Showing papers by "Minyue Fu published in 2011"

Distributed Consensus With Limited Communication Data Rate

Abstract: Communication data rate and energy constraints are important factors which have to be considered when investigating distributed coordination of multi-agent networks. Although many proposed average-consensus protocols are available, a fundamental theoretic problem remains open, namely, how many bits of information are necessary for each pair of adjacent agents to exchange at each time step to ensure average consensus? In this paper, we consider average-consensus control of undirected networks of discrete-time first-order agents under communication constraints. Each agent has a real-valued state but can only exchange symbolic data with its neighbors. A distributed protocol is proposed based on dynamic encoding and decoding. It is proved that under the protocol designed, for a connected network, average consensus can be achieved with an exponential convergence rate based on merely one bit information exchange between each pair of adjacent agents at each time step. An explicit form of the asymptotic convergence rate is given. It is shown that as the number of agents increases, the asymptotic convergence rate is related to the scale of the network, the number of quantization levels and the ratio of the second smallest eigenvalue to the largest eigenvalue of the Laplacian of the communication graph. We also give a performance index to characterize the total communication energy to achieve average consensus and show that the minimization of the communication energy leads to a tradeoff between the convergence rate and the number of quantization levels.

Identification of ARMA models using intermittent and quantized output observations

Abstract: This paper studies system identification of ARMA models whose outputs are subject to finite-level quantization and random packet dropouts. A simple adaptive quantizer and the corresponding recursive identification algorithm are proposed and shown to be optimal in the sense of asymptotically achieving the minimum mean square estimation error. The joint effects of finite-level quantization and random packet dropouts on identification accuracy are exactly quantified. The theoretic results are verified by simulations.

Preview Control of Dual-Stage Actuator Systems for Superfast Transition Time

Abstract: This paper introduces a preview control design method to reduce the settling time of dual-stage actuators (DSAs). A DSA system is comprised of two actuators connected in series, a primary (coarse) actuator, and a secondary (fine) actuator. The objective of the proposed design is to account for the redundancy of actuators and use the information of future reference levels in order to compute a pair of inputs to be applied before the output transition time. Experimental results show that the proposed design method significantly reduces the output transition time when compared to conventional forms of DSA control design.

Kalman filtering with intermittent observations: Bounds on the error covariance distribution

Abstract: When measurements are subject to random losses, the covariance of the estimation error of a state estimator becomes a random variable. In this paper we present bounds on the cumulative distribution function of the covariance of the estimation error for a discrete time linear system. We also show that the bounds can be arbitrarily tight if sufficient computational power is available. Numerical simulations show that the proposed method provides tighter bounds than the ones available in the literature.

Kalman filtering for a class of degenerate systems with intermittent observations

Abstract: This paper addresses the performance of a Kalman filter when measurements are intermittently available, i.e., network transmission problems. More specifically, we present a method to determine whether the expected value of the estimation error covariance is bounded for a given stochastic network model. The method applies to very general network models and for a class of degenerate systems. It can be easily adapted to non-degenerate systems, recovering known results on the critical value. The main result follows from the convergence conditions on a series that describes the bounds on the expected error covariance.

Hybrid Filterbank ADCs With Blind Filterbank Estimation

Abstract: The hybrid filterbank architecture permits implementing accurate, high speed analog-to-digital converters. However, its design requires an accurate knowledge of the analog filterbank parameters, which is difficult to have due to the nonstationary nature of these parameters. This paper proposes a blind estimation method for the analog filterbank parameters, which is able to cope with nonstationary input signals. This is achieved by using the notion of averaged input spectrum. The estimated parameters are used to reconstruct the samples in a least mean squares (LMS) sense. The proposed LMS design generalizes existing approaches by dropping the bandlimited assumption on the input signal. Instead, it assumes that the input has an arbitrary power spectrum which is adaptively estimated. Numerical experiments are presented showing the good performance of the blind estimation stage and the clear advantage of the proposed LMS design.

A unified framework for mean square stability of Kalman filters with intermittent observations

Abstract: This paper presents a unified framework for analysis of the stability of the expected error covariance (EEC) of Kalman filters subject to intermittent observations. A brief literature review summarizing some of the most important results in the area is provided. We state a method in the most general form, making no assumptions on the network model and only minor assumptions on the system. Then, as we adopt particular network models and assume some particular system structures, we recover most of the known results in the literature, which can be seen as a special case of our approach. Tight necessary and sufficient conditions for the EEC to be bounded are given for most cases, except for general degenerate systems, where only sufficient conditions are given in a closed form.

Optimal PMU placement for power system state estimation with random communication packet losses

Abstract: Phasor measurement units (PMUs) become important to state estimation for power systems by providing globally synchronized measurements of real-time phasors of voltage and currents with a high sampling rate. However the large quantities of measurement data produced by PMUs brings a serious burden to the communication system, which aggravates communication constraints such as the packet loss rate. In this paper, a novel optimization criterion for choosing PMU placements is proposed, considering random communication packet losses. Based on this criterion, a simplified optimal solution searching algorithm is given. Finally numerical simulations are given to test the validity of this algorithm. The dependence of the optimal PMU placement solution on the packet loss rate is indicated as well.

Power system dynamic state estimation with random communication packets loss

Abstract: In this paper power system dynamic state estimation problem is studied considering random communication packet losses. Two sorts of stochastic processes, i.i.d. process and Markov process, are respectively utilized to model two different communication packet losses cases. The first case only considers packet losses rate, and the second case includes both of packet losses rate and recovery rate. The degradation of the performance of state estimation caused by communication packet losses is analyzed on IEEE 14 buses test system, and numerical results are given.

Quantized control for stochastic system

Abstract: This paper considers the coarsest quantization control problem. Different from the previous works [4] where the systems are restricted to be deterministic, we focus on the feedback quantization control for general stochastic systems with multiplicative noises. It is showed that the coarsest quantizer that stabilize the multiplicative-noise stochastic system in mean-square sense is logarithmic, and the quantization density of which is larger than the results obtained in [4] for deterministic systems for the deterioration of multiplicative noises. Also, we explore that the solvability of the quantizer density is related to a special stochastic linear control problem.

Horizontal gene transfer in herpesviruses identified by using support vector machine.

Abstract: Horizontal gene transfer (HGT) is the probable origin of new genes. Identification of HGT-introduced genes would be helpful to the understanding of the genome evolution and the function prediction of new genes. In this study, a method using support vector machine (SVM) was used to distinguish horizontally transferred genes and non-horizontally transferred genes of mammalian herpesviruses based on the atypical composition identification, with accuracy higher than 95% within a reasonable length of time by using just a common PC. This identified 302 putative horizontally transferred genes, 171 genes being identified for the first time. Although most putative transferred genes are of unknown function, many genes have been discovered or predicted to encode glycoproteins or membrane proteins.

Improved servomechanism control design - Dynamically damped case

Abstract: Time optimal control (TOC) for servomechanism is not a practical controller due to the chattering phenomenon that occurs on the presence of noise and model uncertainty. Maybe the most popular attempt to transform this controller in a practical one comes from the so called Proximate Time Optimal Servomechanism (PTOS). This approach starts with a near time optimal controller and then switches to a linear controller when the system output approaches the target. While the chattering phenomenon is avoided, this comes at an expense in performance generated by the so called “acceleration discount factor”. This paper will present a controller that makes use of dynamic damping in order to push the acceleration discount factor arbitrarily close to one, thus practically eliminating the conservatism present in the PTOS. Experimental results support the proposed design.

Delay modeling and estimation of a wireless based network control system

Abstract: In this paper, a wireless based network control system is considered. Poisson process models are proposed to approximate the transmission delays for both single-hop and multi-hop wireless transmissions. State estimation is done in the real wireless network and a simulated network using the Poisson process model. The estimate errors in the two cases match well, which verifies the validity of the delay model.

High-bandwidth control design for a piezoelectric nanopositioning stage

Abstract: To achieve fast and accurate tracking of a wideband reference trajectory, the piezoelectric (PZT) actuator requires a high-bandwidth control system, which is however restricted by the resonant mode of the PZT positioning stage. In this paper, we study two resonant compensation techniques to damp the resonant mode for increased servo bandwidth. First, we present a feedback control system using a conventional notch filter (NF). Subsequently, we develop a complex lead compensator (CLC) using the phase-stabilized compensation method. Unlike the NF that is aimed at reducing the resonant peak gain, the CLC specializes in shaping the phase of open-loop system at the resonant frequency. The analysis shows that the closed-loop bandwidth achieved by the CLC is around four times higher than that of the NF without sacrificing the stability margin. Finally, we propose a multi-resonant filter (MRF) to suppress periodic tracking errors by significantly attenuating the gains at specified frequencies in the sensitivity function. The experimental results verify that the CLC is superior in disturbance compensation and periodic trajectory tracking as compared to the NF, and the add-on MRF can greatly reduce the tracking error.