Identifying the complexity of air traffic is an important task in air traffic management. Most current algorithms are usually tested using some macro-indexes, such as aircraft density, aircraft clusters, stranded degree, and so on. In this paper, the air traffic situation is described from the perspective of complex networks: aircraft in airspace are regarded as nodes and edges form within Airborne Collision Avoidance System (ACAS) communication ranges. The dynamic air traffic situation is studied by selecting topological characteristic indexes such as loop numbers, node strength, average clustering coefficient, betweenness centrality and network efficiency. On this basis, Independent Component Analysis (ICA) is used to recognize air traffic complexity online and treat the smooth traffic as a training data set. The congestion is recognized according to the changes of SPE-statistic, I²-statistic and I_e²-statistic. The simulation results show that the proposed method has the ability to identify air traffic complexity well.

In a titled orbital milling process, the cutting tool is set as tilted against the axis of the hole with a small angle, which makes the revolving motion of cutting tool in conventional helical milling process be changed to a conical pendulum motion. It reduces the axial drilling forces and improves the hole-making quality. In this paper, the tilted orbital milling method is adopted to make a hole for the Carbon Fiber Reinforced Polymer (CFRP)/titanium alloy laminated structures. The material removal rates in the region of entrance and exit of hole, the ratios of the material removed by peripheral cutting edge and frontal cutting edge, and the velocity zero point in the frontal edge are analyzed. Compared with the conventional helical milling process, there is no sharply change of the material removal rate in the entrance and exit of hole, its peripheral cutting edge milling ratio is increased, and the the velocity zero point in the frontal edge is not cut. The axial cutting forces and cutting temperatures are analyzed through experiments. The results show that the axial cutting force and temperature decrease significantly in the tilted orbital milling process. The morphology of the holes is checked through Scanning Electron Microscope (SEM), and the results show that the delamination of CFRP in the region of entrance is eliminated and there is no obvious defect in the boundary region of CFRP and titanium alloy. The results of this research show that the tilted orbital milling method is helpful to improve the hole-making quality of CFRP/titanium alloy laminated structures and has a potential application in the industry.

Aimed at the problem that it is impossible to install high-precision sub-Inertial Measurement Unit (IMU) on every sub-array in engineering practice, a global estimation method based on the least square fitting for distributed Position and Orientation System (POS) is proposed. Firstly, a transfer alignment error model from master system to slave system considering one-dimensional flexibility deformation is built. Secondly, Kalman Filter (KF) is used to estimate motion parameters of array antenna of the sub-arrays with sub-IMUs. Finally, the least square fitting is used to estimate the motion parameters of the sub-arrays without sub-IMU. The semi-physical simulation results show that the proposed method can realize the global estimation of the motion parameters of array antenna, and the estimation precision of the motion parameters of sub-arrays without sub-IMU is almost the same as that of its neighboring sub-arrays.

The generation of lift and thrust mainly depends on the formation of low-pressure vortices above the arc groove on the leading edge of the fan-wing, which makes the lift and thrust have a strong coupling relationship. How to decouple and control the lift and thrust is the key to the further engineering application of the fan-wing. When the geometric parameters of the fan-wing airfoil are determined, the leading edge opening angle has the greatest influence on the aerodynamic performance. Therefore, the method of installing leading edge winglets on the leading edge of a base fan-wing airfoil was considered to change the opening angle of the leading edge. By means of numerical simulation, the effects of single, double and three leading-edge winglets on lift and thrust of the fan at different installation angles, inflow velocities and angles of attack are compared and analyzed. The results show that the angle control of the leading edge winglet installation angle can not only improve the lift and thrust of the fan-wing, but also control the strength and position of the low-pressure vortices to meet the requirements of active control of the aerodynamic force of the fan-wing, so the attitude control of the fan-wing aircraft can be realized.

For the aerospace engineering application of Distant Retrograde Orbit (DRO), a calculation method and orbit characteristics are studied, and the main perturbation factors of DRO in actual force environment are analyzed to provide a theoretical foundation for DRO's precise modeling and nominal orbit design. Firstly, the effectiveness of the stream function method in calculating the DRO periodic orbit family is verified by simulation examples. Secondly, the DRO periodic orbit family is calculated by adjusting the Jacobi constant, and the DRO orbits with different resonance ratios are obtained. The simulation results show that the trajectory of DRO with an integer resonance ratio in the Earth-Moon inertial coordinate system is a closed curve, while DRO orbits with non-integer resonance ratio are not closed. Finally, the main perturbation factors affecting DRO stability are analyzed by orbit extrapolation. The simulation results show that solar gravitation and lunar orbit eccentricity are the main perturbation factors that affect stability of DRO. In the dynamic model, the standard ephemeris is used to represent the motion state of the planet. When the integration time reaches more than 10 days, the model error is about kilometer-scale. Therefore, in the large spatial scale of the Earth-Moon system, the ephemeris model can be used to analyze the motion state of the DRO approximately in the real force environment, which could provide a basis for mission orbit design.

In order to investigate the forming limit of TA15 titanium alloy at high temperature and clarify the influence of parameters in the constitutive equation on the forming limit, the constitutive relationship of TA15 titanium alloy in high temperature environment was established considering the high temperature softening phenomenon, and meanwhile, the forming limit of TA15 titanium alloy plate at high temperature was obtained by high temperature forming limit test platform and theoretical predicted through applying the M-K instability theory, respectively. The theoretical results indicate that the major strain under the plane strain state increases from 0.18 to 0.33 when the temperature increases from 800℃ to 880℃. Based on the M-K instability theory and the established high temperature constitutive model, the influence of the parameters in the constitutive equation on the forming limit is analyzed. The results show that, increasing the values of hardening index and the rate sensitivity factor, and decreasing the value of the softening factor can increase the strain hardening rate, and consequently the strain state in the groove region is delayed approaching the plane strain state. Therefore the position of the forming limit curve in strain space is improved. At the same time, the theoretical calculation results show that the influence of the strain rate sensitivity factor on the left side of the forming limit curve is greater than that on the right side, and it is attributed to the fact that the effect of the strain rate sensitivity factor on the strain hardening rate under different strain size is different.

This paper investigates a Multi-Objective Immune Optimization Algorithm (MOIOA) based on danger theory to solve the problem of nonlinear Multi-Objective Probabilistic Constrained Programming (MOPCP) with unknown noise information. In the design of the algorithm, adaptive sampling methods are used to estimate each chance constraint's probability and objective values, while each evolving population is divided into infected, susceptible and uninfected sub-populations in terms of one specific immune response mechanism contained by danger theory. The capability of global and local search can be enhanced, relying upon simulated binary crossover, adaptive mutation probability and polynomial mutation strategy. Numerical experiment results show that the proposed multi-objective algorithm has high efficiency and has some advantages over seven comparative methods with regard to solution quality. It has application potential to complex engineering problems.

Radio interferometers can achieve high spatial resolution imaging by combining multiple groups of visibility data measured over long periods of time. However, the variable information of temporally variable source is missing. A image reconstruction algorithm of varied sources by sparse baseline aperture synthesis based on sparse constraint on direct sum of background and inter-frame difference is proposed. The brightness temperature at initial moment and the brightness temperature difference of adjacent moments are taken as the vector of solution to seek, and the brightness temperatures at different moments are the sums of them, which leads to the measuring equation of the brightness temperature at initial moment and the difference. Transient source images at different moments are reconstructed by solving the sparsity of brightness temperature at initial moment and brightness temperature difference of adjacent moments. The results of numerical experiments show that the proposed method matches the best on transient source in a local background and outperforms the existing methods on varying source in a global background.

The use of light scattering signals to achieve the simultaneous inversion of the fractal morphology and particle size distribution parameters of soot aggregates have important application value in flame radiation heat transfer simulation and pollution control. The direct model of inversion is based on the Rayleigh-Debye-Gans Polydisperse Fractal Approximation (RDG-PFA) light scattering theory. Two signal schemes were investigated: multi-angle scattering, multi-angle scattering and collimated transmittance. Before the inversion, by comparing the residual fitness value distributions of the two signal schemes, it is found that the simultaneous use of scattering and transmission signals effectively reduces the ill-posedness of the inverse problem. The inversion process is based on the Covariance Matrix Adaptive Evolutionary Strategy (CMA-ES) algorithm, which has a strong local search capability and provides a guarantee for fast and stable inversion of each target parameter. The final inversion results demonstrate the feasibility and universality of the method in a large search space. And it is also proved that the combination of multi-angle scattering and collimated transmittance effectively improves the inversion accuracy of the target parameters.

By the employment of quartz compensation chamber and high-speed camera, the visualization experimental study on the compensation chamber of a propylene loop heat pipe was implemented, which mainly focused on the variation of state of working fluid in the compensation chamber with the effect of the working fluid inventory and heat transfer capacity, and the effect of working fluid inventory on the heat transfer performance of the loop heat pipe. The results indicate that the optimal working fluid inventory for the loop heat pipe with volume of 51.4 mL is about 19.7 g. The liquid levels inside the compensation chamber are lower than the bayonet when the fluid inventory is less than the optimal one, intense two-phase heat exchange between the evaporator and the compensation chamber is confirmed by the observation of obvious condensation and flow of the liquid on the outer surface of the bayonet, and the condensation rate and flow velocity increase with the rise of heat transfer capacity; the heat transfer thermal resistance of the loop heat pipe decreases and the heat transfer capacity below 280 K increases with the rise of the fluid inventory. With an optimal fluid inventory, the liquid level inside the compensation chamber immerses the bayonet and is close to the top of the evaporator core, and thus the best performance is obtained: a maximum power of 40 W that can be transferred below 280 K and a corresponding thermal resistance of 2 K/W. Liquid levels inside the compensation chamber are higher than the top of the evaporator core when the fluid inventory is more than the optimal one. The heat transfer thermal resistance increases and the heat transfer capacity below 280 K decreases with the rise of the fluid inventory. The liquid distribution inside the compensation chamber and evaporator core has considerable effect on the heat leak between the evaporator and the compensation chamber, which is a significant factor for the influence of working fluid inventory on the performance of loop heat pipe.

In this paper, a helicopter trim calculation method based on the CFD/CSD loose coupling strategy is proposed. The CSD solver and rotor CFD solver use the blade elastic axis and pitch axis as media to exchange aerodynamic load and response data through linear interpolation method. In the coupling strategy of this paper, CFD and CSD solver operate in the time domain respectively, and exchange the data once per revolution. The aerodynamic force calculated by CFD solver is used to correct the aerodynamic input of aeroelastic analysis in trim calculation, until the trim parameters and CFD aerodynamic force no longer change in the iterative process, and the coupling trim solutions are obtained. In this paper, the SA349/2 helicopter is taken as an example to calculate the forward flight state. The results show that the coupling method in this paper converges rapidly and has good stability. The comparison between the calculation results and the measured flight values verifies the effectiveness of the proposed method, and it has a good ability to capture the aerodynamic curve of blade and the blade vortex distraction.

Light field imaging technology is an emerging non-contact temperature measurement technology. Aiming at the important role of light field imaging reconstruction algorithm for flame temperature field measurement, the flame temperature field reconstruction algorithm based on light field imaging was studied. A flame temperature field reconstruction method based on the light-field imaging technique is proposed, the generalized sourced multi-flux method is used as the calculation method of direct problem, and Landweber algorithm is applied to reconstruct the 3D temperature field of absorbing and scattering flame based on the flame light-field imaging model. The Least-Square QR (LSQR) decomposition algorithm is also introduced to our study as a comparison to verify the performance of Landweber algorithm. Effect of measurement errors on the computational accuracy is studied. The reconstruction results demonstrate that the temperature field can be reconstructed reasonably by these two methods, and even with 5% measurement error, the mean reconstruction relative errors are 0.91% and 0.92% respectively, which are acceptable. The comparative results show that the Landweber algorithm has the similar calculation precision as LSQR algorithm, but the calculation time of the Landweber algorithm is one tenth of that of LSQR algorithm, and thus the Landweber algorithm is much more efficient than LSQR algorithm.

In target tracking task for multi-agent formation, the agent will lose the target when it is blocked by obstacles in the environment and external disturbances can affect the time-varying formation tracking control for multi-agent systems. This paper studies the time-varying formation tracking and collision avoidance control for second-order multi-agent systems under the simultaneous existence of these two factors. A switching topology control strategy based on target tracking priority is adopted to achieve continuous tracking of the target in the obstacle occlusion environment. A formation tracking controller including the disturbance compensation term is designed based on active disturbance rejection theory. First, an active disturbance rejection time-varying formation target tracking control protocol is proposed for multi-agent systems with switching topologies based on consensus methods, and a formation command generation method based on tracking differentiator is presented. Then, an algorithm is designed to determine the control coefficient matrix, and the stability of the system under the protocol is analyzed and proved. Moreover, a collision avoidance control protocol is designed based on artificial potential field method. Finally, the active disturbance rejection time-varying formation target tracking and collision avoidance control protocol is proposed considering the occlusion of target tracking by obstacles in the environment. The simulation results show that the control protocol designed in this paper still has good control effect when the above two factors exist.

The logistics drone airport in the city needs the ability to take off and land multiple UAVs at the same time. An airport model with single inlet and single exit, route crossing and multiple aprons is designed for small and medium sized vertical takeoff and landing UAVs, and a flow control method based on graph theory to ensure flight safety is proposed. According to the flight status of drones in the field, arrival and departure time of drones are controlled, and flight routes within the airport are planned to ensure the flight safety of drones. The simulation results show that the flow control method can ensure safe and orderly flight of UAVs in the airport airspace. Arrival and departure operating capacity of airports with different sizes were tested separately. When the area of a single apron is fixed, the greater the number of aprons the larger the arrival capacity, but the smaller the arrival and departure capacity. Therefore, when designing an airport, it is necessary to properly plan the number of tarmac in a given airport area.

Single-frequency RTK technology is widely used in high-accuracy mapping, unmanned driving and other fields. A single-frequency RTK dynamic accuracy detection method is proposed to accurately quantify the dynamic positioning accuracy. It is easy to operate, suitable for a wide area and does not need additional auxiliary equipment. Using this method, we first establishes some reference lines on the ground, then measures RTK static and dynamic combined data along the reference lines in the run-stop mode. The static data are fitted by the total least squares algorithm to get the called checking lines, which are then used as the references for evaluating the dynamic positioning accuracy. The reliability of the proposed method is also checked using the same sets of the test data. During accuracy assessment, the average deviation from the dynamic point to the checking lines is used as the dynamic accuracy index, and the reliability index of the method is characterized by the difference between the distances of two reference lines and those of two checking lines. The experimental results show that the proposed method has high reliability and can accurately quantify the dynamic positioning accuracy of single-frequency RTK of about 2-5 cm.

Craters are the most significant topographic features on the surface of celestial bodies. The traditional method of craters identification is mainly to study the dichotomy of positive and negative samples of small craters, with low efficiency and accuracy. This paper takes large craters under the macroscopic view of the planet as the research object, combines the knowledge of digital image processing and neural network, creates a crater sample library of different data sources to study the influence of data source on network model generalization ability, and proposes a more efficient crater multi-classification identification method. Based on the Non-Maximum Suppression (NMS) algorithm, a higher precision crater detection algorithm is proposed. Through parameter optimization and experimental verification, the multi-scale and multi-classification craters automatic recognition network framework based on deep learning constructed in this paper achieves a high accuracy rate, with the recognition rate up to 0.985 on homologous verification set and 0.863 on heterogeneous verification set, and effectively improves the redundancy of detection box and false detection in target detection.

To solve the attitude tracking problem of transport aircraft with sensor faults and actuator faults, a backstepping fault-tolerant control method based on extended state observer is proposed. The state observer and controller are designed separately. The extended state observer with neural network is designed to estimate the flight states, sensor faults and actuator faults simultaneously. On this basis, the flight states are replaced by the estimates, the control law of attitude tracking control is designed based on backstepping control technique, and the command filter is also introduced into the fault-tolerant controller which can improve the control performance.The final bounded convergence of closed-loop system tracking errors is derived and proved using Lyapunov stability analysis. The simulation results show that the proposed method is effective and can solve the attitude angle tracking problem of transport aircraft in the presence of sensor faults and actuator faults.

The variation rule determination of oxygen concentration in the free space of fuel tank is the basis of the design of inerting system.However, there are many factors that influence the oxygen concentration in the free space of fuel tank. At present, people still lack the necessary analysis and calculation for the actual phenomenon of the diurnal temperature changes. Taking the central wing fuel tank in a certain airplane as the research object, based on FAR25 airworthiness clauses of diurnal temperature changes, this paper builds theory simulation model. The model is verified by experimental data, and the corresponding relationship between oxygen concentration of free space of fuel tank and temperature difference between day and night is discussed. The influence of the diurnal temperature range, oil load, initial concentration, dissolved oxygen precipitation and other factors on the fuel tank free space oxygen concentration is analyzed, and the initial oxygen concentration limit of fuel tank before shutdown at night that satisfies the requirement of airworthiness clauses is proposed. The study results show that the variation range of day and night temperature, oil load, initial oxygen concentration and other factors have different effects on the oxygen concentration in the free space of fuel tank. The initial oxygen concentration limit of the fuel tank before shutdown should be lower than the minimum oxygen concentration limit of 0.5% - 1%. The research results will be of good reference to the design of inerting system and the calculation of average flammability exposure time of fuel tank.

Digital measurement technology has been gradually applied in aircraft large-scale component measurement, which also brings about many issues, and the fundamental one is how to plan the number and layout of measurement points to well describe the curve or surface. Aimed at solving the issue of complicated curve or surface measurement points planning, a novel difference measurement points planning method based on deterministic representation is proposed. The curve's deterministic representation is acquired by using Non-Uniform Rational B-Splines (NURBS) theory, and the particle swarm optimization is adopted to optimize the control points and corresponding weights factors to construct the high-precision fitting curve. Two different measurement point planning strategies considering curvature and measurement uncertainty are designed, and combined with surface characteristics, a complete and high efficiency measurement point planning process is constructed. The measurement points planning algorithm proposed in this paper is programmed based on the CAA, and by taking test piece as experiment object, the feasibility of the algorithm and the effectiveness of the system are verified.

To solve the problem of oxygen mole fraction changing with time in the upper space of oil tank of oxygen consuming inerting system. A mathematical model was established for the reaction process of the Green On-Board Inert Gas Generation System (GOBIGGS), and the inerting process of the GOBIGGS and the On-Board Inert Gas Generation System (OBIGGS) was simulated by CFD method. The simulation results were compared with the experimental data, which verifies the accuracy of the simulation results. The research results show that, when the flow rate of the exhaust gas from the fuel tank in the GOBIGGS is identical with that of the Nitrogen Enriched Air (NEA) in the OBIGGS, GOBIGGS not only has shorter inerting time than OBIGGS, but also reduces the oxygen mole fraction of the fuel tank to a lower level. The inerting effect of the GOBIGGS is similar to that of the OBIGGS of NEA0 (100%N₂) under the same flow rate. GOBIGGS makes the upper oxygen mole fraction greater than the lower oxygen mole fraction in the gas phase space of the fuel tank, while OBIGGS is the opposite.

The traditional image stitching detection algorithm manually constructs the stitching features by researchers. With the advancement of technology and the continuous development of image processing technology, the limitations of the features of manual construction, such as weak robustness and difficult positioning, are gradually manifested. Aimed at this kind of problem, this paper proposes to construct a Convolutional Neural Network (CNN) by means of fixed pre-convolution kernel, and detect the image tampering area by feature self-learning. Through experiments and research, it is found that the features of the mosaic tampering area of the spliced tamper image can be learned by the CNN model. Prior to the CNN model, the convolution kernel uses a high-pass filter and the activation function uses an Exponential Linear Unit (ELU), which makes the CNN model be capable of identifying features such as splicing and tampering with image edge traces. The detection results show that the positioning accuracy for the falsification image's tampering area is 84.3% in the IEEE IFS-TC image forensics training set and the detection true negative rate of the tampering area is 96.18%.