The design of mini-missiles(MMs)presents several novel challenges.The stringent mission requirement to reach a target with a certain precision imposes a high guidance precision.The miniaturization of the size of MMs m...The design of mini-missiles(MMs)presents several novel challenges.The stringent mission requirement to reach a target with a certain precision imposes a high guidance precision.The miniaturization of the size of MMs makes the design of the guidance,navigation,and control(GNC)have a larger-thanbefore impact on the main-body design(shape,motor,and layout design)and its design objective,i.e.,flight performance.Pursuing a trade-off between flight performance and guidance precision,all the relevant interactions have to be accounted for in the design of the main body and the GNC system.Herein,a multi-objective and multidisciplinary design optimization(MDO)is proposed.Disciplines pertinent to motor,aerodynamics,layout,trajectory,flight dynamics,control,and guidance are included in the proposed MDO framework.The optimization problem seeks to maximize the range and minimize the guidance error.The problem is solved by using the nondominated sorting genetic algorithm II.An optimum design that balances a longer range with a smaller guidance error is obtained.Finally,lessons learned about the design of the MM and insights into the trade-off between flight performance and guidance precision are given by comparing the optimum design to a design provided by the traditional approach.展开更多
To better complete various missions, it is necessary to plan an optimal trajectory or provide the optimal control law for the multirole missile according to the actual situation, including launch conditions and target...To better complete various missions, it is necessary to plan an optimal trajectory or provide the optimal control law for the multirole missile according to the actual situation, including launch conditions and target location. Since trajectory optimization struggles to meet real-time requirements, the emergence of data-based generation methods has become a significant focus in contemporary research. However, due to the large differences in the characteristics of the optimal control laws caused by the diversity of tasks, it is difficult to achieve good prediction results by modeling all data with one single model.Therefore, the modeling idea of the mixture of experts(MoE) is adopted. Firstly, the K-means clustering algorithm is used to partition the sample data set, and the corresponding neural network classification model is established as the gate switch of MoE. Then, the expert models, i.e., the mappings from the generation conditions to the optimal control law represented by the results of principal component analysis(PCA), are represented by Kriging models. Finally, multiple rounds of accuracy evaluation, sample supplementation, and model updating are conducted to improve the generation accuracy. The Monte Carlo simulation shows that the accuracy of the proposed model reaches 96% and the generation efficiency meets the real-time requirement.展开更多
A new efficient coupling relationship description method has been developed to provide an automated and visualized way to multidisciplinary design optimization (MDO) modeling and solving. The disciplinary relation mat...A new efficient coupling relationship description method has been developed to provide an automated and visualized way to multidisciplinary design optimization (MDO) modeling and solving. The disciplinary relation matrix (DRM) is proposed to describe the coupling relationship according to disciplinary input/output variables, and the MDO definition has been reformulated to adopt the new interfaces. Based on these, a universal MDO solving procedure is proposed to establish an automated and efficient way for MDO modeling and solving. Through a simple and convenient initial configuration, MDO problems can be solved using any of available MDO architectures with no further effort. Several examples are used to verify the proposed MDO modeling and solving process. Result shows that the DRM method has the ability to simplify and automate the MDO procedure, and the related MDO framework can evaluate the MDO problem automatically and efficiently.展开更多
The distribution-free P-box process serves as an effective quantification model for timevarying uncertainties in dynamical systems when only imprecise probabilistic information is available.However,its application to ...The distribution-free P-box process serves as an effective quantification model for timevarying uncertainties in dynamical systems when only imprecise probabilistic information is available.However,its application to nonlinear systems remains limited due to excessive computation.This work develops an efficient method for propagating distribution-free P-box processes in nonlinear dynamics.First,using the Covariance Analysis Describing Equation Technique(CADET),the dynamic problems with P-box processes are transformed into interval Ordinary Differential Equations(ODEs).These equations provide the Mean-and-Covariance(MAC)bounds of the system responses in relation to the MAC bounds of P-box-process excitations.They also separate the previously coupled P-box analysis and nonlinear-dynamic simulations into two sequential steps,including the MAC bound analysis of excitations and the MAC bounds calculation of responses by solving the interval ODEs.Afterward,a Gaussian assumption of the CADET is extended to the P-box form,i.e.,the responses are approximate parametric Gaussian P-box processes.As a result,the probability bounds of the responses are approximated by using the solutions of the interval ODEs.Moreover,the Chebyshev method is introduced and modified to efficiently solve the interval ODEs.The proposed method is validated based on test cases,including a duffing oscillator,a vehicle ride,and an engineering black-box problem of launch vehicle trajectory.Compared to the reference solutions based on the Monte Carlo method,with relative errors of less than 3%,the proposed method requires less than 0.2% calculation time.The proposed method also possesses the ability to handle complex black-box problems.展开更多
The flight dynamics model of air-breathing hypersonic vehicles （AHVs） is highly nonlinear and multivariable cou- pling, and includes inertial uncertainties and external disturbances that require strong, robust, and ...The flight dynamics model of air-breathing hypersonic vehicles （AHVs） is highly nonlinear and multivariable cou- pling, and includes inertial uncertainties and external disturbances that require strong, robust, and high-accuracy controllers. In this paper, we propose a linear-quadratic regulator （LQR） design method based on stochastic robustness analysis for the longitudinal dynamics of AHVs. First, input/output feedback linearization is used to design LQRs. Second, subject to various system parameter uncertainties, system robustness is characterized by the probability of stability and desired performance. Then, the mapping rela- tionship between system robustness and LQR parameters is established. Particularly, to maximize system robustness, a novel hybrid particle swarm optimization algorithm is proposed to search for the optimal LQR parameters. During the search iteration, a Chernoff bound algorithm is applied to determine the finite sample size of Monte Carlo evaluation with the given prohabilily levels. Finally, simulation results show that the optimization algorithm can effectively find the optimal solution to the LQR parameters.展开更多
Morphing aircraft can meet requirements of multi-mission during the whole flight due to changing the aerodynamic shape,so it is necessary to study itsmorphing rules along the trajectory.However,trajectory planning con...Morphing aircraft can meet requirements of multi-mission during the whole flight due to changing the aerodynamic shape,so it is necessary to study itsmorphing rules along the trajectory.However,trajectory planning considering morphing variables requires a huge number of expensive CFD computations due to the morphing in view of aerodynamic performance.Under the given missions and trajectory,to alleviate computational cost and improve trajectory-planning efficiency formorphing aircraft,an offline optimizationmethod is proposed based onMulti-Fidelity Kriging(MFK)modeling.The angle of attack,Mach number,sweep angle and axial position of the morphing wing are defined as variables for generating training data for building the MFK models,in which many inviscid aerodynamic solutions are used as low-fidelity data,while the less high-fidelity data are obtained by solving viscous flow.Then the built MFK models of the lift,drag and pressure centre at the different angles of attack andMach numbers are used to predict the aerodynamic performance of the morphing aircraft,which keeps the optimal sweep angle and axial position of the wing during trajectory planning.Hence,themorphing rules can be correspondingly acquired along the trajectory,aswell as keep the aircraftwith the best aerodynamic performance during thewhole task.The trajectory planning of amorphing aircraft was performed with the optimal aerodynamic performance based on the MFK models,built by only using 240 low-fidelity data and 110 high-fidelity data.The results indicate that a complex trajectory can take advantage of morphing rules in keeping good aerodynamic performance,and the proposed method is more efficient than trajectory optimization by reducing 86%of the computing time.展开更多
Cooperative autonomous air combat of multiple unmanned aerial vehicles(UAVs)is one of the main combat modes in future air warfare,which becomes even more complicated with highly changeable situation and uncertain info...Cooperative autonomous air combat of multiple unmanned aerial vehicles(UAVs)is one of the main combat modes in future air warfare,which becomes even more complicated with highly changeable situation and uncertain information of the opponents.As such,this paper presents a cooperative decision-making method based on incomplete information dynamic game to generate maneuver strategies for multiple UAVs in air combat.Firstly,a cooperative situation assessment model is presented to measure the overall combat situation.Secondly,an incomplete information dynamic game model is proposed to model the dynamic process of air combat,and a dynamic Bayesian network is designed to infer the tactical intention of the opponent.Then a reinforcement learning framework based on multiagent deep deterministic policy gradient is established to obtain the perfect Bayes-Nash equilibrium solution of the air combat game model.Finally,a series of simulations are conducted to verify the effectiveness of the proposed method,and the simulation results show effective synergies and cooperative tactics.展开更多
This paper tackles the formation-containment control problem of fixed-wing unmanned aerial vehicle(UAV)swarm with model uncertainties for dynamic target tracking in three-dimensional space in the faulty case of UAVs’...This paper tackles the formation-containment control problem of fixed-wing unmanned aerial vehicle(UAV)swarm with model uncertainties for dynamic target tracking in three-dimensional space in the faulty case of UAVs’actuator and sensor.The fixed-wing UAV swarm under consideration is organized as a“multi-leader-multi-follower”structure,in which only several leaders can obtain the dynamic target information while others only receive the neighbors’information through the communication network.To simultaneously realize the formation,containment,and dynamic target tracking,a two-layer control framework is adopted to decouple the problem into two subproblems:reference trajectory generation and trajectory tracking.In the upper layer,a distributed finite-time estimator(DFTE)is proposed to generate each UAV’s reference trajectory in accordance with the control objective.Subsequently,a distributed composite robust fault-tolerant trajectory tracking controller is developed in the lower layer,where a novel adaptive extended super-twisting(AESTW)algorithm with a finite-time extended state observer(FTESO)is involved in solving the robust trajectory tracking control problem under model uncertainties,actuator,and sensor faults.The proposed controller simultaneously guarantees rapidness and enhances the system’s robustness with fewer chattering effects.Finally,corresponding simulations are carried out to demonstrate the effectiveness and competitiveness of the proposed two-layer fault-tolerant cooperative control scheme.展开更多
A synchronous control of relative attitude and position is required in separated ultraquiet spacecraft, such as drag-free, disturbance-free, and distributed spacecraft. Thus, a twistorbased synchronous sliding mode co...A synchronous control of relative attitude and position is required in separated ultraquiet spacecraft, such as drag-free, disturbance-free, and distributed spacecraft. Thus, a twistorbased synchronous sliding mode control is investigated in this paper to solve the control problem of relative attitude and position among separated spacecraft modules. The twistor-based control design and the stability proof are implemented using the Modified Rodrigues Parameter（MRP）.To evaluate the effectiveness of the proposed control method, this paper presents a case study of separated spacecraft flying control considering the mass uncertainty and external disturbances. In addition, a simulation study of the Proportional-Derivative（PD） control is also presented for comparison. The results indicate that the twistor-based sliding mode controller can ensure global asymptotic stability. The states converge fast with ultra-precision and ultra-stability in both the attitude and position. Moreover, the proposed twistor-based sliding mode control system is robust to the mass uncertainty and external disturbances.展开更多
In the event of a significant natural disaster or a local conflict,the demand for a regional satellite navigation becomes imperative.The navigation can provide accurate position information and navigation augmentation...In the event of a significant natural disaster or a local conflict,the demand for a regional satellite navigation becomes imperative.The navigation can provide accurate position information and navigation augmentation services for regional emergency operations.To satisfy the requirements,a formation configuration of low-Earth orbit(LEO)regional navigation satellites is proposed innovatively.The strategy offly-around formation consisting of four satellites for LEO regional navigation is determined and a comprehensive configuration design method for the formation is presented,encompassing the determination of configuration parameters,the establishment of dynamic equations and the presentation of performance indicators,including duration of ground coverage,positioning accuracy Geometric Dilution Precision(GDOP)and fuel consumption.The effects of formation radius and orbit altitude on the performance indicators are analyzed,respectively.Based on the above investigations,a method to enhance regional navigation performances by using genetic algorithm(GA)guided by penalty function is introduced.The rationality and feasibility of the formation configuration are verified through simulation studies.展开更多
With the growing efficiency of the use of unlicensed spectrum,the challenge of ensuring spectrum security has become increasingly daunting.Spectrum managers aim to accurately and efficiently detect and recognize anoma...With the growing efficiency of the use of unlicensed spectrum,the challenge of ensuring spectrum security has become increasingly daunting.Spectrum managers aim to accurately and efficiently detect and recognize anomaly behaviors in the spectrum.In this study,we propose a novel framework for spectrum anomaly detection and localization by spectrum interpolation recovery.Spectrum interpolation recovery refers to the recovery of the rest of the spectrum distribution based on a part of the spectrum distribution,which is achieved through a masked autoencoder(MAE)model with a core of multi-head self-attention(MHSA)mechanism.The spectrum interpolation recovery method restores the region where the masked abnormal signals are present,yielding anomaly-free results,with the difference between the restored and the masked representing the anomaly signals.The proposed method has been demonstrated to effectively reduce model-induced over-recovery of anomalous signals and dilute large-scale generation errors caused by anomalies,thereby improving the detection and localization performance of anomaly signals,and improving the area under the receiver operating characteristic curve(AUC)and the area under the precision-recall curve(AUPRC)by 0.0382(3.68%)and 0.1992(68.90%),respectively.On a designed dataset containing 3 variables of interference-to-signal ratio(ISR),signal-to-noise ratio(SNR),and anomaly type,the total recall of anomaly detection and localization at a 5%false alarm rate reached 0.8799 and 0.5536,respectively.Furthermore,a comparative study among different methods demonstrates the effectiveness and rationality of the proposed method.展开更多
文摘The design of mini-missiles(MMs)presents several novel challenges.The stringent mission requirement to reach a target with a certain precision imposes a high guidance precision.The miniaturization of the size of MMs makes the design of the guidance,navigation,and control(GNC)have a larger-thanbefore impact on the main-body design(shape,motor,and layout design)and its design objective,i.e.,flight performance.Pursuing a trade-off between flight performance and guidance precision,all the relevant interactions have to be accounted for in the design of the main body and the GNC system.Herein,a multi-objective and multidisciplinary design optimization(MDO)is proposed.Disciplines pertinent to motor,aerodynamics,layout,trajectory,flight dynamics,control,and guidance are included in the proposed MDO framework.The optimization problem seeks to maximize the range and minimize the guidance error.The problem is solved by using the nondominated sorting genetic algorithm II.An optimum design that balances a longer range with a smaller guidance error is obtained.Finally,lessons learned about the design of the MM and insights into the trade-off between flight performance and guidance precision are given by comparing the optimum design to a design provided by the traditional approach.
基金Defense Industrial Technology Development Program (JCKY2020204B016)National Natural Science Foundation of China (92471206)。
文摘To better complete various missions, it is necessary to plan an optimal trajectory or provide the optimal control law for the multirole missile according to the actual situation, including launch conditions and target location. Since trajectory optimization struggles to meet real-time requirements, the emergence of data-based generation methods has become a significant focus in contemporary research. However, due to the large differences in the characteristics of the optimal control laws caused by the diversity of tasks, it is difficult to achieve good prediction results by modeling all data with one single model.Therefore, the modeling idea of the mixture of experts(MoE) is adopted. Firstly, the K-means clustering algorithm is used to partition the sample data set, and the corresponding neural network classification model is established as the gate switch of MoE. Then, the expert models, i.e., the mappings from the generation conditions to the optimal control law represented by the results of principal component analysis(PCA), are represented by Kriging models. Finally, multiple rounds of accuracy evaluation, sample supplementation, and model updating are conducted to improve the generation accuracy. The Monte Carlo simulation shows that the accuracy of the proposed model reaches 96% and the generation efficiency meets the real-time requirement.
基金supported by the National Natural Science Foundation of China(51505385)Shanghai Aerospace Science and Technology Innovation Foundation(SAST2015010)the Defense Basic Research Program(JCKY2016204B102)
文摘A new efficient coupling relationship description method has been developed to provide an automated and visualized way to multidisciplinary design optimization (MDO) modeling and solving. The disciplinary relation matrix (DRM) is proposed to describe the coupling relationship according to disciplinary input/output variables, and the MDO definition has been reformulated to adopt the new interfaces. Based on these, a universal MDO solving procedure is proposed to establish an automated and efficient way for MDO modeling and solving. Through a simple and convenient initial configuration, MDO problems can be solved using any of available MDO architectures with no further effort. Several examples are used to verify the proposed MDO modeling and solving process. Result shows that the DRM method has the ability to simplify and automate the MDO procedure, and the related MDO framework can evaluate the MDO problem automatically and efficiently.
基金supported by the major advanced research project of Civil Aerospace from State Administration of Science,Technology and Industry of China.
文摘The distribution-free P-box process serves as an effective quantification model for timevarying uncertainties in dynamical systems when only imprecise probabilistic information is available.However,its application to nonlinear systems remains limited due to excessive computation.This work develops an efficient method for propagating distribution-free P-box processes in nonlinear dynamics.First,using the Covariance Analysis Describing Equation Technique(CADET),the dynamic problems with P-box processes are transformed into interval Ordinary Differential Equations(ODEs).These equations provide the Mean-and-Covariance(MAC)bounds of the system responses in relation to the MAC bounds of P-box-process excitations.They also separate the previously coupled P-box analysis and nonlinear-dynamic simulations into two sequential steps,including the MAC bound analysis of excitations and the MAC bounds calculation of responses by solving the interval ODEs.Afterward,a Gaussian assumption of the CADET is extended to the P-box form,i.e.,the responses are approximate parametric Gaussian P-box processes.As a result,the probability bounds of the responses are approximated by using the solutions of the interval ODEs.Moreover,the Chebyshev method is introduced and modified to efficiently solve the interval ODEs.The proposed method is validated based on test cases,including a duffing oscillator,a vehicle ride,and an engineering black-box problem of launch vehicle trajectory.Compared to the reference solutions based on the Monte Carlo method,with relative errors of less than 3%,the proposed method requires less than 0.2% calculation time.The proposed method also possesses the ability to handle complex black-box problems.
基金the National Natural Science Foundation of China （No. 11672235）
文摘The flight dynamics model of air-breathing hypersonic vehicles （AHVs） is highly nonlinear and multivariable cou- pling, and includes inertial uncertainties and external disturbances that require strong, robust, and high-accuracy controllers. In this paper, we propose a linear-quadratic regulator （LQR） design method based on stochastic robustness analysis for the longitudinal dynamics of AHVs. First, input/output feedback linearization is used to design LQRs. Second, subject to various system parameter uncertainties, system robustness is characterized by the probability of stability and desired performance. Then, the mapping rela- tionship between system robustness and LQR parameters is established. Particularly, to maximize system robustness, a novel hybrid particle swarm optimization algorithm is proposed to search for the optimal LQR parameters. During the search iteration, a Chernoff bound algorithm is applied to determine the finite sample size of Monte Carlo evaluation with the given prohabilily levels. Finally, simulation results show that the optimization algorithm can effectively find the optimal solution to the LQR parameters.
基金This study was co-supported by the National Defense Fundamental Research Funds of China(No.JCKY2016204B102 and JCKY2016208C001).
文摘Morphing aircraft can meet requirements of multi-mission during the whole flight due to changing the aerodynamic shape,so it is necessary to study itsmorphing rules along the trajectory.However,trajectory planning considering morphing variables requires a huge number of expensive CFD computations due to the morphing in view of aerodynamic performance.Under the given missions and trajectory,to alleviate computational cost and improve trajectory-planning efficiency formorphing aircraft,an offline optimizationmethod is proposed based onMulti-Fidelity Kriging(MFK)modeling.The angle of attack,Mach number,sweep angle and axial position of the morphing wing are defined as variables for generating training data for building the MFK models,in which many inviscid aerodynamic solutions are used as low-fidelity data,while the less high-fidelity data are obtained by solving viscous flow.Then the built MFK models of the lift,drag and pressure centre at the different angles of attack andMach numbers are used to predict the aerodynamic performance of the morphing aircraft,which keeps the optimal sweep angle and axial position of the wing during trajectory planning.Hence,themorphing rules can be correspondingly acquired along the trajectory,aswell as keep the aircraftwith the best aerodynamic performance during thewhole task.The trajectory planning of amorphing aircraft was performed with the optimal aerodynamic performance based on the MFK models,built by only using 240 low-fidelity data and 110 high-fidelity data.The results indicate that a complex trajectory can take advantage of morphing rules in keeping good aerodynamic performance,and the proposed method is more efficient than trajectory optimization by reducing 86%of the computing time.
基金supported by the National Natural Science Foundation of China(Grant No.61933010 and 61903301)Shaanxi Aerospace Flight Vehicle Design Key Laboratory。
文摘Cooperative autonomous air combat of multiple unmanned aerial vehicles(UAVs)is one of the main combat modes in future air warfare,which becomes even more complicated with highly changeable situation and uncertain information of the opponents.As such,this paper presents a cooperative decision-making method based on incomplete information dynamic game to generate maneuver strategies for multiple UAVs in air combat.Firstly,a cooperative situation assessment model is presented to measure the overall combat situation.Secondly,an incomplete information dynamic game model is proposed to model the dynamic process of air combat,and a dynamic Bayesian network is designed to infer the tactical intention of the opponent.Then a reinforcement learning framework based on multiagent deep deterministic policy gradient is established to obtain the perfect Bayes-Nash equilibrium solution of the air combat game model.Finally,a series of simulations are conducted to verify the effectiveness of the proposed method,and the simulation results show effective synergies and cooperative tactics.
基金the National Natural Science Foundation of China(61933010)the Natural Science Basic Research Plan in Shaanxi Province of China(2023-JC-QN-0733).
文摘This paper tackles the formation-containment control problem of fixed-wing unmanned aerial vehicle(UAV)swarm with model uncertainties for dynamic target tracking in three-dimensional space in the faulty case of UAVs’actuator and sensor.The fixed-wing UAV swarm under consideration is organized as a“multi-leader-multi-follower”structure,in which only several leaders can obtain the dynamic target information while others only receive the neighbors’information through the communication network.To simultaneously realize the formation,containment,and dynamic target tracking,a two-layer control framework is adopted to decouple the problem into two subproblems:reference trajectory generation and trajectory tracking.In the upper layer,a distributed finite-time estimator(DFTE)is proposed to generate each UAV’s reference trajectory in accordance with the control objective.Subsequently,a distributed composite robust fault-tolerant trajectory tracking controller is developed in the lower layer,where a novel adaptive extended super-twisting(AESTW)algorithm with a finite-time extended state observer(FTESO)is involved in solving the robust trajectory tracking control problem under model uncertainties,actuator,and sensor faults.The proposed controller simultaneously guarantees rapidness and enhances the system’s robustness with fewer chattering effects.Finally,corresponding simulations are carried out to demonstrate the effectiveness and competitiveness of the proposed two-layer fault-tolerant cooperative control scheme.
基金supported by the National Natural Science Foundation of China(Nos.51675430,11402044,and U1537213)
文摘A synchronous control of relative attitude and position is required in separated ultraquiet spacecraft, such as drag-free, disturbance-free, and distributed spacecraft. Thus, a twistorbased synchronous sliding mode control is investigated in this paper to solve the control problem of relative attitude and position among separated spacecraft modules. The twistor-based control design and the stability proof are implemented using the Modified Rodrigues Parameter（MRP）.To evaluate the effectiveness of the proposed control method, this paper presents a case study of separated spacecraft flying control considering the mass uncertainty and external disturbances. In addition, a simulation study of the Proportional-Derivative（PD） control is also presented for comparison. The results indicate that the twistor-based sliding mode controller can ensure global asymptotic stability. The states converge fast with ultra-precision and ultra-stability in both the attitude and position. Moreover, the proposed twistor-based sliding mode control system is robust to the mass uncertainty and external disturbances.
基金supported by the National Natural Science Foundation of China under Grant Nos.52305117 and 52075446.
文摘In the event of a significant natural disaster or a local conflict,the demand for a regional satellite navigation becomes imperative.The navigation can provide accurate position information and navigation augmentation services for regional emergency operations.To satisfy the requirements,a formation configuration of low-Earth orbit(LEO)regional navigation satellites is proposed innovatively.The strategy offly-around formation consisting of four satellites for LEO regional navigation is determined and a comprehensive configuration design method for the formation is presented,encompassing the determination of configuration parameters,the establishment of dynamic equations and the presentation of performance indicators,including duration of ground coverage,positioning accuracy Geometric Dilution Precision(GDOP)and fuel consumption.The effects of formation radius and orbit altitude on the performance indicators are analyzed,respectively.Based on the above investigations,a method to enhance regional navigation performances by using genetic algorithm(GA)guided by penalty function is introduced.The rationality and feasibility of the formation configuration are verified through simulation studies.
基金supported in part by the National Natural Science Foundation of China(grant numbers 52075446 and 51675430)CASC Application Innovation Program(grant number 6230111005).
文摘With the growing efficiency of the use of unlicensed spectrum,the challenge of ensuring spectrum security has become increasingly daunting.Spectrum managers aim to accurately and efficiently detect and recognize anomaly behaviors in the spectrum.In this study,we propose a novel framework for spectrum anomaly detection and localization by spectrum interpolation recovery.Spectrum interpolation recovery refers to the recovery of the rest of the spectrum distribution based on a part of the spectrum distribution,which is achieved through a masked autoencoder(MAE)model with a core of multi-head self-attention(MHSA)mechanism.The spectrum interpolation recovery method restores the region where the masked abnormal signals are present,yielding anomaly-free results,with the difference between the restored and the masked representing the anomaly signals.The proposed method has been demonstrated to effectively reduce model-induced over-recovery of anomalous signals and dilute large-scale generation errors caused by anomalies,thereby improving the detection and localization performance of anomaly signals,and improving the area under the receiver operating characteristic curve(AUC)and the area under the precision-recall curve(AUPRC)by 0.0382(3.68%)and 0.1992(68.90%),respectively.On a designed dataset containing 3 variables of interference-to-signal ratio(ISR),signal-to-noise ratio(SNR),and anomaly type,the total recall of anomaly detection and localization at a 5%false alarm rate reached 0.8799 and 0.5536,respectively.Furthermore,a comparative study among different methods demonstrates the effectiveness and rationality of the proposed method.
