Abstract:The drag-free control of the spacecraft is one of the key technologies to realize the ultra-static and ultra-stable operation of science platform for gravitational wave space detection. At present, the dynamics and control of the spacecraft system have been studied deeply by various research institutions at domestic and overseas, and different detection tasks have been proposed according to the requirements of different detection frequency bands. The design and control of spacecraft formation were introduced and analyzed in detail according to the exploration mission, and the principles and theoretical methods of drag-free and attitude control, high-precision inertial sensors and actuators involved were deeply dissected. The overall situation of demonstration and verification of drag-free spacecraft in orbit for space-based gravitational wave detection was detailed and analyzed. On this basis, the key problems required to be solved in the research of follow-up related were presented, and the research hotspots and trends in the methods of future dynamics and control of drag-free spacecraft systems were also pointed out.

Abstract:Aiming at the cooperative acquisition control problem of geocentric very high orbit constellations, a cooperative acquisition control strategy was designed on the basis of the virtual formation method, and a three-pulse burnup optimal trajectory planning algorithm was used to coordinate the trajectory of the conformation acquisition. And the adaptive whole integral sliding mode controller was combined to track and control the transfer trajectory of the satellites. Taking the three-star constellation configuration capture at an orbital altitude of 100 000 km as an example for simulation verification, simulation results show that this strategy can be effectively applied to the configuration control of the constellation with very high geocentric orbit. It can make the satellites in the constellation reach their nominal positions at the same time with less burn up, and at the same time has high accuracy.

Abstract:In order to acquire the accurate adjustment of the orbit shape and height of the detector, a new method of spacecraft continuous thrust control based on radial force equilibrium flight was proposed. The dynamic polar coordinate model of continuous thrust equilibrium flight was established, and the analytical orbit solution under special conditions was derived, the boundary conditions were further analyzed, and the control law of continuous thrust was given. Using this equilibrium flight control theory, the optimal control strategy for orbit capture can be constructed. Considering the thrust level of thrusters, the integrated adjustment of orbit shape, orbit height and orbit phase can be adjusted through one or more control processes. Numerical simulations show that the space gravitational wave detector with micro thruster can achieve high-precision orbit capture by using the orbit control method of equilibrium flight. This method has the advantages of analytical control process, small calculation, simple and practical.

Abstract:An on-orbit drag-free control technique for spaceborne gravitational wave detection missions was discussed. Based on the analysis and design of a possible future geocentric orbit detection mission, the relative motion dynamics and coupling characteristics between the spacecraft and mass blocks of an on-orbit drag-free system with with two test masses were modeled. At the same time, the performance index and perturbation of the drag-free system in the mission were preliminarily analyzed, and a relative translational control law based on frequency domain H_∞ optimal control theory was designed. Numerical simulation results show that when the test masses of the two-test-mass on-orbit drag-free system are arranged according to the breathing angle of laser rangefinder, without a fixed tracking point strategy and without suspension control input along the non-sensitive axis, the spacecraft can achieve tracking of the reference point while meeting the frequency domain performance index of the system. At the same time, the time domain displacement of each test mass can be controlled to the micron level, thus obtaining the pure gravitational reference required by the mission.

Abstract:Starting from the mechanism of SWTBLI(shock wave/turbulent boundary layer interaction) and the urgent need for flow control, the research progress of the adaptive control techniques in the SWTBLI from four aspects was summarized, namely adaptive vortex generator, adaptive bump, adaptive micro jets and adaptive secondary recirculation jet. Analysis shows that developing adaptive flow control techniques, combined with AI technology, and accelerating intelligent control schemes can be used as an important technical means for wide-speed flight of the new generation hypersonic vehicle. Specifically, it is to realize local flow acceleration/deceleration, aerodynamic thermal protection, aerodynamic control and other functions in different areas of the hypersonic vehicle by adjusting external excitation, and to establish a control feedback loop according to flow field parameters, so it can adaptively adjust the structure of the local flow field to meet the actual needs of engineering.

Abstract:For the calculation difficulty directly of the circular error probability test, a truncated sequential probability circle test method was proposed. Taking the target point as the center, two concentric circles were drawn, in which the area within the small circle was the acceptance domain, the area outside the big circle was the rejection domain, and the middle area was the continuing test domain. On this basis, the basic decision rules in sequential testing were defined, and a calculation model for the risk of both parties and the average number of experiments was provided. Two different decision threshold optimization calculation models were proposed, and the acceptance and rejection domain radii were determined by solving the optimization problem. The various parameters in the inspection scheme under two different decision threshold calculation methods were analyzed using numerical simulation. Among them, the calculation model with the minimum average number of experiments as the optimization objective has better engineering practicality.

Abstract:To overcome the difficulty that DVM(discrete velocity method) for solving the Boltzmann equation has extremely slow convergence speed and high computational resource consumption in the slip flow and early transition flow regimes, an acceleration scheme which coupling the mesoscopic/macroscopic equations in the full flow area was proposed. Boltzmann model equation could be solved based on the DVM using finite difference method at the mesoscopic level, and the moment equations could be solved based on the semi-implicit method for pressure linked equations using finite volume method at the macroscopic level. Fast convergence characteristic of the Navier-Stokes-Fourier/R26 moment equations in the low Knudsen number regime was fully utilized to accelerate mesoscopic equation. The distribute function could be reconstructed from the high-order Hermite polynomial function, so that the data transfer between the macroscopic and the mesoscopic equations could be done. Simulation results indicate that:the acceleration scheme, coupling the mesoscopic/macroscopic equations in the full flow area, shows great acceleration performance in the slip and early transition regime, which is able to save up to 95.28% computational time cost. However, the acceleration performance decreases significantly in the middle and large Knudsen number regime.

Abstract:The subcritical problem of wing flutter system with freeplay nonlinearity was addressed by introducing the NES(nonlinear energy sink) technology to improve the critical velocity of the system′s limit cycle oscillation. A dynamic model of flutter system with NES control for wings with freeplay nonlinearity was established, and the suppression effects of NES parameters such as mass ratio, frequency ratio, damping ratio, and relative position on the limit cycle oscillation of the flutter system were analyzed. The influence of NES parameters on the critical speed of the limit cycle oscillation of the flutter system was also investigated. The results show that a larger damping ratio can stabilize the system at a smaller natural frequency ratio but required a more stringent location requirement for the NES, i.e., closer to the leading edge of the wing. On the other hand, a smaller damping ratio results in a smaller NES mass required to stabilize the limit cycle response of the flutter system. When the NES location is closer to the leading edge of the wing, increasing the natural frequency ratio significantly improve the suppression effect of limit cycle oscillation, and increasing the mass ratio also significantly improve the suppression effect and critical speed of limit cycle oscillation. Additionally, a smaller damping ratio of the NES results in better suppression of limit cycle oscillation in the flutter system.

Abstract:A series of direct-connected tests were conducted on a rotating detonation ramjet with a cylindrical isolator-combustor configuration, under the conditions of a total temperature of 860 K and Ma=2 inflow. The influence of the combustor leading edge expansion angle(θ=30°,45°,60°, 90°) on the propagation characteristics, operating range, and pressure distribution of detonation waves was investigated. The results indicate that the combustion mode is consistently deflagration when the expansion angle of the leading edge of the combustor is 90°. As the expansion angle gradually decreases, the combustion mode transitions towards sawtooth and hybrid mode (including single wave stage). When the expansion angle of the leading edge of the combustor is 30°, the rotating detonation exhibits the widest self-sustaining operating range and highest combustor pressure. Additionally, as the expansion angle decreases, the lower limit of the equivalence ratio for achieving the hybrid mode decreases. At the same time, the impact of combustion modes on the inflow was analyzed and it was found that the periodic high-frequency pressure oscillation in the sawtooth wave/hybrid mode combustor could cause the positioning of leading edge of the shock train in the isolator to move upstream. The hybrid mode has the most significant impact on supersonic inflow.

Abstract:In order to solve the problem of trajectory maneuver adjustment caused by large deviation of flight trajectory after midcourse penetration of ballistic missile, an optimization model of maneuver adjustment timing strategy was established. A reverse sequence Q learning algorithm for maneuver adjustment was designed, and a Tile coding approximator encoding was used to encode the state characteristics space, and the space was linearly approximated. A reverse-order update strategy mechanism combining Q learning algorithm and Monte Carlo method was constructed, the optimal timing of missile maneuvering adjustment was trained. The simulation results show that the strategy obtained by training 10 000 generations of reinforcement learning algorithm can reliably control the adjustment decision of flight trajectory after missile penetration with the minimum maneuver times under given scenario parameters, which verifies the effectiveness of the method.

Abstract:An optimal design of SpaceFibre node transmission system was proposed to meet the high bandwidth, high reliability and light weight requirements for the SpaceFiber link between satellite payloads. This design adopts an incremental calculation method based on frame accumulation to reduce the redundancy of computing circuits, and satisfies the timing requirements for error detection operations through a four-stage pipeline architecture. The module balances the use of hardware resources by sharing CRC(cyclic redundancy check). To improve reliability, the design adopts a two-layer control strategy with complete response. Resource-optimized FSM(finite-state machine) and storage architecture are built to simplify the implementation of the retransmission algorithm. A verification system at board-level with two nodes was implemented in XC7Z100FFG900-2 FPGA(field programmable gate array). The results show that the design satisfies the functions stipulated by the standard. Compared with similar design schemes, the maximum frequency is increased by 1.5 times, and supports up to 6.25 Gbit/s transmission rate. And resources of lookup tables and storage are reduced, while register resources are similar. The design provides a reference for the development of high-speed and reliable SpaceFibre codecs with independent intellectual property rights.

Abstract:In order to solve the problem of imbalance between positive and negative samples in liquid rocket engine fault diagnosis, and to enable adaptive fault detection during engine steady working state, a anomaly detection model based on fast incremental one-class support vector machine was established. Feature engineering method was adopted to extract features from sensor-obtained multivariate time series. Through incremental leaning, the one-class support vector machine model was improved and applied to liquid rocket engine anomaly detection. The one-class support vector machine detection model was endowed with adaptability for various engine individuals and multiple working conditions, while increasing computing speed. The analysis results of multiple hot test data show that the model is effective, fast-training and practically valuable.

Abstract:In order to further advance the development of visual speech learning, the task definition and research significance of visual speech generation was expounded and the difficulties and challenges were deeply analyzed in this field. Besides, the current status and development level of visual speech generation research was introduced, and the recent mainstream methods were sorted, classified and commented based on the difference of generation frameworks. At the end of the paper, the potential problems and possible research directions of visual speech generation were discussed.

Abstract:Joint design of transmit beamforming and receiving filter for airborne MIMO(multiple-input multiple-output) radar in the background of signal-dependent clutter was studied. Airborne MIMO radar transmit beamspace STAP(space-time adaptive processing) signal model was established. In order to ensure the target detection performance in clutter environment, the SINR(signal-to-interference-plus-noise ratio) was maximized and a new iterative optimization algorithm based on the MM(maximization-minimization) framework was designed to solve the joint design problem. By properly finding the lower bound of the objective function, the convergence speed of the proposed algorithm was effectively improved and the running time of the proposed algorithm was reduced. In addition, compared with the traditional phased array radar and MIMO radar, the optimized transmit beamforming and receive filter can significantly improve the output SINR. Simulation results demonstrate the effectiveness of the proposed method.

Abstract:Based on 3D-TCAD simulations, the sensitivity region of SET(single-event transient) effect in 22 nm FDSOI (fully depleted silicon-on-insulator) devices were investigated. A comparison was made between the methods of using a single transistor and using an inverter to study the sensitivity region of device SET, in order to analyze the influence of heavy ion strike position on the SET sensitivity of 22 nm FDSOI devices in actual circuits, and to explain it from the perspective of charge collection mechanism. In depth analysis reveals that the parasitic bipolar amplification effect is sensitive to the position of heavy particle strike, which is the reason for the different sensitivity of SET in different regions of the device. The increased sensitivity of the drain caused by a constant voltage source connected to the drain of a single transistor is the reason why the SET sensitive area of the device in the single transistor and inverter is different. The research method of studying SET sensitive regions of the devices under FDSOI process was improved. The simulation result of inverter is more in line with the actual situation than single transistor, which will provide theoretical guidance for SET hardening.

Abstract:To grasp the film cooling characteristics of the turbine blade tip in transonic flow, the pressure sensitive paint technique was used to experimentally obtain the effects of the tip clearance gap and mass flow ratio on the film cooling performance of the blade tip. The results indicate that, under conditions of low mass flow ratio, increasing the tip clearance height effectively enhances the film coverage in the mid-chord region of the blade tip. However, when the mass flow ratio is high, changes in the tip clearance height have an insignificant effect on the distribution of film cooling efficiency in this region. Under conditions of small tip clearance heights, the cooling coverage in the mid-chord region deteriorates gradually as the mass flow ratio increases. In contrast, with larger tip clearance heights, an improvement in cooling coverage is only observed when the mass flow ratio increases from 0.1%+0.05% to 0.14%+0.07%.

Abstract:Aiming at the problem of the low dynamic range of extreme exposure images, such as underexposure and overexposure images, a detail perception image fusion algorithm based on multi-scale decomposition was proposed. After detail enhancement, the underexposed image was roughly fused with the overexposed image. Wavelet transform was used to perform multi-scale decomposition on the brightness component of the image after detail enhancement, and a special high-frequency and low-frequency fusion strategy was designed to realize the fine fusion of brightness components. The hue and saturation components of the coarse fusion image were recombined with the brightness components of the fine fusion image to obtain the final fusion result. Based on abundant testing data, the experimental results show that the visual effect of this method is excellent. The index of averaged MEF-SSIM is 0.985 4, and the index of averaged SSIM is 0.650 8, which are superior to the existing mainstream algorithms.

Abstract:The transponder deception jamming with DRFM(digital radio frequency memory) is highly coherent with real radar echoes, which makes it difficult for radar to distinguish real radar echoes and jamming. To address this issue, a DRFM-based deception jamming detection method based on Hough transform was proposed. The jamming signal model based on linear frequency modulation was established and the spectrum of the jamming harmonics was analyzed subsequently. Then, the short-time Fourier transform and two-dimensional constant false alarm rate detector were used to extract the features of the jamming signal. The Hough transform was used to complete deception jamming detection. The proposed method is based on the characteristics of DRFM deception jamming itself, and does not depend on prior information and application scenario. Moreover, it has low computational complexity and good detection performance under the condition of low signal-to-noise. The effectiveness of the proposed method is verified through simulations.

Abstract:Increasing transparency of modern warfare has brought serious challenges to the survival of missile launching vehicle on the battlefield. Around the launcher anti-damage ability of analysis and evaluation technology, the threat of damage and damage mechanism faced by launcher were analyzed, current research status of four types of damage(shock wave, kinetic energy, thermal, and electromagnetic) from the perspective of the application of anti-damage capability assessment were elaborated, and different damage types in the evaluation of anti-damage capability and application directions in the analysis were putted forward; based on the vulnerability analysis space theory, the development process of equipment vulnerability analysis was summarized. It was considered that the logical transfer relationship from physical space to performance space was the key to the vulnerability analysis of launchers at present. Research trends of vulnerability criteria were introduced from the physical criteria, performance criteria and classification criteria, and the key criteria for vulnerability of launchers was proposed. Research concept and the main problems faced by analytical methods of the anti-damage capability of the launcher were explicitly described. Research conclusions can provide references for the analysis and evaluate related research of the anti-damage capability of the missile launching vehicle.

Abstract:NL2SQL(natural language to structured query language) task aims to translate natural language queries into SQL(structured query language) executable by the database. A Chinese cross-domain NL2SQL algorithm enhanced by auxiliary tasks was proposed. Core idea was to perform multi-task training and improve the accuracy of the model by adding auxiliary tasks in the decoder and combining the prototype model. Auxiliary task was designed by modeling the database schema into a graph, predicting the dependency relations between the natural language queries and the nodes in the database schema graph, and explicitly modeling the dependency relations between the natural language query and the database schema. Through the improvement of auxiliary tasks, the model can better identify which tables/columns in the database schema are more effective for predicting the target SQL for specific natural language queries. Experimental results on the Chinese NL2SQL dataset DuSQL show that the algorithm after adding auxiliary tasks has achieved better results than the prototype model, and can better handle cross-domain NL2SQL task.

Abstract:Aiming at the problem of scarcity of expert knowledge required by knowledge-driven demand forecasting model and insufficient interpretability of data-driven demand forecasting model, a fuzzy prediction model of spare parts demand driven by data and knowledge was proposed. Based on the fuzzy clustering algorithm, the numerical data was clustered into a rule base with simple structure and strong interpretability. The domain expert knowledge was represented as a Mamdani-type rule base by utilizing fuzzy logic. On this basis, a new type of intelligent computing theory—fuzzy network theory was introduced, the two types of rule bases were merged into an initial prediction model. A genetic algorithm was employed to optimize the fuzzy set parameters of the model′s rule base to enhance the model′s predictive accuracy. Compared with the fuzzy clustering algorithm, the proposed model has advantages in interpretability and accuracy.

Abstract:To solve the increasingly prominent cost of traditional path lookup strategy in the Linux kernel, Staged Lookup was proposed to accelerate path lookup by dynamically caching hot directories to reduce file access latency.The essence of Staged Lookup lay in caching frequently used directory entries, thus avoiding the repetitive traversal of paths from the root node. Unlike the retrieval operations that start from the root node, Staged Lookup expanded the search strategy to allow for forward or backward path lookup from the most recently cached directory entry. A prototype of Staged Lookup was deployed on Linux kernel versions 3.14 and 5.4 and subjected to real system tests. Experimental data indicates that Staged Lookup can achieve performance improvements of up to 46.9% compared to traditional path lookup methods.

Abstract:Since existing low-illumination image enhancement networks have insufficient ability to perceive and express feature information of different scales, a low-illumination image enhancement network model based on pyramid asymptotic fusion was proposed. The network performed multiple down-sampling operations on the image to form a feature pyramid. It fused the feature maps at different scales by adding skip connections to three different branches of the feature pyramid. Fine recovery module further extracted the refined information, and restored the feature map to a normal light image. Results indicate that, the network model not only effectively enhances the brightness of the overall low-illumination image, but also maintains the detailed information and clear edge contours of the objects in the image. Moreover, it can effectively suppress the dark noise, and make the overall enhanced image realistic and natural.

Abstract:In view of the structural residual noise in the weighted nuclear norm minimization algorithm and the inability to maintain the edge structure of the image, a denoising method that minimizes the weighted kernel norm and improves the wavelet threshold was adopted. The total variation model to perform preliminary denoising of the noise image, and the noisy image to subtract the preliminary denoised image were used. An improved wavelet threshold function was used to denoise the noise difference image obtained after subtraction. The denoised residual image was superimposed with the preliminary denoised image, and the superimposed image was finally denoised using an iterative weighted kernel norm minimization algorithm based on the residual noise level. Compared with the more popular denoising algorithms, the PSNR and SSIM processed by this algorithm are improved, the texture structure of the image can be maintained, and the effect is better in a high-noise environment.