Abstract:Ocean wave energy, which is much rich on the earth, is one of the highest density energy among all renewable energies such as wind and solar energy. However, traditional ocean energy harvesters are unpractical because of their complicated conversion structures, low transmission efficiency，and weak anti-corrosion and anti-shock capacities. Thus, a wave energy harvesting device for promoting the tran-sferring efficiency, reliability and stability, which is called direct energy harvester, is more and more emphasized and has become a research focus. This paper summarizes in detail the research status, characteristics, main classes, charging control strategies and power management systems of direct ocean wave harvesters. Moreover, the development tendencies of direct ocean wave harvesters are prospected.

Abstract:Traditional full-view 3D reconstruction systems are expensive and complex. To solve the above problems, a new algorithm for reconstructing an omnidirectional 3D scanning system is proposed based on a single Kinect and a rotating platform. The algorithm involves point cloud preprocessing, registration, minimizing global error between coordinate frames and color rectification. Firstly, RGB-D data are acquired by a Kinect sensor and preprocessed, and then a matching method is developed under the swivel table constraint for coarse registration. Further, the iterative closest point (ICP) algorithm is utilized for fine registration. Finally, for the loop closure caused by cumulative error and the color difference induced by different shooting angles, a global error correction and a color correction algorithm are conducted to improve the accuracy of the reconstruction results. Experimental results show that the reconstruction method can achieve full-view reconstruction of 3D objects and is superior to the KinectFusion method of Microsoft in accuracy.

Abstract:The modified compressive sampling matching pursuit(MCSMP) algorithm is proposed to solve the deficiency of the reconstruction of power quality disturbance signal based on compressive sampling matching pursuit(CoSaMP) algorithm. In the selection stage of the candidate sets, the MCSMP adopts the fuzzy threshold method instead of the fixed number compared with the CoSaMP, and uses the change of the correlation between the adjacent iterative sensing matrix and the residual error as the stop condition, which reduces the burden for the clipping of the backtrace process and the unnecessary iterations, and improves the efficiency of the algorithm. Simulation results show that: MCSMP algorithm is better than CoSaMP algorithm both in reconstruction performance and reconstruction time.

Abstract:In order to solve the problems that the traditional chaotic detection system is easily affected by the transition zone and the detection system is affected by the noise, an inverse phase transition chaos detection method based on high-order cumulants and Duffing oscillator is proposed. Firstly, the Lyapunov index method is used to calculate the critical threshold γ_d of the detection system, so that the the periodic driving force of the detection system is equal to γ_d. Secondly, the signal to be detected is preprocessed by calculating its high-order cumulant, which can reduce the noise power and obtain the amplitude variation of the harmonic signal. Then, the processed signal to be detected is input to the detection system, and the Lyapunov index is used to get the periodic driving force amplitudes corresponding to the reverse phase change. Finally, the amplitude of the signal to be detected and the detected signal-to-noise ratio are calculated according to the difference of the amplitude of the periodic driving force before and after the inverse phase transition. Simulation results show that the proposed method can be used for the detection of weak sinusoidal signals with a signal-to-noise ratio of -50.97 dB, which has a better detection effect than the traditional Duffing oscillator detection system.

Abstract:To improve the performance of BP decoding algorithm of digital fountain codes on wireless channels, a delta decoding algorithm based on the ripple set is proposed. The algorithm analyzes the likelihood ratio threshold of the variable nodes. When the likelihood ratio of the variable node is greater than the threshold, it can be successfully decoded in advance. On the other hand, when the overhead is increased, we can delete those variable nodes that have been decoded, and decode the nodes that have not achieved the decoding threshold, to further reduce the amount of calculation. The simulation shows that the performance of the new algorithm is as good as the traditional BP decoding algorithm, but the decoding efficiency is greatly improved.

Abstract:This paper investigates a wireless?powered relay network, in which a source transmits its signal to a destination with the aid of a decode?and?forward (DF) relay. Contrary to conventional DF relay networks, we consider the scenario that the DF relay has no sufficient embedded energy supply, and it is equipped with an energy harvesting unit and rechargeable battery. As such, it can accumulate the energy harvested from the source’s signals before help forwarding the information. A harvest—transmit—store (HTS) model with in?block time switching (TS) is developed such that TS can be implemented within one transmission block. By modeling the finite?capacity battery of the relay as a finite?state Markov Chain with a two?stage state transition, we derive the closed?form expression for the outage probability of the proposed HTS model with in?block TS over Nakagami?m fading channels. Numerical results validate our theoretical analysis and show that the proposed HTS model with in?block TS significantly outperforms the HTS model without in?block TS in terms of outage probability and successful rate.

Abstract:For up?link multi?user multiple?input multiple?output relay systems, with the aim of minimizing the bit error rate, a linear precoding scheme design is proposed based on the minimum mean?squared error (MMSE) rule in the presence of the feedback delay and channel estimation error. The proposed precoding algorithm decomposes the constrained optimization problem into three sub?convex optimization problems that are solved separately. The transmitter equipped with single antenna is subject to independent distribution, and it can be assumed that the precoding matrix is a diagonal matrix. The convex optimization problem of relay matrix is transformed into semi?definite programming problem (SDP), which can be solved by CVX toolbox. The equalizer matrix can be introduced by the gradient?based line search algorithms. Then the optimal solution is obtained by using joint iteration. The simulation results show that the proposed scheme can achieve better system performance.

Abstract:In order to adapt to demands in future wireless communication system including equipment diversity, high data rate, low latency and low power consumption,new waveform has become one of the keytechnologies studied in 5G wireless communication system. Universal filtered multicarrier (UFMC) is a widely researched 5G candidate waveform. However, the existing research only involves with UFMC?SISO scenario, and feasibility and performance of UFMC?MIMO remain vacant. Considering MIMO is a necessary scenario in 5G, it has great significance to study UFMC?MIMO. A feasible UFMC?MIMO system scheme including the transmitter, the receiver and the precoding realizing algorithm is proposed. It is proved that this UFMC?MIMO scheme can recover transmitted data correctly by mathematical derivation and its performance is verified by simulation. Simulation results show that UFMC?MIMO outperforms OFDM?MIMO no matter in AWGN channel or multipath channel, and it has great advantage over robustness to relative carrier frequency offset,which validates that UFMC?MIMO transmission scheme has the capacity to be utilized in 5G wireless communication system.

Abstract:Overlapping peaks and low amplitude peaks in peak detection stage will lead to a high false detection rate. Therefore, a peak enhancement step is added before the peak detection stage to improve the resolution of overlapping peaks and increase the amplitude of the low amplitude peaks. The method in the model is to combine the classical nonlinear diffusion with the derivative spectra. In other word, the signal after the derivative spectrum enhancement is used as the initial signal of the classical nonlinear diffusion, and the enhanced signal is the result of the diffusion. As a test of the proposed mode, the performance of signal enhancement by the proposed model is compared with non-enhancement one, then compared with the performance of other signal enhancement methods. Results show that the proposed model is effective. Finally, the proposed model is applied to the enhancement of MALDI mass spectrometry.

Abstract:To deal with the speech segmentation of TV-drama which has large coherent text transcriptions but no time-stamps, an automatic semi-supervised speech segmentation algorithm is proposed in the paper. Firstly, the original text transcriptions are used to build a biased language model, then the model is applied to the TV-drama speech recognition in a semi-supervised way, and finally, the resulting automatic speech decoding hypothesis are well combined with the traditional segmentation methods to improve the performances of speech segmentation. These traditional methods are usually based on the distance metric, model classification and the phone recognizers. Experimental results on the British TV-drama “Doctor Who” database demonstrate that, the proposed approach can achieve significant performance improvement over traditional baseline algorithms. Meanwhile, the proposed approach allows high quality segmentation and the associated transcription alignments for the large coherent TV-drama speech recordings.

Abstract:The random interfering noise contained in seismic record, if not removed properly, will inevitably pose a threat on the fracture development zone prediction accuracy because of greatly disturbing the key edge detection algorithm used in predicting step. Therefore, it is necessary to remove noise from seismic data and improve the quality of original seismic data. In this study, the independent component analysis(ICA) denoising technique, a blind source separation method, is used to decompose the seismic data into different levels of background and target reflection response of reservoir, and effectively make a distinction between effective signal and the random noise,which makes the processing result better than the conventional denoising algorithm. The processing ensures that the signal information basically does not suffer any losses and proffers a better lateral consistency in waveform characteristics. The field results show that, by applying the denoising method to the seismic data before edge detection, a robust fracture prediction result of fracture development zone distribution is achieved corresponding to the regional characteristics of fracture development, and it is in accordance with the drilling results of fracture development characteristics. This study improves the reliability of the fracture prediction of igneous rock zones.

Abstract:The conventional application of mud transmission rate (about 1 b/s) is unable to meet the demand for real?time uploading of measurement parameters with the rapid development of logging while drilling(LWD) technology, while data compression technology is an effective method to improve the transmission rate. By analyzing the characteristics of the mud pulse signal and the experience of the field application, a data compression algorithm based on frame-to-frame difference is designed and applied in data coding and analytic processing. The practical application results show that the algorithm can effectively improve the transmission rate and save the data’s updating time, thus having a high popularize prospect.

Abstract:Identifying signal peptide accurately is significant for protein research and localization. This paper presents a new method to extract high discriminant features for signal peptide sequence. Firstly, features based on compressed sensing are extracted by projecting a high-dimensional sequence onto a low-dimensional space, which remove redundant data while preserving the important information. And then dynamic time warping (DTW) algorithm is introduced to create the new features. The features extracted by the new method can reflect the important information of amino acid composition, sequence order and structure in the signal peptide, and also can nonlinearly align the different regions of signal peptide in the time dimension. Therefore the effective feature expression of the signal peptide for machine learning algorithm is provided. Experimental results show that the recognition accuracies with the extracted features are 99.65%, 98.05% and 98.56% respectively in the three datasets Eukaryotes, Gram+ bacteria and Gram- bacteria. Moreover, the new method can be simply applied to the identification of several biological sequences.

Abstract:Feature selection by picking a small size of important features out of the feature space facilitates learning algorithms to perform more accurately and more efficiently on the datasets. Considering the universal existence of relevance between features in real datasets, this paper proposes an unsupervised feature selection framework in which the feature correlating to each other form a network structure and the importance of each of them is measured by degree centrality index of a complex network. The bigger the degree centrality of a feature in this network, the higher the rank of its importance. At the end we select a given number of features with the highest ranks. This framework allows more flexibility on handling feature importance and feature redundancy. Later the proposed method will be compared to classical selection/extraction techniques on six high?dimensional datasets. Experiments demonstrate the advantages of our model on both continuous and discrete datasets.

Abstract:A key problem of language identification (LID) is how to design effective representations which are specific to language information. Recent advances in deep neural networks (DNNs) have led to significant improvements in language identification. The acoustic feature extracted from a structured DNN which is discriminative to phoneme or tri-phone states can significantly improve the performance. End-to-end schemes also show its strong capability of modelling in recent years. A novel end-to-end convolutional neural network (CNN) LID system is proposed, called language identification network (LID-net), taking advantage of neural networks (NNs) with the capability in feature extraction and discriminative modelling, which can extract units that discriminant to languages, and we call them LID-senones, thus can extract effective utterance representation with pooling layer. Evaluations on NIST LRE 2009 show improved performance compared to current state-of-the-art deep bottleneck feature with total variability (DBF-TV) method, can achieve 1.35%, 12.79% and 29.84% relative equal error rate (EER) improvement on 30, 10 and 3 s utterances and receive over 30% relative gain in C_avg on all durations.

Abstract:Outlier detection aims at detecting abnormal values from observational data, which has been used in various files. In outlier detection, normal data are generally embedded in some kind of intrinsic structure that is not suitable for characterizing outliers. Hence, how to effectively utilize the difference in structure between normal data and outliers will contribute to the identification of outlier. Then, a novel label propagation-based outlier detection algorithm is proposed in this paper. To characterize the above intrinsic structure, the graph model is adopted for implementing multiple label propagations. Thus, the difference in structure between normal data and outliers will be identical to the difference of label confidence between them. Furthermore, the statistical characteristic of the label confidences associated to those normal data is explored to give the final ensemble decision on the abnormity of the input test data. The experimental results have validated the effectiveness of the proposed method.

Abstract:Rough set theory is an effective method for dealing with uncertain boundary objects. The rough K-means clustering algorithm which combines rough set with K-means is simple and efficient. Though it can deal with clustering boundary elements, it has some drawbacks, for instance, the original rough K-means clustering algorithm is sensitive to the initial center, the set-up of empirical weigh ignores data difference, the unreasonable threshold setting engenders fluctuation of clustering results. To tackle these drawbacks, this paper proposed an improved rough K-means clustering algorithm combined with ant colony algorithm. The improved algorithm is optimized for rough K-means clustering by using random probability selection strategy and pheromone update of positive and negative feedback mechanisms in ant colony algorithm, and using dynamic threshold adjustment algorithm and associated weights method. Finally, the UCI’s Iris set, Balance-scale set and Wine set are used for verification of the algorithm. The results show that this algorithm exhibits a higher clustering accuracy.

Abstract:When extracting vehicle vibration acceleration from smart phone through software developed, the data quality of the vibration acceleration of the vehicle should be guaranteed in order to correctly evaluate the track smoothness and vehicle running comfort. This paper establishes an abnormal value recognition model based on the methods of probability-statistics and wavelet, in which median filter and wavelet filter are used to eliminate the random errors caused by the performance stability of mobile sensors and the change of testing environment. The effect of two filtering methods on the random error of mobile phone detection data is verified by combining with the detection data from Chengdu metro. The analysis result shows that the abnormal value location of mobile phone detection data can be accurately extracted based on the abnormal value recognition model. The mobile phone detection data truly reflect the vibration response of the car body by eliminating random errors caused by external environmental changes using methods of median filter and wavelet filter, so the model can be used to correctly evaluate the track smooth state and vehicle operating comfort.

Abstract:Fast Fourier transform (FFT) is a key block in the field of digital signal processing (DSP). A low hardware-cost 128-point FFT for UWB system is presented in this paper. Mixed radix-24-23 algorithm is adopted, and single-path delay feedback (SDF) architecture is used for hardware implementation. A novel cascade canonical signed digit (CSD) multiplier is proposed for the complex multiplication of W128i instead of the common booth multiplier, which can significantly reduce the hardware-cost. Based on QUARTUS PRIME tool with Cyclone 10 LP, the proposed scheme is developed, and the compilation report shows that the proposed scheme has the least hardware-cost and power consumption compared with the existing schemes.

Abstract:Most α wave studies only analyze α wave differences in different brain regions, which provides insufficient information to study their neural mechanisms. Aiming at this problem, a α wave independent component energy analysis method is proposed, and the main area of α wave generation is traced combining with the source localization algorithm. Firstly, the main active zone of α wave is analyzed by calculating α wave power in each brain area after pre-treatment. Then the independent components of decomposition from FastICA algorithm are source localized. Finally, the relationship between α wave in main brain area of action and every independent component is analyzed by α wave independent component energy analysis method. Testing results on six 26-year-old right-handed male subjects show that the main action areas of α wave in relaxed state are the left and the right occipital regions, followed by the right and the left posterior iliac regions. Although there is no distinct differences in α wave power of the left and the right occipital regions, the main α wave energy comes from different neural sources, which is located at the left brain near the left occipital area and the right brain near the right occipital area, respectively. Results of two-way repeated analysis of variance show that these two different neural sources have effect on α wave of the left and the right occipital regions.