Abstract:Multiphase flows widely exist in modern industries. The velocity measurement of multiphase flow is the key to ensure safety as well as improve the quality and efficiency of the production processes. Cross-correlation velocimetry is one of the main methods for the velocity measurement of multiphase flows. However, the classical cross-correlation algorithm is difficult to meet the demand of real-time and accuracy due to the complexity of industrial multiphase flows and thus has some limitations in application. In this paper, recent progress of the method is reviewed and the improvements that have been made to the method are elaborated and analyzed in terms of calculation efficiency and accuracy. Evaluation results show that both the adaptive cross-correlation method and the frequency domain method can obtain accurate results with less computational effort. Besides, the frequency domain method can also obtain the most stable measurement results. Based on the state-of-the-art cross-correlation velocimetry, possible future research is discussed for further development of the method.

Abstract:Tangential turbine flowmeter is applied to oilfield cementing engineering, the impeller structure will have a direct impact on the metering effect. In order to improve the measurement accuracy of flowmeter and the reliability of impeller structure, the impeller structure parameters are optimized. Combining response surface with crowd search algorithm, a three factor four level orthogonal test is designed, and the linearity error of flowmeter instrument coefficient K is established by response surface method δ and the quadratic regression model of flowmeter impeller radius, impeller edge thickness and impeller inclination angle. Taking the minimum linearity error as the optimization objective, the impeller structure parameters are optimized based on crowd search algorithm. The optimal parameter combination of impeller structure is obtained. After optimization, the linearity error of instrument coefficient is reduced from 3. 106 0% to 0. 445 2%. The measurement accuracy and stability of cementing mud flowmeter are improved, which provides theoretical support for subsequent optimization design.

Abstract:In response to the application requirements of phase-shifting arrays in massive MIMO and wireless backhaul systems for the fifth-generation mobile communications, in this paper the topology of the reflection-type phase shifter ( RTPS ) is compared and analyzed, the design method of the four-element dual-adjustable (FEDA) load topology and the adjustment method of phase conversion are proposed, and the influence factors of phase shift step and insertion loss are analyzed. Moreover, the digital tunable capacitor (DTC) is used for RTPS tuning design. The experimental results show that the RTPS in this paper has a measured phase shift range greater than 360°, a phase shift step less than 12°, an insertion loss less than 1. 8 dB and the size only 10 mm× 8 mm in the full frequency range of 4. 4~ 5. 0 GHz. In summary, this RTPS has the advantages of low insertion loss, miniaturization, large bandwidth, high precision and easy control.

Abstract:According to the requirements on manipulator robot in the electromagnetic compatibility measurement, this paper proposes the design methodology and control strategy of 6-DOF manipulator in emission source microscopy (ESM) task. The task constraints, the structure of the manipulator, orientation of the antenna and the integrity of scanning surface are modeled, and the ESM sampling cylinder is generated according to the principle of area optimization. The improved RRT∗ algorithm based on joint search strategy (RRT∗-JSS) realizes the optimal planning of the joint position state of the manipulator under a given scanning plane. In the simulation experiments, above method and strategy enable the manipulator to achieve a reduction of 68. 96% in the average joint change angle, under the scanning path loss of 4. 3%. In the verification experiment based on Elfin5 manipulator and patch-shaped radiation source, the error between the actual obtained ESM imaging and the simulation results is below 0. 35 dBm. Above simulation and experimental results prove the effectiveness of the studies, which is expected to apply in the field of EMI measurement.

Abstract:The construction of a smart grid cannot be separated from advanced sensing and measurement technologies. Compared with the traditional giant magneto-resistive current sensor with an opening in the magnetic concentrating ring, this paper proposes a sealed magnetic concentrating ring applied to the giant magneto-resistive current sensor, including the shape and structure of the sealed magnetic concentrating ring, and the air gap opening. The design of the size and the degree of unevenness, and the selection of the materials used in the sealed magnetic concentrator. Using COMSOL Multiphysics to simulate the model of the sealed magnetic concentrator, the relationship between the magnetic saturation degree of the sealed magnetic concentrator and the traditional magnetic concentrator and the uniformity of the magnetic field at the center of the air gap was compared and analyzed. The simulation results show that the magnetic concentrating ability of the sealed magnetic concentrator is 764 times that of the traditional open magnetic concentrator, and its structure can make the giant magneto-resistive sensor have lower hysteresis and better sensitivity than the traditional open magnetic ring. It can be used for real-time, accurate and reliable current measurement in the field; small current experiments have verified that the magnetic concentrating effect of the sealed magnetic concentrating ring is remarkable under weak magnetic fields. The magnetic field generated by the current of 0. 2 A and below acts on the giant magnetoresistive chip after being concentrated by the sealed magnetic concentrating ring. The output voltage of the giant magnetoresistive chip after sealing is more than 5 times that before sealing.

Abstract:A picosecond event timer based on high-speed ring oscillator is studied and designed. The rising edge of the signal representing the event is used to trigger the high-speed ring oscillator to generate the clock pulse signal synchronized with the event. The sinusoidal reference signal is sampled and processed by the all phase FFT algorithm to greatly improve the precision of event timing measurement. Experimental results show that in the case of a sinusoidal reference signal with 10 MHz frequency, a 14 bit ADC with 140 MHz sampling frequency, and the all phase FFT with operation point number N = 8 192, we have achieved the single-shot time interval measurement precision of 3. 16 ps rms and the time stability are better than ±0. 31 ps/ h. The results are in good agreement with the error budget based on the theoretical analysis, reaching picosecond-level event timing measurements.

Abstract:Since the flexible NFC tag antenna is usually in a deformed working state, in order to solve the problem of frequency offset and the performance degradation caused by the frequency offset in this working state, and to improve its applicability in the application of wireless human health monitoring, a flexible NFC tag antenna with human skin adaptation and excellent performance is designed. The antenna uses gallium indium liquid metal as a flexible conductive material, combined with PDMS flexible substrate material to enhance the stretchability and flexibility of the NFC tag antenna. Based on the principle of miniaturized NFC antenna design, this paper proposes a novel non-equal spacing structural design scheme to increase the working bandwidth of the antenna, maintain relatively stable electrical parameters, resonant frequency as well as communication distance under different deformed states and improve the working stability. The simulation and experimental results show that the working bandwidth below -10 dB of the non-equal spacing NFC tag antenna with an external size of 25 mm× 25 mm and a line spacing growth rate of 200% can reach 0. 87 MHz. Compared with an equal line spacing antenna with the same area, the maximum increase in working bandwidth is up to 52. 6%. Aiming to the particularity of the application environment of this flexible NFC tag antenna, in the stretching working state with a stretching degree of 30% and the bending working state with a curvature radius of 5 mm, the center frequency deviation of the antenna is 3. 2% and 2. 4%. The antenna communication distance is always no less than 30 mm, and it shows good working stability under deformed working states.

Abstract:Bandwidth and efficiency are two important indicators of power amplifiers design, how to make the power amplifier meet the design requirements of wide bandwidth and high efficiency at the same time has been one of the hot spots and difficulties in the research of amplifiers. For the problems mentioned above, this paper proposes a continuous class F ultra-wideband amplifier, based on hardware intelligence and continuous power amplifier theory, expanding impedance matching space with continuous theory, which promote traditional class F power amplifier to continuous class F power amplifier. And the scattered frequency bands are integrated into a whole through reconfigurable technology, to broaden its working bandwidth effectively. By making the physical objects and testing, in the range of 0. 9~ 4 GHz, the power added efficiency (PAE) is greater than 72%, the average gain is about 12. 5 dB, and the saturated output power is 41 dBm. This paper combines the advantages of both reconfigurable technology and continuum theory. The designed power amplifier has high efficiency, wide operating band, and circuit flexibility, which can adapt well to the requirements of 5G wireless communication system.

Abstract:In view of the problem that the early performance degradation point of rolling bearing is difficult to monitor, a method based on improved VMD and comprehensive characteristic indexes performance degradation assessment was proposed. Firstly, the parameters of VMD were optimized by Kullback Leibler divergence (KL-divergence), the bearing vibration signals were decomposed by the optimized VMD, and the modal components that sensitive to degradation characteristics were screened by wasserstein distance ( WD) method. Then singular value decomposition ( SVD) was performed to obtain singular value characteristics. Secondly, the comprehensive characteristic index of rolling bearing degradation was composed by combining the entropy energy rate ( EER) and confidence value (CV). Finally, SVDD model was used to calculate performance degradation index to realize early weak fault detection and performance degradation assessment. The validity of the proposed method was verified by using the bearing life cycle experiment data, the detection results of early performance degradation points are earlier than other degradation assessment methods, which provides a new idea for performance degradation assessment of rolling bearings.

Abstract:According to the current development trend of wireless communication systems, a miniaturized dual-unit ultra-wideband (UWB) multiple-input multiple-output (MIMO) antenna with high isolation is designed, and its overall size is only 18 mm×22 mm× 1 mm. The antenna’ s radiation patch adopts a rectangular structure etched with a hexagonal star, and a “ w” shape and a small rectangle are cut out on the upper and lower positions on the right side of the antenna to realize the broadbandization of the antenna. The high isolation of the antenna is achieved by using defective ground (DGS) structure and etching two semicircular slots of equal radius at the top of the ground plate. The simulation results are basically consistent with the actual measurement results. The final operating bandwidth of the antenna is 2. 75~ 10. 64 GHz (relative bandwidth reaches 117. 85%), and the isolation within the bandwidth range is greater than 22 dB, and the envelope correlation coefficient (ECC) is less than 0. 015. Therefore, the designed antenna is not only extremely small in size, simple in structure, but also good in performance, and can be widely used in wireless communication systems.

Abstract:In order to accurately evaluate the interior noise, a sound quality evaluation method of interior noise based on time-frequency energy characteristics is proposed. First, the noise signal is adaptively decomposed by variational modal decomposition, and a series of intrinsic modal function components are obtained. Then, the instantaneous intensity and weighted energy of each component are obtained through Hilbert transform, which are used as the time-frequency energy characteristics of the noise signal. On this basis, a sound quality evaluation model based on genetic algorithm optimal back propagation (GA-BP) neural network is established with the time-frequency energy characteristics and the psychoacoustic parameters as the input. The proposed method is used to evaluate the interior noise of a vehicle. The correlation between the results and the subjective evaluation results reach 93. 7%, and the relative error is less than 8. 0%, which accurately reflects the sound quality of the vehicle interior noise. The proposed method enjoys a high accuracy and has a good application prospect in the practice of vehicle sound quality development.

Abstract:Traditional operation and maintenance knowledge base does not have the ability to identify the failure phenomena in the image. Therefore, the knowledge base cannot handle the problem of unstructured data. To tackle this issue, based on fault classification networks in deep learning, an improved capsule network feature extraction structure based Caps-DRFN algorithm is proposed, which can realize automatic classification of operation and maintenance images of electromechanical equipment. Firstly, aiming at the multi-noise problem of operation and maintenance images, the deep residual shrinkage networks (DRSN) are introduced to improve the feature extraction performance of the model on noisy data. Subsequently, for the multi-scale problem of actual shooting operation and maintenance images, through the combination of the feature pyramid networks (FPN) algorithm, the Caps-DRFN realizes image multiscale feature fusion and improves the accuracy of model classification. Finally, the vector neuron is constructed by using the capsule structure, and the digital capsule of the classification structure is obtained through the feature transmission method of dynamic routing. The model realizes the fault classification of electromechanical equipment. The experimental results show that compared with the traditional capsule network algorithm, the accuracy of the proposed Caps-DRFN algorithm based on feature fusion is increased by 15% and it is more robust.

Abstract:Due to the rapid development of human body communication technology and broad application prospects, this paper proposes a biometric identification method based on human body communication technology. In the UWB frequency band and HBC frequency band suitable for human body communication, the path loss of the human body communication link is used as a biological feature. First, use the support vector machine for identification according to the measured values of different links; then use the C-SVM and Nu-SVM methods with different kernel functions for identification under the data set of 11 links; finally select 8 ~ 10 GHz UWB sub-band is recognized to improve the calculation speed. The results show that: the longer the link communication distance, the higher the recognition rate. The C-SVM with Gaussian kernel has the best recognition effect in the UWB frequency band, with a recognition rate of 96. 41%, AUC of 0. 999 1 and EER of 0. 017 2%. By selecting the sub-band to reduce the calculation time to 0. 142 s, the speed is significantly improved.

Abstract:Aiming at the problem of low recognition rate of traditional target detection algorithm for inner wire joint sealing surface defects, an improved YOLOv4 algorithm was proposed to detect the defects. Firstly, k-means++ clustering algorithm is used to optimize the parameters of the anchor frame of the target sample, and improve the matching degree between the anchor frame and the feature map; Secondly, the SENet attention mechanism module is introduced into the backbone network to strengthen the key information of the image, suppress the background information of the image, and improve the confidence of the defect that is not easy to identify; after that, the SPP module is added to the neck of the network to enhance the acceptance domain of the backbone network output features and separate the important context information; Finally, using the collected data set of inner wire joint sealing surface defects to train the original YOLOv4 and the improved YOLOv4, and the performance of models were tested respectively on test set. The experimental results show that the performance of YOLOv4 is good, but some small targets are missed; The improved model has excellent detection performance for small target defects, the mean average accuracy (mAP) reaches 87. 47%, which is 10. 2% higher than the original YOLOv4, and the average detection time is 0. 132 s, which realize the rapid and accurate detection of inner wire joint sealing surface defects.

Abstract:Aiming at the low efficiency and stability of the traditional two-dimensional bar code positioning algorithm in complex industrial and logistics transportation scenarios, a two-dimensional bar code positioning algorithm based on lightweight CenterNet network is proposed, a lightweight CenterNet detection algorithm is proposed. In view of the size change of two-dimensional bar code in the actual situation, CSPDarknet53-tiny is used as the backbone network and modified SPP module is added to improve the accuracy of the network. The upsampling and detection head of CenterNet are lightweight transformed. 5 × 5 depth separable convolution is used to replace ordinary convolution. The change strategy of learning rate during training adopts cosine annealing learning rate to prevent over fitting. The experimental results show that the positioning accuracy is only 0. 64% lower than YOLOv4 tiny. It not only avoids the problems that the accuracy of the traditional algorithm is greatly affected by the background and the robustness is not strong, the real-time reasoning speed also reaches 124 fps, which can be better used for all kinds of two-dimensional bar code location under low hardware configuration.

Abstract:In order to solve the uneven illumination of phase contrast microscopic cell images, achieving the goals of accurate and fast cell image segmentation and automatic counting, this paper proposes a method of image illumination equalization combined with double threshold segmentation and counting. The method chunks the image, adjusts the brightness based on the ratio of sub-blocks to the average gray level of the whole image, and optimizes the ratio by using Gaussian function weights to eliminate the “square effect” caused by image chunks, and determines the optimal standard deviation by the relationship curve between image entropy and Gaussian function standard deviation. The images were segmented using a double-threshold method to optimize the morphology of the restored cells by combining cavity filling and area constraints. The algorithm was tested using the C2C12 phase-difference microscopic cell image dataset, in which the segmentation accuracy, recall and F-value of the high cell density images were 0. 966 2, 0. 967 8 and 0. 967 0, respectively, which were significantly better than other comparative methods. The results showed that the method could achieve light equalization when processing the light inhomogeneous phase contrast cell images of different cell densities, with less image information loss, high-accuracy cell-segmentation and counting results.

Abstract:The most important task in pedestrian trajectory prediction is to establish a pedestrian trajectory interaction model. Aiming at the lack of semantic information about time and speed in the model, a spatial-temporal graph network algorithm combined with speed control is proposed to establish pedestrian interaction model and predict trajectory. The overall model adopts the conditional generative adversarial networks architecture, in which the speed prediction module is used to predict the future speed of pedestrians, and the control condition of the conditional generative adversarial networks. The speed information is explicitly introduced into the pedestrian trajectory prediction to avoid the influence of large deviation speed on the trajectory. A spatial-temporal information fusion module is designed in the generator. While extracting the motion features of pedestrian trajectory sequence and paying attention to its spatial interaction, it explicitly encodes the temporal correlation of pedestrian sequence. Finally, the trajectory interactive features combined with space-time information and speed information are decoded to complete the trajectory prediction. In addition, considering the shortcomings of the existing evaluation methods, the average collision times is used as the evaluation of trajectory rationality. The model is verified on the public datasets ETH and UCY. The experimental results show that the proposed algorithm can better complete the pedestrian trajectory prediction, with an average displacement error of 0. 40 m and a final displacement error of 0. 79 m.

Abstract:Ultra-wideband signals have the advantages of low power consumption and anti-attenuation in human body communication technology. Aiming at the problem of the correlation between shadow fading and motion state in the human body channel modeling work, this article establishes a human body model based on the human body's different motion postures and the electromagnetic characteristics of tissues, and analyzes the dynamic human body channel shadow fading characteristics under ultra-wideband signals. First, by establishing 80 data transmission links to analyze the body surface propagation characteristics of the human body in different states, a second-order exponential attenuation relationship between body surface propagation distance and path loss is given. Secondly, the influence of changes in the electrical characteristics of the dynamic model organization on the path loss is analyzed. Finally, the average bit error rate is used to study the performance difference of UWB system under dynamic shadow fading. The results show that the antiloss performance of the UWB frequency band under the dynamic human body model is better than that of the HBC frequency band, and the intensity of shadow fading is affected by the motion state more than the HBC frequency band. The research provides a theoretical basis for dynamic human shadow fading intensity distribution in the modeling and application of ultra-wideband signals in human body communication.

Abstract:To solve the problem of serious spectrum performance loss of hybrid precoding in millimeter wave multiple input multiple output (MIMO) systems, a hybrid precoding algorithm based on alternating optimization orthogonal matching pursuit is proposed in this paper. Firstly, an analog precoding matrix under constant mode constraint is used to determine the good initial connection state between the stationary phase displacement phaser and the antenna to improve the iterative convergence rate. Then, the optimal candidate simulation precoding matrix is constructed according to the connection state to solve the global optimal index vector. Finally, the digital precoding matrix composed of the optimal index vector is fed back to the dynamic network, which can dynamically optimize and update the connection state of the phase shifter and the antenna array alternately. At the same time, the proposed algorithm only needs a small number of fixed phase shifters to achieve a good balance between spectrum performance and complexity. Simulation results show that compared with other existing algorithms, the proposed algorithm has higher spectral efficiency, higher iterative convergence rate and lower complexity, especially when the number of RF links is greater than the number of data streams, the improvement of spectral efficiency is more significant.

Abstract:A lightweight action recognition network with fused attention is proposed to deal with the three problems of the traditional 3D convolutional neural network: large number of parameters, information redundancy and insufficient extraction of temporal information. First, in order to lighten the network parameters and fuse short-medium-long temporal information, an efficient residual block is developed to replace two cascaded 3×3×3 convolutions; second, by extending the channel attention mechanism, a temporal attention mechanism is derived, and both of the two mechanisms are integrated into the proposed network to suppress the influence of redundant information on recognition results; finally, experiments are conducted on the UCF101 dataset to verify the effectiveness of the network. The results show that the proposed action recognition network has a computational cost of 8. 9 GFlops, a parameter amount of 18. 0 M, and a recognition accuracy rate of 94. 8%, which reveals a high recognition accuracy with a low cost computation in comparison with other behavior recognition networks.

Abstract:Currently, reliable detection and accurate counting of firearm projectiles is one of the difficult points of gun and ammunition management. To improve the accuracy and reliability of projectile detection algorithm based on acceleration signals, we propose a new time-domain feature extraction method for firearm firing signals: The time-domain segmental feature extraction method, which avoids the problem that time-domain features are overly dependent on acceleration transient spikes. Firstly, various statistical features of the sample signals of gunshot acceleration in the time and frequency domains have been extracted. Then machine learning classification algorithms K-nearest neighbors, logistic regression, support vector machines, decision trees and random forests are used for gunshot recognition modeling. Finally, the effects of various single features on the performance of gunshot recognition models are explored and compared. The experimental results show that the extracted main fluctuation domain area feature have the optimal discrimination and can achieve more than 99% classification accuracy on most machine learning algorithms.

Abstract:In order to solve the problem that atmospheric lidar detection is easily interfered by noise and the signal-to-noise ratio (SNR) of distant signals drops rapidly, according to the long sequence characteristics of the lidar detection, an improved variable weighted Kalman filter method for lidar detection is proposed. A constant term is added to the variable weighted coefficient in the algorithm. Therefore, the changing weighted coefficient can be provided for the long sequence measurements values at different time in the improved variable weighted Kalman filter algorithm. The correction effect of the new measurement is enhanced and the influence of the old measurement on the optimal estimation is reduced in this algorithm. The algorithm is verified by the actual atmospheric lidar measurements in different weathers. Compared with the other three Kalman filtering algorithms, under cloudy weather, the SNR of lidar detections are improved by nearly 4. 9, 3. 7 and 2. 5 dB, respectively. The inversion error of aerosol extinction coefficient is reduced by 57%, 26% and 4% respectively. In cloudless day, the signal to noise ratio of lidar echo signals are improved by nearly 5. 5, 4. 4 and 3. 4 dB respectively. The inversion error of aerosol extinction coefficient is reduced by 53%, 25% and 3% respectively. The inversion accuracy of atmospheric aerosol optical properties are improved using this algorithm. An effective method for fine detection of aerosol microphysical parameters and practical application of lidar is provided.

Abstract:Numerical simulation is used to study the optimal values of the balancing hole number and the diameter of hole opening circle of a balanced orifice plate. It introduces the method and procedure to obtain optimal balancing hole number and circle diameter ratio are presented for the case of pipe diameter D= 100 mm and equivalent diameter ratio of β = 0. 6. It is found that the optimal hole number is N= 10, after comparing the discharge coefficient, pressure loss and momentum difference for different hole numbers with a fixed circle diameter ratio of K= 0. 67, and the optimum circle diameter ratio is K= 0. 67, after comparing the above-mentioned evaluation indexes for different circle diameters with a fixed balancing hole number of N = 10. The optimal mechanical parameters are given for various combinations of five pipe sizes within DN50 to DN1000 and six equivalent diameter ratios. The optimal balancing hole number is 8 or 10, and the optimal center-circle diameter ratio K is between 0. 64 and 0. 73 ( in most cases, it is between 0. 66 and 0. 70). In the paper, the concept of momentum difference is proposed to evaluate the performance of a balanced orifice plate, and it is proved that ΔM reaching to the minimum value can be used as an auxiliary criterion for achieving an optimal mechanical parameter. The results of the study are of reference value for the mechanical design of a balanced orifice plate.

Abstract:In order to identify PD types in GIS effectively and ensure the safe and stable operation of equipment, a PD type in GIS identification method based on EEMD energy moment and ISSA-SVM algorithm is proposed. Firstly, a GIS partial discharge experiment platform that can produce four PD effects is built to obtain four PD signals. Then, EEMD and energy moment are used to decompose the modes and extract the feature vectors of the four PD signals. Finally, ISSA-SVM algorithm is used to identify GIS PD types. Experiment results show that the proposed method can identify different PD types in GIS effectively, and the recognition accuracy is improved by 16. 7% and 8. 5% respectively compared with PSO-SVM and SSA-SVM algorithm. The effectiveness and superiority of the proposed PD type identification method in GIS are verified by the experiment.

Abstract:Aiming at the coupling of vibration signals and inaccurate fault diagnosis of planetary gearbox, a fault diagnosis method of planetary gearbox based on feature fusion and ResNet is proposed. Firstly, the collected analog fault vibration signals such as planetary gear crack, wear, sun gear broken tooth and composite fault are decomposed by MEEMD and VMD to screen and determine the effective components respectively. Then, the selected effective features are fused and classified by using traditional CNN network and ResNet. The results show that the ResNet has higher classification accuracy, up to more than 95%. Finally, the classification accuracy of data before and after feature fusion is compared by using ResNet. The accuracy before fusion was only 91. 16%, which was lower than 97. 18% of after fusion. Thus, this method is very effective for coupling vibration signal processing and fault diagnosis of planetary gearbox.

Abstract:To address the problems of poor differentiation of soft fault features of power electronic circuits and not easy to diagnose, a fault diagnosis method of variational modal decomposition ( VMD) combined with an improved sparrow search algorithm ( ISSA) optimized extreme learning machine (ELM) is proposed. Firstly, the acquired fault signals are decomposed into the intrinsic modal components (IMF) by VMD, and the twelve-dimensional time-domain parameters of the linearly reconstructed IMF are extracted as the feature vectors for fault diagnosis. Secondly, in order to improve the accuracy of ELM in fault diagnosis, ISSA is proposed to optimize the parameters of ELM and establish ISSA-ELM classification model. ISSA is improved by three strategies such as initializing the population with Iterative mapping, introducing adaptive inertia weight factor at the discoverer position update, and introducing levy variation operator to perturb at the solution position to get a new solution to improve the algorithm performance. In the 8-class benchmark function test, ISSA has improved the convergence speed and finding accuracy than the other 4 intelligent algorithms, and the accuracy of VMD combined with ISSA-ELM reaches more than 99% in the soft fault diagnosis of 150 W Boost circuit.

Abstract:Traditional A ∗ algorithms plan paths that run the risk of hitting obstacles and create too many redundant nodes. For these problems, a path planning with improved search 8-neighborhood node strategy A ∗ algorithm is proposed. Firstly, the node search condition is improved in the A ∗ algorithm so that the obtained nodes maintain a safe distance from the surrounding obstacle nodes. Secondly, the redundant nodes on the improved path are removed using the vertical distance limit method, and the critical nodes are retained. Finally, a smooth path is obtained by fitting a B spline curve to the key nodes to achieve a smooth path. By conducting several experiments and comparisons in multiple obstacle map environments of different scales. The results show that the proposed improved A ∗ algorithm, compared to the traditional A ∗ algorithm, maintains an average of 0. 46 path nodes to obstacle nodes and reduces path nodes by an average of 66. 8%, effectively improving the efficiency and safety of the robot.

Abstract:With the improvement of living standards, people are paying more and more attention to health. In particular, the portable physiological monitoring device such as smart wristband that adapts to fast-paced life is favored by people. Photoplethysmography (PPG), as a non-invasive human pulse collection method, is widely used in such device. The human pulse contains a lot of physiological information. In order to extract and analyze this information, the method of machine learning is generally used to establish a mathematical model. However, such methods require a large amount of long-term pulse data to improve the accuracy of physiological parameter models. In response to the problem, this article uses Colaboratory to establish a GRU neural network model and together with LSTM to predict the pulse wave data and adjust the main parameters that affect the performance of the model. The automated machine learning tool AutoML_Alex analyzes the pulse wave data and establishes the LightGBM network based on the best ones, which can be used as a baseline model with reference value. Use large amounts of pulse wave data collected from different individuals to build three different models, compare with MAPE, LSTM is 0. 879%, single-layer GRU is 0. 852%, LightGBM is 0. 842%, and four-layer GRU model is 0. 828%, and apply different models to different individual predictions. It is found that the stability of the single layer in the GRU model is better in the application of different individuals. The results show that we can establish a GRU network model based on short-term pulse wave data of different individuals, predict long-term pulse wave data, and then monitor the human body's arteriosclerosis and other physiological conditions, while providing technical and data support for portable physiological monitoring equipment.