2023年度优秀论文
为进一步提升学术影响力，激励和培育优秀作者，国防科技大学学报编辑部组织了2023年度优秀论文评选。编辑部综合论文学术质量、学术影响力和网络投票结果评选出10篇优秀论文。
关键词：暂无
2022, 44(4):180-189.
DOI: 10.11887/j.cn.202204020
Abstract:
Existing convolutional neural networks, which are used for radio frequency fingerprints recognition, process time-sequenced IQ (in-phase and quadrature) signals as images directly, resulting in low recognition accuracy and high computation complexity. IQCNet(convolutional neural network structure based on IQ correlation features), an efficient convolutional network structure, was proposed. IQCNet firstly extracted IQ correlation features and time domain features, then obtained the average value of each channel features through adaptive average pooling, and finally used only one fully connected layer for classification. Experimental results under a variety of channel conditions show that IQCNet improves recognition accuracy greatly with lower computation complexity compared with traditional convolutional neural networks.
Existing convolutional neural networks, which are used for radio frequency fingerprints recognition, process time-sequenced IQ (in-phase and quadrature) signals as images directly, resulting in low recognition accuracy and high computation complexity. IQCNet(convolutional neural network structure based on IQ correlation features), an efficient convolutional network structure, was proposed. IQCNet firstly extracted IQ correlation features and time domain features, then obtained the average value of each channel features through adaptive average pooling, and finally used only one fully connected layer for classification. Experimental results under a variety of channel conditions show that IQCNet improves recognition accuracy greatly with lower computation complexity compared with traditional convolutional neural networks.
2022, 44(4):52-59.
DOI: 10.11887/j.cn.202204006
Abstract:
When the robot is running in a complex environment with densely distributed obstacles, the DWA (dynamic window approach) algorithm is prone to obstacle avoidance failure or unreasonable planning. In this regard, an improved DWA planning algorithm based on MOPSO(multi-objective particle swarm optimization) was proposed. Based on the establishment of multi obstacle environment coverage model, a method was put forward for judging obstacle-dense areas in complex environments. And the original DWA algorithm was improved by optimizing the sub-evaluation functions. On these basis of the improved MOPSO algorithm, the adaptive change of DWA weight coefficients were transformed into a multi-objective optimization problem. According to the requirements of path planning, the safety distance and speed can be set as the optimization goals, moreover, the corresponding multi-objective optimization model was established. The results of a series of simulations show that this method enables the robot to effectively pass through the dense area of obstacles while taking account of the safety and speed of operation, and has better path planning effect.
When the robot is running in a complex environment with densely distributed obstacles, the DWA (dynamic window approach) algorithm is prone to obstacle avoidance failure or unreasonable planning. In this regard, an improved DWA planning algorithm based on MOPSO(multi-objective particle swarm optimization) was proposed. Based on the establishment of multi obstacle environment coverage model, a method was put forward for judging obstacle-dense areas in complex environments. And the original DWA algorithm was improved by optimizing the sub-evaluation functions. On these basis of the improved MOPSO algorithm, the adaptive change of DWA weight coefficients were transformed into a multi-objective optimization problem. According to the requirements of path planning, the safety distance and speed can be set as the optimization goals, moreover, the corresponding multi-objective optimization model was established. The results of a series of simulations show that this method enables the robot to effectively pass through the dense area of obstacles while taking account of the safety and speed of operation, and has better path planning effect.
2022, 44(3):93-103.
DOI: 10.11887/j.cn.202203012
Abstract:
Stealth aircraft has gradually become a great power, and will continue to play an important role, making stealth technology become the key technology of aircraft design. RCS(radar cross section) measurement of stealth aircraft is a necessary step to design, manufacture and maintain stealth aircraft. The basic process of RCS measurement of stealth aircraft was reviewed from three aspects:RCS test of scaled model, outdoor RCS test of full-scale aircraft and indoor near-field test of full-scale aircraft, the theoretical basis of RCS near-field measurement of stealth aircraft was summarized, and the near-field RCS measurement technology with imaging diagnosis function was analyzed intensively. The development trend and key technologies of stealth aircraft RCS measurement were prospected, which is helpful to have a general understanding of RCS measurement of stealth aircraft and grasp the development direction of RCS measurement.
Stealth aircraft has gradually become a great power, and will continue to play an important role, making stealth technology become the key technology of aircraft design. RCS(radar cross section) measurement of stealth aircraft is a necessary step to design, manufacture and maintain stealth aircraft. The basic process of RCS measurement of stealth aircraft was reviewed from three aspects:RCS test of scaled model, outdoor RCS test of full-scale aircraft and indoor near-field test of full-scale aircraft, the theoretical basis of RCS near-field measurement of stealth aircraft was summarized, and the near-field RCS measurement technology with imaging diagnosis function was analyzed intensively. The development trend and key technologies of stealth aircraft RCS measurement were prospected, which is helpful to have a general understanding of RCS measurement of stealth aircraft and grasp the development direction of RCS measurement.
2021, 43(4):76-84.
DOI: 10.11887/j.cn.202104010
Abstract:
In order to realize the high maneuverability of the electromagnetic catapult and the high power density of the linear catapult motor of the land-based unmanned aerial vehicle, a multi-population genetic algorithm with elite retention (MPGAER) was proposed to optimize the maximum power density of motor. It takes flux density and current density as constraints, and optimizes the structure parameters of the motor with its strong search ability and fast convergence speed, and compares with the initial scheme based on magnetic circuit method and the optimization results of traditional genetic algorithm. The results show that, compared with the initial scheme design, MPGAER can reduce the motor mass by 6.25%, increase the power density by 10%, and improve the dynamic performance of the motor; the motor power density of the MPGAER optimized design is higher than the genetic algorithm design result, which indicates that the proposed method effectively solves the problems of easy convergence to the local best advantage and poor optimization effect in the optimization process.
In order to realize the high maneuverability of the electromagnetic catapult and the high power density of the linear catapult motor of the land-based unmanned aerial vehicle, a multi-population genetic algorithm with elite retention (MPGAER) was proposed to optimize the maximum power density of motor. It takes flux density and current density as constraints, and optimizes the structure parameters of the motor with its strong search ability and fast convergence speed, and compares with the initial scheme based on magnetic circuit method and the optimization results of traditional genetic algorithm. The results show that, compared with the initial scheme design, MPGAER can reduce the motor mass by 6.25%, increase the power density by 10%, and improve the dynamic performance of the motor; the motor power density of the MPGAER optimized design is higher than the genetic algorithm design result, which indicates that the proposed method effectively solves the problems of easy convergence to the local best advantage and poor optimization effect in the optimization process.
2021, 43(3):98-106.
DOI: 10.11887/j.cn.202103013
Abstract:
Ice accretion affects flight safety severely. Intelligent ice accretion prediction is an important basis and support for the intelligent anti-icing and de-icing system design and safety design of aircraft. To solve the problem that there are multiple values of ice thickness in the normal direction of the same position of the airfoil surface with complex ice shape, a graphical ice accretion prediction method based on the transposed convolution neural network was proposed. The corresponding neural network structure, loss function and data specification of the prediction model were designed. The input of the prediction model was the data from flight and atmospheric conditions which directly affect ice accretion. The output of the prediction model is gray-scale image of ice shape. Ice shape data set was generated through numerical simulation method based on NACA0012 airfoil. To ensure the credibility of the generated data, wind tunnel test data was used to verify the numerical simulation method. Prediction model was established with five input parameters:liquid water content, median volumetric diameter, freezing time, temperature and flight speed, and was then trained and validated in simulation. The simulation results show that the proposed method can predict the ice shape quickly, and the main features of the predicted ice shape such as outline, upper and lower limits, position of ice horn and thickness fit well with the results of numerical simulation.
Ice accretion affects flight safety severely. Intelligent ice accretion prediction is an important basis and support for the intelligent anti-icing and de-icing system design and safety design of aircraft. To solve the problem that there are multiple values of ice thickness in the normal direction of the same position of the airfoil surface with complex ice shape, a graphical ice accretion prediction method based on the transposed convolution neural network was proposed. The corresponding neural network structure, loss function and data specification of the prediction model were designed. The input of the prediction model was the data from flight and atmospheric conditions which directly affect ice accretion. The output of the prediction model is gray-scale image of ice shape. Ice shape data set was generated through numerical simulation method based on NACA0012 airfoil. To ensure the credibility of the generated data, wind tunnel test data was used to verify the numerical simulation method. Prediction model was established with five input parameters:liquid water content, median volumetric diameter, freezing time, temperature and flight speed, and was then trained and validated in simulation. The simulation results show that the proposed method can predict the ice shape quickly, and the main features of the predicted ice shape such as outline, upper and lower limits, position of ice horn and thickness fit well with the results of numerical simulation.
2021, 43(2):25-32.
DOI: 10.11887/j.cn.202102004
Abstract:
The importance measure analysis can find out the important feature variables of model, which can effectively reduce the variable dimension and decrease the computation time. The relationship between the important measure of random forest and the variance-based global sensitivity measure was explored, which can give a novel way to solve variance-based main sensitivity index Si and total sensitivity index STi. The importance measure of single and group variables based on random forest were established to improve the corresponding measure index system. Several examples are given to verify the validity of the proposed important measures and the correctness relation derivation about variance-based sensitivity indices.
The importance measure analysis can find out the important feature variables of model, which can effectively reduce the variable dimension and decrease the computation time. The relationship between the important measure of random forest and the variance-based global sensitivity measure was explored, which can give a novel way to solve variance-based main sensitivity index Si and total sensitivity index STi. The importance measure of single and group variables based on random forest were established to improve the corresponding measure index system. Several examples are given to verify the validity of the proposed important measures and the correctness relation derivation about variance-based sensitivity indices.
2021, 43(1):155-162.
DOI: 10.11887/j.cn.202101020
Abstract:
Aiming at the characteristics of complex composition of aviation equipment system-of-systems, the fault tree analysis was introduced into the evaluation problem of aviation equipment′s structure contribution rate to system-of-systems, and an evaluation method based on fault tree analysis was proposed. The aviation equipment system-of-systems structure was constructed from the three aspects of operation equipment, information support equipment and support equipment. Furthermore, the fault tree of aviation equipment system-of-systems structure was constructed and the criticality importance index of bottom event was used to calculate the aviation equipment′s structure contribution rate to system-of-systems. And the evaluation model of aviation equipment′s structure contribution rate to system-of-systems based on criticality importance was established. Taking an aviation equipment system-of-systems as an example, the results show that the proposed method can evaluate the aviation equipment′s structure contribution rate to system-of-systems accurately and provide methodological support for short board determination and optimization of aviation equipment system-of-systems structure.
Aiming at the characteristics of complex composition of aviation equipment system-of-systems, the fault tree analysis was introduced into the evaluation problem of aviation equipment′s structure contribution rate to system-of-systems, and an evaluation method based on fault tree analysis was proposed. The aviation equipment system-of-systems structure was constructed from the three aspects of operation equipment, information support equipment and support equipment. Furthermore, the fault tree of aviation equipment system-of-systems structure was constructed and the criticality importance index of bottom event was used to calculate the aviation equipment′s structure contribution rate to system-of-systems. And the evaluation model of aviation equipment′s structure contribution rate to system-of-systems based on criticality importance was established. Taking an aviation equipment system-of-systems as an example, the results show that the proposed method can evaluate the aviation equipment′s structure contribution rate to system-of-systems accurately and provide methodological support for short board determination and optimization of aviation equipment system-of-systems structure.
2021, 43(1):127-138.
DOI: 10.11887/j.cn.202101017
Abstract:
As an important carrier of group intelligence, robot clusters have also received a lot of research and development and more and more applications. Robot cluster path planning technology has also developed rapidly as a core key technology. The technical development status of robot cluster path planning was comprehensively and deeply researched. The basic computing architecture applying to different cluster sizes, scalability requirements, communication requirements, and algorithms of the requirements was innovatively summarized. The basic computing architecture consists of redundant computing architecture, distributed computing architecture, and hierarchical computing architecture. From the perspective of scalability and applicability, the path planning method which takes advantage of the most suitable robot clusters, including bionics, artificial potential field method, geometric method, classical search method and evolutionary learning method, was classified. For the cluster simulation verification research, it provided seven free downloadable or open source robot cluster simulation verification platforms.
As an important carrier of group intelligence, robot clusters have also received a lot of research and development and more and more applications. Robot cluster path planning technology has also developed rapidly as a core key technology. The technical development status of robot cluster path planning was comprehensively and deeply researched. The basic computing architecture applying to different cluster sizes, scalability requirements, communication requirements, and algorithms of the requirements was innovatively summarized. The basic computing architecture consists of redundant computing architecture, distributed computing architecture, and hierarchical computing architecture. From the perspective of scalability and applicability, the path planning method which takes advantage of the most suitable robot clusters, including bionics, artificial potential field method, geometric method, classical search method and evolutionary learning method, was classified. For the cluster simulation verification research, it provided seven free downloadable or open source robot cluster simulation verification platforms.
2021, 43(1):109-116.
DOI: 10.11887/j.cn.202101015
Abstract:
With the rapid development of urban subway construction, many city subway lines have been networked.Considering the public welfare when making ticket price, most cities in China calculate the ticket price between the two stations of the line network by the shortest path or the least stations now. the traditional Dijkstra algorithm was improved based on the traditional Dijkstra algorithm. The traditional Dijkstra algorithm and the improved Dijkstra algorithm were respectively used to calculate the shortest walking route of Changsha metro line 1~5 network. Results show that the improved Dijkstra algorithm not only effectively improves the efficiency of the algorithm and overcomes the long-time defect of the traditional algorithm, but also partly eliminate the accumulated errors between lines and improve the ductility of the network.
With the rapid development of urban subway construction, many city subway lines have been networked.Considering the public welfare when making ticket price, most cities in China calculate the ticket price between the two stations of the line network by the shortest path or the least stations now. the traditional Dijkstra algorithm was improved based on the traditional Dijkstra algorithm. The traditional Dijkstra algorithm and the improved Dijkstra algorithm were respectively used to calculate the shortest walking route of Changsha metro line 1~5 network. Results show that the improved Dijkstra algorithm not only effectively improves the efficiency of the algorithm and overcomes the long-time defect of the traditional algorithm, but also partly eliminate the accumulated errors between lines and improve the ductility of the network.
2021, 43(1):41-48.
DOI: 10.11887/j.cn.202101006
Abstract:
In view of the path planning of the unmanned AUV (autonomous underwater vehicle) in the complex sea battlefield environment, the environment model is complex and there are many constraints. A relatively perfect sea battlefield environment model including battlefield shape, enemy threats, obstacles, and sea current field was established. The AUV navigation time and threat time were the shortest as the optimization goal, and an oscillation-based type was given. The AUV global path planning method of the oscillation IWO(invasive weeds optimization) algorithm was compared with other three path planning algorithms, such as the standard IWO algorithm, the full-oscillation IWO algorithm and the PSO (particle swarm optimization) algorithm respectively. Simulation results show that the proposed method which has strong searching ability and excellent robustness, can effectively plan the navigation path which meets the performance requirements for the AUV in the complex sea battlefield environment.
In view of the path planning of the unmanned AUV (autonomous underwater vehicle) in the complex sea battlefield environment, the environment model is complex and there are many constraints. A relatively perfect sea battlefield environment model including battlefield shape, enemy threats, obstacles, and sea current field was established. The AUV navigation time and threat time were the shortest as the optimization goal, and an oscillation-based type was given. The AUV global path planning method of the oscillation IWO(invasive weeds optimization) algorithm was compared with other three path planning algorithms, such as the standard IWO algorithm, the full-oscillation IWO algorithm and the PSO (particle swarm optimization) algorithm respectively. Simulation results show that the proposed method which has strong searching ability and excellent robustness, can effectively plan the navigation path which meets the performance requirements for the AUV in the complex sea battlefield environment.
