Abstract:In order to understand the selfpulsation characteristics of the gas/liquid openend swirl coaxial injector, the atomization test in the atmospheric environment is carried out by adopting the shadow method and the dynamic measurement system. The exit of injector nearby presents a periodic spray process significantly with upstream transient pressure responding repeatedly, while selfpulsation happens. Moreover, the frequency of former pressure, acoustic pressure and spray graphics are highly consistent, ranging from 1 583 Hz to 5 634 Hz. Through analyzing flow rate discipline, its concluded that the injection pressure always presents increasing due to the interaction of gas phase and liquid phase, whether the atomization field displays stable or selfpulsation. As well, liquid flow has stronger influence on the injection pressure than gas flow. The results show that selfpulsation frequency increases with gas phase and liquid phase mass flow, and is relatively sensitive to liquid phase mass flow. As the liquid mass flow growing to critical level, the selfpulsation disappears. Then the selfpulsation frequency keeps improving slowly with gas mass flow increasing. The openend swirl coaxial injector has the upper and lower liquid phase boundary condition generated selfpulsation, and the range widens as gas phase increases. So the gas phase promotes selfpulsation, however liquid phase hampers the development of the selfpulsation. When the liquid swirl injectors wall is thickened, the liquid film thickness increases and the dimensionless gas core shrinks. Furthermore, the selfpulsation frequency of the new coaxial injector is larger than that of the original one.

Abstract:Wake vortex field formed in the process of flight is an important factor exerting an influence on the flight safety. In order to study the impact of wake vortex on the aircraft below at altitudes above 12,500 m, a physical model of aircraft wake vortex is established by a fast simulation method of wake vortex, and the sinking height of wake vortices at different flight altitudes is also calculated by Monte Carlo simulation. The difference in the sinking height of vortex core at high altitude and midlow altitude is analyzed. Then, the sinking height of highaltitude wake vortices are calculated under condition of different aircraft weights and atmospheric turbulence intensities. The result shows that compared with those in mediumand lowaltitudes, wake vortices in highaltitudes have larger sinking height on an average increment of 42.4~49.7 m. Reducing the aircraft weight can reduce the vertical impact range of the highaltitude wake vortex. After the atmospheric turbulence intensity exceeds 1.2×10-4 m2/s3, the sinking height of highaltitude wake vortex core changes slowly. These results give a reference to some lights on the reduction of vertical separation of highaltitude wake vortex.

Abstract:In view of the safety conflict caused by goaround on the intersecting runway related operation mode, an entire conflict domain of intersecting runway is designed to keep the flight separation between approach goaround and departure aircraft. The entire conflict domain of cross runway conflict is defined as front window and rear window. The collision risk assessment models of the front window and the rear window of the direct and noncross converging runways are established in turn, and then the influence of the runway cross angle on the lateral, longitudinal, vertical directions, and the overall risk is studied. The direct cross convergence runway group of Wichita airport and noncross converging runways group of Tianfu airport are selected as an example. The results show that the entire conflict domain of Tianfu and Wichita airport ranges respectively from 11.89 and 10.93 km before the approach runway entrance to 6.44 and 6.85 km after the entrance. Compared with the timebased wake interval constraint in the regulations, the entire conflict domain of the cross runway can finally determine a safe interval between go around and departure aircraft, and can also provide a reference to the separation of the future domestic cross runway configuration under related operation mode.

Abstract:In order to study the jet characteristics of airborne plasma igniter, the formation and development process of plasma jet is recorded by schlieren camera technology, and the influence of the transverse inflow velocity on the jet characteristics of airborne plasma igniter is studied. The results show that under condition of the transverse inflow, the increase of the incoming flow velocity makes the deflection angle of the jet great. When the transverse inflow velocity Vm=5 m/s, the efflux Angle θ=15.9°; when the transverse inflow velocity Vm=20 m/s, the efflux Angle θ=77.1°; when the inflow velocity increases by 15 m/s, the efflux Angle increases by 3.8 times. At the same time, when the incoming flow velocity increases, the loss rate of plasma jet momentum and heat increase, resulting in the decrease of plasma jet penetration depth, jet velocity, jet area and jet area change rate.

Abstract:In order to meet the needs of flight test trajectory measurement, a highperformance flight test target trajectory measurement system is constructed. On the basis of this, an improved adaptive Kalman filtering method is proposed based on the multimachine learning to deal with the abnormal data in actual flight measurement. This method is to take the traditional unscented Kalman filter (UKF) as a basis. Firstly, BP neural network improved by genetic algorithm (GABPNN) is introduced to improve the UKF algorithm, realizing the regulation and correction of UKF global error, and improving the estimation accuracy of UKF. Furthermore, the outlier resistant technology is used to eliminate the isolated and spotted outliers in measurement, realizing the GABPNN The purpose of UKF's further improvement, and improving the robustness of filtering. Finally, the simulation used to verify the effectiveness of the new algorithm and the experimental analysis of the actual flight measurement data (the actual data obtained through the established trajectory measurement system) shows that the algorithm is valid.

Abstract:Being a new radar system in which frequency diverse array (FDA) combines with multiple input multiple output (MIMO) structure, the waveform diverse array radar has great application prospects in the field of main lobe interference suppression. In this paper, the FDAMIMO team interference suppression method based on logarithmic sparse array is analyzed. Firstly, the logarithmic sparse array is introduced into FDAMIMO system, and then the multimain lobe problem caused by the basic FDA “s” beam pattern is improved by using nonlinear frequency offset, and then, RCB algorithm is used to correct the steering vector mismatch in the process of interference suppression. Finally, the simulation results show that when the target and the jamming are separable in the range dimension but close in the angle dimension, the proposed method can achieve the effective null of the jamming while maintaining the main lobe gain, providing a new idea for the effective suppression of team interference.

Abstract:Microwave correlation imaging extends from the idea of quantum intensity correlation imaging to the microwave field, and this not only solves the problems of highresolution staring imaging and complex motion compensation in traditional radars, but also has the characteristics of high resolution and strong antiinterference ability, and extensive attention. In view of the high sampling number and poor reconstruction quality of traditional microwave correlation imaging reconstruction algorithms, this paper proposes a reconstruction method of microwave correlation imaging based on convolutional neural network and residual network. The radar receiver echo data are used as the input of the network, after the initial reconstruction, the trained feature extraction network and the image enhancement network are used for feature extraction and feature enhancement in order to perform image reconstruction. The algorithm in this paper is compared with pseudoinverse algorithm and compressed sensing algorithm. The simulation results show that compared with the existing optimal microwave correlation reconstruction compressed sensing algorithm, the image reconstructed by this algorithm has more 〖JP2〗advantages. At the same time and without sacrificing image quality, the expending time to execute a single image reconstruction program is about 0.06 s,〖JP〗 improving the speed of image reconstruction and having great significance to engineering applications.

Abstract:Aimed at the problems that the existing malicious code detection model is iow in recognition rate of malicious code and its variants, and the amount of parameters is too large, a method of Malicious code family detection model is proposed based on GhostDenseNetSE is proposed in combination with the lightweight convolutied Ghost, densely connected network DenseNet and the channel domain attention mechanism SE. This model serves as a type of compressing the model volume and improving the recognition rate. the standard convolutional layer in DenseNet is replaced with a lightweight Ghost module, and the channel domain attention mechanism is introduced to assign different weights to the feature channels to extract the key features of malicious code and improve model checking accuracy. The experimental results on the Malimg data set show that the models recognition accuracy of malicious code families can reach 9914%, Compared with AlexNet and VGGNet, the recognition accuracy increases by 1.34% and 2.98% respectively, and the amount of model parameters is lower. This algorithm not only improves the classification accuracy, but also reduces the complexity of the model. The algorithm has important engineering value and practical significance in malicious code detection.

Abstract:Being intelligence and high integration, solid state plasma reconfigurable antenna based on SPiN diode can be extensively used in modern communication system. A SPiN diode is designed on a high resistivity silicon wafer, and solid state plasma region is formed when a forward bias is applied on this diode, which greatly improves the antenna integration and reconfigurability. The simulation and experiment results show that the diode dimensions and boundary layers are optimized to trap carriers in welldefined channels approaching high concentration levels exceeding 1 018 cm-3. Besides, the antenna’s reconfigurability is demonstrated by turning different sections on or off to change the plasma region of the monopole. In this paper, there are two resonant frequencies at 8.83 GHz and 9.71 GHz are obtained, and the maximum gains of this monopole at two working states are 2.04 dB and 1.68 dB, respectively. Other radiation parameters of this antenna also show good performance, providing a new method for the design and application of highperformance reconfigurable antennas.

Abstract:Aimed at the problems that integrated navigation system is rapid divergent in error when global navigation satellite system (GNSS) signal is interrupted, most of the current methods are to adopt fully connected neural network, but this can only deal with the mapping relationship between input and output at a single moment with being ignorant of the dependence of errors on the past values of strapdown inertial navigation system (SINS), a method of using recurrent neural network (RNN) to assist the integrated navigation system is proposed. RNN can train on the basis of the current and past position and speed samples respectively, so that the neural network can better process the timing signals in the system, and predict the position and speed errors of SINS more accurately. UAV flight test data are adopted to verify the method. The result shows that the navigation accuracy of this algorithm is increased by 77% on average when the satellite signal is interrupted, and this meets the realtime requirements required by navigation. Compared with the traditional RBF neural network assisted integrated navigation system, the root mean square error of its position and velocity is reduced by 39% on average.

Abstract:In order to solve the problem of low detection accuracy of class imbalance data by existing intrusion detection systems based on machine learning, an intrusion detection method based on conditional Wasserstein generative adversarial network (CWGAN) and deep neural network (DNN) is proposed. CWGANDNN improve the class imbalance problem of data sets by generating samples, and the detection efficiency of intrusion detection system (IDS) on minority and unknown classes is increased. Firstly, the data are preprocessed by the variation Gaussian mixture model (VGM) to decompose the mixed distribution of continuous features. And then the CWGAN is used to learn the distribution of original dataset and generate minorityclass data to balance the training dataset, and train the DNN with balanced dataset. Finally, the trained DNN is used for intrusion detection. The experimental results on NSLKDD dataset show that the data generated by CWGAN can improve DNN’s classification accuracy and F1 score by 5%, but AUC decreases by 2%. Compared with other equalization methods, the accuracy, F1 score and AUC of CWGANDNN are improved by at least 3%, 1% and 1%.

Abstract:In view of the confusion problem of few shot learning in communication signal recognition, a communication signal recognition algorithm is proposed based on semisupervised generative adversarial network in this paper. Firstly, this algorithm utilizes a small amount of labeled samples and a large amount of unlabeled samples for training the network with the idea of semisupervised learning. Secondly, an auxiliary classifier in the output layer is added to determine the result and a new loss function and objective optimization are designed to meet the purpose of generating fake data and realizing signal classification. Finally, in order to reduce the complexity of the algorithm and accelerate the convergence speed of the algorithm effectively, different activation functions are chosen by the network and the deconvolution and Dropout are used to replace the pooling layer. The experimental results show that this algorithm is adaptable and computationally, the recognition accuracy is improved by 6%~13% compared with the traditional algorithm, effectively realizing the modulation pattern recognition in few shot learning.

Abstract:Aimed at the problem that adaptability of fuzzy control obstacle avoidance algorithm is poor in the environment with unknown obstacle information, this paper proposes an obstacle avoidance strategy for UAV based on fuzzy neural network. Firstly, the fuzzy controller is designed, and the input variables of the fuzzy controller are optimized. Secondly, the fuzzy neural network is trained by using the relevant data in the ideal obstacle avoidance path. The simulation results show that the method has stronger adaptability, better obstacle avoidance effect in the face of complex obstacles, can shorten the local planning path, can safely reach the target area in the unknown environment, and the motion trajectory is smooth.

Abstract:Aimed at the problems that there are the inconsistency and unreality in tracking and formation control objectives of multiple unmanned systems, an autonomous predictive control algorithm is designed in guaranteeing the consistency and reality of control objectives. Firstly, according to the given tracking and formation control objectives, the cost function is redesigned for each unmanned system by introducing planned control objective, solving the inconsistency in tracking and formation control objectives. Then, some additional constraints of the introduced planned objective are designed to guarantee its reality, and further to achieve the convergence to planned control objective, the terminal constraint set is designed for each unmanned system under condition of predictive control framework. Finally, for each unmanned system, an individual control optimization problem is constructed through combining the redesigned cost function and all designed constraints, and an associated autonomous control algorithm is presented for implementation. With applying the presented algorithm by each unmanned system, a consistent and realizable control objective can be planned based on the given inconsistent and unrealizable tracking and formation control objectives, and can be achieved by the whole multiple unmanned systems. By introducing Linear Matrix Inequalities (LMI) technique, the constructed control optimization problem is casted into LMI form and can be solved through LMI toolbox of MATLAB. The simulation results in MATLAB show that the algorithm is valid.

Abstract:Aimed at the problem that supplier selection method of military enterprises is not comprehensive yet, this paper proposes a supplier selection method for military enterprises based on the prospectregret theory. The requirements of military enterprises and the characteristics of military products are analyzed, and the evaluation indexes suitable for supplier selection of military enterprises are constructed. Considering the hesitation psychology in expert evaluation, probabilistic hesitation fuzzy set is introduced to describe the evaluation information. Based on the prospectregret theory, the risk attitude and regret avoidance psychology of decision makers are described and the evaluation information of the supplier scheme is aggregated.The evaluation results of the schemes are obtained and ranked. The scientificity and rationality of the proposed method are verified by a numerical example and contrast analysis.

Abstract:With the rapid development of science and technology, the traditional manmachine interface of antiaircraft antimissile equipment display and control platform is unable to adapt to the increase of information volume and time pressure in the course of combat. In order to improve the operation performance and reduce the cognitive load of operators, a design concept of “humancentered” should be emphasized more in the process of interface design, and a concept of intuitive interaction design should be integrated into the design process. Firstly, the manmachine interface of each display and control station in the charge cabin is analyzed, the existing problems are pointed out, the layout principle by using fuzzy analytic hierarchy process is quantified, and a layout optimization model is established. Secondly, in consideration of the fuzziness of the interface evaluation index, a multiattribute fuzzy comprehensive evaluation method is proposed to evaluate the optimization scheme. The optimization scheme and the evaluation method are valid.