自动化学报最新文献

This paper discusses the sampled-data consensus problem of multi-agent systems with general linear dynamics and time-varying sampling intervals. To investigate the allowable upper bound of sampling intervals, we employ the property of discretization of sampled-data to identify the upper bound on the variable sampling intervals via a continuous-time model. Without considering the states in the sampling intervals, the decrease of Lyapunov function is guaranteed only at each sampling time. Consequently, it results in a more robust sampling interval which is obtained by verifying the feasibility of LMIs. Subsequently, provided a limited matrix variable, the control gain matrix K is solved by the LMI approach. Numerical simulations are provided to demonstrate the effectiveness of theoretical results.

This paper investigates the consensus control for multi-agent systems subject to external disturbances, input delays and model uncertainties of networks. By defining an appropriate controlled output, we transform this question into a robust H_∞ control problem. Then, we give two criteria to judge the consensusability of closed-loop multi-agent systems and present a cone-complementary linearization algorithm to get the state feedback controller's parameters. Finally, numerical examples are given to show the effectiveness of the proposed consensus protocols.

This paper investigates the consensus problem of discrete-time linear multi-agent systems (DLMASs) with directed switching information topologies and time-varying delays. First, we transform the consensus problem to an asymptotic stability problem of a corresponding time-delayed switched linear system (TDSLS) via a proper linear transformation. Then by using a constructed Lyapunov functional and the average dwell-time scheme, we establish a novel delay-dependent sufficient condition for the solvability of the consensus problem in terms of linear matrix inequalities (LMIs) for two cases, respectively: 1) all of the given information topologies are consensusable; 2) some of the given information topologies are consensusable. Finally, numerical examples are given to show the validness of the established results.

This paper investigates the distributed fusion Kalman filtering over clustering sensor networks. The sensor network is partitioned as clusters by the nearest neighbor rule and each cluster consists of sensing nodes and cluster-head. Using the minimax robust estimation principle, based on the worst-case conservative system with the conservative upper bounds of noise variances, two-level robust measurement fusion Kalman filter is presented for the clustering sensor network systems with uncertain noise variances. It can significantly reduce the communication load and save energy when the number of sensors is very large. A Lyapunov equation approach for the robustness analysis is presented, by which the robustness of the local and fused Kalman filters is proved. The concept of the robust accuracy is presented, and the robust accuracy relations among the local and fused robust Kalman filters are proved. It is proved that the robust accuracy of the two-level weighted measurement fuser is equal to that of the global centralized robust fuser and is higher than those of each local robust filter and each local weighted measurement fuser. A simulation example shows the correctness and effectiveness of the proposed results.

This paper considers the cooperative tracking of linear multi-agent systems with a dynamic leader whose input information is unavailable to any followers. Cooperative iterative learning controllers, based on the relative state information of neighboring agents, are proposed for tracking the dynamic leader over directed communication topologies. Stability and convergence of the proposed controllers are established using Lyapunov-Krasovskii functionals. Furthermore, this result is extended to the output feedback case where only the output information of each agent can be obtained. A local observer is constructed to estimate the unmeasurable states. Then, cooperative iterative learning controllers, based on the relative observed states of neighboring agents, are devised. For both cases, it is shown that the multi-agent systems whose communication topologies contain a spanning tree can reach synchronization with the dynamic leader, and meanwhile identify the unknown input of the dynamic leader using distributed iterative learning laws. An illustrative example is provided to verify the proposed control schemes.

Compressed sensing (CS) is a novel technology to acquire and reconstruct sparse signals below the Nyquist rate. It has great potential in image and video acquisition and processing. To effectively improve the sparsity of signal being measured and reconstructing efficiency, an encoding and decoding model of residual distributed compressive video sensing based on double side information (RDCVS-DSI) is proposed in this paper. Exploiting the characteristics of image itself in the frequency domain and the correlation between successive frames, the model regards the video frame in low quality as the first side information in the process of coding, and generates the second side information for the non-key frames using motion estimation and compensation technology at its decoding end. Performance analysis and simulation experiments show that the RDCVS-DSI model can rebuild the video sequence with high fidelity in the consumption of quite low complexity. About 1 ~ 5 dB gain in the average peak signal-to-noise ratio of the reconstructed frames is observed, and the speed is close to the least complex DCVS, when compared with prior works on compressive video sensing.

In this paper, we present a system for real-time performance-driven facial animation. With the system, the user can control the facial expression of a digital character by acting out the desired facial action in front of an ordinary camera. First, we create a muscle-based 3D face model. The muscle actuation parameters are used to animate the face model. To increase the reality of facial animation, the orbicularis oris in our face model is divided into the inner part and outer part. We also establish the relationship between jaw rotation and facial surface deformation. Second, a real-time facial tracking method is employed to track the facial features of a performer in the video. Finally, the tracked facial feature points are used to estimate muscle actuation parameters to drive the face model. Experimental results show that our system runs in real time and outputs realistic facial animations. Compared with most existing performance-based facial animation systems, ours does not require facial markers, intrusive lighting, or special scanning equipment, thus it is inexpensive and easy to use.

Unsupervised feature selection is fundamental in statistical pattern recognition, and has drawn persistent attention in the past several decades. Recently, much work has shown that feature selection can be formulated as nonlinear dimensionality reduction with discrete constraints. This line of research emphasizes utilizing the manifold learning techniques, where feature selection and learning can be studied based on the manifold assumption in data distribution. Many existing feature selection methods such as Laplacian score, SPEC (spectrum decomposition of graph Laplacian), TR (trace ratio) criterion, MSFS (multi-cluster feature selection) and EVSC (eigenvalue sensitive criterion) apply the basic properties of graph Laplacian, and select the optimal feature subsets which best preserve the manifold structure defined on the graph Laplacian. In this paper, we propose a new feature selection perspective from locally linear embedding (LLE), which is another popular manifold learning method. The main difficulty of using LLE for feature selection is that its optimization involves quadratic programming and eigenvalue decomposition, both of which are continuous procedures and different from discrete feature selection. We prove that the LLE objective can be decomposed with respect to data dimensionalities in the subset selection problem, which also facilitates constructing better coordinates from data using the principal component analysis (PCA) technique. Based on these results, we propose a novel unsupervised feature selection algorithm, called locally linear selection (LLS), to select a feature subset representing the underlying data manifold. The local relationship among samples is computed from the LLE formulation, which is then used to estimate the contribution of each individual feature to the underlying manifold structure. These contributions, represented as LLS scores, are ranked and selected as the candidate solution to feature selection. We further develop a locally linear rotation-selection (LLRS) algorithm which extends LLS to identify the optimal coordinate subset from a new space. Experimental results on real-world datasets show that our method can be more effective than Laplacian eigenmap based feature selection methods.

This paper provides a new adaptive algorithm for single-step prediction by modeling the potential field of a one dimension Schrödinger wave equation using neural network. This new architecture is referred to as the recurrent quantum neural network (RQNN). The RQNN can filter the signal embedded with non-stationary noise without any priori knowledge of the shape of the signal and statistical properties of the noise. We compared the simulation results of the RQNN with a classical adaptive stochastic filter-RLS. It is shown that the RQNN is much more efficient in denoising signals embedded with Gaussian stationary, non-Gaussian stationary and Gaussian non-stationary noise such as DC, sinusoid, staircase and speech signals. The RQNN can enhance the signal to noise rate (SNR) by 20 dB, which is more than 10 dB given by the traditional technology when it denoising sinusoid signal.

Abstract Interest point detection is a fundamental issue in many intermediate level vision problems and plays a significant role in vision systems. The previous interest point detectors are designed to detect some special image structures such as corners, junctions, line terminations and so on. These detectors based on some simplified 2D feature models will not work for image features that differ significantly from the models. In this paper, a scale invariant interest point detector, which is appropriate for most types of image features, is proposed based on an iterative method in the Gabor based energy space. It detects interest points by noting that there are some similarities in the phase domain for all types of image features, which are obtained by different detectors respectively. Firstly, this approach obtains the positions of candidate points by detecting the local maxima of a series of energy maps constructed by Gabor filter responses. Secondly, an iterative algorithm is adopted to select the corresponding characteristic scales and accurately locate the interest points simultaneously in the Gabor based energy space. Finally, in order to improve the real-time performance of the approach, a fast implementation of Gabor function is used to accelerate the process of energy space construction. Experiments show that this approach has a broader applicability than the other detectors and has a good performance under rotation and some other image changes.