Search

Printed Circuit Board Assembly (PCBA) plays an important role in 3C (Computer, Communication and Consumer Electronics) industry. The object detection of electronic components is the key technology for flaw inspection and quality control of PCBA. YOLOv3 has been widely used in object detection. Image pre-processing can eliminate some distractions and prominent some feature to improve the detection accuracy. The tailored YOLOv3 has a smaller computational cost and transfer learning can accelerate the training speed and improve inspection accuracy.

SMT (surface mounted technology) is the most popular technology and process in the electronic assembly industry. Chip resistors are widely used in this technology. Printing characters recognition of chip resistors can inspect the specification of the resistors and check whether the chip resistors are mounted on the correct location. The traditional image segmentation methods are used to extract the printing characters from the images of chip resistors. And then the extracted characters are input an artificial neural network model and recognized. The combination of image segmentation and neural network has high accuracy and efficiency.

The generative adversarial network (GAN) provides a different way for SAR data augmentation. The traditional GAN model is mainly based on the Jensen-Shannon (JS) divergence or Wasserstein distance. The former faces mode collapse, while the latter is not suitable for multi-category image generation. In this paper, an improved model based on WGAN-GP is proposed. An encoder is used to learn the features of real samples as the input of the generator to control training to a certain extent and make the generated image quality better. In addition, a pre-trained classifier is introduced as the constraint of the generator to ensure the generated images have the correct category information. MSTAR dataset is used to verify the generation capability of the proposed model. The results show that the proposed model has the stable generation capability to provide high-quality SAR images as a supplementary training dataset, which could assist in achieving good classification accuracy.

In order to fulfill the market requirements of a well- known citrus market, this paper uses artificial neural network technology to identify the maturity of the citrus fruits, especially the citrus fruits in Xinhui, China. The proposed method has the characteristics of non-contact, low cost, easy implementation and good performance, which can be effectively used in remote monitoring. Also, the AI recognized maturity is practically significant to agricultural production in the recording of citrus growth and forecasting of market sales.

An activation function is a key component of artificial neural networks (ANNs). It has a great impact on the performance and convergence of neural networks. In this work, a self-adapting methodology is proposed to explore the optimal adaptive activation functions for various tasks based on S-shaped or ReLu-shaped activation functions, which are regulated only by introducing several parameters. To verify the effectiveness of the proposed methodology, a series of comparison experiments are performed with MLP, CNN and RNN network structure on the benchmark datasets of image, text and audio. The experimental results are encouraging, and show that the proposed methodology can locate the optimal activation functions for various tasks. Nevertheless, the obtained functions are competitive and the improvements on network performance are significant compared with other popular activation functions, such as ELU, PReLU, ReLU, and Sigmoid.

The design of a compact circular polarization (CP) electrically small antenna (ESA) is carried out and presented in this paper. The designed antenna consists of a planar circuit board (PCB) on which two 3-dB hybrid couplers are fabricated for providing left-hand circular polarization (LHCP) and right-hand circular polarization (RHCP) waves, and two ceramic substrates covered by thin metal films with the high dielectric constants of 20.5 and 39 and the sizes of 25x25x6 mm and 18×18×4 mm, respectively. This compact CP ESA can realize a coverage of ±75° elevation angle with gain above -4dB in the frequency band of 1518 - 1525 MHz for RHCP and in the band of 1668 - 1675 MHz for LHCP, and a coverage of ±60°elevation angle with gain above - 2dB in the band of 1575.42±1.023 MHz for RHCP.

Electrically small antennas (ESAs) have been attracting many researchers owing to their compact sizes and potential applications in modern wireless communication systems. Considering that ESAs have high reactance, low radiation resistance and narrow bandwidth, metamaterials (MTMs) structures can be applied to the design of wideband and miniaturized antennas to overcome the above limitations. In this paper, the achievements in MTMs inspired ESAs are summarized, the research status is analyzed, and the future development trend of wideband MTMs inspired ESAs is predicted.

As an important industry that transforms the country's economic development mode, the new energy plant station must ensure the security of its network information. This paper mainly studies the monitor technology of unprotected cross-area access behavior in new energy plant stations, and proposes a horizontal boundary unprotected bypass monitoring scheme to identify the key elements of network flow under the protection device. A private network remote operation and maintenance monitoring method based on network topology automatic discovery technology is also proposed. At last, it proposes a violation outreach monitoring technology based on improved agent, which can discover the illegal outreach in the form of dual network cards. The experiment proves that the proposed algorithm can realize the identification of unprotected cross-area access. In addition, the effectiveness of the remote network operation and maintenance monitoring method and the violation outreach monitoring technology based on improved agent are verified.

Organ models are important for medical training and surgical planning. With the fast development of additive fabrication technologies, including 3D printing, the fabrication of 3D organ phantoms with precise anatomical features becomes possible. Here, we develop the first high-resolution kidney phantom based on soft material assembly, by combining 3D printing and polymer molding techniques. The phantom exhibits both the detailed anatomy of a human kidney and the elasticity of soft tissues. The phantom assembly can be separated into two parts on the coronal plane, thus large renal calculi are readily placed at any desired location of the calyx. With our sealing method, the assembled phantom withstands a hydraulic pressure that is four times the normal intrarenal pressure, thus it allows the simulation of medical procedures under realistic pressure conditions. The medical diagnostics of the renal calculi is performed by multiple imaging modalities, including X-ray, ultrasound imaging and endoscopy. The endoscopic lithotripsy is also successfully performed on the phantom. The use of a multifunctional soft phantom assembly thus shows great promise for the simulation of minimally invasive medical procedures under realistic conditions.

Wavelength division multiplexing passive optical network (WDM-PON) with centralized lightwave in optical line terminal (OLT) is recognized as the cost effective way in engineering application, for each optical network unit (ONU) do not have to do the laser light source configuration. In the transmission link between OLT and each ONU of the centralized lightwave WDM-PON, bilateral signal including downstream and upstream should be transmitted properly. In most reports from literature, different formats of signal modulation, including differentiated phase shift keying (DPSK), frequency shift keying (FSK), and on-off keying (OOK), are usually chosen in downlink and uplink to realize 2 or more types of service data transmission separately in a single wavelength channel. Consider that OOK format has the simplest way in system design, in this paper, a centralized lightwave WDM-PON approach has been proposed which employs OOK in both modulation and re-modulation process with only differences in extinction ratio. By simulation, the feasibility of such approach has been proved and the optimal extinction ratio for both downlink and uplink has been given.

Soft materials are being adopted in robotics in order to facilitate biomedical applications and in order to achieve simpler and more capable robots. One route to simplification is to design the robot's body using `smart materials' that carry the burden of control and actuation. Metamaterials enable just such rational design of the material properties. Here we present a soft robot that exploits mechanical metamaterials for the intrinsic synchronization of two passive clutches which contact its travel surface. Doing so allows it to move through an enclosed passage with an inchworm motion propelled by a single actuator. Our soft robot consists of two 3D-printed metamaterials that implement auxetic and normal elastic properties. The design, fabrication and characterization of the metamaterials are described. In addition, a working soft robot is presented. Since the synchronization mechanism is a feature of the robot's material body, we believe that the proposed design will enable compliant and robust implementations that scale well with miniaturization.

Microscopic artificial swimmers hold the potential to enable novel non-invasive medical procedures. In order to ease their translation towards real biomedical applications, simpler designs as well as cheaper yet more reliable materials and fabrication processes should be adopted, provided that the functionality of the microrobots can be kept. A simple single-hinge design could already enable micro-swimming in non-Newtonian fluids, which most bodily fluids are. Here, we address the fabrication of such single-hinge microrobots with a 3D-printed soft material. Firstly, a finite element model is developed to investigate the deformability of the 3D-printed microstructure under typical values of the actuating magnetic fields. Then the microstructures are fabricated by direct 3D-printing of a soft material and their swimming performances are evaluated. The speeds achieved with the 3D-printed microrobots are comparable to those obtained in previous work with complex fabrication procedures, thus showing great promise for 3D-printed microrobots to be operated in biological fluids.

In this paper, a wireless electrogoniometer is described using a pair of ultra-wide band (UWB) wireless smart sensor nodes that are interfaced to low power 3-axis accelerometers. A high resolution SAR ADC as well as a low power asynchronous continuous sampling event-driven ADC are integrated for the digitization of the input signal under different operation modes. An anti-self-locking circuit is included in the asynchronous event-driven ADC to improve the robustness. A tri-channel UWB transceiver is designed particularly for the data transmission of the 3-axis accelerometers. A power consumption of 4.6pJ/bit is achieved for the transmitter at 10Mbps while operating under 1.2V power supply. The receiver power consumption can be as low as 0.32nJ/bit at 10Mbps under 1.8V supply. To demonstrate one application of the device, the electrogoniometer was used to quantify the in-vivo response of a somatosensory neuron to joint angle changes.

Biomedical applications of micro or nanorobots require active movement through complex biological fluids. These are generally non-Newtonian (viscoelastic) fluids that are characterized by complicated networks of macromolecules that have size-dependent rheological properties. It has been suggested that an untethered microrobot could assist in retinal surgical procedures. To do this it must navigate the vitreous humor, a hydrated double network of collagen fibrils and high molecular-weight, polyanionic hyaluronan macromolecules. Here, we examine the characteristic size that potential robots must have to traverse vitreous relatively unhindered. We have constructed magnetic tweezers that provide a large gradient of up to 320 T/m to pull sub-micron paramagnetic beads through biological fluids. A novel two-step electrical discharge machining (EDM) approach is used to construct the tips of the magnetic tweezers with a resolution of 30 μm and high aspect ratio of ~17:1 that restricts the magnetic field gradient to the plane of observation. We report measurements on porcine vitreous. In agreement with structural data and passive Brownian diffusion studies we find that the unhindered active propulsion through the eye calls for nanorobots with cross-sections of less than 500 nm.

Heat conduction analogous process has ever been introduced into information filtering named standard heat conduction (SHC) method, resulting in a highly personalized but less accurate recommendation. In order to improve the recommendation accuracy, different algorithms have been proposed, with typical examples to be the highly accurate mass diffusion (MD) method, and a both highly accurate and highly diverse biased heat-conduction method (BHC). These previous algorithms have not considered the rating effect, where ratings essentially depict how users like objects. In this article, we propose a heterogeneous heat conduction method (HHC), by taking the ratings as the weight of heat conduction, which thus generates a heterogeneous heat diffusion pattern. Experimental results obtained from the Movie Lens dataset show that, the HHC greatly enhances the recommendation accuracy against the SHC, with the improvement percentage to be 46.32%, and also elevates the recommendation accuracy against the MD as well as the BHC. Moreover, the HHC simultaneously outperforms the MD, and even the BHC in recommendation diversity.

On-line monitoring of flame stability is crucial as it is closely related to plant safety, combustion efficiency, and pollutant emissions. This paper presents a new method for the stability monitoring of a burner flame using its medial axis. Flame video clips are acquired on a combustion test rig under different combustion conditions. Each frame of the flame video clips is processed and the medial axis of the flame is extracted. The processed medial axes are then superimposed for statistical analysis. The test results show that, although the shape of the flame is randomly fluctuated even under the same combustion condition, some of the statistical shape parameters are relatively stable for the given burner condition. These parameters can be used to indicate the stability of the flame.

In process monitoring, faults might be completely detected by T2 or SPE or both of them. However, one index rather than two indices is preferred to monitor a process in practice. The number of principal components (PCs) in principal component analysis (PCA) impacts the fault detection performance obviously. In this paper, we focus on developing a selection criterion for the number of the PCs based on the best detection performance of the combined index and discussing the uncertainty of PCA model through cross validation. The simulation demonstrates that one index has a good detection performance and simplifies the detection system.

We consider the problem of passive haptic rendering and interfacing of multiple degree-of-freedom (DOF) virtual proxy (VP), which has nonlinear Lagrangian dynamics and interacts with deformable virtual objects. For this, we solve the followings: 1) how to extend our recently-proposed non-iterative passive mechanical integrator (NPMI [1]) to simulate this nonlinear Lagrangian VP haptically-fast and discrete-time passively; 2) how to utilize virtual coupling technique to interface this NPMI-simulated VP and haptic devices while guaranteeing (sampled-data) passivity; and 3) how to passively render the interaction between VP and linear visco-elastic deformable virtual objects, while enhancing passivity at contact-on/off switchings. Some experimental results are also presented to support theory.

Initial configuration effect is investigated for a both highly accurate and highly diverse biased heat conduction method. According to the individual object degree, we assign a heterogeneous initial resource for each object. Experimental results obtained from the MovieLens dataset show that, the proposed method outperforms the standard heat conduction method by 47.33%, and also outperforms an accurate mass diffusion method by 24.04% in recommendation accuracy. Especially, even compared with an excellent hybrid method of heat conduction and mass diffusion, and the original biased heat conduction method, the manifested method further enhances both the recommendation accuracy and the diversity.

The shape of a burner flame is an important characteristic closely linked to the burner configuration, fuel type, and other operational parameters of a combustion process. This paper presents an algorithm to extract the 3-D medial axis of a burner flame and use it as a descriptor for the flame shape characterization. The algorithm is based on the 2-D medial axes of two flame images captured from two orthogonal directions of the burner using a flame imaging system. Since the data of the extracted features are much less than those of the original image, the computing process is simplified and efficient. Experimental results demonstrate that the flame 3-D medial axis determined using the proposed algorithm reflects its dynamical movement. Together with the radial parameter of the defined medial axis, a range of 3-D geometric parameters of the flame can be estimated.

To meet the increasingly stringent standards on pollutant emissions, oxy-coal combustion technologies are being proposed for both existing and new coal-fired power plants. However, there is lack of research to characterize this new type of combustion flame due to the complicity of its structure and background noise. This paper presents a new computer algorithm to identify the combustion region in oxy-coal flame images through the combination of image enhancement, wavelet denoising, and watershed transformation. The algorithm overcomes the limitations of the thresholding method which often wrongly detects the bright background and ignores the dark combustion region. Results on oxy-coal flame images obtained on an oxy-coal combustion test facility are reported to demonstrate the effectiveness of the algorithm proposed.

The quantitative characterization of combustion flames has long been an important but challenging issue both in combustion research and industrial applications. The shape of a flame is recognized as one of most useful characteristics in the adjustment of combustion parameters. Although medial axis is a useful shape representation of a flame, limited work has been reported on how to extract the medial axis of the flame. This paper describes an efficient algorithm, Horizontally-Cut Medial Axis (HCMA) algorithm, to detect the medial axis of an elongated flame. Experimental results indicate that the algorithm can determine the medial axis efficiently without any complex pruning process. Potential applications of flame medial axes on the flame shape representation are also discussed.

Hawkins' T H 2 statistic has not been analyzed in PCA thoroughly for ill-conditioning problem. This paper provides an analysis of Hawkins' T H 2 statisic by analyzing characters of industy process and the calculation accuracy of modern computor, which shows that Hawkins' T H 2 statistic is applicable for industy process monitoring. And a data simulation experiment indicates that Hawkins' T H 2 statistic has similar efficiency as SPE statistic.

Sperm motility analysis is an indispensable test in male infertility diagnosis. The current microscopic method cannot select the sperm of specific velocity for further in vitro Fertilization (IVF) usage or research purposes. Here, we report a novel and simple microfluidic device capable of balancing sperm's swimming speed with local flow velocity. The results suggested that sperm motility was quantitatively graded and sperm with different average straight line velocity (VSL) 40.5 ± 3.0, 27.2 ± 0.6 and 18.7 ± 0.6 µm/s respectively were automatically trapped in corresponding regions. Sperm motility classification (SMC) and sperm trapping may facilitate both the diagnosis of male infertility and further studies on sperm physiological behavior.

Digital image processing is playing an increasingly important part in imaging based flame monitoring systems. A crucial step in flame image processing is to detect the flame edge, a dividing boundary between the area where there is thermo-chemical reaction and those without. The determination of flame edges is a precursor to flame image processing and measurement of flame parameters. Several known edge detection methods have been tested to identify flame edges but the results achieved are disappointing. As a result of recent work in feature analysis of flame images, a novel flame edge detection method has been developed, which can detect the flame edges effectively and efficiently.

As a promising web application paradigm, web services have been widely adopted as e-business solutions. However, UDDI, the discovery mechanism of Web services, has the limitation of lack of semantic inference support, and the search mechanism based on keyword and categorization information leads to a limited performance. The problem can be addressed by a semantic approach for service matching based on semantic similarity of service properties. To integrate the matching mechanism into UDDI, we propose a semantic matchmaker system. We present the architecture of the system and demonstrate the key processes of the matchmaker for service publishing and discovery.