2,312 publications from this institution
The classical Kalman filtering (KF) algorithm has recently been extended to interval linear systems with interval parameters under the same statistical assumptions on noise, where the new algorithm is called Interval Kalman Filtering (IKF) scheme. The IKF algorithm has the same structure, and preserves the same optimality, as the classical KF scheme but provides interval-valued estimates. If the interval system has confidence description about the distribution of its interval values, we can further incorporate the IKF scheme with fuzzy logic inference, so as to develop a new filtering algorithm, called Fuzzy Kalman Filtering (FKF) algorithm. This algorithm preserves the same recursive mechanism of the KF and IKF, but produces a scalar-valued (rather than an interval-valued) estimate at each iteration of the filtering process. To compare the FKF to the IKF, computer simulation is included, which shows that the FKF is also robust against system parameter variations.
In this paper, a new adaptive control scheme, that employs a Gaussian radial basis function network with output weights updated on-line, is developed for regulating a class of uncertain chaotic systems. Theoretical analysis guarantees that the controlled uncertain chaotic systems can asymptotically track target orbits within arbitrarily small tolerance bounds. The uncertain Duffing-Holmes system is used as an example for illustration.
Capacity bounds are calculated for the differential chaos shift keying (DCSK) system with non-coherent detectors, including differentially coherent detector and energy detector, over additive white Gaussian noise (AWGN) and Rayleigh fading channels. Through theoretical analysis and numerical simulations, it is found that the capacity characteristics of the non-coherent DCSK system are significantly different from those of the coherent case. It is shown that there are different capacity bounds corresponding to different spreading factor values. Optimal code rates are found, which minimise the required bit signal-to-noise ratio for a reliable communication system, and they are increasing with the increase of the spreading factor value over an AWGN channel while decreasing over a Rayleigh fading channel. As compared with the AWGN channel, the performance over a Rayleigh fading channel is much more sensitive to the code rate. The new results are useful as benchmark for designing capacity-approaching codes for the non-coherent DCSK system.
Summary This paper introduces the motion‐planning approaches to solve the distributed consensus problems via sampling measurements. First, for first‐order multiagent systems, a class of sampled‐data–based algorithms are developed with arbitrary sampling periods, which solve the asymptotic consensus problem under both directed fixed and random switching topologies. Then, a new kind of distributed consensus algorithms is designed based on sampling measurements for second‐order multiagent systems. Under both the directed fixed and periodical switching topologies, asymptotic consensus problems of second‐order multiagent systems can be solved by using the proposed algorithms. Compared with existing continuous‐time consensus algorithms, one of remarkable advantages of proposed algorithms is that the sampling periods, communication topologies, and control gains are decoupled and can be separately designed, which relaxes many restrictions in controller designs. Finally, some numerical examples are given to illustrate the effectiveness of the analytical results.
<p style='text-indent:20px;'>The dynamics of a class of one-dimensional polynomial maps are studied, and interesting dynamics are observed under certain conditions: the existence of periodic points with even periods except for one fixed point; the coexistence of two attractors, an attracting fixed point and a hidden attractor; the existence of a double period-doubling bifurcation, which is different from the classical period-doubling bifurcation of the Logistic map; the existence of Li-Yorke chaos. Furthermore, based on this one-dimensional map, the corresponding generalized Hénon map is investigated, and some interesting dynamics are found for certain parameter values: the coexistence of an attracting fixed point and a hidden attractor; the existence of Smale horseshoe for a subshift of finite type and also Li-Yorke chaos.
No abstract is provided for this article.
No abstract is provided for this article.
No abstract is provided for this article.
No abstract is provided for this article.
In this paper, stability, robust stabilization and H ∞ control of singular-impulsive systems are studied. Some new fundamental properties are derived for switched singular systems subject to impulse effects. Applying the Lyapunov function theory, several sufficient conditions are established for exponential stability, robust stabilization and H ∞ control of the corresponding singular-impulsive closed-loop systems. Some numerical examples are given to demonstrate the effectiveness of the proposed control and stabilization methods.
The aim of this paper is to introduce the special session on Recent Advances in Complex Networks: Theories and Applications at ISCAS 2010 by giving a brief outline of the subject and presenting some recent advances in a field of our own interest, i.e., Control and Synchronization of Complex Networks, with special attention to adaptive synchronization and control strategies.
We study the effect of heterogeneous neuron and information transmission delay on stochastic resonance of scale-free neuronal networks. For this purpose, we introduce the heterogeneity to the specified neuron with the highest degree. It is shown that in the absence of delay, an intermediate noise level can optimally assist spike firings of collective neurons so as to achieve stochastic resonance on scale-free neuronal networks for small and intermediate αh, which plays a heterogeneous role. Maxima of stochastic resonance measure are enhanced as αh increases, which implies that the heterogeneity can improve stochastic resonance. However, as αh is beyond a certain large value, no obvious stochastic resonance can be observed. If the information transmission delay is introduced to neuronal networks, stochastic resonance is dramatically affected. In particular, the tuned information transmission delay can induce multiple stochastic resonance, which can be manifested as well-expressed maximum in the measure for stochastic resonance, appearing every multiple of one half of the subthreshold stimulus period. Furthermore, we can observe that stochastic resonance at odd multiple of one half of the subthreshold stimulus period is subharmonic, as opposed to the case of even multiple of one half of the subthreshold stimulus period. More interestingly, multiple stochastic resonance can also be improved by the suitable heterogeneous neuron. Presented results can provide good insights into the understanding of the heterogeneous neuron and information transmission delay on realistic neuronal networks.
Digital filtering techniques are widely used in many scientific and industrial endeavors such as digital telephony and communications, television and facsimile image processing, electrical and speech signal processing, radar, sonar, and space control systems. The...
This paper introduces an optimal fuzzy proportional-integral-derivative (PID) controller. The fuzzy PID controller is a discrete-time version of the conventional PID controller, which preserves the same linear structure of the proportional, integral, and derivative parts but has constant coefficient yet self-tuned control gains. Fuzzy logic is employed only for the design; the resulting controller does not need to execute any fuzzy rule base, and is actually a conventional PID controller with analytical formulae. The main improvement is in endowing the classical controller with a certain adaptive control capability. The constant PID control gains are optimized by using the multiobjective genetic algorithm (MOGA), thereby yielding an optimal fuzzy PID controller. Computer simulations are shown to demonstrate its improvement over the fuzzy PID controller without MOGA optimization.
