reported so far, the authors have performed the overflow stability analysis of linear systems but very little (if any) work has been reported on the overflow oscillation elimination (for nonlinear systems). In the present work, sector conditions derived from saturation constraint along with Lipschitz condition are used with a suitable Lyapunov function for the stability analysis of nonlinear digital systems under overflow. The validity and efficacy of these criteria are tested by using examples from real nonlinear physical systems, including Moon chaotic system’s observer and recurrent neural network.

[1]  Ligang Wu,et al.  Model Approximation for Fuzzy Switched Systems With Stochastic Perturbation , 2015, IEEE Transactions on Fuzzy Systems.

[2]  Peng Shi,et al.  Two-Dimensional Dissipative Control and Filtering for Roesser Model , 2015, IEEE Transactions on Automatic Control.

[3]  Yves Rolain,et al.  Fast measurement of quantization distortions in DSP algorithms , 2004, Proceedings of the 21st IEEE Instrumentation and Measurement Technology Conference (IEEE Cat. No.04CH37510).

[4]  Riccardo Scattolini,et al.  Robust Stability Analysis of Nonlinear Discrete-Time Systems With Application to MPC , 2012, IEEE Transactions on Automatic Control.

[5]  Mario A. Rotea,et al.  Optimal realizations of finite wordlength digital filters and controllers , 1995 .

[6]  Hj Hans Butterweck,et al.  Finite wordlength effects in digital filters , 1989 .

[7]  Lennart Harnefors,et al.  Suppression of overflow limit cycles in LDI all-pass/lattice filters , 2000 .

[8]  Rajesh Rajamani,et al.  Real-Time Estimation of Rollover Index for Tripped Rollovers With a Novel Unknown Input Nonlinear Observer , 2014, IEEE/ASME Transactions on Mechatronics.

[9]  M. Nasir Uddin,et al.  Development and Implementation of a Nonlinear-Controller-Based IM Drive Incorporating Iron Loss With Parameter Uncertainties , 2009, IEEE Transactions on Industrial Electronics.

[10]  Yongduan Song,et al.  A novel approach to output feedback control of fuzzy stochastic systems , 2014, Autom..

[11]  Haranath Kar,et al.  Elimination of overflow oscillations in digital filters employing saturation arithmetic , 2005, Digit. Signal Process..

[12]  Vimal Singh,et al.  Stability analysis of 2-D digital filters described by the Fornasini-Marchesini second model using overflow nonlinearities , 2001 .

[13]  Wei Xing Zheng,et al.  New Stability Criterion for Fixed-Point State-Space Digital Filters With Generalized Overflow Arithmetic , 2012, IEEE Transactions on Circuits and Systems II: Express Briefs.

[14]  Choon Ki Ahn,et al.  Hankel Norm Performance of Digital Filters Associated With Saturation , 2017, IEEE Transactions on Circuits and Systems II: Express Briefs.

[15]  Tao Shen,et al.  Stability criterion for a class of fixed-point digital filters using two's complement arithmetic , 2013, Appl. Math. Comput..

[16]  Jun Wang,et al.  A general projection neural network for solving monotone variational inequalities and related optimization problems , 2004, IEEE Transactions on Neural Networks.

[17]  Haranath Kar,et al.  An improved criterion for the global asymptotic stability of fixed-point state-space digital filters with combinations of quantization and overflow , 2014, Digit. Signal Process..

[18]  Tingwen Huang,et al.  Exponential stabilization of delayed recurrent neural networks: A state estimation based approach , 2013, Neural Networks.

[19]  Jiahui Wang,et al.  Stability analysis of discrete-time switched nonlinear systems via T-S fuzzy model approach , 2016, Neurocomputing.

[20]  Choon Ki Ahn.  Overflow Oscillation Elimination of 2-D Digital Filters in the Roesser Model with Wiener Process Noise , 2014, IEEE Signal Processing Letters.

[21]  V. Singh.  Elimination of overflow oscillations in 2-D digital filters employing saturation arithmetic: an LMI approach , 2005, IEEE Signal Processing Letters.

[22]  Jun Wang,et al.  Real-time synthesis of linear state observers using a multilayer recurrent neural network , 1994, Proceedings of 1994 IEEE International Conference on Industrial Technology - ICIT '94.

[23]  Wei Xing Zheng,et al.  Dissipativity-Based Sliding Mode Control of Switched Stochastic Systems , 2013, IEEE Transactions on Automatic Control.

[24]  Vassilis Paliouras,et al.  Low-Power Logarithmic Number System Addition/Subtraction and Their Impact on Digital Filters , 2013, IEEE Trans. Computers.

[25]  C. Ahn.  Criterion for the elimination of overflow oscillations in fixed-point digital filters with saturation arithmetic and external disturbance , 2011 .

[26]  Ligang Wu,et al.  Stochastic stability analysis for 2-D Roesser systems with multiplicative noise , 2016, Autom..

[27]  Carl R. Knospe,et al.  Feedback linearization of an active magnetic bearing with voltage control , 2002, IEEE Trans. Control. Syst. Technol..

[28]  Vimal Singh,et al.  Stability analysis of 2-D digital filters with saturation arithmetic: an LMI approach , 2005, IEEE Transactions on Signal Processing.

[29]  Ligang Wu,et al.  Reliable Filter Design for Sensor Networks Using Type-2 Fuzzy Framework , 2017, IEEE Transactions on Industrial Informatics.

[30]  Muhammad Tufail,et al.  On elimination of overflow oscillations in linear time-varying 2-D digital filters represented by a Roesser model , 2016, Signal Process..

[31]  Weifeng Liu,et al.  Extended Kernel Recursive Least Squares Algorithm , 2009, IEEE Transactions on Signal Processing.

[32]  Haranath Kar.  An improved version of modified Liu-Michel's criterion for global asymptotic stability of fixed-point state-space digital filters using saturation arithmetic , 2010, Digit. Signal Process..

[33]  Vimal Singh,et al.  Robust stability of 2-D discrete systems described by the Fornasini-Marchesini second model employing quantization/overflow nonlinearities , 2004, IEEE Transactions on Circuits and Systems II: Express Briefs.

[34]  A. Michel,et al.  Stability analysis of state-space realizations for two-dimensional filters with overflow nonlinearities , 1994 .

[35]  Priyanka Kokil,et al.  An improved criterion for the global asymptotic stability of fixed-point state-space digital filters with saturation arithmetic , 2012, Digit. Signal Process..

[36]  Vimal Singh.  LMI Approach to Stability of Direct Form Digital Filters Utilizing Single Saturation Overflow Nonlinearity , 2007, IEEE Transactions on Circuits and Systems II: Express Briefs.

[37]  James Lam,et al.  Multi-Bound-Dependent Stability Criterion for Digital Filters With Overflow Arithmetics and Time Delay , 2014, IEEE Transactions on Circuits and Systems II: Express Briefs.

[38]  E. Dubois,et al.  Design of multidimensional finite-wordlength FIR and IIR filters by simulated annealing , 1995 .

[39]  Tian-Hua Liu,et al.  Design and Implementation of an Online Tuning Adaptive Controller for Synchronous Reluctance Motor Drives , 2013, IEEE Transactions on Industrial Electronics.

[40]  Ali Zemouche,et al.  Observer Design for Lipschitz Nonlinear Systems: The Discrete-Time Case , 2006, IEEE Transactions on Circuits and Systems II: Express Briefs.

[41]  Haranath Kar,et al.  An improved criterion for the global asymptotic stability of 2-D state-space digital filters with finite wordlength nonlinearities , 2014, Signal Process..

[42]  Peng Shi,et al.  Generalized Dissipativity Analysis of Digital Filters With Finite-Wordlength Arithmetic , 2016, IEEE Transactions on Circuits and Systems II: Express Briefs.

[43]  Narayanan Kumarappan,et al.  Day-Ahead Deregulated Electricity Market Price Forecasting Using Recurrent Neural Network , 2013, IEEE Systems Journal.

[44]  Haranath Kar.  Asymptotic stability of fixed-point state-space digital filters with combinations of quantization and overflow nonlinearities , 2011, Signal Process..

[45]  Huijun Gao,et al.  Extended State Observer-Based Sliding-Mode Control for Three-Phase Power Converters , 2017, IEEE Transactions on Industrial Electronics.

[46]  Seong-Whan Lee,et al.  A new recurrent neural-network architecture for visual pattern recognition , 1997, IEEE Trans. Neural Networks.

[47]  Bryan A. Tolson,et al.  A New Formulation for Feedforward Neural Networks , 2011, IEEE Transactions on Neural Networks.

[48]  Choon Ki Ahn.  $l_{2} - l_{\infty}$ Elimination of Overflow Oscillations in 2-D Digital Filters Described by Roesser Model With External Interference , 2013, IEEE Transactions on Circuits and Systems II: Express Briefs.

[49]  Panajotis Agathoklis,et al.  Coefficient sensitivity and structure optimization of multidimensional state-space digital filters , 1998 .

[50]  Yunlong Liu,et al.  Input-to-state stability for discrete-time nonlinear switched singular systems , 2016, Inf. Sci..

[51]  Nonperiodic modes in two-dimensional (2-D) recursive digital filters under finite wordlength effects , 1989 .

[52]  Choon Ki Ahn.  Two new criteria for the realization of interfered digital filters utilizing saturation overflow nonlinearity , 2014, Signal Process..

[53]  Zhong-Ping Jiang,et al.  On Uniform Global Asymptotic Stability of Nonlinear Discrete-Time Systems With Applications , 2006, IEEE Transactions on Automatic Control.

[54]  P K Houpt,et al.  Roundoff noise and scaling in the digital implementation of control compensators , 1983 .

[55]  Yongduan Song,et al.  Fault Detection Filtering for Nonlinear Switched Stochastic Systems , 2016, IEEE Transactions on Automatic Control.

[56]  Lorenz T. Biegler,et al.  Stability of a class of discrete-time nonlinear recursive observers , 2010 .

[57]  Lucas C. Cordeiro,et al.  Formal Non-Fragile Stability Verification of Digital Control Systems with Uncertainty , 2017, IEEE Transactions on Computers.

[58]  Vimal Singh.  Elimination of overflow oscillations in fixed-point state-space digital filters using saturation arithmetic: An LMI approach , 2006, Digit. Signal Process..

[59]  Jaeyong Chung,et al.  Bit-Width Optimization by Divide-and-Conquer for Fixed-Point Digital Signal Processing Systems , 2015, IEEE Transactions on Computers.

[60]  Vimal Singh,et al.  Robust stability of 2-D digital filters employing saturation , 2005, IEEE Signal Processing Letters.

[61]  Irza Arif,et al.  Toward Local Stability Analysis of Externally Interfered Digital Filters Under Overflow Nonlinearity , 2017, IEEE Transactions on Circuits and Systems II: Express Briefs.

[62]  Choon Ki Ahn,et al.  Some new results on the stability of direct-form digital filters with finite wordlength nonlinearities , 2015, Signal Process..

[63]  Peng Shi,et al.  Dissipativity analysis for fixed-point interfered digital filters , 2015, Signal Process..

[64]  Yutaka Maeda,et al.  Simultaneous perturbation learning rule for recurrent neural networks and its FPGA implementation , 2005, IEEE Transactions on Neural Networks.