Search

Joint human activity localization and recognition has broad application prospects in human-computer interaction, virtual reality, smart healthcare system, security monitoring and robotics. Ultra-wideband (UWB) is an emerging technology adopted in real-time location system (RTLS) and has shown satisfactory performance in the task of human activity localization. However, few studies have been carried out to simultaneously recognize human activities based on UWB RTLS, which limits the use of UWB RTLS in many applications. In this study, we develop a RTLS based on UWB for the joint task of activity localization and recognition. A compressed sensing-based activity recognition approach is proposed for the task of activity recognition and several machine learning methods are designed to further improve the activity localization accuracy for the task of activity localization. The experimental results show that our UWB RTLS achieves good performance in this joint task.

High-precision indoor positioning has nowadays emerged as a critical function for many applications. However, many existing indoor positioning systems either fail to achieve a high positioning accuracy or are very easily affected by indoor environments composed of many obstacles, preventing them from satisfying many application requirements. Ultra wideband (UWB) has recently drawn extensive attention in the field of indoor positioning due to its great ability to achieve high ranging and localization accuracy while minimizing the effect of multipath interference. Meanwhile, advanced artificial intelligence and signal processing techniques have been explored to improve the precision and performance of indoor positioning system. In this paper, a high-precision UWB indoor positioning system integrating artificial intelligence and signal processing techniques is designed. And field tests are also conducted to validate the design of the system.

The application of compressed sensing and machine learning in power quality disturbance (PQD) classification has drawn more and more attention. This paper presents an adaptive compressed sensing and machine learning (ACSML) method to classify both single PQD and combined PQDs with the consideration of the correlation and the sparsity properties in the PQD signals. This method first uses random projection to reduce the dimensionality of the PQD signals. Meanwhile, a simplified near neighbor algorithm is proposed to reduce the size of the required PQD signal training dataset. The PQD classification problem is finally solved using an adaptive compressed sensing classification algorithm. Experiment results show that the proposed ACSML method achieves higher classification accuracy and faster classification speed in classifying PQDs than the other existing compressed sensing based PQD classification methods.

Indoor positioning system plays an important role in smart grid. Although GPS is the predominant outdoor positioning technology, it is unsuitable to be used in many fields of smart grid for three main reasons: first, signals sent from GPS could easily get blocked by solid materials such as metal or brick; second, the complex electromagnetic interference induced by electrical circuits greatly affects GPS signals; third, GPS can only achieve meter-level real time positioning accuracy, which is far from sufficient for many requirements of smart grid applications. Some other indoor positioning technologies, such as Bluetooth, Wi-Fi, ultrasound, infrared and RFID, fail in either the positioning accuracy, the positioning range, or the positioning speed required in many smart grid applications. Therefore, this paper proposes a real time indoor positioning system for smart gird based on a more promising technology, ultra-wideband (UWB). UWB is suitable for real-time localization in smart grid because UWB has short radio frequency pulse duration and wide bandwidth, which can minimize the effects of multipath interference and allow for high-resolution ranging and easier material penetration. In addition, since high-accuracy position information is required in many smart grid fields, a comprehensive framework integrating several artificial intelligence techniques, including outlier detection, line-of-sight/non-line-of-sight classification, filter design, range measurement correction and maximum likelihood localization estimation, is also proposed to further improve the positioning accuracy. At last, the performance of this system is verified through a series of experiments.

Compressive sensing related methods have achieved success in power system fault classification. However, those methods, either only focus on sparsity without considering the correlation within power system fault signals, or emphasize too much on the correlation but neglect the sparsity property of power system fault signals, resulting in the decrease of classification precision and the increase of the number of training samples needed. In addition, those methods do not explore other machine learning techniques at the same time to further improve the classification accuracy and speed. To overcome the drawbacks and limitations of those methods, this paper proposes an adaptive compressive sensing and machine learning classification framework to quickly and accurately classify power system faults considering both the correlation and the sparsity properties inherited in power system fault signals. First, random projection is applied to reduce the dimensionality of the fault signal. Second, a k-nearest neighbor based algorithm is designed to reduce the number of fault signals needed in the training dataset. Finally, the classification of power system faults is formulated as solving an adaptive L1-norm and L2-norm combined objective function using an alternating direction optimization method. Simulation results and analyses show that the proposed framework has better performance than the existing compressive sensing based power system faulty classification methods in both accuracy and speed.

Canonical correlation analysis (CCA) is a popular and powerful technique for two-view dimension reduction and feature extraction. But, CCA is not able to directly handle more than two view data and has a rigorous assumption that all the samples from two different views are paired. However, practical multiple view data are often semi-paired. To address this problem, we in this paper propose a novel semi-paired multiview dimension reduction approach, which takes cross-view neighborhood relationship among semi-paired data and within-view global structure information into consideration. The proposed approach can not only deal with multiview (more than two) data, but also take sufficient advantage of unpaired multiview data and then mitigate overfitting effectively caused by the limited paired data. Experimental results on two benchmark data sets demonstrate the effectiveness of our proposed method.

Anti-washout underwater concrete (AWC) is widely used in nondrainage strengthening, however, there still exist some problems with it, such as high strength loss and poor interfacial bond in practical engineering application. Based on the study of self-stressed concrete (SSC), a research on the mix ratio for the C30 self-stressed anti-washout underwater concrete (SSAWC) was carried out in this paper in hope of solving the above problems, specifically, by adding an expansive agent to the AWC. The parameters, such as strength, fluidity, anti-dispersity, and expansibility, were picked as target indices in determination of the mix ratio. The orthogonal test design and range analysis were used to determine the reasonable mix ratio and study the influence of various parameters on the performance of SSAWC. The experimental program conducted includes a series of strength, fluidity, anti-dispersity, and expansibility tests on 18 groups of specimens. The results show that C30 SSAWC has an excellent performance using the optimal mix ratio. Compared with AWC, the expansibility and self-stress of the SSAWC can be easily observed, and the compressive strength ratio of the SSAWC casted in water to that casted in air is much bigger. This implies that SSAWC is applicable to the nondrainage strengthening.

We investigate facial expression recognition using state-of-the-art classification models. Recently, CNNs have been extensively used for face recognition. However, CNNs have not been thoroughly evaluated for facial expression recognition. In this paper, we train and test a CNN model for facial expression recognition. The performance of the CNN model is used as benchmark for evaluating other pre-trained deep CNN models. We evaluate the performance of Inception and VGG which are pre-trained for object recognition, and compare these with VGG-Face which is pre-trained for face recognition. All experiments are performed on publicly available face databases, namely, CK+, JAFFE and FACES.

The initial rotational stiffness of connections is important for the design of semirigid frames. However, most current studies on steel beam to concrete-filled steel tube column connections...

The bolted T-stub connection joining beam with column is being widely applied. To enhance the energy dissipation capacity of conventional T-stub connections, two rectangular plates are proposed to be inserted between the T-stub and column, so that the T-stub flange can yield both under tensile and compressive loads. This study put forward a mechanical model of a new T-stub connection with inserted plates and investigated important factors that could affect its mechanical behavior through experimental tests. Thirty specimens were designed with different configurations that differed according to the existence or absence of inserted plates, the fabrication method and the width of inserted plates. These configurations were tested under axial and cyclic loading conditions, and results showed that the proposal aiming to improve the energy dissipation capacity was feasible. The mechanical model presented coincided with the test observation and data. The advent of two inserted plates elevated the load bearing capacity, stiffness and ductility of connections under compression, whereas in tension the properties were not substantially enhanced. The welded T-stub connections outperformed those cut from standard section steel. The energy dissipated by connections with inserted plates was about 150% of that by traditional connections without inserted plates. Only within a reasonable range can the increment of plate width promote the energy dissipation capacity of T-stub connections. The optimum width of plates in terms of energy consumption accounted for around 31% of the overall width of connections.

Five square reinforced concrete (RC) stub columns were fabricated and experimental tested under an axial compression load in order to study the effectiveness of strengthening by using prestressed carbon fiber reinforced polymer (CFRP) strips. The specimens tested included one un-strengthened and four strengthened. Three specimens were strengthened by using prestressed strips, and the other one strengthened by using non-prestressed was used to compare. The prestress of the strip was achieved by tensed using a special device before it was pasted to the column. The experimental data such as the failure modes, load-displacement curves and strain distribution of CFRP strips were obtained. The effects of prestress magnitude of CFRP strips on the ultimate bearing capacity and deformation capacity of the specimens were analyzed. The results show that the prestress of CFRP strips can effectively increase the ultimate capacity and delay the crack occurrence of the specimens. And ultimate strain, plastic deformation of the specimens were also improved with increase of prestress magnitude of strip.

Local specialties or souvenirs are called ‘shouxin’ in the Lingnan area. Through market research on a number of shouxin enterprises in Zhuhai, this thesis contrasts different products of the same type investigated with multi-dimensional analysis, and summarizes the features and benefits of the various products. Referring to theories and experience of cultural innovation in Taiwan especially the design context “building a brand on the basis of culture and using it in daily life in the form of a specific product”, it proposes that souvenir enterprises in Zhuhai could take humanity as the core of its creative design and culture as the basis. It constructs a development strategy and framework of creative design for local shouxin enterprises in Zhuhai.

We investigate facial expression recognition based on geometric features, and image appearance, using a range of classifier models. First, we evaluate expression recognition of face images based on geometric features, namely, facial landmarks based on the Active Appearance Model, and Action Units based on the Facial Action Coding System. A generalized linear model and a neural network are used to classify face images based on these geometric features. Second, the classification achieved by facial landmarks and Action Units is compared with the classification achieved by convolutional neural network (CNN) which extracts image features from raw pixels. Finally, we use transfer learning technique to evaluate classification using a pre-trained model. All experiments are performed on the state-of-the-art CK+ face database.

This paper investigated on the performance of slender concrete-filled carbon fiber-reinforced polymer (CFRP)–PVC tubular columns under axial compression. Four specimens with different slenderness ratios were firstly carried out to explore their performance and failure modes under axial compressive loads in laboratory. Secondly, a nonlinear finite element model (NFEM) was developed to analyze the relation between loads and deformation, and the analytical values were compared with the experimental results. After the efficiency of NFEM was verified, the parametric analyses varying the number of CFRP layer, concrete compressive strength and the slenderness ratio were conducted by the NFEM, and the ultimate bearing capacity formula as well as the relation model of the stability coefficient (φ) versus slenderness ratios (λ) were proposed. To validate the efficiency, ultimate bearing capacity was compared among the experimental results, NFEM analytical values and the prediction by the proposed formula. The results indicate that not only the developed NFEM can effectively depict the varying process of the loads and deformation in details, but also the proposed ultimate bearing capacity formula is reasonable and reliable in the fast prediction of the bearing capacity for slender concrete-filled CFRP–PVC tubular columns. It is beneficial to widely apply such new type composite columns in pile foundation or corrosion environments.

This paper presents an innovative bolt model suitable for the three dimensional finite element analysis (FEA) of the semirigid beam-to-column bolted connections. The model is particularly useful for the moment-rotation relationship of beam-tocolumn connections, especially in cases where the connectors such as endplates, angles, T-stubs, are not particularly thin. In this paper, the bolt tensile behavior is firstly discussed by using a refined finite element model, in which the complex geometries of both external and internal threads were modeled. Then, the bolt behavior predicted by the commonly used models was compared with that of the refined FEA to appraise the accuracy of these models. The comparison shows most of the models commonly used can not predict accurately the axial stiffness, carrying capacity and ductility of bolt simultaneously. Afterwards, an innovative bolt model was proposed and the model accorded with the refined FEA for single bolts. Finally, the proposed model was applied to analyze the moment-rotation behavior of several experimented and well documented connections with different configurations. The results indicate that the proposed model is feasible and efficient.

We investigate facial expression recognition (FER) based on image appearance. FER is performed using state-of-the-art classification approaches. Different approaches to preprocess face images are investigated. First, region-of-interest (ROI) images are obtained by extracting the facial ROI from raw images. FER of ROI images is used as the benchmark and compared with the FER of difference images. Difference images are obtained by computing the difference between the ROI images of neutral and peak facial expressions. FER is also evaluated for images which are obtained by applying the Local binary pattern (LBP) operator to ROI images. Further, we investigate different contrast enhancement operators to preprocess images, namely, histogram equalization (HE) approach and a brightness preserving approach for histogram equalization. The classification experiments are performed for a convolutional neural network (CNN) and a pre-trained deep learning model. All experiments are performed on three public face databases, namely, Cohn–Kanade (CK[Formula: see text]), JAFFE and FACES.