Search

Controlling and maintaining the orientation of the balloon-borne gondola for high-altitude flight is a prerequisite for ensuring the pointing control of observation instruments. When the balloon-borne gondola is flying in the stratosphere of the atmosphere, the existing external interferences will be converted into the coupling moment to the azimuth control system. Meanwhile, those uncertain factors and the frictional nonlinearity of the control system will also cause a certain magnitude of coupling moment. The existence of such coupling moment largely impacts on the accuracy and stability of the orientation control for the angular momentum exchange devices of the balloon-borne gondola. To address such an issue, this paper proposes and implements a novel type of integrated decoupler device. With this decoupler adopted, the aziDmuth control system could sense the existence of coupling torque and azimuth fluctuations quickly and suppress the influences of external interference, uncertain factors, and system structure nonlinearity on the azimuth control effectively, thereby improving the control accuracy of the azimuth control system. Both simulations and experiments are conducted to verify the effectiveness of the proposed device. The results show that the integration of the decoupler and the controller of the azimuth control system provide the azimuth control of the balloon-borne gondola with high accuracy and stability. Such a decoupler device design has a broad potential and could not only be used for balloon-borne gondola control but also could be applied onto other control systems using angular momentum exchange devices as actuators.

As a key feature of networked control systems (NCSs), the time delays induced by communication medium sharing and data exchange over the system components could largely degrade the NCS performances or may even cause system instability, and thus, it is of critical importance to reduce time delays within NCSs. This paper studies the time-delay reduction problem in distributed NCSs and presents a dual-way data scheduling mechanism for time-delay reductions in delay-bounded NCSs with time-varying delays. We assess the time delays and their influences on the NCSs first with various delay factors being considered and then describe a one-way scheduling mechanism for network-delay reductions in NCSs. Based upon such a method, a dual-way scheduling algorithm is finally proposed for distributed NCSs with different types of transmitted data packets. Experiments are conducted on a remote teaching platform to verify the effectiveness of the proposed dual-way scheduling mechanism. Results demonstrate that, with the stability time-delay bound considered within the scheduling process, the proposed mechanism is effective for NCS time-delay reductions while addressing the stability, control accuracy, and settling time issues efficiently. Such a proposed mechanism could also be implemented together with some other existing control algorithms for time-delay reductions in NCSs. Our work could provide both useful theoretical guidance and application references for stable tracking control of delay-bounded NCSs. Published version

This paper proposes a practically executable path planning method, namely, Pheromones-RRT(PRRT), for robots with a large joint range in a complex environment. To inter-activate with the real world, the point cloud is utilized as the scene information, while for sampling, the pheromones approach is designed to describe the pheromone content carried by each sampling point. During the sampling process, random sampling nodes are performed with a probability of ε, or those nodes with the highest pheromone content in the current sampling tree are selected with a probability of 1-ε and sampled in their neighborhood. To avoid the local minimum problem, the concept of pheromone volatile factor (PVF) is proposed, while in the expansion, double trees are also generated by PRRT in both cartesian and configuration spaces to improve the speed of the algorithm. The pheromone accumulation enables PRRT to certain learning abilities, reducing the randomness of the sampling process. Simulation results show that the proposed method can effectively plan an optimal obstacle avoidance path with satisfactory performances as compared with the RRT-Connect.

With the rapid advancements of various information technologies in recent years, many new forms of e-learning schemes have emerged. Due to the various issues with such e-learning schemes, however, the commonly adopted teaching method is still the traditional face-to-face classroom in practice. In such a teaching mode, the three key elements of all teaching modes, i.e., theory, experiment, and practicing, are separate and independent of each other, which largely degrades the learning efficiency. To address such a practical issue, based on the objectives of professional personnel training for navigation and control, this paper presents a curriculum teaching model that integrates the three learning key elements together. Specifically, this paper shows the architecture of the curriculum construction mode first, and then presents the composition and function of the student dimension, teacher platform dimension, laboratory server dimension and laboratory equipment dimension in the navigation and control course group, respectively. Finally, with the course of “principle of inertial sensor” as a case study, this paper introduces the teaching process of using the teaching mode, which integrates the three key teaching elements together in the course. The practical teaching results show that this proposed teaching mode could not only make the teaching process of engineering courses intuitive and easy to understand, but also enable students to participate in the theorical, experimental, practical teaching and learning process throughout the presented mode. This mode could largely enhance both students’ learning enthusiasm and the efficiency.

As a crucial power amplifying component for electro-hydraulic servo valves, the servo-valve spool valve largely determines the overall performance of an electro-hydraulic servo system. Due to the machining errors during the servo-valve manufacturing process, however, it is extremely difficult to characterize such response characteristics in engineering practice, and thus, there may exist certain nonlinearity between the flow rate and valve spool displacement caused by the spool and valve sleeve overlap. To accurately characterize the relationship between the flow rate and spool displacement to facilitate servo-valve production, an indirect flow measurement-based servo-valve spool valve grinding (FM-SVG) scheme is proposed in this study. Specifically, a valve spool overlap measurement mechanism is presented first based on the valve spool grinding principle, and then a data processing method is devised to eliminate the errors introduced by the fluctuations of differential pressure at the valve ports. Meanwhile, a closed-loop control system is also devised to maintain a constant differential pressure at the servo-valve ports. Finally, the FM-SVG scheme is embedded onto a lab-customized electro-hydraulic servo valve platform, and experiments are conducted to verify the feasibility and effectiveness of the FM-SVG scheme. Results show that, by utilizing the devised control scheme to maintain a differential pressure around 2.00 MP, the proposed grinding scheme helps achieve a grinding accuracy of 2.0 μm. With only grinding process needed, such a grinding scheme helps improve both grinding accuracy and production efficiency in practice.

Nonstationary signal decomposition (SD) is a primary procedure to extract monotonic components or modes from electroencephalogram (EEG) signals for the development of robust brain-computer interface (BCI) systems. This study proposes a novel automated computerized framework for proficient identification of motor and mental imagery (MeI) EEG tasks by employing empirical Fourier decomposition (EFD) and improved EFD (IEFD) methods. Specifically, the multiscale principal component analysis (MSPCA) is rendered to denoise EEG data first, and then, EFD is utilized to decompose nonstationary EEG into subsequent modes, while the IEFD criterion is proposed for a single conspicuous mode selection. Finally, the time-and frequency-domain features are extracted and classified with a feedforward neural network (FFNN) classifier. Extensive experiments are conducted on four multichannel motor and MeI data sets from BCI competitions II and III using a tenfold cross-validation strategy. Results compared with the other existing methods demonstrated that the highest classification accuracies of 99.82% (data set IV-A), 93.33% (data set IV-b), 91.96% (data set III), and 88.08% (data set V) in subject-specific scenarios, while 82.70% (data set IV-A) in the subject-independent framework are achieved for IEFD with FFNN classifiers collectively. The overall exploratory results authenticate that the proposed IEFD-based automated computerized framework not only outperforms the conventional SD methods but is also robust and computationally efficient for the development of subject-dependent and subject-independent BCI systems. This work was supported in part by the National Key Research and Development Program of Shaanxi (2021SF-342), Fundamental Research Funds for the Central Universities (G2018KY0308), China Postdoctoral Science Foundation under Grant (2018M641013), Postdoctoral Science Foundation of Shaanxi Province (2018BSHYDZZ05).

As key components of robot hydraulic drive control system, the hydraulic pipelines (HPs) largely determine the robot performances, and thus, the selection of such components is of critical importance to robot design. In practice, however, there exists a trade-off between the respective requirements on the HPs internal structures and their stiffness properties, since a control system typically requires HPs to be as light as possible, while the breaking stresses like vibrations and pulsed pressures require HPs to be with high stiffness and strength. To help achieve a balance between such HPs requirements, we evaluate the overall stresses experienced by a simply-supported robot HP and estimate the HP fatigue life under the hybrid excitation of water-hammer waves and external vibrations for the first time, to the best of our knowledge. Specifically, we propose a non-steady flowing fluid mathematical model to describe the HP pressures based on water-hammer wave theory and solve it using the method of characteristic (MoC) first, and then present another mathematical model to integrate the external robot link vibrations into such a model. The equivalent stresses for a single-axis pipeline is calculated and its fatigue life is also estimated with this proposed model. Finally, experiments on the pipeline fracture and damage mechanics are conducted to verify the effectiveness of those models. Results show that the practical operation life of the robot HP matches well with the theoretically estimated fatigue life under the hybrid excitation of water-hammer waves and external vibrations.

This paper proposes a new robust adaptive cerebellar model articulation controller (CMAC) neural network-based multisliding mode control strategy for a class of unmatched uncertain nonlinear systems. Specifically, by employing a stepwise recursion-based multisliding mode method, such a proposed strategy is able to obtain the virtual variables and the actual control inputs of each order first, and then it reduces the conservativeness for controller parameter design by adopting the CMAC neural network to learn both system uncertainties and virtual control variable derivatives of each order online. Meanwhile, with the hyperbolic tangent function being chosen to replace the sign function in the variable structured control components, the proposed strategy is able to avoid the chattering effects caused by the discontinuous inputs. The stability analysis shows that the proposed control strategy ensures that both the system tracking errors and the sliding modes of each order could converge exponentially to any saturated layer being set. The control strategy was also applied onto a passive electrohydraulic servo loading system for verifications, and simulation results show that such a proposed control strategy is robust against all system nonlinearities and external disturbances with much higher control accuracy being achieved.

Due to the rapid spread of COVID-19 and its induced death worldwide, it is imperative to develop a reliable tool for the early detection of this disease. Chest X-ray is currently accepted to be one of the reliable means for such a detection purpose. However, most of the available methods utilize large training data, and there is a need for improvement in the detection accuracy due to the limited boundary segment of the acquired images for symptom identifications. In this study, a robust and efficient method based on transfer learning techniques is proposed to identify normal and COVID-19 patients by employing small training data. Transfer learning builds accurate models in a timesaving way. First, data augmentation was performed to help the network for memorization of image details. Next, five state-of-the-art transfer learning models, AlexNet, MobileNetv2, ShuffleNet, SqueezeNet, and Xception, with three optimizers, Adam, SGDM, and RMSProp, were implemented at various learning rates, 1e-4, 2e-4, 3e-4, and 4e-4, to reduce the probability of overfitting. All the experiments were performed on publicly available datasets with several analytical measurements attained after execution with a 10-fold cross-validation method. The results suggest that MobileNetv2 with Adam optimizer at a learning rate of 3e-4 provides an average accuracy, recall, precision, and F-score of 97%, 96.5%, 97.5%, and 97%, respectively, which are higher than those of all other combinations. The proposed method is competitive with the available literature, demonstrating that it could be used for the early detection of COVID-19 patients.

Chronic kidney disease (CKD) is one of the public health threats around the world, which may cause serious health problems like cardiovascular disease, kidney failure, or even more serious as premature death. Although CKD usually could be managed by general internists, such a way of treatment can be applied only when significant symptoms appear, which is very slow. It has also been reported that CKD could be characterized by means of glomeruli, and classified by the stages of disease severity for early treatment. However, due to lack of reliable method to detect the cellular-level microstructures for disease severity characterization, the diagnosis is troublesome, and thus, the treatments might be delayed while the best treatment time could be missed. For early detection of CKD, it is imperative to develop reliable tools to detect and characterize the disease at an early stage with minimal or noninvasiveness. For this research, the micro-optical coherence tomography $\mathbf{(\mu OCT)}$ was assessment it feasibility as a cellular-level structure imaging tool in kidney disease diagnosis at an early stage. Specifically, by measuring the number of glomeruli within a volumetric kidney tissue, a new diagnostic criteria is also established. Imaging results of the kidney specimens as compared their corresponding histology show that the cellular-level glomeruli structures could be identified clearly, and as a basic functional unit of kidney, it could be utilized as a reliable parameter to access the severity of the CKD.

Background Precise identification, discrimination and assessment of central nervous system (CNS) tumors is of critical importance to brain neoplasm treatment. Due to the complexity and limited resolutions of the existing diagnostic tools, however, it is difficult to identify the tumors and their boundaries precisely in clinical practice, and thus, the conventional way of brain neoplasm treatment relies mainly on the experiences of neurosurgeons to make resection decisions in the surgery process. The purpose of this study is to explore the potential of Micro-optical coherence tomography (μOCT) as an intraoperative diagnostic imaging tool for identifying and discriminating glioma and meningioma with their microstructure imaging ex vivo, which thus may help neurosurgeons to perform precise surgery with low costs and reduced burdens. Methods Fresh glioma and meningioma samples were resected from patients, and then slices of such samples were excised and imaged instantly ex vivo with a lab-built μOCT, which achieves a spatial resolution of ~ 2.0 μm (μm). The acquired optical coherence tomography (OCT) images were pathologically evaluated and compared to their corresponding histology for both tumor type and tumor grade discriminations in different cases. Results By using the lab-built μOCT, both the cross-sectional and en face images of glioma and meningioma were acquired ex vivo. Based upon the morphology results, both the glioma and meningioma types as well as the glioma grades were assessed and discriminated. Comparisons between OCT imaging results and histology showed that typical tissue microstructures of glioma and meningioma could be clearly identified and confirmed the type and grade discriminations with satisfactory accuracy. Conclusions μOCT could provide high-resolution three-dimensional (3D) imaging of the glioma and meningioma tissue microstructures rapidly ex vivo. μOCT imaging results could help discriminate both tumor types and grades, which illustrates the potential of μOCT as an intraoperative diagnostic imaging tool to help neurosurgeons perform their surgery precisely in tumor treatment process.

With rapid advancements in artificial intelligence and mobile robots, some of the tedious yet simple jobs in modern libraries, like book accessing and returning (BAR) operations that had been fulfilled manually before, could be undertaken by robots. Due to the limited accuracies of the existing positioning and navigation (P&amp, N) technologies and the operational errors accumulated within the robot P&amp, N process, however, most of the current robots are not able to fulfill such high-precision operations. To address these practical issues, we propose, for the first time (to the best of our knowledge), to combine the binocular vision and Quick Response (QR) code identification techniques together to improve the robot P&amp, N accuracies, and then construct an autonomous library robot for high-precision BAR operations. Specifically, the binocular vision system is used for dynamic digital map construction and autonomous P&amp, N, as well as obstacle identification and avoiding functions, while the QR code identification technique is responsible for both robot operational error elimination and robotic arm BAR operation determination. Both simulations and experiments are conducted to verify the effectiveness of the proposed technique combination, as well as the constructed robot. Results show that such a technique combination is effective and robust, and could help to significantly improve the P&amp, N and BAR operation accuracies, while reducing the BAR operation time. The implemented autonomous robot is fully-autonomous and cost-effective, and may find applications far beyond libraries with only sophisticated technologies employed.

The influences of redundant force exist in electro-hydraulic load simulators utilized for aircraft booster aerodynamic load simulations. To suppress the influences of this force, however, the conventional way of introducing extreme points in feed-forward systems makes the compensation frequency band narrow, while the other way of servo synchronous compensation control is largely affected by the actuator movement amplitude and frequency, which thus worsen the system adaptive abilities. In this paper, we propose a neuron PID based self-adaptive synchronous servo compensation mechanism for redundant force suppression in aerodynamic loading system. We establish a single-neuron PID based self-adaptive algorithm first, and then adopt the algorithm output as a parameter for the servo synchronous compensation, to compensate the vibrations caused by actuator movement amplitude and frequency. Simulation experiments are conducted to verify the effectiveness of the proposed mechanism. Results show that the proposed mechanism achieves better redundant force suppression at a wider band as compared with the conventional feed-forward compensation and the servo synchronous compensation methods, and could be applied for redundant force suppressions in various load simulators for its advantages in structural simplicity and self-adaptive capability.

During the ground test of aircraft hydraulic steering gears, the high precision simulation of the load experienced by steering engine in aircraft flying process is of critical importance. To tackle with the control system non-linearity, parameter time-varying properties as well as the excessive force caused by the position perturbation, a parallel CMAC and PID based control algorithm is proposed. Specifically, a nonlinear mathematical model for the system is established first, and then, the PID controller is utilized to realize feedback control, and meanwhile, provide the mentor signals to the CMAC neural network. While the CMAC neural network itself is a feed-forward control system that can be utilized to reduce the interference caused by the excess forces and time-varying parameters, which thus realize a nonlinear approximation process. Simulations are conducted to verify the effectiveness of the proposed algorithm, and results show that the proposed algorithm is able to suppress the excess forces to be around 22.5% that of the conventional method. Compared to the existing methods, such a proposed algorithm is of higher tracking precision and anti-interference capability and more robust, and thus could improve the system performances and provide references to the practical application of the system.

Optical coherence tomography (OCT) is a noninvasive high-resolution diagnostic imaging modality that plays an increasingly important role in dermatology. Diagnosis of skin diseases using OCT requires both cellular-level high resolution and large area skin coverage. In practice, however, there exists a trade-off between the achievable spatial resolutions and the transverse scanning range. In this study, we report a Micro-OCT (μOCT) system that is capable of providing three-dimensional images of the skin at multiple spatial scales with both cellular-level resolution (1-2 μm) mode and large area (∼ mm2) scanning mode. Specifically, in the cellular-level scanning mode, we achieve a transverse resolution of ∼1.5 μm and an axial resolution of 1.7 μm (n = 1.38) which enables the visualization of cellular-level skin microstructures. While in the large-area scanning mode, the system is capable of covering an en face imaging area reaching up to with a lateral resolution of ∼5.5 μm at a scanning speed of 60 K Alines/s. We experimentally verify the imaging capabilities of such a multiscale μOCT system including in vivo visualization of epidermal cells in the cellular-level scanning mode as well as the internal fingerprints and sweat gland ducts in the large area scanning mode. Micro-anatomical imaging at multiple spatial scales could provide comprehensive information of the skin that is valuable to disease diagnosis. NRF (Natl Research Foundation, S’pore) MOE (Min. of Education, S’pore) NMRC (Natl Medical Research Council, S’pore) Accepted version

Organic semiconductors have potentials for a broad range of applications; however, it is difficult to be integrated with traditional inorganic material to meet the need of further application. Based on low-temperature silicon nitride (SiNx) deposition technique, here we demonstrate a hybrid structure fabricated by directly depositing high-quality SiNx on organic polymer film Poly[2-(2',5'-bis(2"-ethylhexyloxy)- phenyl) -1,4-phenylene vinylene] (BEHP-PPV). Stacked BEHP-PPV/SiNx hybrid structures with different periods are obtained and their optical properties are systematically characterized. Moreover, a group of BEHP/PPV embedded SiNx micro-disk is fabricated and amplification of spontaneous emission (ASE) is observed under optical pumping, further confirming that the gain properties of BEHP/PPV are well preserved. Our technique offers a platform to fabricate organic/inorganic hybrid optical devices compatible with integrated components., 6 pages, 4 figures

Silicon nitride-based integrated photonic devices, including DBRs, waveguides, microdisks, ring resonators, and photonic crystal nanocavities have also been demonstrated, suggesting that they are suitable for operating in the visible to infrared wavelength range.

Random lasing action is demonstrated in a BDMO-PPV/CHDV composite drop-coating film, with multiple narrow peaks around 597 nm when pumping above its lasing threshold of 0.87 mJ/cm2.

The tool-set accuracy and maneuverability are the key factors which impact the processing precision and efficiency of CNC grinding machines. Pointing at the shortage of the study on this field, we raise a method of two-stage feed that combines electric detecting and instantaneous power detecting. In the backward stage, which determines the accuracy of tool-set, the model based on instantaneous power detecting—Park converting model is used to detect the values of three-phase instantaneous power of the sampling points at every other half grid cycles. The weighted judging formula is set up to judge whether the two sets of data suit the formula and count at every half grid cycle when they suit the formula. If the counting value gets to the set value, we consider the tool-set has finished. The precision of this method improves at least one time comparing with other method based on detecting power of the spindle motor. This method has advantages of easy operability ,good reliability and high precision.

High-quality SiNx films with controllable low stress and low optical loss are deposited at ultra-low temperature (75 °C) using inductively coupled plasma chemical vapor deposition (ICP-CVD). Two kinds of integrated photonic structures have been demonstrated that exemplify its viability as a photonic integration platform. A microcavity consists of two distributed Bragg reflectors (DBR) formed by alternating a total of 49 layers of SiNx and SiO2 with a total thickness of about 11.5 μm is grown without any cracks, confirming the excellent stress control in the process. Microring resonators are also fabricated in as-deposited planar SiNx waveguide layer using electron-beam lithography (EBL) and plasma etching. Average waveguide loss of 0.79 ± 0.22 dB/cm has been achieved in the range of 1550-1600 nm for ring radii larger than 40 μm. The ultra-low temperature grown SiNx with properties of low loss and low stress is therefore a promising photonic integration platform for various photonic integration applications.