Abstract:Aiming at the shortcomings of the existing resonant magnetic sensor principle that high quality (Q) factor and low power consumption cannot be achieved at the same time, a low-power resonant magnetic sensor with high Q-factor and digital frequency output is designed by combining the double-ended quartz crystal tuning fork (DETF) resonator with magnetostrictive FeGa plate. Using the decoupling characteristic of the structure of DETF resonator, the stress coupling structure is designed to isolate the magnetomechanical damping of the magnetostrictive material. Thus, the Q-factor of the composite sensor is almost equal to that of the DETF resonator. The magnetostrictive force occurring under magnetic field induces a longitudinal force in the DETF. The magnetic-force-frequency conversion is achieved due to the force sensitive characteristic of the DETF. A low pressure packaged sensors based on the composite structure of FeGa alloy and DETF resonator is fabricated. The test results show that the Q-factor and power consumption of the sensor are about 16 764 and 124. 8 μW, respectively. In the linear region, the sensitivity and resolution are 3. 05 Hz/ Oe and 6mOe, respectively.

Abstract:A novel micro electro-mechanical system (MEMS) scanning mirror driven by double-layer combs is proposed here, which posses with a large size, large deflection angle and low drive voltage. The architecture of the MEMS scanning mirror contains a nested inner and outer double-layer vertical comb. The comb teeth are driven by an electrostatic repulsion in this design, and the optimized Sshape torsion beam effectively reduces the stiffness of the sensitive shaft. The theoretical model of the scanning mirror is firstly analyzed based on the theoretical analysis of the vertical comb drive. The static and dynamic analysis and verification are then carried out by using MAXWELL and ANSYS simulation software. The bulk silicon preparation process of the scanning mirror is further explored. The simulation results demonstrate that a maximum deflection angle of ± 13. 46° can be achieved on this MEMS scanning mirror under a driving voltage of 110 V. In addition, the resonance frequency of the structure reaches to 1. 79 kHz, which is much lower than other high-order modes, leading to effectively suppressing the cross-interference motion of other non-working modes and showing a good working bandwidth.

Abstract:Aiming at the problem of inaccurate measurement results caused by the influence of environmental factors on the line wind speed measured by the time difference method, a data fusion method based on adaptive fireworks-BP neural network ( IFWA-BP) is proposed. Data fusion of linear wind speed information and environmental information is used to reduce the inaccuracy of linear wind speed measurement through multi-source information complementation. The adaptive firework algorithm introduces adaptive inertia weights into the firework algorithm and improves the explosion operator to enhance the global search ability of the firework algorithm, thereby optimizing the optimization process of the weights and thresholds in the BP neural network. In order to compare the fusion effect of the IFWA-BP fusion model, a multi-algorithm fusion model comparison experiment was carried out. The experimental results show that the IFWA-BP fusion model reduces the error of linear wind speed measurement and makes the accuracy of the linear wind speed measurement system reach 98. 48%.

Abstract:Compared with the traditional Lorentz line shape, the Fano resonance spectrum with asymmetric line shape has higher spectral resolution and is especially suitable for sensing applications. In this paper, two air holes are introduced in the coupling region of the silicon-based bus waveguide and the runway micro-ring cavity to make the continuous spectrum produce abrupt phase shift, and coupled with the runway micro-ring resonator to form Fano resonance, and in each of the wider spectrum range. All resonance peaks appear as asymmetrical lines. By optimizing the coupling distance and the offset between the air hole and the center of the coupling region through simulation, a silicon-based Fano resonant device with a maximum spectral resolution of 312. 05 dB/ nm and an extinction ratio of 53. 09 dB is obtained. Under the condition that the refractive index change range of the simulated liquid is 1. 33~ 1. 332, the simulation shows that the refractive index sensing sensitivity is 125 nm/ RIU. The simulation results show that the device proposed in this paper has simple structure, compact size and small manufacturing error, which provides a new idea for the application of silicon photonic devices to highly sensitive integrated biochemical sensors.

Abstract:Systematic calibration technology has the advantages of being independent of turntable accuracy and insensitive to shock absorber deformation, etc. It has been widely used in the self-calibration of middle and high precision strapdown inertial navigation systems. The systematic calibration usually takes the velocity error or the differential of the velocity error in navigation mode as measurement. Combined with Kalman filtering technology, the calculation process of system error parameters can be realized automatically. In the systematic calibration algorithm based on velocity measurement, the estimation accuracy of the bias of gyroscope is limited by the stability of bias of the accelerometer in the calibration process. A variable dimension Kalman filter for angular velocity measurement is presented in this paper. The accurate estimation of the bias of the gyroscope can be achieved only by changing the setting of the filter parameters, and no additional gyroscope bias test is required. At last, multiple repeated calibration experiments have been carried out, and the estimation accuracy of gyro bias in the improved scheme is higher than that in the traditional method. The land vehicular tests also prove the effectiveness of the algorithm.

Abstract:The failure of reciprocating compressor happens frequently for the complex structure and rich excitation sources. Due to the difficulty of designing fault features and relying on experience, the traditional methods have no strong diagnosis ability. The intelligent diagnosis method based on convolutional neural networks ( CNN) can realize end-to-end fault diagnosis without feature extraction. However, there are some problems such as inaccurate extraction of fault features, large number of model parameters and long training time. Therefore, a fault diagnosis method of reciprocating compressor based on PyTorch deep learning framework MPMRNet (multipleprocesses-mini-ResNet) is proposed. In this method, multiple processes are used to load data, ResNet50 is taken as the basic network framework and its depth and width are reduced. Adam and StepLR strategy are used to optimize the network and adjust the learning rate, respectively. And time-frequency images of vibration signals are processed automatically to deeply mine and evaluate sensitive features. Multiple comparison experiments show that this method significantly shortens the training time of the model, reduces the number of model weight parameters from 94. 1 to 0. 58 M, the complexity of the model from 4. 11 to 0. 21 G, and the memory occupancy rate from 37. 08% to 10. 92%, and the fault diagnosis accuracy is up to 98. 28%, the diagnostic ability of the model is obviously improved.

Abstract:Bearing fault signals are usually non-linear and non-stationary. Also, this kind of signal is very weak and is easily concealed by unavoidable background noise and vibration interferences. The mode decomposition method has been proved to be a reliable signal processing method for this kind of signal. Therefore, a fast iterative filtering decomposition method is applied to the bearing fault detection in this paper. The fast iterative filtering decomposition method performs excellently in suppressing mode mixing and noise. More importantly, unlike other mode decomposition techniques, the fast iterative filtering method provides significant computational efficiency, so it can highly improve the computational speed. Both effectiveness and superiority of the proposed method are verified by simulation signals and real-world signals.

Abstract:To further improve the newly proposed improved envelope spectrum via feature optimisation-gram ( IESFO), which may be invalid on early fault weak faulty characteristic extraction of rolling elements bearing and required high frequency resolution, the MOMEDA impact enhancement based on spectral correlation filtering is proposed to extract the fault characteristics of rolling element bearing. Firstly, the optimized demodulation frequency band is determined by IESFO algorithm for band-pass filtering. Then, the impact of bearing failure of the filtered signal is enhanced by the multipoint optimal minimum entropy deconvolution adjusted ( MOMEDA) algorithm. Finally, envelope analysis is performed. The research results based on actual measured signals show that the proposed method can detect the incipient fault information of bearings earlier than most existing methods during the performance degradation process of rolling elements bearing. Another, the proposed method can also be utilized to extract the characteristic frequency of bearing fault under compound fault condition.

Abstract:For a wide variety of domestic appliances, the fault current waveforms among different types of appliances may be similar to normal current waveforms, which leads to the problem that traditional methods of fault arc identification cannot detect effectively, this paper presents a series low voltage fault arc identification method which combines time-frequency domain analysis and random forest which is suitable for a variety of typical loads working independently or mixed. Based on the correlation coefficients between the collected load spectra and the pure resistance load spectra, the loads are divided into switched-supply loads and non-switched-supply loads, then two random forest models are trained to identify the faults. A total of 33 723 sets of normal and fault current samples were collected to verify the proposed detection method, which proves that the proposed method can improve the recognition rate of fault arc.

Abstract:For the low diagnostic rate of traditional machine learning algorithm in bearing fault classification under variable working conditions, this paper proposed a bearing fault diagnosis method based on the combination of joint distribution adaptation ( JDA) algorithm and K-nearest neighbor ( KNN) classification algorithm. Firstly, the time domain features of bearing fault signals under different working conditions are extracted as source domain samples and target domain samples respectively, then calculating the weight of each feature by FLDA method. The feature vectors composed of features with higher weights to adapting joint distribution by JDA method, that is, the source domain samples and target domain samples are mapped to the low-dimensional potential space by kernel function, and the maximum mean discrepancy (MMD) distance is taken as the measurement standard to reduce the marginal distribution and conditional distribution differences between the source domain samples and the target domain samples. Finally, the mapped source domain and target domain samples are used as training data and test data respectively, and the model identification is implemented by KNN classifier, and the bearing fault diagnosis classification under variable conditions is achieved. Compared with the method of PCA, KPCA and TCA, through simulation analysis and experimental verification, the method proposed in this paper significantly improves the accuracy of bearing fault diagnosis under variable working conditions.

Abstract:In this paper, ultrasonic technology is proposed for the inner temperature of clutch case measurement of the slow response, inconvenient installation and large measurement error traditional temperature measurement. The in-depth study of the application of ultrasonic temperature measurement technology in clutch case has significant meanings to solve the problems in traditional temperature methods. Based on the finite element analysis software COMSOL, the physical model of ultrasonic wave propagation is established in the clutch case, and the temperature field of the clutch case is restored by the least square method, which can well solve the problem of complex coupling of physical field inside the clutch case. The results show that the internal temperature of clutch case can be obtained by ultrasonic measurement, and the maximum relative error between the calculation results of flight time and the theoretical results is 0. 26%; The results of temperature field reduction by least square method are consistent with the theoretical results.

Abstract:Aiming at the problem that the mental workload of the workers in the manufacturing workshop is too heavy, causing errors and affecting the efficiency of the process, a mental workload evaluation method is proposed to reduce the risk of workplace injuries and improve the accuracy of operations in the workshop. Design two sets of experiments of 20 events, take the mental load results measured by the PAAS scale and WP scale as the dependent variable, through the NC lathe programming software, simulate the work content of the NC lathe programmer, and compare the PAAS scale and the WP scale. The effectiveness of the table used in the intelligent manufacturing CNC workshop is studied. The experimental result is that the sensitivity of PAAS scale is better than WP scale, and its F value is higher than WP scale 7. 494. In terms of aggregation validity, both scales performed well. It is concluded that the PAAS scale is suitable for the evaluation of the mental workload of the CNC lathe operators in the intelligent workshop. At the same time, the multidimensional characteristics of the WP scale can clearly indicate the influencing factors of the subject’s mental workload.

Abstract:Atomic force microscope (AFM) is an important tool for measuring the surface morphology of material objects. In order to realize the high-speed scanning imaging of AFM, an AFM scanning imaging controller based on a Fractional Feedforward-Feedback PID control algorithm is designed. Use fractional-order iterative learning control ( FOILC) in the feedforward loop to learn the error information of the current period in the tracking process to achieve rapid output convergence along the iterative axis; use fractional-order proportional integral (FOPI) control in the feedback loop Increase the accuracy of high-speed imaging. The trajectory tracking simulation and experimental imaging results show that the algorithm can effectively increase the AFM imaging speed and improve the system's nonlinear impact. When the scanning frequency is 25 Hz, the control accuracy reaches 10 -5 . AFM imaging quality has been significantly improved.

Abstract:In this paper, a novel time-domain channel prediction technique for adaptive OFDM systems is introduced. The novel timedomain channel prediction technique utilizes the RQA to quantify and estimate the local predictability of each time-delay tap in the channel impulse response, and then based on the local predictability of each delay tap, we select those significant time-delay taps in the channel impulse response; finally, the joint echo state network ( JESN) is utilized to predict the channel state information in each significant time-delay tap. In the simulation part, the OFDM system based on the IEEE802. 11ah standard to evaluate the performance. The simulation results show that by the TSS-RQA, those significant time-delay taps in the channel impulse response are accurately selected. In addition, the JESN produces the sparse output weight matrix. Due to the oracle property, the improved prediction performance of the JESN is approximate to 91. 75%, compared to the basic echo state network.

Abstract:Aiming at the problem that missiles are difficult to carry out testability design due to their own characteristics, a multi-signal model for testability design of missile systems is proposed. Based on the FMECA information, the failure mode in the system is determined, and the multi-signal model is used to establish a test model of the missile system. The correlation matrix of the fault-test is obtained and the system testability index is determined. Considering the shortcomings of existing algorithms such as genetic algorithm and binary particle swarm algorithm, a hybrid discrete binary particle swarm-genetic algorithm is proposed to optimize the test, reduced 22 candidate tests to 14, greatly reducing the number of tests. Finally, the experimental results show that the proposed algorithm can meet the accuracy requirements of testability index, and effectively reduce the number of tests and test costs.

Abstract:External noise will affect the results of Lamb wave damage imaging. To solve this problem, a damage factor based on EMDHilbert transform is defined. First, air-coupled ultrasound is used to excite-receive A0 mode Lamb waves in the aluminum plate, and scan them in two orthogonal directions to obtain the received signal. Then, the signal is decomposed by EMD to find the IMF component with the greatest correlation with the original signal. The Hilbert transform is performed to obtain the envelope curve, and the relative change of the envelope curve area is defined as the damage factor. Finally, white noise with different signal-to-noise ratios is added to the scan signal to simulate the noise environment, and damage imaging is performed before and after the simulated noise environment. The results show that the proposed damage factor can effectively form image of the damage in the aluminum plate, and has certain noise resistance. Compared with the imaging result without adding noise signal, the maximum error change is 0. 8 mm.

Abstract:Since it is difficult to overcome the inherent defects of parameter changes and time lag in non-model PID control, in order to optimize the active feedback control of the plasma vertical unstable displacement in the tokamak device. The improved grey GM (1,1) prediction model is used to accurately predict the output current of the EAST fast control power supply based on the cascaded H-bridge topology to optimize the control parameters. The grey GM ( 1,1) prediction model is suitable for small samples and poor information systems, and requires fewer modeling samples and simple calculations. The difference of the predicted fitting sequence leads to the prediction error when the grey GM (1,1) modeling is performed on the upward convex sequence of the output current. Choose a data transformation method that transforms the upward convex sequence axisymmetrically into the upward concave sequence and establishes a non-equidistant grey GM (1,1) prediction model. At the same time, the estimation formula of the forecast time of the non-equal interval sequence is given by using sample points. Based on the improved prediction model, the modeling process of the prediction model is deduced, and the prediction deviation of the two grey GM (1,1) prediction models on the output current of the power supply is compared through simulation, and the improved grey GM (1,1) is verified in an experimental environment. After the improvement, the prediction error rate of the mutation segment is reduced to below 10%. And verify the effectiveness of the improved grey GM ( 1,1) prediction model in an experimental environment.

Abstract:In the process of power quality signal acquisition and transmission, noise interference will be introduced due to the influence of the external environment. It is an important prerequisite for power quality control to effectively denoise and retain the mutation information. An adjustable threshold function is given in this paper. By controlling the adjustable parameter, the new threshold function can be changed between soft and hard threshold functions. The wavelet coefficient energy factor is introduced, and the scale with the largest energy is taken as the characteristic scale. On this scale, the sub interval energy higher than the scale energy is the effective interval. Then a new threshold method using effective interval local threshold is proposed. In addition, considering the propagation characteristics of the wavelet coefficients of noise and signal with different scales, an operator is introduced to modify the threshold. Compared with the traditional global threshold method, the local threshold method based on the effective interval can better reflect the characteristics of wavelet coefficients. The simulation experiments prove that the proposed method not only improves the denoising performance, but also retains the disturbance mutation information. After denoising, the disturbance location is accurate, and the position of the modulus maxima can well reflect the occurrence and end time of the disturbance.

Abstract:In the flight tests of an aircraft performance and quality, obtaining accurate data of control force, control displacement and corresponding surface deflection is a powerful guarantee for studying maneuverability and stability. Aiming at problem that the traditional calibration method cannot accurately reflect relation of the three kinds of control parameters, this paper carries out the research on the joint online calibration technologies for the control parameters’ testing system. Through building wireless-sensor calibration network, designing onsite calibration device, developing calibration software platform, the joint online calibration of multi-parameters was realized. Onboard test results show that the proposed method leads to more than 50% increasement in calibration accuracy and about 67% reduction in calibration time, and the displacement-force curve obtained from calibration is completely consistent with the design curve, which verifies the feasibility and effectiveness of the joint online calibration system.

Abstract:The traditional laser projectile velocity system is susceptible to the impact of high-temperature frontal shock luminescence and space electromagnetic field, resulting in a decrease in the system’ s signal-to-noise ratio. In order to obtain high-quality signal waveforms, a new type of laser projectile velocity measurement system that can be used for light-gas gun is designed, and improved methods of laser light path and photoelectric conversion circuits are proposed in this paper. The use of optical fiber to guide the laser avoids the use of long leads and enhances the system’s anti-interference ability. A solid-state laser with an output power of 40 mW was selected as the system light source, and a composite laser receiving probe with filtering and focusing functions was designed to increase the signal light intensity. The noise and bandwidth of the measurement circuit are analyzed in detail, and the photoelectric conversion circuit is improved and optimized. The signal-to-noise ratio of the measurement circuit is increased to 98 dB. The actual test results show that the signal waveform quality obtained by the laser projectile velocity measurement system is good, and the relative expanded uncertainty of the velocity measurement result is 0. 73%, which meets the requirements of the light-gas gun projectile velocity measurement system.

Abstract:Radial variation and different sections are essential factors affecting the identification of wood by terahertz spectroscopy. To explore the influence of radial deviation and different facets on terahertz spectrum identification of timber, terahertz time-domain spectroscopy technology was used to obtain terahertz spectrum data of Cunninghamia lanceolata and Cryptomeria fortunei samples, a total of 600 sets of wood spectrum data, and then the terahertz spectrum were analyzed. A wood recognition model was established based on BP neural network. The accuracy of model prediction was compared. It was found that the terahertz spectra of wood samples were different due to radial variation and other sections. The prediction accuracy of the model with the same or different radial positions between the predicted sample and the modeling sample had little difference, and the highest accuracy was 96. 25%. When the expected sample and the modeled sample belong to the same and different sections, the difference in accuracy is massive. The results show that the technology based on terahertz time-domain spectroscopy can accurately realize wood recognition, radial variation has little effect on wood recognition, and different sections have a more significant impact on wood recognition.

Abstract:A novel acoustic scene classification (ASC) system based on asymmetric convolution and knowledge transfer is proposed to address the problem caused by limited ASC datasets. Compared with the existing methods which trained models from scratch, the proposed system fine-tunes a pretrained model of other tasks to preserve valid information from the source domain. Besides, targeted at the nature of acoustic features, it adopts asymmetric convolutions to enhance the network capability of feature extraction. Experiments shows that the proposed system outperforms the baseline system by 0. 023. Besides, as shown in the visualization results of convolutional filters, textures of the proposed system are more detailed than other methods. The experiment proves that knowledge transfer can boost model ability of feature representation, and it can further improve system performance by combining with asymmetric convolution.

Abstract:The performance of CN-based similarity index is not satisfied due to only taking into account the local information of a network. The community information contains the network structure features of nodes, which can be adopted to improve the prediction accuracy. Therefore, a community-based link prediction algorithm using the community structure information is proposed to address the problem. Employing community relations and community information of common neighbors, it was developed in an attempt to improve the prediction precision. Firstly, two graph embedding methods--DeepWalk and Node2vec were employed, that is, a deep learning model, i. e. Skip-Gram was adopted to train the nodes’ sequences generated from short random walk and then the acquired embedding vectors of nodes were used in communities division to obtain high quality communities that contain more network topology information. Then, the similarity model of communities was proposed via defining the edge relationship between communities. Finally, the similarity of nodes, the similarity between the communities where the nodes are located, and the community information of the nodes’ common neighbors were integrated into the suggested algorithm to evaluate the link probability of two unknown nodes. Finally, experiments on six real-world networks like USAir are conducted, and the AUC of the suggested method is increased about 2. 3% at most compared with four benchmark algorithms including CN. Thus it shows that community structure information plays an important role when predicting the latent links.

Abstract:One of the important technologies to improve the Reynolds number is to reduce the temperature of the air flow inside the wind tunnel. The air flow working temperature can reach 77 K at least. At this time, it is necessary to carry out thermal protection for the internal equipment of the wind tunnel. In view of the complex layout and various characteristics of the equipment in the wind tunnel, this paper takes the thermal protection and temperature control strategy of the equipment in the wind tunnel as the research object, and designs the temperature control strategy of the typical equipment from the perspectives of heating element selection, temperature control logic, temperature control algorithm and heating mode. A physical test platform was designed and built. The feasibility of temperature control strategy was verified by the physical test platform under two typical temperature conditions of normal temperature and low temperature. The test results show that the comprehensive design of the temperature control strategy can meet the temperature control requirements of different shapes, heat capacities and heat leakage conditions, and the temperature control accuracy is better than 1 ℃ , which can meet the design requirements and effectively ensure the normal operation of the equipment in the wind tunnel.

Abstract:In order to realize the long-distance and low-loss transmission of vehicle high-definition video, the framework and principle of vehicle camera transmission system based on GMSL technology are studied in this paper. Through the exploration of the hardware circuit, hardware codec and video signal coding principle of the vehicle camera transmission system, a vehicle forward-looking camera system is designed in this paper. First of all, the overall environmental framework of the transmission system is simulated according to the research results. Then, the research conclusions of the power supply and the initial coding and decoding of the video signal are applied in practice so that the video signal can be encoded stably and transmitted effectively. Moreover, the process of exploring the serializer/ de-serializer architecture, completing the configuration of the hardware, collecting and summarizing the data waveforms is conducted step by step. Finally, the restored signals are analyzed and image quality is evaluated, therefore, the research of the vehicle camera system based on GMSL is completed. The experimental results show that the camera system can well achieve the functions of serial and de-string to modulate the transmittable video signal over 15 m. After the master unstrings the video signal, a stable and real-time high-definition video image can be obtained at last. The research results show that the vehicle camera system can greatly transmit low-loss video images with resolution of 720 P over long distances. Furthermore, the outcomes of this research can be compatible with the current self-driving cars.

Abstract:In view of the application of augmented reality (AR) technology in the industrial field, an augmented reality system model used in the industrial indoor environment is proposed. The model is mainly for the industry that will not change dramatically in a short period of time in the indoor environment. First, the characteristic points are detected by the Harris detection algorithm on the environmental images collected by the HMD, namely the helmet-mounted display, then the program compares the images to add relevant annotations at the feature points, finally displays the real image containing the annotations on the head-mounted display. This paper mainly introduces the specific steps and implementation methods of the model. The Harris corner detection algorithm has the characteristics of dealing with the change of light intensity and image rotation, thus increasing the robustness of the whole model. The model is applied to the wind turbine nacelle internal environment to verify the feature point detection section according to the feature points by many experimental threshold adjustment feature points in 20, in the part of image matching, more than 85%, the experimental results show that the model achieves the function of augmented reality.

Abstract:In the traditional motor test system, it is impossible to directly test the transient parameters of the motor and the requirements of dynamic arbitrary load loading when the motor is loaded or the problem of high implementation cost. This research uses STM32+FPGA as the core to form the main control of the motor test system Unit, combines with the superior control performance of the ARM processor and the high-speed data processing advantages of FPGA. The FPGA and STM32 are controlled by the FSMC bus communication and data ping-pong algorithm. Through above method, there is a small-size control core board manufactured in the motor test system. It can be embedded in the existing traditional dynamometer system to improve the dynamometer transient parameter test and dynamic motor loading performance. Experimental results show that the system can basically meet the requirements of motor transient test and dynamic loading, and can achieve the fastest data refresh rate of 1 ms for measuring the measured motor’ s speed and torque, and arbitrary waveform loading within 100 M sampling frequency, with a loading error of 0. 8%. It satisfies the motor transient test and dynamic loading requirements.