Search

In order to cooperate the wind farm operators with grasping the operation status of wind power converter, a novel reliability assessment strategy is proposed based on supervisory control and data acquisition (SCADA) multistate parameters prediction of permanent magnet synchronous generator (PMSG) wind turbine. The strategy considers “off-line training, on-line matching and assessment”. The operation reliability of wind power converter is obtained via the analysis and weight computing of confidence level, prediction value and actual value of SCADA multistate parameters. In the “off-line training” part, first, the FP-Growth association algorithm is employed to analyze the confidence levels of SCADA variables to the faults of wind power converter. The variables with high confidence level are defined as SCADA multistate parameters. Afterwards, wavelet packet transform (WPT) and K-means algorithm are employed to decompose, reconstruct, normalize and cluster the time series (off-line data) of multistate parameters under normal operation of wind turbine, to improve the generalization capability of long short term memory (LSTM) prediction model. In the part of “on-line matching and assessment”, the actual value time series (on-line data) of multistate parameters are decomposed and reconstructed by WPT. Each normalized sample is matched to the closest cluster centroid, which is generated in the part of “off-line training”. Then the corresponding LSTM model is conducted to predict based on the each sample. The final prediction value is sum of prediction results of entire samples in clusters. Finally, the reliability of wind power converter is assessed by the proposed strategy. The effectiveness of proposed assessment strategy is verified by the experimental results on a PMSG wind power converter.

Fabric defect detection plays an important role in automated inspection and quality control in textile manufacturing. As the fabric images have complex and diverse textures and defects, traditional detection methods show a poor adaptability and low detection accuracy. Robust principal component analysis (RPCA) model that can be used to separate the image into object and background have proven applicable in fabric defect detection. However, how to represent texture feature of the fabric image more effectively is still problematic in this kind of method. In addition, the use of the traditional RPCA may result in low accuracy and more noises in sparse part. In this article, a novel fabric defect detection method based on multilevel deep features fusion and non-convex total variation regularized RPCA (NTV-RPCA) is proposed. Firstly, the image representation ability is well enhanced through multilevel deep features extracted by a convolutional neural network. Then, the non-convex total variation regularized RPCA is proposed in which total variation constraint significantly reduces the noises in sparse part and non-convex solution is more approximate to the authentic one. Next, multilevel saliency maps generated by the sparse matrixes are fused via RPCA to produce a more reliable detection result. Finally, the defect region is located by segmenting the fused saliency map via a threshold segmentation algorithm. Qualitative and quantitative experiments conducted on two public fabric image databases demonstrate that the proposed method improves the adaptability and detection accuracy comparing to the state-of-the-arts.

The larger capacity of power converters increases with the size of wind turbines over the years, which implies more failures occur in the components of power converter. Meanwhile, the strategies of fault diagnosis of power converter in wind turbine are still under discussion and research. The main purpose of this paper is proposing a strategy containing wavelet transform, feature analysis, judgment and back propagation neural network (BPNN) classification (WT-FA-JD-BP) to identify the single and double power components open-circuit faults accurately, happen in grid-side converter (GSC) of permanent magnet synchronous generator (PMSG) wind turbine systems. First, as the original signals, the three-phase bridge legs voltage of the GSC are collected under different faults. Second, wavelet transform is used to decompose and reconstruct the signals. Third, a new method includes feature analysis and judgment is conducted to amplify the divergence of data obtained by wavelet transform. Finally, the data are used as the inputs of BPNN for decision-making and classification. The simulation and experimental results show that the proposed strategy can classify the single and double open-circuit faults, and the accuracy is higher than that without data feature analysis and judgment.

In recent years, the unmanned aerial vehicles (UAVs) have exhibited significant market potential to greatly reduce the cost and time in the field of logistics. The use of UAVs to provide commercial courier has become an emerging industry, remarkably shifting the energy use of the freight sector. However, due to limited battery capacities, the flight duration of civilian rotorcraft UAVs is still short, hindering them from performing remote jobs. In this case, people customarily utilize ground vehicles to carry and assist UAVs in various applications, including cargo delivery. Most previous studies on vehicle-drone cooperative parcel delivery considered only one UAV, thereby suffering from low efficiency when serving a large number of customers. In this paper, we propose a novel hybrid genetic algorithm, which supports the cooperation of a ground vehicle and multiple UAVs for efficient parcel delivery. Our routing and scheduling algorithm allows multiple UAVs carried by the vehicle to simultaneously deliver multiple parcels to customers residing in different locations. The proposed algorithm consists of a pipeline of several modules: population management, heuristic population initialization, and population education. The performance evaluation results show that the proposed algorithm has significant efficiency over existing algorithms.

Sparse random linear network coding (SRLNC) is a promising solution for reducing the complexity of random linear network coding (RLNC). RLNC can be modeled as a linear operator channel (LOC). It is well known that the normalized channel capacity of LOC is characterized by the rank distribution of the transfer matrix. In this paper, we study the rank distribution of SRLNC. By exploiting the definition of linear dependence of the vectors, we first derive a novel approximation to the probability of a sparse random matrix being non-full rank. By using the Gauss coefficient, we then provide a closed approximation to the rank distribution of a sparse random matrix over a finite field. The simulation and numerical results show that our proposed approximation to the rank distribution of sparse matrices is very tight and outperforms the state-of-the-art results, except for the finite field size and the number of input packets are small, and the sparsity of the matrices is large.

The Internet of Things (IoT) is widely regarded as a key component of the Internet of the future and thereby has drawn significant interests in recent years. IoT consists of billions of intelligent and communicating “things”, which further extend borders of the world with physical and virtual entities. Such ubiquitous smart things produce massive data every day, posing urgent demands on quick data analysis on various smart mobile devices. Fortunately, the recent breakthroughs in deep learning have enabled us to address the problem in an elegant way. Deep models can be exported to process massive sensor data and learn underlying features quickly and efficiently for various IoT applications on smart devices. In this article, we survey the literature on leveraging deep learning to various IoT applications. We aim to give insights on how deep learning tools can be applied from diverse perspectives to empower IoT applications in four representative domains, including smart healthcare, smart home, smart transportation, and smart industry. A main thrust is to seamlessly merge the two disciplines of deep learning and IoT, resulting in a wide-range of new designs in IoT applications, such as health monitoring, disease analysis, indoor localization, intelligent control, home robotics, traffic prediction, traffic monitoring, autonomous driving, and manufacture inspection. We also discuss a set of issues, challenges, and future research directions that leverage deep learning to empower IoT applications, which may motivate and inspire further developments in this promising field.

Rate Compatible Modulation (RCM) has been shown to be capable of achieving smooth rate adaptive transmission in highly dynamic channel scenarios. In this paper, the capacity of RCM is analyzed for different channel conditions, and an important result is obtained, i.e., in low-to-middle signal to noise ratio (SNR), the same achievable capacity in RCMs may be obtained for different weight sets being used. Based on this result, instead of employing fixed-weight sets with large values or big cardinalities for all channel conditions as in conventional RCM, a novel RCM scheme with variable weight sets (RCM-VWS) is proposed by employing different weight sets for different SNR ranges such that the demodulation complexity can be significantly reduced. In RCM-VWS, the SNR value of the received signal is divided into five unoverlapped ranges, which are estimated and encoded into 3-bit information at the receiver. This 3-bit information is delivered back to the transmitter for weight set selection when the SNR value of the received signal changes from one range to another or for some transmission interval. Thus, the cost of the feedback load is trivial in practice. Theoretical analysis and simulations show that compared with conventional RCM with fixed weight sets, the computational complexity of RCM-VWS can be reduced by as much as 97% at SNR

To reduce energy consumption and balance the resource load of physical machines (PMs) in cloud data centers, we present a game-based consolidation method of virtual machines (VMs) with energy and load (applied load) constraints. First, we test every measured value of the resource load using a t-test to filter outliers. Based on these values, the future resource load is forecast using gray theory. Second, all online PMs are grouped by the number of VMs on them and their future load values. Based on the groupings, a pre-processing algorithm for selecting destination PMs is proposed to determine a set of destination PMs for a VM awaiting migration. Finally, we select the final destination PM for the VM using game-based methods aimed at optimizing overall energy consumption. The experimental results show that our method can reduce energy consumption as well as balance loads without unnecessarily increasing the number of VM migrations.

With the increasing development of information technology and emerging of modern mobile devices, researchers have been shedding light on the integration of ubiquitous learning and online games in the construction of smart campus, aiming to enhance the online users' learning performance and perceptions. However, challenges of such blend learning model, for example, how to avoid the learners' distraction from online games, triggered many concerns from software engineers and educators. This paper investigated the factors and driving mechanism to construct a learner's technology engagement (TE) model towards ubiquitous game-based learning (UGL) in the smart campus context. Based on the technology acceptance model and the media system dependency theory, this paper identified the relationships among individual expectation (IE), facility environment, social environment (SE), and TE. A total of 168 participants, who were randomly assigned in the UGL settings, took part in the investigation. SPSS 19.0 was employed to analyze factorial validity, internal consistency reliability, correlation, and multiple regression analyses. The result shows that learners' TE toward UGL was influenced directly by IE (including fragmented learning and infotainment learning) and SE (including peer coaching and parental supports). In addition, the ease of use significantly moderated the fragmented learning and TE in the context of smart campus, while it was an insignificant moderator of infotainment. Besides, the ease of use significantly moderated the coaching of peers and TE, while it failed to moderate the parental supports and TEs. The theoretical and practical implications are finally discussed in terms of the development tendency of the construction of smart campus with UGL model.

Rate compatible modulation (RCM) has high spectrum efficiency and achieves seamless and blind rate adaptation in wide range of channel conditions. However, due to many convolution operations at symbol nodes, the belief propagation decoding algorithm of RCM has a high level of computational complexity. In this paper, we investigate the low complexity algorithms for fast decoding of RCM. Instead of computing the outgoing messages at symbol nodes via multi-level convolutions, we first design a novel two-level computing structure (2L-RCM) for symbol nodes in the probability-domain, each level is composed of one set of multiplications followed by one set of additions. Based on 2L-RCM, we derive Log-2L-RCM decoding algorithm in the log-domain, which converts the multiplications and additions of 2L-RCM into additions and Jacobian logarithms, respectively. Furthermore, we propose some approximate algorithms to reduce the complexity of Jacobian logarithms. In particular, the improved Max-Log-RCM (IMax-Log-RCM) algorithm obtains good performance-complexity trade-off. The simulation results and the numerical analyses show that IMax-Log-RCM achieves only 0.3-dB worse decoding performance than the original decoding algorithm with much fewer additions.

The background of our research is, designing a precise phase detection sensor for high-speed coherent homodyne systems, and combining it with optical phase-locked loops (OPLLs) to achieve coherent demodulation and carrier recovery. For existing phase sensors, there are problems, such as low phase gain, long dead-time, and difficulty in distinguishing directions. Therefore, we proposed a multilevel-loop compound control phase sensor based on exclusive-OR gates. The sensing system consists of three loops. The tunable range of the temperature loop is 30 GHz. Its tuning rate value is 50 MHz/s. In addition, the inner ring consists of a piezoelectric loop and an acousto-optic frequency shifter loop. According to demand, we established a relevant model, and analyzed and calculated the parameters. The reliability and validity of this sensor are substantiated after setting up the experimental environment. The results show that the lock-in range is 30 GHz, the bandwidth is 1.5 MHz, and the real-time performance of the PLL is improved. In a communication test with a rate of 5 Gbps, the realized bit error ratio is 1.55 × 10-8 when the optical power of the signal is -40.4 dBm. Therefore, this sensor system improves detection sensitivity. At the end of this paper, we propose some ideas to improve sensitivity, which will be the focus of our next phase.