SYSU-CMU International Joint Research Institute

Redundancy resolution is a critical problem in the control of robotic manipulators. Recurrent neural networks (RNNs), as inherently parallel processing models for time-sequence processing, are potentially applicable for the motion control of manipulators. However, the development of neural models for high-accuracy and real-time control is a challenging problem. This paper identifies two limitations of the existing RNN solutions for manipulator control, i.e., position error accumulation and the convex restriction on the projection set, and overcomes them by proposing two modified neural network models. Our method allows nonconvex sets for projection operations, and control error does not accumulate over time in the presence of noise. Unlike most works in which RNNs are used to process time sequences, the proposed approach is model-based and training-free, which makes it possible to achieve fast tracking of reference signals with superior robustness and accuracy. Theoretical analysis reveals the global stability of a system under the control of the proposed neural networks. Simulation results confirm the effectiveness of the proposed control method in both the position regulation and tracking control of redundant PUMA 560 manipulators.

In this work, the human parsing task, namely decomposing a human image into semantic fashion/body regions, is formulated as an active template regression (ATR) problem, where the normalized mask of each fashion/body item is expressed as the linear combination of the learned mask templates, and then morphed to a more precise mask with the active shape parameters, including position, scale and visibility of each semantic region. The mask template coefficients and the active shape parameters together can generate the human parsing results, and are thus called the structure outputs for human parsing. The deep Convolutional Neural Network (CNN) is utilized to build the end-to-end relation between the input human image and the structure outputs for human parsing. More specifically, the structure outputs are predicted by two separate networks. The first CNN network is with max-pooling, and designed to predict the template coefficients for each label mask, while the second CNN network is without max-pooling to preserve sensitivity to label mask position and accurately predict the active shape parameters. For a new image, the structure outputs of the two networks are fused to generate the probability of each label for each pixel, and super-pixel smoothing is finally used to refine the human parsing result. Comprehensive evaluations on a large dataset well demonstrate the significant superiority of the ATR framework over other state-of-the-arts for human parsing. In particular, the F1-score reaches 64.38 percent by our ATR framework, significantly higher than 44.76 percent based on the state-of-the-art algorithm [28].

A microgrid is a promising approach to provide clean, renewable, and reliable electricity by integrating various distributed generations and energy storage systems into power systems. However, highly intermittent renewable generations and various load demands pose new challenges to the optimal resource management in a microgrid. This paper proposes a fully distributed control strategy based on the consensus algorithm for the optimal resource management in an islanded microgrid. The proposed strategy is implemented through a multiagent system framework, which only requires information exchange among neighboring agents through a local network. The objective is achieved through a two-level control strategy. The upper control level is a consensus-based optimization algorithm that discovers the reference of optimal power generation or demand while maintaining the supply-demand balance. The lower control level is responsible for reference tracking of the associated component. Simulation results in the IEEE 14- and 162-bus systems are presented to demonstrate the effectiveness of the proposed control strategy.

The atypical face scanning patterns in individuals with Autism Spectrum Disorder (ASD) has been repeatedly discovered by previous research. The present study examined whether their face scanning patterns could be potentially useful to identify children with ASD by adopting the machine learning algorithm for the classification purpose. Particularly, we applied the machine learning method to analyze an eye movement dataset from a face recognition task [Yi et al., 2016], to classify children with and without ASD. We evaluated the performance of our model in terms of its accuracy, sensitivity, and specificity of classifying ASD. Results indicated promising evidence for applying the machine learning algorithm based on the face scanning patterns to identify children with ASD, with a maximum classification accuracy of 88.51%. Nevertheless, our study is still preliminary with some constraints that may apply in the clinical practice. Future research should shed light on further valuation of our method and contribute to the development of a multitask and multimodel approach to aid the process of early detection and diagnosis of ASD. Autism Res 2016, 9: 888-898. © 2016 International Society for Autism Research, Wiley Periodicals, Inc.

Matrix inversion often arises in the fields of science and engineering. Many models for matrix inversion usually assume that the solving process is free of noises or that the denoising has been conducted before the computation. However, time is precious for the real-time-varying matrix inversion in practice, and any preprocessing for noise reduction may consume extra time, possibly violating the requirement of real-time computation. Therefore, a new model for time-varying matrix inversion that is able to handle simultaneously the noises is urgently needed. In this paper, an integration-enhanced Zhang neural network (IEZNN) model is first proposed and investigated for real-time-varying matrix inversion. Then, the conventional ZNN model and the gradient neural network model are presented and employed for comparison. In addition, theoretical analyses show that the proposed IEZNN model has the global exponential convergence property. Moreover, in the presence of various kinds of noises, the proposed IEZNN model is proven to have an improved performance. That is, the proposed IEZNN model converges to the theoretical solution of the time-varying matrix inversion problem no matter how large the matrix-form constant noise is, and the residual errors of the proposed IEZNN model can be arbitrarily small for time-varying noises and random noises. Finally, three illustrative simulation examples, including an application to the inverse kinematic motion planning of a robot manipulator, are provided and analyzed to substantiate the efficacy and superiority of the proposed IEZNN model for real-time-varying matrix inversion.

Energy storage systems (ESSs) are often proposed to support the frequency control in microgrid systems. Due to the intermittency of the renewable generation and constantly changing load demand, the charging/discharging of various ESSs in an autonomous microgrid needs to be properly coordinated to ensure the supply-demand balance. Recent research has discovered that the charging/discharging efficiency of ESSs has remarkable dependence on the charging/discharging rate and state-of-charge of the ESS. This paper proposes a distributed cooperative control strategy for coordinating the ESSs to maintain the supply-demand balance and minimize the total power loss associated with charging/discharging inefficiency. The effectiveness of the proposed approach is validated by simulation results.

In this study, we propose a novel distributed secondary control scheme for both voltage and frequency in autonomous microgrids. By incorporating predictive mechanisms into distributed generations, the secondary voltage control is converted to a tracker consensus problem of distributed model predictive control, with the synchronous convergence procedure for voltage magnitudes to the reference value drastically accelerated at a low communication cost. A sufficient local stability condition with the parameter analysis is established. Thus, a distributed proportional integral method combined with a finite-time observer to estimate the global reference information is presented in the frequency restoration while maintaining accurate active power sharing. Our approach accommodates model uncertainty, plug-and-play capability, and especially robustness against information update intervals, which is essential when the conventional method probably yields toward a poor performance. Meanwhile, the distributed architecture implemented on the local and neighboring information allows for a sparse communication network and eliminates the requirement for a centralized controller. Simulation results are provided to verify the effectiveness of the proposed control methodology.

Although many successful ensemble clustering approaches have been developed in recent years, there are still two limitations to most of the existing approaches. First, they mostly overlook the issue of uncertain links, which may mislead the overall consensus process. Second, they generally lack the ability to incorporate global information to refine the local links. To address these two limitations, in this paper, we propose a novel ensemble clustering approach based on sparse graph representation and probability trajectory analysis. In particular, we present the elite neighbor selection strategy to identify the uncertain links by locally adaptive thresholds and build a sparse graph with a small number of probably reliable links. We argue that a small number of probably reliable links can lead to significantly better consensus results than using all graph links regardless of their reliability. The random walk process driven by a new transition probability matrix is utilized to explore the global information in the graph. We derive a novel and dense similarity measure from the sparse graph by analyzing the probability trajectories of the random walkers, based on which two consensus functions are further proposed. Experimental results on multiple real-world datasets demonstrate the effectiveness and efficiency of our approach.

In this paper, a new Taylor-type numerical differentiation formula is first presented to discretize the continuous-time Zhang neural network (ZNN), and obtain higher computational accuracy. Based on the Taylor-type formula, two Taylor-type discrete-time ZNN models (termed Taylor-type discrete-time ZNNK and Taylor-type discrete-time ZNNU models) are then proposed and discussed to perform online dynamic equality-constrained quadratic programming. For comparison, Euler-type discrete-time ZNN models (called Euler-type discrete-time ZNNK and Euler-type discrete-time ZNNU models) and Newton iteration, with interesting links being found, are also presented. It is proved herein that the steady-state residual errors of the proposed Taylor-type discrete-time ZNN models, Euler-type discrete-time ZNN models, and Newton iteration have the patterns of O(h(3)), O(h(2)), and O(h), respectively, with h denoting the sampling gap. Numerical experiments, including the application examples, are carried out, of which the results further substantiate the theoretical findings and the efficacy of Taylor-type discrete-time ZNN models. Finally, the comparisons with Taylor-type discrete-time derivative model and other Lagrange-type discrete-time ZNN models for dynamic equality-constrained quadratic programming substantiate the superiority of the proposed Taylor-type discrete-time ZNN models once again.

In this paper, a new inequality-based criterion is proposed and investigated for obstacle avoidance of redundant robot manipulators at the joint-acceleration level. By incorporating such a dynamically updated inequality criterion and the joint physical constraints (i.e., joint-angle limits, joint-velocity limits, and joint-acceleration limits), a novel minimum-acceleration-norm (MAN) scheme is presented and investigated for robots' redundancy resolution. In addition, the resultant obstacle-avoidance MAN scheme resolved at the joint-acceleration level is reformulated as a general quadratic program (QP). Moreover, two important “Bridge” theorems are established, which show that such a QP problem can be equivalent to linear variational inequality (LVI) and then to piecewise-linear projection equation (PLPE). An LVI-based numerical method is thus developed and applied for online solution of the QP problem and the inequality-based obstacle-avoidance MAN scheme. Simulative results based on the PA10 robot manipulator in the presence of window-shaped and point obstacles further demonstrate the efficacy and superiority of the proposed acceleration-level inequality-based MAN scheme for obstacle avoidance of redundant robot manipulators. In addition, experimental verification conducted on a practical six-link planar robot manipulator substantiates the effectiveness and physical realizability of the proposed obstacle-avoidance scheme.

A de novo genome assembly algorithm for repeats and nonrepeats" is corrected as follows: The E-size [27] is designed to answer the question: if you choose a location (a base) in the reference genome at random, what is the expected size of the contig or scaffold containing that location?

We present a novel global stereo model designed for view interpolation. Unlike existing stereo models which only output a disparity map, our model is able to output a 3D triangular mesh, which can be directly used for view interpolation. To this aim, we partition the input stereo images into 2D triangles with shared vertices. Lifting the 2D triangulation to 3D naturally generates a corresponding mesh. A technical difficulty is to properly split vertices to multiple copies when they appear at depth discontinuous boundaries. To deal with this problem, we formulate our objective as a two-layer MRF, with the upper layer modeling the splitting properties of the vertices and the lower layer optimizing a region-based stereo matching. Experiments on the Middlebury and the Herodion datasets demonstrate that our model is able to synthesize visually coherent new view angles with high PSNR, as well as outputting high quality disparity maps which rank at the first place on the new challenging high resolution Middlebury 3.0 benchmark.

This paper aims at developing an integrated system of clothing co-parsing, in order to jointly parse a set of clothing images (unsegmented but annotated with tags) into semantic configurations. We propose a data-driven framework consisting of two phases of inference. The first phase, referred as "image co-segmentation", iterates to extract consistent regions on images and jointly refines the regions over all images by employing the exemplar-SVM (ESVM) technique [23]. In the second phase (i.e. "region colabeling"), we construct a multi-image graphical model by taking the segmented regions as vertices, and incorporate several contexts of clothing configuration (e.g., item location and mutual interactions). The joint label assignment can be solved using the efficient Graph Cuts algorithm. In addition to evaluate our framework on the Fashionista dataset [30], we construct a dataset called CCP consisting of 2098 high-resolution street fashion photos to demonstrate the performance of our system. We achieve 90.29% / 88.23% segmentation accuracy and 65.52% / 63.89% recognition rate on the Fashionista and the CCP datasets, respectively, which are superior compared with state-of-the-art methods.

The availability of many heterogeneous but related views of data has arisen in numerous clustering problems. Different views encode distinct representations of the same data, which often admit the same underlying cluster structure. The goal of multi-view clustering is to properly combine information from multiple views so as to generate high quality clustering results that are consistent across different views. Based on max-product belief propagation, we propose a novel multi-view clustering algorithm termed multi-view affinity propagation (MVAP). The basic idea is to establish a multi-view clustering model consisting of two components, which measure the within-view clustering quality and the explicit clustering consistency across different views, respectively. Solving this model is NP-hard, and a multi-view affinity propagation is proposed, which works by passing messages both within individual views and across different views. However, the exemplar consistency constraint makes the optimization almost impossible. To this end, by using some previously designed mathematical techniques, the messages as well as the cluster assignment vector computations are simplified to get simple yet functionally equivalent computations. Experimental results on several real-world multi-view datasets show that MVAP outperforms existing multi-view clustering algorithms. It is especially suitable for clustering more than two views.

In this paper, the discrete economic dispatch problem is formulated as a knapsack problem. An effective distributed strategy based on distributed dynamic programming algorithm is proposed to optimally allocate the total power demand among different generation units considering the generation limits and ramping rate limits. The proposed distributed strategy is implemented based on a multiagent system framework which only requires local computation and communication among neighboring agents. Thus, it enables the sharing of computational and communication burden among distributed agents. In addition, the proposed strategy can be implemented with asynchronous communication, which may lead to simpler implementation and faster convergence speed. Simulation results with a four-generator system and the IEEE 162-bus system are presented to demonstrate the effectiveness of the proposed distributed strategy.

Traditionally, economic dispatch and demand response (DR) are considered separately, or implemented sequentially, which may degrade the energy efficiency of the power grids. One important goal of optimal energy management (OEM) is to maximize the social welfare through the coordination of the suppliers' generations and customers' demands. Thus, it is desirable to consider the interactive operation of economic dispatch and DR, and solve them in an integrated way. This paper proposes a fully distributed online OEM solution for smart grids. The proposed solution considers the economic dispatch of conventional generators, DR of users, and operating conditions of renewable generators all together. The proposed distributed solution is developed based on a market-based self-interests motivation model since this model can realize the global social welfare maximization among system participants. The proposed solution can be implemented with multiagent system with each system participant assigned with an energy management agent. Based on the designed distributed algorithms for price updating and supply-demand mismatch discovery, the OEM among agents can be achieved in a distributed way. Simulation results demonstrate the effectiveness of the proposed solution.

Many of today's cloud datacenters are powered by electricity generated from brown energy (e.g., fuel fossil and oil), which directly translates into severe harm to the environment. To reduce the carbon footprint and the operating cost of datacenters, there is a clear trend to migrate to green datacenters, which are entirely (or partially) powered by renewable energy. However, given a portfolio with multiple off- and on-site power supplies (such as power grid, renewables, and batteries), it is challenging for datacenter operators to conduct efficient power management and thus meet the highly dynamic user demand. In this paper, we proposed an online algorithm, which is called EcoPower, to perform eco-aware power management and load scheduling jointly for geographically distributed cloud datacenters. Our objective is to minimize the time-average eco-aware power cost of cloud datacenters while still ensuring the quality-of-experience (QoE) constraint of user requests. To this end, we formulated the problem into a constrained stochastic optimization problem and apply the Lyapunov optimization theory to design an online control algorithm, which approaches the optimality with explicitly provable upper bounds. We also conducted extensive trace-driven simulations, and our results show that our proposed EcoPower algorithm can achieve a good balance between power cost savings, environment protection, and the user QoE, with the eco-aware power cost being cut down by over 20%. We found that wind dominant, solar complementary is a better strategy for cloud datacenters to integrate renewables into their power supply.

This letter presents a compact fractal loop rectenna for RF energy harvesting at GSM1800 bands. First, a fractal loop antenna with novel in-loop ground-plane impedance matching is proposed for the rectenna design. Also, a high-efficiency rectifier is designed in the loop antenna to form a compact rectenna. Measured results show that an efficiency of 61% and an output dc voltage of 1.8 V have been achieved over 12-kΩ resistor for 10 μW/cm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> power density at 1.8 GHz. The rectenna is able to power up a battery-less LCD watch at a distance of 10 m from the cell tower. The proposed rectenna is compact, easy to fabricate, and useful for various energy harvesting applications.

Traffic control at intersections is a key issue and hot research topic in intelligent transportation systems. Existing approaches, including traffic light scheduling and trajectory maneuver, are either inaccurate and inflexible or complicated and costly. More importantly, due to the dynamics of traffic, it is really difficult to obtain the optimal solution in a real-time way. Inspired by the emergence of vehicular ad hoc network, we propose a novel approach to traffic control at intersections. Via vehicle to vehicle or vehicle to infrastructure communications, vehicles can compete for the privilege of passing the intersection, i.e., traffic is controlled via coordination among vehicles. Such an approach is flexible and efficient. To realize the coordination among vehicles, we first model the problem as a new variant of the classic mutual exclusion problem, and then design algorithms to solve new problem. Both centralized and distributed algorithms are. We conduct extensive simulations to evaluate the performance of our proposed algorithms. The results show that, our approach is efficient and outperforms a reference algorithm based on optimal traffic light scheduling. Moreover, our approach does not rely on traffic light or intersection controller facilities, which makes it flexible and applicable to various kinds of intersections.

A novel planar end-fire circularly polarized (CP) complementary Yagi array antenna is proposed. The antenna has a compact and complementary structure, and exhibits excellent properties (low prolile, single feed, broadband, high gain, and CP radiation). It is based on a compact combination of a pair of complementary Yagi arrays with a common driven element. In the complementary structure, the vertical polarization is contributed by a microstrip patch Yagi array, while the horizontal polarization is yielded by a strip dipole Yagi array. With the combination of the two orthogonally polarized Yagi arrays, a CP antenna with high gain and wide bandwidth is obtained. With a prolile of 0.05λ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sub> (3 mm), the antenna has a gain of about 8 dBic, an impedance bandwidth (|S <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">11</sub> | <; -10 dB) of 13.09% (4.57-5.21 GHz) and a 3-dB axial-ratio bandwidth of 10.51% (4.69-5.21 GHz).