This paper dealt with the fluctuation and instability of contact force that isgenerated between robot end-effector and environment during the process of robotic grinding, polishing and deburring. In order to obtain constant tracking force, the contact force is generated by robot end-effector onto the workpiece surface is analyzed and the mapping relationship between the contact force of robot end-effector in real contact conditions and the known sensor coordinate system was built. Meanwhile, a hybrid force/position control scheme based on adaptive iterative learning algorithm was proposed to compensate the robot end-effector trajectory offset. The control method is composed with two steps. An iterative learning control law was designed based on the impe-dance model of robot-environment in dynamic interaction task. This control law coped with the unknown parameters and disturbances by adding the iterative term to the PD feedback structure. Meanwhile, a Lyapunov energy function was designed to prove the convergence of the control law. The adaptive iterative learning control law was then combined with the force/position hybrid control method to design the constant-force curved-surface-tracking control scheme with robotic manipulator. The experimental results show that after 15 times iteration, the fluctuating range of contact force becomes small gradually and is within ±3 N, which illustrates the effectiveness of the designed control scheme.

In order to solve the flight control problem of the air-breathing hypersonic vehicle, a new design method of neural inversion controller with prescribed performance was proposed. By constructing a prescribed performance function, it is ensured that the velocity tracking error and the altitude tracking error can converge to a desired area according to the prescribed convergence speed, overshoot amount and steady state error, and satisfy the preset transient performance and steady state accuracy of the system. Under the backstepping control design structure, the radial basis function(RBF) neural network was introduced to approximate the model unknown function and uncertainties, which improved the robustness of the control system. Only one parameter of the introduced RBF neural network needed to be updated online, which effectively improved the control accuracy, avoided the "differential expansion problem" in the backstepping control method, and reduced the burden of calculation. Finally, the simulation experiments verify the effectiveness and feasibility of the designed control system.

The various factors affecting the carrier aircraft catapult take-off safety are analyzed in detail. The nonlinear six-degree-of-freedom motion model of the aircraft in climbing phase is established. The effect of the rolling and yaw motion and crosswind on the catapult take-off characteristics is simulated and analyzed, and it is found that the main influences on the rolling and side-slip movement of the carrier aircraft after leaving the ship are the deck rolling motion and crosswind disturbance respectively. The nonlinear dynamic inverse control method is proposed to keep the nonlinear features of the model, make the result more precise, and realize the decoupling control of directional and lateral motion states. The simulation result indicates that the designed lateral control law can ensure that the rolling angle of the aircraft meets the safety criterion of not exceeding 5° within 3 seconds after leaving the ship and that there will be no obvious side-slip phenomenon due to crosswind disturbance, which can guarantee the safe take-off of carrier aircraft.

To improve the dorsal S-shaped inlet flow performance of low-energy inflow.The flow fields of both original and improved dorsal inlet were studied by using improved delayed detached-eddy simulation (IDDES) method, and the mass flow and pressure fluctuation characteristics of the two inlets were compared. The results show that, influenced by the fuselage, airflow turning angle near the inlet lip is very large and this will cause the formation of separation bubbles. The intensity of the separation vortex at the lip of the original inlet is very high. When the separation bubbles with high energy break down at the top of the inlet, massive high-strength vortices are created. These high-strength vortices aggravate the separation near the inlet top and cause fierce pressure fluctuations inside the S-shaped inlet. The amplitude of the sound pressure level can reach 145 dB. The strength of separation bubbles inside the improved inlet issuccessfully decreased, which enlarges the effective flow area of the inlet and increases mass flux. The pressure fluctuations inside the S-shaped inlet decrease and the reduction of sound pressure level is up to 8 dB. Meanwhile, the comprehensive distortion coefficient decreased by 9.5% which improves the characteristic of outlet flow field.

The multi-star simulator dynamically displays star image in the sky map and is the key device for testing a star sensor. As the star spot in the star image smears under dynamic conditions, it is necessary for the star simulator to accurately simulate the smeared star image in real time. First, the convolution surface model of the smeared star spot is established. This model is a convolution of the point spread function and the weight function along the star spot trajectory, and can represent any smeared star spot moving on the image plane. Then, a pixelating algorithm based on the convolution surface is proposed to improve the real-time performance of the star image simulation. The algorithm separates the trajectory into tiny arc segments, calculates pixel values of star spot at each arc segment, and sums all the spots. The pixelating algorithm turns multiple integrals into simple operations, and increases the simulation speed by nearly ten times, with a refresh rate of 30 Hz for multi-star simulator.

To guarantee the safety of low-altitude airspace, a target recognition method based on motion model was proposed for the non-cooperative UAV target in low-altitude airspace, as an extension of the application of the existing target tracking algorithm, which could detect the UAV and reject the false alarms such as flying birds with radar data. Firstly, multiple motion models were established to simulate the movement of UAV and flying bird targets. Secondly, the targets were tracked with multiple motion models and the appearance probabilities of these models were estimated. Thirdly, the transformation frequency between target motion models was measured by the mean variance of the appearance probabilities of multiple models in continuous time domain. By processing the simulation data and the measured data of the airport low-altitude surveillance radar, the method can track the UAV target in cluttered environment and eliminate the flying bird target, further verifying its effectiveness and practicability.

In order to study the non-destructive testing of micro-crack orientation angle of metal materials, the research of the ultrasonic sum frequency nonlinear effect about the micro-crack orientation of metallic materials is carried out. In theory, the relationship between the ultrasonic nonlinear frequency characteristic coefficient and the orientation angle of micro-crack is established. Then, the results of finite element simulation and calculation show that with the gradual increase of the orientation angle of the micro-cracks, there is a clear positive correlation trend between the ultrasonic nonlinear frequency characteristic coefficient and the micro-crack orientation angle, and compared to the secondary nonlinear coefficient, the sum frequency nonlinear coefficient is more sensitive to micro-crack orientation detection. At the same time, from the perspective of the average ultrasonic wave energy density, for example, the sound intensity, the sound intensity of the sum frequency component will increase with the increase of the orientation angle of the micro-crack, and the sound intensity of the second harmonic component will not change substantially. The ratio of the sound intensity of the sum frequency component is also significantly higher than that of the second harmonic component. The calculation results of the ultrasonic intensity are basically consistent with the simulation results, which proves the correctness of the theoretical model. Finally, through the design experiments, the use of simulated cracks to verify the validity of the model provides an effective means for the detection of micro-crack orientation of metallic materials.

Icing will destroy the dynamic performance of the aircraft and cause the safety envelope shrink, which seriously affects the flight safety. It is of great significance to study the changes of nonlinear stability region of the icing aircraft for reducing flight accidents. In this paper, the NASA's GTM is taken as the object aircraft. First, the dynamic model of longitudinal channel under icing condition is established based on polynomial fitting of the aerodynamic parameters and the icing factor model. Then, the change of flight state under different icing conditions and control commands is studied by bifurcation analysis method which used to guide flight manipulation. Considering the limitation of bifurcation analysis method, the nonlinear stability region of flight system is determined by differential manifold theory. And the nonlinear stability region is regarded as flight safety boundary. Finally, considering the icing condition, the bifurcation analysis method and differential manifold theory are combined to guide manipulation. Furthermore, the time domain validation of the manipulation is carried out. The results show that icing will shrink the safety boundary, and a slight disturbance may contribute to flight state outside the safety boundary. Moreover, with the increasing degree of icing, the stability of the aircraft will even change and the flight state will be difficult to maintain within the original safety boundary. At this moment, the flight state can be brought to the new safety boundary by changing the pilot's manipulation instruction. The research results are helpful for flight safety manipulation and boundary protection.

Mixed uncertainties of random variables and fuzzy variables are ubiquitous in the parameters of the current mechanism products, but the existing fuzzy reliability model mainly aims at static problems, which cannot describe the time-dependent problem with mixed uncertainty. This paper proposes a reliability modeling and computation method of the fuzzy time-dependent mechanism based on the advanced envelope function through the kinematic error analysis of mechanism and considering the fuzziness of failure criterion and the variables. First, fuzzy criterion can be transferred into random variables in the limit state function. Then, the cut set of fuzzy theory can be used to deal with the fuzzy and random variables, and thus the fuzzy time-dependent reliability model is built. After that, the advanced envolope function is used to calculate the time-dependent reliability of the mechanism. Finally, the feasibility of the method is verified by the motion error issue of four-bar linkage. The results show that the method has high computational accuracy.

In order to satisfy the special requirement for fire control of air-to-air missiles under unmanned air combat conditions, the problem of launchable area of air-to-air missile based on target maneuver estimation is presented. Based on the missile-target tracking escape countermeasure strategy, the target maneuver estimation model is designed. According to the relative position information of the missile and the target, the estimation of the target escape maneuver mode is realized. Based on a variety of practical constraints, a missile dynamics model is constructed. A launchable area boundary solving strategy based on the golden section search strategy is designed to achieve a fast and accurate search for the boundary value of the launchable area. The simulation results show that the air-to-air missile has a greater probability of hitting the target in the launchable area based on the target maneuver estimation proposed in the paper. The presented launchable area is more suitable for the dramatic change of the target escape maneuver in the close air combat, which is beneficial to the full play of the missile tactical performance.

Aimed at the problem of temperature field uniformity of proton exchange furnace, combined with the structural characteristic of the furnace, temperature control algorithm was developed based on FLUENT user-defined function (UDF). Based on this, various heating and temperature control schemes were proposed. The FLUENT software was used to simulate temperature field of the furnace under different schemes. The relationship between the temperature field uniformity and the placement of the sensor and the height of the heating wire under different temperature control methods was analyzed to find the best scheme. The results show that the temperature field uniformity is best when three temperature controllers are used, the positions of three sensors are respectively arranged in the middle of three heating wires, and the height of heating wire is arranged 4 times the length of the uniform temperature zone. The maximum temperature deviation in the uniform temperature zone is 0.03℃. For a given structure of the vertical furnace, increasing the height of the heating wire and optimizing the design of the sensor layout and temperature control method of the furnace body can improve the temperature field uniformity. This method provides ideas for optimizing temperature field uniformity of the same type of electric heating furnace.

The high-aspect-ratio flexible wing has become the main structural type of emerging aircraft with the increasing demand and performance improvement of aircraft. The wing type holds the inherent characteristics of high lift-to-drag ratio, large deformation and low weight, and the geometric nonlinear effect is obvious. However, the high aspect ratio will lead to larger wing deformation, resulting in nonlinear aeroelastic behavior. To evaluate the nonlinear aeroelastic behavior and reduce the risk and cost of the design, it is necessary to design a scaling model and conduct wind tunnel test with a scaling model to represent the aeroelastic characteristics of real aircraft. Based on this purpose, traditional linear scaling approaches are applied first. Two linear scaling methods, stiffness-mass coupled matched modal response and stiffness-mass decoupled matched modal response, and continually optimizes the design parameters of scaled model structure to meet the target values. Then, a new method named the nonlinear static deformation and mode collaborative optimization of the dynamic finite element model is proposed, which employs two different optimization subroutines to match the nonlinear static response and the mode shapes according to the full model equivalent static loads. The results show that, compared with the traditional linear scaling model, the nonlinear aeroelastic behavior of the full-size aircraft can be reproduced better by using the geometric nonlinear scaling method.

The life monitoring of the landing gear structure plays an important role in ensuring its safety and economy. However, it is difficult to accurately predict the fatigue damage of the individual spectrum due to the complex high-low load interaction of the ground spectrum. Therefore, based on the equivalent damage calculation method, the paper analyzes the similarity between the individual spectrum and the reference spectrum by analyzing the individual spectrum in the life monitoring, and then analyzes the applicability of the damage calculation method. An approach for comparing the similarity of landing gear load spectrum based on dynamic time warping method in time series analysis is proposed, and fatigue test under the reference spectrum and 4 individual spectra is conducted. By analyzing the relationship between the damage calculation error and the similar distance of the landing gear load spectrum, the rationality of the load spectrum similarity discrimination approach is verified.

For the complex fault modes in the redundant system composed of servo-controlled hydraulic actuator (SHA) and electromechanical actuator (EMA), the fault diagnosis method based on bond graph model can be used to diagnose many parameter faults in the system. Firstly, the behavioral model of SHA/EMA redundant system was established, from which the diagnostic bond graph model was transformed according to the causality inversion method, and then the analytical redundancy relation (ARR) was derived to calculate system residuals, and the fault signature matrix (FSM) was created as a basis for fault isolation. Several typical faults were selected for simulation verification. Behavioral model and diagnostic model were coupled to diagnose isolated faults, and fault parameters were estimated through ARR to diagnose inseparable faults. The results show that both isolated and inseparable faults are successfully isolated, and this method is verified to be effective and feasible for fault diagnosis of SHA/EMA redundant system.

In order to overcome the defect that the phased array beam has only angular resolution, the two-dimensional range-angle correlation of the beam is realized by adding a very small frequency increment relative to the carrier frequency between the array elements in the frequency diverse array. Three kinds of reception signal processing mechanism were introduced, and the theoretical deduction and analysis were carried out. The simulation shows the practicability of the two mechanisms. Aimed at target steering vector estimation and real target steering vector mismatch problem in the presence of the pointing error, the closed solution of the corrected steering vector is given by using robust Capon beamforming(RCB) algorithm. And in the two kinds of signal processing mechanism, the beampattern is simulated. The simulation results show that the RCB algorithm can form a high gain in the target position and form a null in the interference position. The effectiveness of the algorithm in the frequency diverse array is verified.

In order to establish dynamic and static modulus correlation analysis model of subgrade with taxiing aircraft, this paper considers the subgrade stress level of taxiing aircraft, compaction degree and moisture content of typical subgrade, and common frequency of taxiing aircraft. Dynamic and static triaxial tests are conducted to analyze the influence of stress level, compaction degree, moisture content and frequency on the dynamic and static modulus of subgrade. The results show that dynamic and static modulus are positively correlated with the compaction degree and confining pressure, and negatively correlated to the moisture content. Especially, the variation amplitude of dynamic modulus is significant when the frequency is less than 3 Hz. Meanwhile, based on dynamic and static modulus measurement databases, dynamic and static modulus conversion system is established and verified with multifactor comprehensive function and taxiing aircraft. The research provides the reference for pavement design, construction, detection and maintenance.

Due to the limitations of wind tunnel test conditions, the real flight environment of Mars reentry aircraft is difficult to fully be simulated, so it is necessary to establish the relationship of Mars reentry vehicle between windtunnel conditions and real flight. In this investigation, based on the published data of the literatures, the numerical method is used to study the extrapolation methods between the wind tunnel test and the real flight of Mars re-entry vehicle with the shape of the Pathfinder. The results show that under the conditions of high enthalpy air wind tunnel and conventional air wind tunnel test, the dimensionless pressure and pressure coefficient of near the stagnation point of the windtunel model can be used as a correlation parameters between the wind tunnel that and the flying conditions. However, the heat flow, the dimensionless heat flow and Stanton number of wind tunnel test cannot be directly used as correlation parameters under the conditions of high enthalpy air wind tunnel test and conventional air wind tunnel test. Under the high enthalpy CO₂ wind tunnel test conditions, the pressure coefficient, dimensionless pressure and Stanton number near the windtunnel model's stagnation point can be used as extrapolation parameters, but not the heat flow and dimensionless heat flow of the wind tunnel test are directly used as correlation parameters to obtain the performance parameters of the aircraft under flight conditions.

By studying and analyzing the vulnerability detection method of network program, a vulnerability detection method based on program modeling was proposed, and this method is aimed at the binary network program vulnerability in the C/S structure. The method analyzes the network program architecture, extracts the key system functions in different types of network programs, and develops the program modeling and detection system execution module. The technique of selective symbolic execution is adopted for detection, the semantics of hook function and the operation of trigger are customized by means of function hooks, and the execution process of symbolic data and guiding symbol is introduced. Based on this technology, a network program vulnerability detection system is realized. This system can identify the target web application using the I/O model. According to the different types of target program, it can call the different detection modules and use selective symbolic execution technology to implement the automated vulnerability detection process. The experimental results show that, compared with the existing detection tools, the system is more targeted in the vulnerability detection of network programs with higher vulnerability detection rate and good scalability.

Scale invariant feature transform (SIFT) algorithm is widely used in the field of computer vision because of its excellent robustness. In order to solve the problem of low real-time performance of computation-intensive SIFT algorithm on CPU, a fast SIFT hardware architecture is proposed based on field programmable gate array (FPGA), with reduced complexity by optimizing the feature descriptor extraction part of the algorithm. By reducing the bit width of gradient information (including gradient amplitude and gradient direction), optimizing the generation of the Gauss weight coefficients, simplifying the calculation of the three linear interpolation coefficients and simplifying the computation process of the histogram index of the gradient amplitude, the proposed design avoids complex computations such as exponent, trigonometric function and multiplication, and reduces the complexity of hardware architecture and hardware resource consumption. The experimental results show that the proposed low-complexity fast SIFT hardware architecture can speed up by about 200 times compared to the software implementation. Compared with the related research, the speed is improved by 3 times and the stability of the feature descriptor is increased by more than 18%.

Aircraft sheet metal part is an important part of the aircraft structure, which has the characteristics of large quantity and shape variety. The time consumed in the production process of aircraft sheet metal parts occupies a large proportion in the process of aircraft development. Current production process through the functions provided by the existing CAD software cannot meet the requirements of modern aircraft design and manufacture in terms of efficiency and quality. Research and development of relevant automatic design and manufacturing system for aircraft sheet metal parts have become an urgent demand. Recognition and extraction of basic features of parts based on B-rep model are the basis and premise for subsequent related process planning and manufacturing. Aimed at this, this paper proposes the basic feature model of parts and presents a feature recognition algorithm based on same-side face. That is, with the STEP data as input, the web faces on both sides of parts are selected. Using the methods of attribute adjacency graph (AAG) construction, effective-adjacent faces recognition and complete recognition of the correlative faces, the correlative faces at all levels are recognized step by step to construct the same-side faces on both sides. Finally, the basic features and their adjacency graph are constructed by matching of the same-side unit. In order to ensure the integrity of feature faces, the definition and recognition method of fragmentary face defects in 3D part model are presented. Examples are given to illustrate the feasibility and effectiveness of the approach.

For efficient countermeasure against fault injection attacks, a mixed-grain parity-code-based fault detection approach was proposed. Traditional parity-code-based fault detection approach assigns a parity bit per n bits, representing the parity of the n-bit word. As n decreases, the number of parity bits increases, leading to increased resource usage and fault detection rate. To achieve tradeoff between fault coverage and resource usage, the fine-grain parity code (small n) was applied to the data processed in the fault-sensitive parts or critical parts of circuits, and the coarse-grain parity code was applied to other parts of circuits. The approach was applied to RC5 encryption algorithm to explain the principle and application of the mixed-grain parity-code-based fault detection technology, and to theoretically analyze the fault coverage and resource usage of different grain solutions. The experimental results show that, compared to the RC5 circuit with one parity bit per 32 bit, the mixed-grain parity-code-based detection approach improves the fault coverage by 29.44% and increases resource usage slightly by 2.48%.

As a small and low-cost sun sensor, the photodiode, combined with the earth sensor, can determinate full three-axis satellite attitude. However, the photodiode is sensitive to the surrounding light sources, such as the earth, which limits its application. The mathematical model of the earth's albedo is complicated. To solve this problem, a simplified measurement model of the photodiode, describing the effects of the earth's albedo as the dynamic bias and deviation as the mixed-Gaussian noise, is established first. Then parameters in the model are online estimated and updated with windowing and random weighting strategies. To improve the accuracy of the parameter estimation and robustness of the algorithm, the multi-scale factors are used to estimate the influence of albedo on each photodiode, and the Huber function is introduced to prevent the outliers. The experimental results show that the high-precision satellite attitude estimation can be achieved by the new measurement model and the unscented Kalman filter (UKF) algorithm, and the three-axis attitude accuracy can arrive at 0.2°-0.3°.

In practical engineering, a system is usually complex and repairable. Its reliability model is not a series model but a mixed model of series, parallel, bypass and voting. The equal-distribution method is used mostly to allocate mission reliability for complex system at present. The allocation result using this method is always not accurate. Therefore, a mission reliability allocation method for repairable system considering the effects of maintenance, fault logic and other factors was proposed in this paper. First, the failure rate was calculated to remove the effect of maintenance. Then the scoring allocation method considering failure logic was adopted to allocate mission reliability. After that, the spare part coefficient was used to correct the allocation result. Finally, a case was analyzed to validate the method, which could provide an easy and effective way to allocate mission reliability for repairable systems in practical engineering.

Laser frequency stabilization is a common request in the development of quantum sensors. The frequency of the laser often needs to be locked on the detuning far away from the resonance, which is key to improve the sensitivity and accuracy of quantum sensors. Aimed at far off-resonance laser frequency stabilization, a method of frequency stabilization based on a Fabry-Perot (F-P) cavity to transfer the stability from a previously locked laser is proposed. A laser frequency was stabilized by saturated absorption spectrum method as a reference to lock the length of F-P cavity, and the F-P cavity was locked according to a lock-in principle. The highly stable length of F-P cavity, as a benchmark, helped the target laser wavelength to be locked on 767.001 nm, which is as far as 150 GHz from the resonance. The results show that the laser frequency drift is 1 MHz/h after locked. The method solves the problem of far off-resonance laser frequency stabilization, which is of great significance to engineering practice and scientific research.