Abstract:Aircraft relay can quickly build an end-to-end communication network, enhance the coverage and transmission capacity of the communication system, strengthen the signal receival and improve the data transmission speed, and improve the reliability of the communication system. This paper summarizes the development status of UAV relay technology and UAV relay in wireless optical communication at home and abroad, and conducts in-depth research on the key technologies of UAV relay in wireless optical communication. The channel characteristics and their influence on communication and suppression techniques are analyzed, the advantages and disadvantages of different modulation and demodulation methods and the specific methods of beam capture, tracking and alignment in the communication process are compared, and the fast alignment of the beam is introduced in detail. Finally, the development direction of UAV relay in wireless optical communication is prospected.

Abstract:Coil-type magnetic flux detection technology has been widely used in clinical diagnosis applications. However, adjacent magnetic samples’ detection signals are prone to aliasing. Most of the existing researches used algorithm processing and sample space isolation to avoid the mutual interference of signals, which limits the development of this technology in the field of point-of-care testing. In view of the above problems, this study uses mathematical physical modeling combined with finite element simulation to analyze the structure of the coil-type magnetic flux sensor, and optimizes the sensor structure. The final experimental results show that the optimized sensor can effectively distinguish the adjacent magnetic signal sources with a distance of 4 mm. In the immunochromatographic quantitative detection of gastrin-17 in clinical samples, the detection range reached 11 pg / mL ~ 110 ng / mL. The above result confirm that the optimized sensor structure has the capability of multiplex synchronous multiplexing detection.

Abstract:The utilization of electromagnetic ultrasonic sensor arrays for pipe defect detection not only improves the signal-to-noise ratio, sensitivity and resolution of electromagnetic ultrasonic detection signals, but also enhances the intuitiveness and flexibility of electromagnetic ultrasonic detection. In this paper, the working principle of the Lorentz force-based periodic permanent magnet arraytype electromagnetic ultrasound transducer (PPM-EMAT) for excitation of ultrasound guided waves was referred and the mechanism of defect localization and imaging was used by the TFM and the SCF. Then the FE model was built to verify the process of quasi-T (0,1) mode guided wave propagation in a pipeline structure. Finally, the developed multi-channel electromagnetic ultrasonic inspection system was used to perform actual inspection of stainless-steel pipelines with defects and verify the simulation results. The experimental results show that the developed system can detect multiple through-hole defects in the pipeline specimen, and the longitudinal positioning error can be controlled below 1. 5%, which verifies that the array electromagnetic ultrasonic sensor pipeline inspection method can realize the defect imaging and defect positioning of the pipeline

Abstract:To determine the influence of the structural parameters of the cylindrical gear flowmeter and the physical characteristics of the fluid medium on its performance, and to determine the optimal assembly clearance, a simulation method based on the six-degree-offreedom motion model is proposed. Based on this method, the effects of different assembly clearances and different fluid viscosities on the performance of the DN16 cylindrical gear flowmeter are studied. The results show that the linearity error gradually decreases with the decrease of the assembly clearance. When the tip clearance is 140 μm and the gear end clearance is 100 μm, the linearity error reaches the optimal value of 0. 13%. The linearity error decreases with the increase of fluid viscosity. When the viscosity is 42. 7 mm 2 / s, the linearity error is only 0. 03%.

Abstract:RF MEMS switches have the advantages of simple fabrication process and easy integration. However, the electrostatic drift and frequent mechanical collisions caused by the charge accumulation of dielectric film currently result in serious reliability problems that hinder the stability improvement of its embedded terminal RF system. Therefore, the structure of the dielectric suspension film is used to improve the charge accumulation problem, and the switch mechanical structure limit to realize the dynamic buffer of the switch to reduce the high-frequency mechanical collision damage. At the same time relying on the tab contact structure to reduce electrostatic drift. The theoretical models of dielectric film charging and switch life are established and the switch life is predicted. The results show that the designed switch life exceeds 12 900 hours. Compared with the existing RF MEMS switches, the proposed switch structure improves the lifetimes by 253 and 166 times, respectively, and greatly improves the electrostatic drift problem when the bipolar plate spacing and the metal beam-dielectric film spacing are equal, respectively. The isolation of -41. 31 dB, loss of -0. 25 dB, and response time of 50 μs at 52. 2 GHz operating frequency provide a theoretical model for high-performance, high-reliability, and long-life RF switches.

Abstract:In the uncertainty evaluation of wind speed measured by wind speed sensors, the traditional method is to simplify the measured wind speed model and use guide to the expression uncertainty in measurement(GUM) to evaluate. However, GUM is not suitable for complex model. In order to study reliable method for the uncertainty evaluation of measured wind speed, GUM and Monte Carlo method (MCM) were used for uncertainty evaluation. On the basis of comparative analysis of results, the applicability of GUM was verified using MCM evaluation. The results show that under the simplified model, the difference between GUM and MCM evaluation is small, but only when the standard uncertainty is taken as one significant digit, GUM evaluation method is verified and evaluation is consistent; MCM evaluation under the actual measurement model are similar in envelope shape compared to GUM evaluation under simplified model, but the best estimate of measured wind speed is significantly larger, GUM evaluation method cannot be validated; When changing the distribution of some input variables, the two methods evaluated best estimated values of measured wind speed are very close. However, the inclusion interval of GUM evaluation is significantly wider than that of MCM, the probability distribution difference is significant, and GUM evaluation method cannot be validated. Therefore, appropriate evaluation methods should be selected according to the complexity of model, the distribution of input quantities and accuracy of measurement. If the distribution of input quantities follows normal distribution and measurement accuracy is not high, GUM can be used for evaluation. On the contrary, MCM is recommended to evaluate to improve the accuracy and reliability of observation results.

Abstract:To improve the quality of multi-strand carbon fiber wires, a method for detecting latent defects in multi-strand carbon fiber wires based on high-precision temperature sensors is studied. Design a high-precision temperature sensor using Brillouin optical timedomain reflection technology, and implant a single fiber into multiple carbon fiber wires to achieve temperature collection by injecting a narrow spectral light source. Using heterodyne detection method to process narrow-spectrum light sources and obtain temperature collection results of multi-strand carbon fiber wires. Using an average processing method to filter out internal noise in temperature data. Based on the filtered temperature data, the inverse problem model of the temperature field of the multi strand carbon fiber conductor is established, and the Newton Iterative method is used to solve the model, so as to obtain the thermal conductivity of the multi strand carbon fiber conductor. Experiments have shown that this method can effectively collect the temperature of multiple carbon fiber wires and successfully filter out 99% of internal noise. The thermal conductivity of detecting defect free multi stranded carbon fiber wires is 235 W/ m·k. Based on this, this method can also effectively detect latent defects in wires based on the difference in thermal conductivity compared to defect free wires. It is worth noting that the method performs better in detecting latent defects when the load current of the wire increases.

Abstract:Aiming at the strong nonlinearity and high uncertainty in the flight process of launch vehicle, and the significant impact of thrust descent faults on the reliability and safety, an interpretable machine learning model based on attention mechanism is proposed to improve the accuracy and robustness of thrust descent fault detection, fault engine location, fault degree estimation, and trajectory prediction after faults. The attention layer is used to extract the features of the high-dimensional time series flight monitoring data, and the feature matrix is used to express the high-dimensional time series data succinctly. Then the self-attention and fully connected network are used to predict the position and degree of thrust descent, the feature vector is decoded by the long-term and short-term memory unit to realize the accurate prediction of flight trajectory. The proposed integrated model is tested on the thrust descent data set to verify the effectiveness. The results show that the accuracy of the proposed model is 96. 0% for the fault location, the accuracy is 94. 7% for the fault severity estimation, and the average trajectory prediction error is 0. 94%. The proposed model has good application effect in thrust descent fault modes.

Abstract:The problem of determining the parameters of variational mode decomposition (VMD) is analyzed, and a sea clutter denoising method based on optimized variational mode decomposition (VMD) is proposed. The whale optimization algorithm (WOA) was used to optimize the number of modes K and penalty parameters, and the original sea clutter signal was decomposed adaptively to remove the modal component with low variance contribution rate (VCR). The modal component dominated by noise was screened by combining with fuzzy entropy and processed by Savitzky-Golay (SG) filtering. The least square support vector machine (LSSVM) was used to predict the sea clutter signal and verify the denoising effect for the signal reconstructed by superposition of filtered component and useful component. Simulation results show that the proposed algorithm can effectively suppress noise interference, and the root mean square error (RMSE) after denoising is 0. 000 29, which is two orders of magnitude lower than the root mean square error (RMSE) before denoising is 0. 012 3.

Abstract:To study the wireless channel of the fifth generation mobile communication system (5G) and obtain accurate wireless channel characteristics and channel model of the 5G networks, a channel measurement platform based on the test model (TM) signal of 5G new radio is proposed. The TM wireless channel measurement platform including the whole architecture and algorithm of the receiving signal process is designed and constructed by using the software radio devices and the high-performance computing equipment. The measurement performance of the TM signals is evaluated in terms of the flatness of the power spectrum, the peak to average power ratio, and the correlation. According to the results, TM signals with better measurement performance are selected as the measurement waveforms. The proposed channel measurement platform is validated based on three different methods including direct connection verification, multipath emulation verification and air interface verification. The measurement platform not only achieves accurate channel measurement, but also obtains transmission performance such as error vector magnitude and bit error rate.

Abstract:With the development of human body communication ( HBC), the study and application of human electroencephalogram signals is becoming more and more extensive. Existing research is limited to non-invasive electroencephalogram (EEG) signals and limited to invasive electrocorticography (ECoG) signals. In this study, a finite difference time domain (FDTD) approach was used to build a human model and explore the transmission properties of ECoG signals. First, the signal transmission that gain at different distances between the receiver electrodes were further analyzed in the frequency range 10 MHz to 10 GHz to determine the optimal transmission frequency band for ECoG signals. Quantitatively analyze the electromagnetic energy absorbed by the human body through specific absorption ratio (SAR) to evaluate the safety of the model. Secondly, we analyze the path loss of transceivers at different distances and show the second-order exponential decay relation between the distance and the path loss. Finally, the shadowing effect of the channel is investigated. It is shown that the channel transmission gain reaches its maximum -45. 63 dB when the carrier frequency band is around 1 600 MHz and the distance between the two electrodes of the receiver is 10 mm. The resulting path loss model conforms to a second-order exponential distribution, which can more accurately describe the transmission properties of the channel.

Abstract:The wireless sensing by multi-sensors and online monitoring technology for rail vehicle traction motor bearing is one of the key technologies to ensure the reliable operation. However, existing methods suffer from the problems of large data volume transmission difficulties and visualization of small characteristic data is not obvious. Therefore, sparse online monitoring method for rail vehicle traction motor bearing fault characteristics is proposed in this paper. The minimum entropy deconvolution method of particle swarm optimization multi-point optimal adjustment ( PSO-MOMED) is used to extract the fault characteristic signal of motor bearing under background noise. The method of discrete cosine transform (DCT) compression sensing is used to acquire the bearing characteristics with a small number of multi-sensors. Meanwhile, sparse visualization of bearing fault features is enhanced based on high order frequency weighted energy operator (HFWEO). Furthermore, the effectiveness of the proposed method is verified by setting up a test bench and measuring a line on site. The experimental results show that when the signal-to-noise ratio is -10 dB, PSO-MOMED method can extract fault characteristic signals more effectively than the traditional method. In the case of 90% compression, the frequency components of bearing fault features can be clearly characterized from the sparse perception signals of traction motor bearing fault features. It effectively

Abstract:To realize intelligent water management and control and flood disaster early warning, it is necessary to accurately sense the change of water level information in real time. Because the prior technology cannot meet the requirements of water level identification in complex and harsh environments such as night, fog, rainy day, floating object occlusion, light shadows, etc. , an intelligent water level detection method based on improved YOLOv5 and RankSE was proposed. Firstly, the YOLOv5 algorithm was improved by the multi-level feature fusion method which strengthens small-scale features, to strengthen the ability of capturing small targets. Secondly, integrating the RankSE module further enhances the perception of small targets. Finally, a new solution of water level elevation was proposed, which can obtain accurate water level elevation information only by using part of water gauge anchor frame information, which greatly improved the robustness of the detection method. The research results show that the accuracy of water level detection in this paper reached 98. 5%, which was 8. 4% higher than the original algorithm. The water level elevation could be automatically and accurately identified in complex and harsh environments. The maximum error was only 0. 11 m. The research results effectively improve the accuracy of water level detection in complex and harsh environments.

Abstract:Taking the standard gauge block as the experimental object, to address the problems that the chamfer features of the parts are prone to wide edges in the parts image under the illumination of ring light source, and the inaccurate image focusing is caused by manual focusing, which lacks the objectivity of camera focusing, etc. , a high-precision focusing method of parts image based on improved gradient weighting is proposed. Firstly, the illumination method of the strip light source arranged at a 45-degree angle is adopted to eliminate the wide edges of the chamfer features in the parts image. Secondly, the feature edge points of image are extracted by the adaptive segmentation threshold based on the improved Otsu. Then, the gradient values of edge pixels are obtained based on the 4- direction Sobel operator. Then, according to the grayscale distribution difference between the pixel and its 8 neighboring pixels, the gradient weighting coefficient of the pixel is obtained. Finally, the sharpness evaluation of image is completed by the improved function of gradient-weighted focus evaluation, thus, the accurate focus image is obtained and the high-precision measurement of size is realized. The experimental results show that the proposed method is more accurate than the traditional high-precision measurement method, and the relative error with the manual measurement is less than 0. 002 4%. The improved focus evaluation function in this paper is improved by 75 times in sharpness ratio, 5 times in sensitivity factor and double in steepness on average compared with the traditional evaluation functions

Abstract:Since terahertz waves can penetrate non-metallic and non-polar materials, terahertz technology can make up for the shortcomings of laser technology and realize perspective ranging of the inside of the target. Terahertz perspective ranging technology based on chirp continuous wave is proposed, which realizes the non-contact measurement of the thickness and dielectric constant of multilayer dielectric materials, further expands the application range of terahertz chirp continuous wave, and provides a new way for the measurement of the dielectric constant of materials. Taking three common materials as the detection target, method is verified by using terahertz linear frequency modulated continuous waves in four frequency bands: 0. 11~ 0. 17 THz, 0. 17~ 0. 22 THz, 0. 22~ 0. 33 THz, and 0. 33~ 0. 50 THz, and it is proved that the measurement accuracy increases with the increase of the signal bandwidth, with errors below one percent. In addition, taking the ceramic material used to make radomes as the inspection target, applying the principle of terahertz perspective ranging and using the two-dimensional scanning frame to obtain all the information of this sample, perspective imaging and positioning of internal defects are realized.

Abstract:As the main information source of intelligent substation, electronic transformer needs to be checked regularly, and ensuring time synchronization is the most critical link in the process of transformer calibration. Based on the previously designed real-time collaborative monitoring system for CT side cloud, this paper constructs a multi-CT side cloud collaborative measurement system. Both standard CT and verified CT are integrated with the monitoring system to form a CT side cloud collaborative measurement unit. Each measurement unit transmits the data collected by CT remotely in real time to the cloud storage, then uniformly exported to the edge computing side for CT verification calculation. Theoretical clock errors and link delay errors of data uploading to the cloud between each measurement unit are analyzed, and a master/ slave clock error model of CT calibration is established. A master/ slave clock error correction algorithm based on KF and RBF-PID is proposed for CT calibration to ensure time synchronization during CT calibration. The proposed algorithm is verified by simulation. Under different initial conditions, the corrected errors of master and slave clocks are all less than 200 ns, and the convergence time is about 0. 4 s. Finally, the feasibility and practicability of the proposed algorithm are verified by experiments. The calibration results of the CT corrected by the error of the master and slave clock are more accurate, with the phase error of about 0. 45″ and the amplitude error of about 0. 003 58%, meeting the requirement of 0. 1 accuracy level.

Abstract:Traditional ceramic defect detection mainly relies on manual visual inspection or magnifying glass observation. In order to solve the problems of low detection efficiency and strong subjectivity of results, this paper proposes a ceramic surface defect detection algorithm based on deep learning. According to the specific situation of the defects on the surface of the ceramic cup, a small target detection layer is added on the basis of the YOLOv5 target detection model, at the same time, the position attention mechanism is used for feature reconstruction to improve the detection accuracy, and high-precision defect detection is achieved. According to the actual production of ceramic double-layer cup data acquisition training, and reasoning for each batch of data, the final average detection accuracy reached 95. 4%. The improved YOLOv5 defect detection model in this paper has the advantages of higher accuracy and faster recognition speed, which can greatly reduce the loss of human and material resources and time cost in ceramic quality inspection.

Abstract:As one of the important components in the antenna, the ceramic antenna PIN needle, its size deviation will directly affect the product quality of the antenna. In order to realize the fast and precise online detection of PIN needles, an online detection device for PIN needles was designed and developed, and a PIN needle size detection method based on an improved sub-pixel algorithm was proposed. First, start to collect images and perform pixel equivalent calibration, perform distortion correction on the image, obtain ROI area, and image preprocessing, then use the sub-pixel edge detection algorithm based on the improved Sobel operator and Gaussian peak position estimation to extract edge points, and use the least squares method, then the edge points are fitted into a pair of parallel straight lines, and the pixel width between the lines is calculated, and the diameter of the measured PIN needle at the ROI area is obtained by converting the pixel equivalent. The experimental results show that the average relative error of this method is less than 0. 25%. While ensuring ±0. 005 mm detection accuracy, its average time is reduced by 64. 32% compared to traditional sub-pixel detection algorithms based on Gaussian fitting.

Abstract:An impedance control strategy based on non-singular fast terminal sliding modes is proposed to solve the problem of contact control during cooperative transportation of double cable-driven aerial manipulators. Firstly, considering the flexibility effect of the cabledrive mechanism, the dynamics model of the dual-machine collaboration system taking into account the flexibility of the joints is established, and the impedance equation in the task space is derived. Then, a cooperative impedance control strategy is designed to accelerate the system state volume and convergence speed with the use of non-singular fast terminal sliding mode surfaces, and to weaken the output torque jitter effect using power functions. At the same time, the stability of the controller is analysed in the Lyapunov framework. Finally, the effectiveness of the proposed control strategy is verified by controller comparative performance simulation and cooperative transportation simulation. The results show that the proposed controller has fast response speed, high control accuracy and excellent stability, and can effectively reduce the system chattering.

Abstract:In response to the problem of distributed cooperative control of high-speed trains under uncertain factors and external interference, a high-speed train speed tracking control strategy based on the super-twisting sliding mode consensus algorithm is proposed. Firstly, considering the external interference, basic resistance, and coupling forces between carriages that the train is subjected to, a multi-agent model for high-speed train is established. Secondly, the consistent sliding mode function is designed according to the displacement and speed information of adjacent carriages, and the super-twisting algorithm is introduced to reduce control input chattering. Finally, design a distributed second-order sliding mode control law and use Lyapunov theory to verify the stability of the algorithm. Conduct a simulation study using the actual parameters of a high-speed train and adding external interference, simulation is carried out with the method proposed in this paper, as well as the ordinary consensus, PID consensus, and sliding mode consensus methods. The results show that compared with the other three algorithms, the proposed algorithm can enable the train unit to quickly and accurately track the target speed curve, with a speed error within (-0. 8~ 1. 1)×10 -3 m/ s, while keeping the distance between adjacent train cars within a safe range. Moreover, the control input is relatively smooth, and the algorithm has good robustness against external disturbances.

Abstract:To accurately extract phase from the noisy interferograms, an adaptive wrapped Kalman filter phase unwrapping algorithm based on Zernike polynomial fitting is proposed in the paper. The phase map is modeled as Zernike polynomial fitting, the fitting coefficients are accurately calculated using wrapped Kalman filter with a quadratic difference weighting strategy for the phase map, and the measurement noise covariance matrix is adaptively adjusted by the innovation of the predicted and measured states without setting different observed noises according to the interferogram. The results of simulation and experiment show that the proposed method can effectively deal with the noisy interference fringe and retrieve the phase map, and is superior to the least square method and Kalman filter for phase unwrapping of noisy interferograms, which has the advantage of good robustness and no need for pre-filtering and manual intervention.

Abstract:In order to reduce the influence of weather factors on the prediction accuracy of the output value of photovoltaic power generation, it is proposed a prediction model incorporating the clustering algorithm ( KDGMM), the improved variational modal decomposition (VMD) and the stochastic configuration network ( SCN), starting from both cluster analysis and signal decomposition. Firstly, the meteorological data are classified into sunny, cloudy and rainy days by KDGMM clustering, and for the problem that it is difficult to predict accurately on cloudy days, gray correlation analysis (GRA) is used to select similar days, and secondly, the L􀆧vy northern goshawk optimization (LNGO) algorithm is introduced to optimize VMD to get the optimal parameters, so as to reduce the nonsmoothness of PV power on cloudy days. Finally, the SCN prediction model is constructed to predict the PV power data and output its prediction results. Through experimental analysis, the root mean square error (RMSE) and the mean absolute percentage error (MAPE) of the proposed method are only 1. 44 and 1. 3%, and the R 2 index for goodness of fit is as high as 0. 99. Compared with other prediction methods, the proposed method has higher prediction accuracy

Abstract:Aiming at the problem of small target and low detection accuracy of printed circuit board surface defects, Multi-CR YOLO, a printed circuit board surface defect detection network, is designed to meet the premise of real-time detection speed and effectively improve the detection accuracy. Firstly, the backbone feature extraction network Multi-CR backbone, which consists of three Multi-CR residual blocks, performs feature extraction for small target defects on printed circuit boards. Secondly, the SDDT-FPN feature fusion module enables the feature fusion from the high level feature layer to the low level feature layer, and at the same time strengthens the feature fusion for the feature fusion layer where the small target prediction head YOLO Head-P3 is located, to further enhance the expressive ability of the low level feature layer. The PCR module strengthens the feature fusion mechanism of the different scales of the backbone feature extraction network and the feature layer of the SDDT-FPN feature fusion module, and prevents the fusion mechanism between the modules. The C5ECA module is responsible for adaptively adjusting the feature weights and adaptively paying attention to the requirement of small target defect information, which further improves the adaptive feature extraction capability of the feature fusion module. Finally, the three YOLO-Head are responsible for predicting small target defects for different scales. The experiments show that the detection mAP of the Multi-CR YOLO network model reaches 98. 55%, the model size is 8. 90 MB, which meets the lightweight requirement, and the detection speed reaches 95. 85 fps, which meets the application requirements of real-time detection of small-target defects.

Abstract:Aiming at the problem that the cable arrangement accuracy of the current cable arrangement methods for ocean winch is easily affected by environmental interference and contact friction and wear, the cable arrangement method is used by the non-contact visual detection method to measure the cable arrangement clearance which directly characterize the quality of the cable arrangement, and a subpixel measurement method for marine winch cable arrangement clearance based on improved Zernike moments is proposed in the paper. Firstly, the two-dimensional image gradient information is vertically projected into the one-dimensional waveform, and a vertical projection cable arrangement clearance location method based on Scharr gradient information is proposed to quickly locate the clearance area. Secondly, an improved sub-pixel edge detection method based on Zernike moments is developed by the main gradient direction interpolation to enhance the sub-pixel edge detection accuracy. Finally, a least-squares curve fitting method based on DBSCAN is used to fit multiple non-intersecting curves for measuring the cable arrangement clearance size. Experimental results show that the proposed method has higher accuracy and measurement precision by comparing with sub-pixel detection methods based on fitting method, interpolation method, and traditional Zernike moment, its measurement error is less than 0. 1 mm and relative error is less than 8. 60%, which provides a prerequisite guarantee for achieving accurate cable arrangement.

Abstract:A fault location method based on GraphSAGE (graph sample and aggregate, GSA) algorithm is proposed in this paper. The morphological black hat operation is performed on the power-side bus voltage of the distribution network as the fault detection criterion to start the fault location algorithm. The GSA model is used to independently mine topology and zero sequence current features, and function mapping is established according to node features and labels to evaluate the running state of the line to achieve fault location. Based on the PSCAD/ EMTDC simulation platform, an IEEE 33-node simulation model is constructed to acquire data resources and validate the proposed method. Reliable fault location results are obtained by applying the proposed method. Furthermore, in distribution networks with topological changes, the model can obtain reliable fault localization results without retraining, which verifies the robustness and adaptability of the method to topological changes