Abstract:In order to master the atomization characteristics of oscillating spray and its relationship with unstable combustion, research and development of gas-liquid coaxial swirl injectors and liquid-liquid coaxial swirl injectors, which are widely used in liquid rocket engines, are reviewed from three aspects of excited oscillation, self-pulsation, and interactions of spray oscillation with combustion stability, so as to deepen our understanding of them. According to the review, for the excited oscillation, the research on the upstream pressure oscillation caused by flow oscillation of supply system under normal temperature and pressure conditions is relatively sufficient. However, the research on the backpressure oscillation caused by pressure oscillation in the combustion chamber and the research on the excited oscillation of spray under supercritical conditions are lack. For the self-pulsation, the current research is focused on the liquid-centered coaxial swirl injectors. There are few studies on the gas-centered coaxial swirl injectors and liquid-liquid coaxial swirl injectors. The optical diagnosis technology is far from enough to extract the single combustion flow field information.

Abstract:Ni-based GH3536 superalloy smooth surface and microporous structured surface with hydrophobic properties were used as the object of study, and the early stage frosting characteristics of the surface were studied. The impact of surface temperature, relative humidity, and surface wettability on frost crystal shape and growth rate were investigated quantitatively. Results indicate that with the decrease of the surface temperature, the initial time of frost crystal formation is shortened, the growth rate of frost crystals is enlarged, and four types of frost crystal shapes are captured. The increase of relative humidity can increase the number of frost crystal and change frost crystal morphology, while the condensation and freezing time of droplets are independent of it. For the same frost crystal morphology, increasing the relative humidity has a certain promotion effect on the frost crystal growth, while when the frost crystal morphology changes, the change in relative humidity will result in significant differences in frost crystal height and growth rate. The hydrophobic surface made of micro-holes structures can evidently delay the formation of the crystal by increasing the nucleation barrier and contact thermal resistance. However, the impacts on frost crystal growth and shape are negligible.

Abstract:To reveal the flow field characteristics of dual synthetic jets impinging plate, the large eddy simulation method was used to numerically study the flow of dual synthetic jets impinging plate. The Lagrangian coherent structure of the flow field was identified on the basis of the finite-time Lyapunov exponents, and the results were compared with the vorticity results in the Euler frame. It is found that the vortex structure of the jet core area of dual synthetic jets is exceeding complex and rich under the alternating action of periodical jets, and there is a pair of stable vortex structure far away from the core. Lagrangian coherent structures are well corresponding to the vorticity, provided guidance for the layout design of dual synthetic jets impingement cooling. Proper orthogonal decomposition analysis is performed on the flow field of dual synthetic jets. The results show that the first-order mode is approximately symmetrical about the central axis of the exit of the actuator, and its energy accounts for 35% of the total energy, and the first 6 modes account for 80%. According to the characteristics of the flow field reflected by the first 6 modes, the flow field of dual synthetic jets impinging plate has a high degree of symmetry.

Abstract:For a full utilization of the probe after the lunar exploration, the extended mission using the probe for flybys of multi small celestial bodies from the Moon or Earth orbit was proposed. Combined with the flight status of the probe after the lunar exploration mission, the energy requirement for escaping from the Moon or the Earth was emphatically analyzed, together with other necessary conditions for the flyby mission, such as the brightness of small celestial bodies, the capacity of onboard equipment, etc. The methods of selecting small celestial bodies and trajectory design were studied. Aiming at the problems of large search space, large amount of computation and hard optimization of multi target sequences caused by large number of targets, a multi-layer optimal search algorithm for small celestial bodies was designed. Taking Chang′e-5 as an example, the search yielded several feasible solutions for small celestial bodies′ flyby detection. The result shows that the orbiter can approach up to five asteroids under given constraints, including the large Asteroid 12923. The study conclusion can be used for the Chang′e-5 extended mission, and provide beneficial reference for the future missions of the Moon and small celestial bodies.

Abstract:In order to effectively solve the shortcomings of the existing gas ejection source, as well as meet the needs of large mass flowrate and miniaturization of an injection system, a gas generator with multi-point injection structure using air and alcohol as propellants was developed based on the comprehensive analysis of various gas injection source schemes, and the experimental research under various working conditions was carried out. Research results show as bellows:the multi-point injection scheme greatly improves the space utilization rate, which effectively meets the miniaturization requirements of ejection system. Spray performance is good, and the spray cone angle and particle diameter are superior to the index requirements. Ignition reliability is high, which solves the shortcomings of low ignition reliability of the existing gas generator in the field. The operation curve indicates a rapid ignition, stable combustion and uniform temperature field of gas generator. The excess air coefficient range for stable operation is 2.52～4.34, showing a wide mass flow-rate range. In addition, efficient film cooling on the combustion liner inner wall effectively ensures the long time operation of the gas generator.

Abstract:In order to improve the uniformity of the neutral atom distribution in the discharge channel of the low-power Hall thruster, the finite element method was used to optimize the key structural parameters of the anode gas distribution ring. Aiming at the double-chamber anode gas distribution ring structure, the influence of key parameters such as the volume ratio of the buffer chambers and the number of diversion holes between the buffer chamber partitions on the uniformity of anode gas supply was analyzed. The research results show that with the increase of the volume ratio, the difference of the anode gas distribution ring outlet pores decreases rapidly and then stabilizes. When the ratio k=1.0, the average difference rate and the maximum difference rate are 1.77% and 3.79%, respectively; when the number of diversion holes between the buffer chamber partitions increases from 8 to 14, the difference rate of the air outlets shows the characteristics of the bathtub curve, and when number of the diversion holes is 10, the average difference rate and the maximum difference rate are 1.8% and 3.8%, respectively. The research results can provide theoretical support for the engineering design of the anode gas distribution ring of the Hall thruster.

Abstract:The ground proximity warning envelope in TSO-C151b standard is a general model design without considering the characteristics of the aircraft. In order to accurately give the warning envelope of excessive descent rate of a certain aircraft, the alarm mechanism was analyzed by combining the aerodynamic characteristics and control characteristics of the aircraft, and a six degree of freedom simulation program was established. Based on the aircraft engine data, aerodynamic data and quality characteristics data, simulation models and operation procedures were established to simulate the operation response process of each rudder surface, as well as the motion attitude and flight path of the aircraft in real time. The loss height pulled up by the aircraft at different descent rates was calculated, and the ground proximity warning envelope was designed. The actual flight test data verified that the model design is reasonable and the calculation results are accurate. The alarm envelope obtained by simulation is compared with TSO-C151b, and the suggestions for the use of the ground proximity warning envelope are given to ensure the flight safety of the aircraft.

Abstract:Aiming at the impact of maintenance scheduling on equipment training and reliability in equipment support, the maintenance work of multiple ships supported by the detachment-level repair station was abstracted into a single maintenance station to support the dispatching and distribution of maintenance forces for multiple systems, the variable vacation strategy of repairman was introduced to describe it. Take the common n out of k system in the equipment structure as the research object, aiming at the problem of excessively strict model constraints caused by typical distributions such as exponential distributions in previous studies, continuous Phase-type distribution was used to describe the system-dependent random variables and build system reliability analysis model.Finally, the model applicability was verified using numerical application, and it was analyzed the influences of whether the repairman was on vacation, the changes of system components numbers, and the effect of repairman vacation rate and maintenance rate on system reliability. Numerical example results show that the reliability model can be effective retrieval the influence of maintenance forces scheduling on the reliability of k-out-of-n system, it can provide theoretical basis and practical reference for the reasonable arrangement of the number of repairman vacation and the optimal configuration of the number of system components.

Abstract:Micro-soft robots with high compliance, low energy consumption and high energy density have wide application prospects in complex environments such as pipeline maintenance and battle field reconnaissance. Energy and actuator play a decisive role in the motion modes and performances of micro-soft robots. In order to make more researchers understand the research progresses of the existing flexible driving technology and their energy sources, the typical driving methods based on physical energy driving, chemical energy driving and biological hybrid driving were summarized and analyzed. The shortcomings and future development of the existing flexible driving technology and their energy sources were discussed and summarized, which can provide reference for the development and improving performance of the flexible drive technology of soft robots in the future.

Abstract:In order to predict the unsteady thrust of the interaction line spectrum of the pump jet propeller rotor and the periodic leading vane wake, ignored the thickness of the pump jet propeller rotor blades, and simplified the pump jet rotor into an annular cascade. According to the sheet theory, the radius that is the segment of the pump jet rotor was intercepted at r, and the change of the flow field parameters along the segment radial was ignored. Therefore, the annular cascade segment can be considered as a planar cascade, and the periodic inflow is considered based on the interaction between the planar cascade and the harmonic wave. The unsteady excitation force of the interaction line spectrum between the front guide vane segment and the rotor segment is derived. The unsteady thrust line spectrum prediction formula is obtained through the circumferential integration of the rotor segment. The validity of the formula is verified through numerical and experimental methods. A design parameter impact analysis was conducted, and it was found that when the ratio of the current guide vane rotor spacing to the chord length of the front guide vane is greater than 1, the rotor rear guide vane spacing has almost no impact on the excitation force line spectrum thrust of a single rotor blade.

Abstract:Aiming at the problem of electromagnetic vibration and noise of motor, it was found that improving the sinusoidal characteristic of air gap flux density does not necessarily reduce the vibration and noise of motor. In view of this, the order and frequency of the main force wave causing the vibration and noise of the motor were analyzed, and the specific functional relationship between the third order harmonic of air gap flux density and the vibration and noise of the motor was explored. In order to analyze the order and frequency of the main force wave which caused the vibration and noise of the motor, the force wave table was eliminated based on the analytical method, and the main noise source was determined. In order to explore the influence of the third order harmonic of air gap flux density on motor vibration and noise, the mathematical models of air gap flux density and radial force wave were established, and the optimal amplitude of the third order harmonic of air gap flux density was solved by genetic algorithm. The analysis results show that improving the sinusoidal air gap flux density does not necessarily reduce the vibration noise of the motor, and the third order harmonic of air gap flux density has an optimal value in a certain range to minimize the vibration noise of the motor. The analysis results provide a reference for the design of low noise motor.

Abstract:In order to solve the problem that the fault state of motor bearings is difficult to identify, which results in low diagnosis accuracy, a motor bearing fault diagnosis model was proposed by combining the signal feature extraction and XGBoost (extreme gradient boosting algorithm). The optimized variational mode decomposition was used to obtain the IMF(intrinsic mode function) components of the vibration signal, and the multi scale entropy value of each IMF component was calculated on the basis of the multi-scale entropy theory for feature reconstruction. The WOA (whale optimization algorithm) was improved by introducing the selection, crossover and mutation operation of genetic algorithm. The improved WOA algorithm was used to optimize the hyperparameters of XGBoost, and obtain the hyperparameter combination that helped the XGBoost achieve the best classification effect. The vibration signals of 7 different types of faults were reconstructed and input into the optimized XGBoost model for fault diagnosis. Experimental results show that the motor bearing fault diagnosis accuracy of the proposed GWOA-XGBoost model can reach 97.14%, which significantly improves the performance compared with the traditional diagnosis method.

Abstract:Predicting submarine buoyancy motion is a prerequisite for exploring and studying the laws of submarine buoyancy motion control and submarine buoyancy safety. The numerical calculation method and calculation process of submarine buoyancy motion were introduced in detail. The turbulence model and discrete scheme in the numerical calculation method were optimized. The uncertainty of numerical calculation was analyzed and the results of numerical calculation were verified based on the model floating test′s results, what′s more, it was found that the reliability of numerical calculation was proved and the numerical calculation results were in good agreement with the experimental results, thus verifying the feasibility of the numerical calculation method. The numerical calculation method and test can provide reference and reference for the research of submarine floating movement.

Abstract:The mainstream FL(federated learning) methods require gradient interaction and the ideal assumption of the independently identically distribution, which brings additional communication overhead, privacy leakage, and data inefficiency. Therefore, a new FL framework called MAFML (model agnostic federated mutual learning) was proposed. MAFML only used a small amount of low-dimensional information (for example, the soft labels output of the neural network in the image classification task) for sharing to achieve cross-participants “mutual learning and mutual education”. Moreover, MAFML did not need a shared global model, users can customize their own private models without restricting the model structure and parameters. At the same time, MAFML used a general approach for avoiding gradient interference so that each participant′s model could be well generalized to other domains without reducing the performance of its own domain data. Experiments on multiple cross-domain datasets show that MAFML can provide a promising solution for alliance business facing the “competition and cooperation” dilemma.

Abstract:In order to solve the defects of edge artifacts and semantic discontinuity when inpainting large and irregular distortions in an image with fine texture and complex background, a residual connection image restoration algorithm based on parallel generation convolution was proposed. The damaged image was inputted into a two-column parallel convolutional structure inpainting network to obtain two image components with different sizes of reception fields. The two image components were combined by the shared decoding and the L2 loss of the output was calculated to optimize the network. The output of the coarse network was sent into the fine inpainting network which contained the residual connection and the attention mechanism to fuse context information and improve the ability of repairing fine details. The global and local discriminators and visual geometry group net were used to calculate the loss and to optimize the overall discrimination network and enhance the global and local consistency of the inpainting result. The performance of the proposed algorithm was validated on internationally recognized databases, and experimental results show that the proposed algorithm can effectively repair large and irregular missing areas under complex background and fine texture, improve the authenticity and integrity of image details, semantics and structure. Its peak signal-to-noise ratio and structural similarity are superior to the state-of-the-art.

Abstract:The behavior recognition network based on 2D convolutional usually integrates classification results of multiple video frames to recognize different behaviors, but it can′t extract space-time feature using the 2D convolution kernels. To solve this problem, MTSC (multi time-scale convolution) was proposed based on TSM(temporal shift module), which contained convolution kernels of different scales to fuse the space-time feature from different time scales. By controlling the position that inserting MTSC into ResNet50 network and the parameter setting of MTSC, the optimal behavior recognition network based on MTSC was discussed. Using the PyTorch training model, an experimental study was conducted on a large open source dataset, Something-Something v2. The results show that the behavior recognition network based on MTSC achieves 59.47% Top-1 accuracy, and outperform TSM and other behavior recognition networks.

Abstract:Visual system is an important part of autonomous unmanned system. The raindrops attached to the surface of the camera lens will produce artifacts in the image, resulting in the degradation of image quality, which will significantly affect the performance of the visual system. The detection and removal methods of attached raindrops in recent years were comprehensively and deeply researched. The essence of the problem was condensed, and the existing raindrop imaging models were summarized. Different technical methods were sorted out from three directions:model-based, data-driven and camera system-based, then the development of deep network model was summarized from two aspects of network architecture and loss function. The existing datasets of attached raindrops were summarized, and the performance of some algorithms was compared through the experimental results. The main problems in the task of raindrop detection and removal were discussed, and the possible development direction in this field was prospected.

Abstract:All-pole model constructs a linear model of signal, and various parameters of the estimated signal can be obtained by fixing the order of the model and calculating it, so as to realize signal prediction. Before sub-band fusion, coherence processing is required and can be solved by means of coherence function and so on. Since the root-MUSIC algorithm is not robust in pole selection in a low signal-noise ratio environment, it will cause the model to judge the order incorrectly. In order to solve the problem that the order estimation of subband fusion is inaccurate when the signal-to-noise ratio is low in the all-pole model, a noise suppression method was proposed. The main diagonal singular value matrix of Hankel matrix was weighted to eliminate the noise component, and the overall forward prediction matrix was used to obtain the pole value and pole amplitude of the multi-subband fusion signal, so that the multi-subband fusion signal at low signal-to-noise ratio was estimated. The results show that the proposed method has better estimation results than the traditional pole-model scheme when the signal-to-noise ratio is -20 dB to 10 dB in simulation environment.

Abstract:A spectrum map construction method based on general regression neural network fitting and clustering Kriging was proposed, in which the path-loss and shadowing components were estimated by general regression nerual network for trend-surface fitting to improve the construction accuracy. In order to improve construction efficiency and guarantee the accuracy meanwhile, monitoring data clustering and optimal neighborhood selection were utilized to reduce the amount of calculated data. The proposed method can realize the accurate and fast construction without prior information by only using limited amount electromagnetic environment monitoring data. A spectrum map prototype verification system was designed and implemented, real measured data from vehicle-based collection system was utilized for testification of the feasibility and performance of the proposed method.

Abstract:In order to deal with the threat of strong electromagnetic pulses to electronic information systems, an ultra-wideband strong ESS(energy selection surface) with transceiver compatibility was designed, which can effectively expand the operating bandwidth and frequency of the ESS, and can provide no less than 14 dB protection effect in L、S、C-band. The ESS was a periodic structure, with each unit containing a pair of arrow shaped structures and a switching diode. It can adaptively switch the reflection or transmission state according to the energy density of external irradiated electromagnetic waves, thereby achieving strong electromagnetic protection and compatibility with working signal transmission and reception. Simulation research shows that the insertion loss of this new ESS in the L、S、C band is less than 1 dB, and its protection effectiveness reaches 22 dB. Finally, the performance of the designed sample was verified in the waveguide. Results show that the average insertion loss of the sample in the waveguide is 1 dB, and the protection effectiveness reaches 14 dB, which preliminarily verifies the low insertion loss and high protection effectiveness of the designed structure.

Abstract:The topology of 5L-ANPC(five-level active-neutral-point-clamped) converter has a simple structure, which can be directly supplied by the common direct current bus, and is easy to realize energy feedback. It is very suitable for the 5～10 kV propulsion variable frequency speed regulation in the integrated power system of ship. The topology and basic working principle of 5L-ANPC were briefly introduced. Based on the analysis of four commutation circuits of 5L-ANPC topology and considering the assembly and adaptability of the snubber circuit, a single capacitor integrated snubber circuit was designed. The over-voltage problem of internal switch in 5L-ANPC topology with snubber circuit using traditional voltage zero crossing strategy was analyzed and an improved output voltage zero crossing switching strategy was proposed,which ensure the safety of switching device and avoid the abnormal voltage jump. The effectiveness and correctness of the improved strategy are verified by simulation and experiment.

Abstract:To address the problem that the boundary conditions affect the electric field signal characteristics when the boundary element method is used to derive the electric field distribution of corrosion in submarines, resulting in the phenomenon of singular peaks in the propeller area, the measured polarization curves of marine 921 alloy steel and nickel bronze propeller materials were used as the modeling boundary conditions, and the electric field characteristics under constant potential and nonlinear boundary conditions were analyzed. By establishing the equivalent circuit of hull propeller electrochemical impedance under the parameters of impedance spectrum, the reason of the singular peak in the propeller area of submarine corrosion electric field was analyzed. The simulation results show that the characteristic distribution of submarine electric field and the phenomenon of singular peak in propeller area are related to the electrochemical polarization state of submarine material, and setting the nonlinear polarization boundary parameters reasonably can achieve the effect of weakening the singular peaks and smoothing the corrosion of the electric field model.

Abstract:In order to master the false alarm laws of dual frequency electromagnetic radiation in band to radar, the stepped frequency continuous wave radar was taken as the research object, based on the theoretical analysis and effect test, the level change law, waveform characteristics and location law of false alarm target were studied. Results show that:under the effect of dual frequency electromagnetic radiation in the band, the tested radar can generate at most two “hill type” false alarm targets with random positions without considering intermodulation interference, and the distance difference is related to the frequency difference of dual frequency interference. The two components of the dual frequency interference signal suppress each other, that is, if the intensity of one component is constant, with the enhancement of another component, the level of the false alarm target formed by the former decreases slowly and the final deceleration is basically constant, while the level of the false alarm target formed by the latter increases gradually until the level is constant. If the strength of the two components of the dual frequency interference signal increases in the same proportion, the levels of both two false alarm targets rise approximately linearly at the initial stage, and then the growth rate gradually decreases to zero.

Abstract:Effective identification of important nodes in power system is helpful to improve robustness of system and reduce the probability of accidents by applying additional protection or changing the topology of important nodes under limited resources. Inspired by the web page sorting algorithm, an algorithm called E-SALSA (electrical stochastic approach for link structure analysis) was proposed for evaluating important nodes in power system. Taking into account the influence of power system topology, power flow and other factors on nodes, this algorithm can effectively reflect the true situation of power system, and its features are more in line with the background of power system. In the IEEE300 node power system, the E-SALSA algorithm was compared with the electrical median algorithm and the MBCC-HITS (model based on co-citation hypertext induced topic search) algorithm by using the two indexes of the scale of load loss and the maximum subgroup size. The results show that the E-SALSA algorithm has advantages over electrical median algorithm in both indicators. Compared with MBCC-HITS algorithm, E-SALSA algorithm can use all factors more comprehensively on the impact of nodes, which further proves its rationality and effectiveness.