For the inverse problem of electromagnetic tomography (EMT), the pathological and ill posed problems of the sensitivity matrix are discussed. A new electromagnetic tomography image reconstruction algorithm is proposed for this situation. Firstly, the principal component analysis (PCA) is used to reduce the dimension of the sensitivity matrix, and then the singular value decomposition (SVD) is used to calculate the generalized inverse matrix to reconstruct the image. After the covariance matrix of the sensitivity matrix is obtained, we need to compute the number of eigenvalues that the covariance matrix should retain. Then the maximization of the image correlation coefficient algorithm is proposed to solve it by using the unique multi-sample characteristics of the sensitivity matrix. It is more reasonable for sensitivity matrix to remove redundant information. And it improves the stability of the solution as far as possible without losing imaging feature information. When the actual data is used for imaging, this algorithm needs only one matrix multiplication, which provides the possibility for fast real-time imaging. In conclusion, compared with the traditional single step algorithm and iterative algorithm, the proposed algorithm has obvious advantages in both imaging quality and speed.

The research on the flow fluid in the annulus of coaxial rotating conical cylinders originates from the classical Taylor Couette flow in the annulus of two coaxial rotating conical cylinders. This paper uses flow visualization experiments and PIV to display and quantify the internal flow field in the annulus of conical cylinders, and the periodic trends of vortex motions are studied to explore the effect of in-annulus Reynolds stress distribution and water level on the flow field. The study indicates that the periodic trends of vortex column move downward over time, and a series of alternate clockwise and anticlockwise vortices are uniformly distributed inside the annulus where pulse number is 100 pulse/s. There exists the alternating vortex fracture periods where pulse number is between 200-500 pulse/s. Under three different water levels, there exists obvious periodic fracture, but the periodic time and number of vortices are different. There are two types of flow (up-convex outward flow and down-concave inward flow) in the mean flow field, and the type of flow is decided by centrifugal force and static pressure. Radial normal stress dominates in Reynolds stress distribution and mainly exists in the middle of the annulus.

Aimed at the problem of unclear description of the conduction relationship of complex equipment risks, a risk conduction uncertain random multi-transfer parameter graph evaluation and review technique (UR-MTPGERT) model is constructed. First, based on the opportunity theory, the moment function of uncertain random variables is defined, and then a multi-transter parameter UR-MTPGERT network is constructed. Second, to describe the micro-risk information of complex equipment systems, analytic parameters are introduced in the model including the degree of risk, the importance of risk primitives and the degree of relevance of risk paths. Then, when solving the moment function, the Delphi method is used to process the expert empirical data to obtain the empirical uncertainty distribution, and the maximum entropy model is used to process the random data. The probability density function is obtained. The matrix analysis technique is introduced to solve the problem of difficult network topology analysis. On this basis, the network parameters are calculated. Finally, the safety analysis of a certain type of aircraft is carried out. The results show that the model can clearly reflect the relationship between risk elements and provide reference for the risk analysis, prediction and safety control of complex equipment.

Electro-hydrostatic actuator (EHA), one of the key subsystems in the more electric aircraft, can reduce the size and weight of airborne equipment because of its highly compact design. However, the heat dissipation is weakened simultaneously, which can easily lead to an extremely rapid temperature increase and loss of function. In view of the drawbacks of traditional 1D thermal model, a "3D-1D-3D" thermal modeling method was proposed in order to research on the temperature rise of EHA driven by oil-cooled motor. Firstly, the conversion and transfer of energy in EHA were analyzed, aiming to find out the heat source and its diffusion. Further, the "3D-1D-3D" thermal modeling method was proposed considering that the parameters varied with time. Secondly, the 3D thermal model of EHA motor was established in ANSYS including the heat generation module and dissipation module, while the other components, for instance piston pump, hydraulic cylinder, valves, pressurized tank, etc., were studied with 1D thermal models. Finally, the whole "3D-1D-3D" thermal model of EHA was integrated on AMESim platform, and validated by experimental results. The results from simulations and experiments reveal the law of temperature rise in EHA, which provides the theoretical guidance for thermal design of EHA.

Two-bed onboard oxygen generation system is facing the problem of large pressure fluctuation of output gas and high oxygen concentration at low altitude during the practical application process. In order to solve these problems, three-bed onboard oxygen generation system needs to be developed. According to the system control logic, the control mode that the electronic controlled pneumatic servo valve worked circularly was adopted. The control design of three-bed onboard oxygen generation system has been proposed. The control method of segmented regulation was proposed, in which different cycle time was applied depending on the altitude. Under different input pressure and flow conditions, the experiments of the three-bed onboard oxygen generation system were performed to investigate the influence of the cycle time on the oxygen concentration of product gas. The study results show that when the cycle time for low altitude and high altitude is determined to be 9 s and 6 s, respectively, and the segmented altitude is determined to be 3.5 km, the system control design is suitable for the three-bed onboard oxygen generation system, and meets the requirements for the control design of three-bed oxygen generation system.

As a typical safety critical system, spacecraft is generally a complex system, which could produce a large amount of test data during the comprehensive testing process.When querying these test data, the existing B/S data query technology obtains data from the database server for each query, which greatly consumes the database server resources, takes up a lot of network bandwidth, and results in pooroverall performance of the system and poor user experience.Based on the classical Web cache replacement algorithm GDSF, this paper proposes a Web cache replacement algorithm GDSF-STW which is suitable for B/S architecture data query system by analyzing the characteristics of test data of the safety critical system and the behavior of user query.Based on the classical Web cache replacement algorithm GDSF, this algorithm introduces the time decay model in data mining and adopts the idea of sliding time window to improve the cache hit rate, system performance, and user experience. Finally, the experimental results show that the GDSF-STW has a better hit rate by comparing the GDSF-STW with the classical algorithms such as LRU, LFU, LFU-DA and GDSF.

Based on the mean impact value (MIV) which is an impact factor weight analysis method, the weights of fifteen impact factors on corrosion rate of four types of carbon steels (Q235, Ste355, St12, and Q450) were evaluated. The impact factors were corrosion duration, temperature, humidity, sunshine hours, precipitation, wind speed, sea salt ion concentration, SO₂, HCl, NO₂, H₂S, sulfation rate, NH₃, water soluble dust fall, and non-water soluble dust fall. These fifteen factors cover three important categories: corrosion duration, climatic factor, and environmental factor. The results show that when the regulation rate of MIV increases from 5% to 25% gradually, for different carbon steels, the weights of impact factors were similar, while the degrees of sensitivity were a little different; in the three categories, the climatic factor is of the largest impact on corrosion rate, followed by corrosion duration. The weights of SO₂ and dust fall are larger among the environmental factors. The most important three factors influencing the corrosion rate are mean relative humidity, sunshine hours, and mean temperature.

A method of aeroelastic optimization design with consideration of panel stiffness matching was developed for the composite wing of large aircraft. The optimization was performed based on the sensitivity algorithm, and the objective was to minimize the structural mass subject to the constraints of panel stiffness matching, flutter speed, deformation at wingtip, design allowable and manufacturability. The composite wings were designed in the case of critical load conditions. The influences of various panel stiffness matching requirements on optimal design results were studied and they were compared with the conventional optimal design results. The results indicate that the structural weight will increase with consideration of panel stiffness matching. However, it has an advantage in local buckling design, damage tolerance design and manufacturing of large composite panel. The optimal design results can be significantly affected by the design ranges of panel stiffness matching, so these design ranges should be properly determined according to the requirements of design and manufacturing. The design allowable of compression is a crucial constraint of the aeroelastic optimization design for composite wing.

The ranging accuracy of time of arrival (TOA) based indoor positioning system is significantly affected by multipath and non-line-of-sight (NLOS) of wireless channel in indoor environment. And these effects result in large measurement error and positioning error. In this paper, the optimization of distance is defined as a nonlinear programming problem. Based on the detection of line of sight (LOS) and NLOS, TOA ranging error model and geometric constraints between target and base stations are used to define the initial values, objective functions and constraint conditions for sequential quadratic nonlinear programming method effectively calibrates the positioning distance value. The typical TOA range error model is used for simulation. Field validation uses wireless positioning nodes with TOA ranging functions in the office environment. The results show that the ranging accuracy of the proposed algorithm is much higher than original range value and the other traditional distance mitigation algorithms, which verifies the effectiveness of the proposed algorithm.

The cables connecting avionics equipment are important coupling path for electromagnetic interference. When modeling a cable, the terminal common-mode impedance of the cable is a crucial input parameter. Because the quantity of cables in avionics system is very large, it is beneficial for quickly modeling cables' coupling through improving the test efficiency of cable terminal common-mode impedance. Therefore, a method to quickly test the terminal common-mode impedances of cables is proposed. First, according to the theory of multi-conductor transmission line, the multi-core cable can be equivalent to a single conductor when the common-mode current is analyzed. In addition, a cable bundle can be equivalent to a multi-conductor transmission line, and a vector network analyzer and a current probe can be used to measure the voltage reflection coefficients at two optional positions of each cable. Then, based on the established model of multi-conductor transmission line, the terminal impedance equations are constructed. Finally, the numerical iterative algorithm is used to solve the equations, and the terminal common-mode impedance of each cable is extracted. Compared with the existing methods, the testing efficiency and accuracy are improved.

Through analyzing unmanned combat aerial vechicle (UCAV) advantage probability and task joint threat and defining task time, the task allocation model for UCAVs with multi-constraint dynamic task allocation is built up, which takes target value damage, UCAV attrition and task expending time as the performance indexes, and the improved grey wolf optimization (GWO) algorithm is used to solve the model. Aimed at the flaw of early convergence from the original algorithm, the GWO algorithm is improved by proposing a self-adaptive adjustment strategy and a step-out local optimum strategy, using quadratic curve control method. According to the characteristics of UCAVs dynamic cooperative task allocation, target task sequence coding is designed to present the UCAVs dynamic task allocation method based on self-adaptive GWO (SAGWO) algorithm. Finally, the simulation results for static and dynamic task allocation show that the task allocation method based on SAGWO algorithm is valid, and compared with other algorithms, the optimizing process is rapid and accurate.

The existing elastomeric damper models commonly introduce dynamic amplitude parameter for applying to wide amplitude situation. It is inconvenient to time domain analysis of helicopter rotor-fuselage coupled dynamic stability on account of the dynamic amplitude changing in time domain. Aimed at this problem, the calculation methods of dynamic amplitude parameter were given for single and double frequency excitation cases while the lagging damping ratio is little. The amplitude curves calculated by the method describe the response amplitudes well while the system is in the state of convergence, neutral stable, or divergence. The improved model of elastomeric damper was used for nonlinear time domain analysis of helicopter ground resonance. The calculation method of excitation moment at blade was given for exciting the regressive lagging mode responses accurately. For different rotor speeds and complex modulus states, the response amplitudes excited by the excitation moment determined by the moment calculation method were compared with the desired values, and the maximum error is under 6%. After the regressive lagging mode responses are analyzed, it is known that the regressive lagging mode responses decay faster than the linearization results, and its modal damping increases in time domain due to the elastomeric damper nonlinearity while system is stable.

The uncertainty of dynamic characteristic parameters of pressure sensors is an important index to characterize its dynamic measurement performance. A method is proposed to evaluate the uncertainty of the sensor's dynamic characteristic parameters. Firstly, a shock tube dynamic calibration system is used to generate step pressure to excite the pressure sensor, and output signal of pressure sensor is obtained. Secondly, a preprocessing method based on empirical mode decomposition (EMD) is applied to reduce the influence of noise on output signals. Thirdly, an adaptive least squares method is performed to establish the mathematical model for pressure sensor based on its input and output signals, and the dynamic characteristic parameters in both time and frequency domains can be derived from the model. Finally, in consideration of the small sample feature of the parameter sequences in the repeated calibration experiments, the bootstrap method is applied to calculate the expanded uncertainty and relative uncertainty of these parameters. A set of dynamic calibration experiments for a pressure sensor are carried out with shock tube system. The uncertainties of dynamic characteristic parameters in both time and frequency domains are calculate and the results are compared with the existing methods. The experimental results show that the proposed method makes up the defect of the Bessel method in evaluating the sequence with small sample size. The relative errors of uncertainty results between the proposed method and Monte-Carlo method are less than 10%. It demonstrates that the proposed method works effectively in evaluating the uncertainty of pressure sensor's dynamic characteristic parameters. In addition, the analysis of the relative uncertainty evaluation results of the dynamic characteristic parameters in both time and frequency domains show that the work frequency band and overshoot are susceptible to the noises, and it can provide significant reference for the improvement of the dynamic calibration experiment conditions.

To allow small data center to work under the office environment to reduce the operating cost, an integrated design of noise reduction and refrigeration is proposed by combining the vapor compression refrigeration system with the server.Based on the verified simulation method of two key component, axial flow fan and evaporator, an airflow organization simulation model for this complex system is developed to predict the airflow and heat transfer. Nonuniformity evaluation indexes based on information entropy and variance are set up to evaluate the influence of different fan arrangement modes on the uniformity of server temperature field. The strategy to deal with larger heat density is analyzed. The results show that the nonuniform momentum drive of axial flow fan leads to uneven flow and heat transfer in the evaporator. The designed noise reduction and refrigeration system is able to maintain the exhaust temperature of the server between 21.6℃ and 22.2℃。Increasing the number of cooling fans can improve the uniformity of temperature field. When heat density increases, increasing the ventilation rate of the server can effectively reduce the exhaust temperature.

In consideration of the low diagnosis accuracy for avionics functional module fault, a new offline localized clustering multi-kernel extreme learning machine (LCMKELM) diagnosis model is proposed in this paper by combining the capabilities of multi-resolution interpretation and local feature self-adaptive representation from localized multi-kernel learning (LMKL) with the characteristic of high-performance operation from extreme learning machine (ELM). In order to avoid overfitting issue, affinity propagation (AP) clustering is used to make full use of the underlying localities in the training data and effectively reduce the computational complexity. Considering that the updating of localized kernel weights in dual optimization form of kernel ELM (KELM)is a difficult quadratic nonconvex problem, gating function M₁ and M₂ are respectively constructed to approximate localized weights by analyzing the clustering characteristics in input space and feature space. The proposed method is applied to actual fault diagnosis task of rotary transformer excitation generating circuit, and the experimental results show that the proposed method has the lower false alarm rate and missing alarm rate in comparison with four state-of-the-art multi-kernel learning algorithms, and meanwhile the diagnosis accuracy is averagely increased by 3.80% when M₁ gating model is used, and increased by 5.98% when M₂ gating model is used. Moreover, compared with canonical LMKL algorithms, the proposed method obtains similar training time cost, but it has less testing time cost.

In order to investigate the influence of deformation of gas film in gas-dynamic hemispherical bearings on the output of three-floated gyroscopes in the platform initial navigation system subject to 3-DOF specific forces, a mathematical model is established to calculate the static error by solving Reynolds equation. Firstl, Reynolds equation is modified to describe gas flow in hemispherical bearings considering the effect of gas rarefaction. Secondl, it is solved by finite difference method to obtain the pressure distribution, and the relationship between load and rotor displacement is used to calculate the gyroscope error. Finally, by regression analysis, a static error model of the gyroscope with gas-dynamic hemispherical bearings is obtained. To simplify the ternary regression analysis to binary regression analysis, the circumferential angle between interference torque and specific force, and the radial interference torque are introduced as intermediate parameters. Numerical results show that the radial interference torque increases with the increase of axial specific force. With the increase of radial specific force, the radial interference torque increases when the radial specific force is small, and decreases when the interference torque is large. Interference torque is 1.35-1.55 rad ahead of specific force in radial direction. The proposed static error model can predict the gyroscope static error caused by rotor displacement with any specific force below 300 m/s².

The transformable wheel is an effective way to enhance the wheeled mobile robots' obstacle navigation ability in harsh terrain. However, traditional single-structure transformable wheels can only adapt to specific environment and unable to simultaneously satisfy the environments with varieties of harsh terrain. The four kinds of multistage transformable wheel were proposed, which combined a kind of planar polygon with round wheel type, expended wheel type and wheel-leg type structure. At the same time, four different structures' advantage and disadvantages were compared. Giving out the corresponding design method, the length of the linkage was also optimized to propose a better solution. Through the effective shifting between the structures depending on different terrain, wheeled mobile robots' obstacle navigation ability were improved while also the adaptability of different terrains were enhanced to broaden the versatility of the mechanism.

Aimed at the aluminum alloy welding effect on corrosion sensitivity, we have applied loading on the 5A06 aluminum alloy with the welds in different directions (vertical, cross) and studied the effects of stress loading on the corrosion behavior of the alloy in 3.5%NaCl solution at 50℃ and the electrochemical corrosion behavior of the welded joints. The results show that after being soaked for 110 days in brine solution, heat-affected zone is firstly corroded and both sides of the corrosion pits on the surface exhibited step and dimple features respectively, because of the different material structure. The stress corrosion sensitivity of the cross weld specimen is relatively low and the stress concentration exists at the corrosion pit tip with a sliding step. Vertical weld specimen has higher sensitivity to stress corrosion with the corrosion pits across the surface along the weld. Both sides of the corrosion pits on the vertical weld specimen tear parallelly. The electrochemical polarization curves indicate that the corrosion potential of the heat-affected zone, base-metal zone and weld zone are -1.369 V, -0.791 V and -0.740 V, respectively.

The atomic magnetometer and Raman cooling need to lock the frequency of the laser on the detuning of several gigahertz away from the resonance. The laser frequency stabilization technique in Faraday rotation spectroscopy can stabilize the laser frequency on the large detuning away from the resonance. But, in this method, changing the frequency lock points can be complex and has high latency. We present a far off-resonance laser frequency stabilization method that can fast and accurately adjust the frequency lock points in the range of tens to hundreds of megahertz based on the Faraday rotation spectroscopy. Based on this method, the frequency lock point whose detuning is -6.2 GHz is precisely shifted by 130 MHz, and we obtain a frequency drift of 3.3 MHz/h and a root mean square fluctuation of 0.6 MHz/h. This satisfies the detuning and frequency stability requirements of atomic magnetometer. In this paper, the influence of temperature on the frequency stabilization method is analyzed, the physical constant in the detuning equation is measured to estimate the frequency of the stabilization point, and the temperature regulation and acousto-optic modulator (AOM) regulation are combined to improve the long-term stable and precise control of the laser frequency on the large detuning of off-resonance.

A symmetric non-equally spaced protocol named adjustable-parameter dynamical decoupling (APDD) was proposed to individually map ¹³C nuclear spin in diamond. The principles of the APDD sequence are in detail numerically calculated and simulation analyzed. The ¹³C nuclear spin addressing accuracy of APDD protocol was compared with Carr-Purcell-Meiboom-Gill (CPMG) and XY4 protocol, which are most widely used in diamond atom quantum manipulation. Results show that the ¹³C nuclear spin addressing accuracy of APDD protocol was 6.27 times higher than CPMG and XY4 sequences. Furthermore, τ₁/τ ratio ranging from 0.51 to 0.58 is the best operating condition for APDD protocol. The APDD protocol will pave the way towards manipulation of the nuclear spins surrounding the NV^-color center in diamond, which has significant applications in quantum information and quantum detector.

In order to verify the effectiveness of the formation control algorithms of UAV swarm in the actual environment, a distributed UAV swarm formation control demonstration and verification system consisting of UAV platform, two sets of ground station and data link was built under outdoor conditions. The control system of UAV was divided into executive layer and decision layer on the basis of the idea of hierarchical control and encapsulation. The PIX autopilot was encapsulated in the executive layer, and the control strategy of heterogeneous swarm can be realized only with autopilot parameter modification and without corresponding control strategy development for different UAV platforms. When the different control algorithms need to be verified, control algorithm for decision layer can be modified so that the system has strong adaptability and scalability. Two sets of ground control stations and data links are adopted, which can achieve the multi-UAV control under a variety of interaction topology or in the case of communication failure and ensure the system's robustness and security. The demonstration and verification system's function and performance are verified with the leader-follower synergy formation control algorithm.

Acceleration overlord during short-distance takeoff might affect the operation performance of pilot. To avoid potential risks, a pilot-cockpit multi-rigid-body model was established to simulate the man-machine dynamic response under chest-to-back (Gx) acceleration loads. Firstly, using 3D CAD software, the multi-body pilot dummy was established according to the 3% male body measurement data. Also, the cockpit model including human-seat and control system (throttle stick and steering axle) was established according to the design parameters. The dynamic response between pilot and control sticks under a standard simulated Gx acceleration curve is simulated using the multibody dynamics simulation software ADAMS. The simulation result shows that the overload acceleration can transmit through the human body to the sticks. The transmission force generated by 5G acceleration acting on the steering axle and throttle stick are 128 N and 211 N, both of which have exceeded the designed effective force threshold, leading to a potential risk of misoperation. Further, considering the simulation results and the existing foreign fighter design, we proposed an effective posture for the pilot to avoid misoperation, which was validated by the simulation in this paper. In conclusion, the proposed method provides an effective technology to avoid the misoperation in short-distance takeoff. The results would give a reference for the man-machine safety design.

Based on the structural design of the Lorentz force magnetic bearing (LFMB), a kind of torquer's non-circular error compensation method is put forward to enhance the sensitivity accuracy of magnetically suspended control sensitive gyroscope (MSCSG) to gyro carrier attitude. First, for a new type MSCSG with double spherical envelope rotor, the structure features of MSCSG and the attitude angular velocity measurement principle of the gyro carrier are introduced and the radius error model of MSCSG torquer, the interference torque model of rotor's deflection and the measurement error model of the gyro carrier's attitude angular velocity are set up. Then, the roundness of torquer is measured through the experiment, and data fitting is conducted by MATLAB to obtain the non-circular characteristic of the torquer. The non-circular characteristic is described by Legendre polynomial series, and the measurement error of the gyro carrier's attitude angular velocity caused by the torquer's non-circular error is compensated effectively. Finally, the effect of error compensation is verified by the simulation and the results show that the compensation method makes the measurement error of the gyro carrier's attitude angular velocity reduce by 83.5%. In addition, the proposed method can solve the common problems related to LFMB gyro.

To solve the problem of 3D reconstruction in textureless environment like railway, the confidence coefficient is proposed to combine TOF and stereo vision system effectively. Through the joint calibration, the coordinate relationship between TOF and stereo vision systems is established. Then by projecting the points in TOF to left camera in stereo vision system, the disparity map of TOF is obtained. After image segment and surface fitting the disparity map is up-sampled and its resolution is equal to that of stereo images. According to the confidence coefficients of different systems, the system weight values of data fusion are defined. Finally, the proposed method is evaluated with Middlebury dataset, and the results show that the accuracy has been raised twofold or more, and the resolution of disparity map is equal to that of stereo images as well.

Solid-gelled propellant gas generator is a new and special gas generator, and is the core component of the variable-depth missile ejection system. In order to study the temperature distribution of storage space of solid-gelled propellant gas generator, the source term method, the dynamic mesh method, the RNG k-ε turbulent model and the discrete ordinates (DO) radiation model were used to numerically simulate the working process of the solid-gelled propellant gas generator. And temperature distributions under different flow rates of gelled propellant were studied and compared. The results show that most of heat that the gelled propellant obtained comes from the high-temperature gas in the extrusion chamber, and the main heat transfer happens in the axial direction. As the extrusion flow rate of gelled propellant increases, the highest and average temperatures of the gelled propellant storage space increase. The results of this study can provide a reference for the thermal protective design and improvement of this type of solid-gelled propellant gas generator.

Lightweight materials such as high-strength steel, aluminum alloy or magnesium alloy applied as substitutions of plain steel to manufacture of bearing structures on vehicle have been an obvious tendency. Using industrial aluminum alloy as material, a new light bus frame was designed in this paper, and then finite element simulations under multiple typical conditions were implemented based on the reasonable finite element model. First, all the components of the frame were designed, assembled and checked in CATIA. Secondly, the frame-suspension finite element model was established in ANSYS Workbench after model simplification, connections simulation, etc. Finally, finite element simulations under totally five kinds of typical conditions such as full-load bending and emergency braking were accomplished, and the results indicate that the strength and stiffness of this new bus frame can satisfy the requirements of the typical conditions.