Abstract:A network intrusion risk identification method based on deep learning is proposed.Firstly,it preprocesses the col-lected network data,extracts statisticals and power spectral features of the data,and then integrate is the extracted features.The en-tropy value of the integrated features is calculated to determine abnormal data in the network data.A loss function is set,regulariza-tion terms to construct the loss function is introduced to reduce model complexity,the extracted features are screened,and the corre-sponding risk factors are calculated.Combined with the probability of risk factor occurrence,it calculates the intrusion risk value of network data and divide the network intrusion risk level.The experimental results show that the designed identification method has an average misidentification rate of 7.2％and high identification accuracy in practical applications.Meanwhile,the computational ef-ficiency of the designed identification method consistently remains above 90％,demonstrating higher efficiency and faster identifica-tion of network intrusion risks.

Abstract:The workload of a certain ship vehicle project is large,the technical difficulty is high,and there are many uncer-tain factors.The project schedule is difficult to accurately estimate,and it is prone to schedule adjustments,making project sched-ule management difficult.On the basis of traditional PERT method,variance adjustment factor is introduced to improve the variance calculation of PERT method.The improved PERT method is used to determine the project schedule,establish a simulation model of the project,and Monte Carlo method is used to simulate and analyze the project schedule and risks.The simulation analysis results show that Monte Carlo method can simulate the project schedule based on the improved PERT method,predict and evaluate the proj-ect schedule,and verify the operability and effectiveness of the improved PERT method.Through sensitivity risk analysis,the risk items in the project schedule can be analyzed and determined,providing more accurate decision-making information for project de-cision-makers to carry out project management work more targetedly.

Abstract:The current transformation of China's military construction is shifting from informatization to intelligence,where in-formation plays an irreplaceable and crucial role in winning future wars.This paper investigates knowledge verification techniques in the automatic construction of knowledge repositories.During the process of building knowledge repositories,conflicting triples can be identified.Furthermore,in research related to knowledge extraction,it is observed that both encyclopedic and open-source data do not guarantee correctness and novelty in all cases,sometimes resulting in conflicting data.Therefore,it is necessary to adopt rele-vant methods to validate and integrate knowledge,ensuring the accuracy of knowledge in the repository by distinguishing between truth and falsehood,and identifying errors.This paper primarily focuses on verifying triples in the knowledge repository,utilizing the method of Markov Logic Networks and entity correlation strength to validate the credibility of knowledge,ultimately leading to improvements in F1 score and precision.

Abstract:The whole angle mode hemispherical resonant gyroscope has extremely high dynamic characteristics,but the distur-bance error of the elliptical parameters of the harmonic oscillator is caused by the uneven circumferential damping of the harmonic oscillator and the mismatch of circuit channel gain caused by the processing technology.In order to solve this problem,a distur-bance error compensation method is proposed.Firstly,the ellipse parameter model of harmonic oscillator with uneven circumferen-tial damping and the influence of demodulation program on ellipse parameters are derived when the channel gain is mismatched.Sec-ondly,a time-sharing circuit and compensation scheme which can suppress the disturbance error of ellipse parameters of harmonic oscillator are proposed.Finally,the method is used to compensate the disturbance error of ellipse parameters of harmonic oscillator,and the effectiveness of the method is verified by experiments.The experimental results show that the long-axis energy fluctuation of the gyro is reduced from 5％to 0.875％,which is 82.5％lower than the same period last year.The output nonlinearity is reduced from 907.44 ppm to 168.04 ppm,which is 81.47％lower than the same period last year.And the scale factor nonlinearity is reduced from 998 ppm to 215.56 ppm,which is 78.4％lower than the same period last year.

Abstract:Instrument landing system is the key navigation equipment for precision approach and landing in civil aviation air-ports.With the rapid development of China's civil aviation industry and the promotion of national science and technology innovation policies,the localization of ILS has become an important strategic goal to achieve a strong country in science and technology and civ-il aviation.Firstly,this paper analyzes the important position of instrument landing system in aviation field and the necessity of local-ization from the course of localization of instrument landing system.Then,the characteristics of instrument landing system at home and abroad are analyzed in detail,and the key factors restricting the localization of instrument landing system are revealed.On the basis of this research,the challenges faced in the process of localization of instrument landing system are further discussed,includ-ing the dependence on imports of core components,insufficient innovation ability and imperfect industrial chain.Finally,the coun-termeasures and suggestions are systematically put forward in view of the above challenges.

Abstract:In recent years,instant messaging software has developed rapidly and become the main information tool in people's daily social life.However,in the work environment,email is still considered a more formal form of communication and is mainly used to send formal reports and notifications.However,spam emails containing advertising,violence,pornography and even Trojan viruses take up our mailbox resources and affect its reliability,causing inconvenience to our work.Therefore,this paper investigates a temporal convolutional neural network(TCN)model based on improved random forest feature selection to accurately identify the content and information of spam emails.Based on the TCN classification model,the model incorporates the Random Forest(RF)with the sparrow optimisation algorithm(SSA)for dimensionality reduction of the original features,which is used to remove poorly correlated and unimportant features and retain effective features with a certain contribution rate to obtain the optimal subset of fea-tures.In order to verify the reliability of the model,comparative experiments are carried out on the Spambase dataset at UCI to anal-yse and evaluate the model based on five indices,which are F1 score,precision,recall,accuracy and number of training parame-ters.The experiments show that the model can effectively identify spam emails,and the optimal feature subset improves the recogni-tion accuracy while reducing the complexity of the model,and the evaluation metrics are all better than the experimental comparison model.

Abstract:In recent years,ship fires have occurred frequently,and coastal ports have an important strategic position both eco-nomically and militarily.With the continuous development of economy,the number of cargoes docked at coastal ports has in-creased,and the sources of danger have also increased,and the fire safety of ships and ports has attracted increasing attention.Therefore,this paper summarizes and sorts out the research on fire control layout planning at home and abroad.From the aspects of fire risk assessment,fire station layout planning and fire emergency plan,this paper summarizes and analyzes the current research status of fire control layout planning in China's ship ports,and finds that there are some limitations in fire control layout planning,such as low fire control standards,few studies on the location of fire stations aiming at port linear requirements,and poor ship-shore coordination.

Abstract:To solve the problem of"relying on experience and weak dynamic adaptability"in the evaluation of joint firepower strike effectiveness in military chess deduction scenarios,this study relies on the advantages of deep learning in processing complex data and modeling.Firstly,combined with the actual background of"multi firepower platform collaboration and dynamic changes in strike situation"in joint firepower strikes,the characteristics and correlation logic of the evaluation data are analyzed.Then this study designs a basic model that is suitable for practical use,using the strike plan as input and rule calculation evaluation results as labels to construct a sample set and complete the training of the intelligent evaluation model.There are two verification methods,one is to compare the consistency between the output results of the intelligent model and the commander's experiential cognition.The sec-ond is to use the RMSE index to analyze the performance deviation between the real results of the test set,the output values of the in-telligent model,and the expert adjusted evaluation results.The results indicate that the deviations of the three types of results are within an acceptable range,which confirms the feasibility and practical value of the design ideas in the evaluation of joint firepower strike effectiveness,and can provide technical support for the optimization of joint firepower strike schemes and intelligent deci-sion-making.

Abstract:Display consoles are mostly used for long-term information reading and command control,which easily causes fa-tigue.The ergonomics level of display console has been paid more attention.In order to compare different standards for the display console screen angle,and find the most ergonomic angle,this research studied changes in the back muscle fatigue of operators at different angles of display console screens.Firstly it compares two commonly-used angles of a display console screen,which are-15°and+15°,and uses 0°for a comparison experiment.Then it collects the value of electromyographic signals of the fatigue of back muscles,the integral electromyography,root median square and mean power frequency are chosen for analysis.It is finally found that fatigue accumulation is most obvious with a display console screen angle of-15°.There is little difference in fatigue be-tween the remaining two angles.Part of the data also shows a trend of voluntary relief of fatigue.The results of this study provide a reference for the information layout design of display consoles and the ergonomic standard setting of display console screens.

Abstract:Aiming at the problems of the current risk identification method of equipment procurement contract execution,the traditional semi-structured interview method is improved from the aspects of simplifying the method process.Aiming at the problems of single scale and strong subjectivity in risk evaluation of equipment procurement contract performance,the meaning of scale is set from two aspects of severity and possibility,and the product method is proposed to calculate the comprehensive weight of indicators under multiple scales.Combining risk value and index weight to calculate risk degree value,a risk assessment method based on risk matrix-AHP is constructed,and the effectiveness of the method is verified by case analysis.

Abstract:With the proliferation of security vulnerabilities in Internet-of-Things(IoT)systems,conventional intrusion-de-tection approaches are increasingly inadequate for real-time and accurate detection of botnet attacks.Existing methods,when de-ployed on resource-constrained devices,struggle to capture the temporal context of periodic and bursty attacks,resulting in models that are both computationally intensive and prone to high false-positive rates.To address these limitations,this paper proposes a compact yet effective intrusion-detection framework specifically tailored for botnet attacks in IoT environments.First,this paper de-rives a concise representation of the N-BaIoT dataset through traffic-pattern analysis,time-series modeling,and centralizes feature extraction.Second,it designs a lightweight detection architecture that synergistically combines a pruned DenseNet with an en-hanced LSTNet,enabling rapid identification of botnet behaviors.Experimental results on the N-BaIoT dataset demonstrate a detec-tion accuracy of 95％,while maintaining high precision and recall across diverse operational scenarios.These findings substantiate the proposed method's efficacy in real-world deployments and its robustness to previously unseen attack patterns.

Abstract:Collaborative navigation algorithm of one single datum node based on network adjustment is proposed,aiming at the issue that satellite navigation of manned/unmanned vehicle cluster is susceptible to be interfered or suppressed.Based on the po-sition of available GNSS navigation,the incremental position of the inertial navigation/Doppler log integration,and the mutual dis-tance measurement between different nodes,a network adjustment algorithm is used to estimate the position of each node in the clus-ter,in order to provide high-precision position for the inertial navigation correction.It is completed that absolute position is estimat-ed for manned/unmanned vehicle clusters when only one node obtains satellite navigation information.Finally,numerical simulation experiments are conducted to verify the effectiveness of the algorithm proposed in this paper.

Abstract:Nanosilver has a high melting point and good electrical properties after sintering,which is an ideal material for the encapsulation of high-temperature resistant devices.Through process exploration,hot press bonding of nanosilver as sealing rings is successfully realized to form compliant hermetically sealed interconnections.Sealing rings with a width of 400 μm are prepared by a dispensing machine.The optimization of nanosilver hot press bonding process parameters is explored,combined with Ar(5％H2)plasma pretreatment method,and the control variable method is used to investigate the effects of bonding pressure,bonding time,and bonding temperature on the bonding quality of nanosilver,and it is obtained that the shear force and airtightness are positively correlated with the bonding pressure,temperature,and time,so as to obtain the bonding parameters suitable for the high-tempera-ture pressure sensors of graphene.The subsequent sensor processing is carried out under the optimal conditions,and the successful-ly bonded samples are tested and characterized,and the bonding interface is relatively flat,with high densities and good airtight-ness.By comparing the electrical characteristics of the chip before and after bonding,the resistance value of graphene is basically unchanged,and this bonding method has high stability.A new method is provided for the low-cost,fast efficiency and high-strength hermetic packaging of graphene MEMS pressure sensors.

Abstract:Based on the Seebeck effect and Fourier's law,two different structures of thermopile-type high-temperature heat flux sensors are designed,and the sensors are physically prepared on 10×10 mm2 square and 10 mm diameter circular Al2O3 ceramic substrates,respectively.Two kinds of thermopile structures,serpentine and annular,are prepared by screen-printing process,and high-pressure spraying is used to make step-distributed and centralized thermal resistance layers correspondingly,and the two kinds of heat flux sensors are tested and compared.The results show that the peak voltage of the circular heat flux sensor is 0.223 2 mV,which is higher than that of the square heat flux sensor of 0.164 1 mV,and the circular heat flux sensor has a better perfor-mance.In a heat flux environment of 48.8 kW/m2～193.2 kW/m2,the output voltage of the circular heat flux sensor reaches 0.225 6 mV,with a sensitivity of 1.34 μV/(kW/m2),a repeatability error of less than 2.34％,and a stable output performance,which pro-vides a solution for the difficult problem of surface heat flux monitoring of miniature,high-temperature components.

Abstract:This paper introduces a test method based on the digital-analog hybrid ATE time measurement unit(TMU),and it is found that the conventional test method will lead to the existence of measurement errors in the TMU,and in order to solve this problem,a digital-analog hybrid ATE programmable oscilloscope method is proposed.To high-speed photocoupler fall time as the object of study,test results show that the optimization of the test value before the existence of a large error,the optimization of the fall time test value and high-speed photocoupler detailed specification of the typical value is basically consistent with the test value of a number of measurements is stable,and has been applied to a number of high-speed integrated circuits in the screening test.

Abstract:The ship industrial control network is composed of sensor nodes,actuators,buses,etc.It has the characteristics of large node layout,difficult to avoid security vulnerabilities,and complex operating status,which poses a threat to the stable opera-tion of industrial processes.Considering the essential information of operational data,a monitoring method for industrial control net-work security operation status based on typical variable probability principal component analysis is proposed.Firstly,this method us-es one-way ANOVA to select the main variables that affect operational performance.Then,typical variable analysis is used to recon-struct the original space into a state matrix and a residual matrix,and a concurrent monitoring model for industrial control networks based on probabilistic principal component analysis is established on this basis.

Abstract:In order to assess close air support,the paper provides the method of analytic hierarchy process and fuzzy comprehensive evaluation.In this method,the analytic hierarchy process is used to construct evaluation index system.The first level proposes and analyzes 4 primary factors of close air support.Moreover,the second level includes 10 evaluating indicators.The index weight is assessed by AHP,and then the final value of close air support is gotten by the method of fuzzy comprehensive evaluation.

Abstract:In response to the problems of limited communication bandwidth,frequent communication interruptions,and large transmission delays in distributed cross domain combat at sea,which can easily lead to difficulties in data exchange and sharing in the collaborative perception of manned and unmanned mixed group situations,the design concept of"collaborative perception and distributed accounting"is adopted.A"dual chain layered and smart contract driven"cloud edge integrated situation collaborative perception architecture is proposed,which can be compatible with various cloud edge participants such as manned and unmanned platforms,achieve timely secure and reliable exchange of battlefield situation information,and provide support for the generation and guarantee of distributed collaborative combat situations.

Abstract:This paper presents a spatial spectrum estimation algorithm of Root-MUSIC in the presence of gain-phase errors.The root-finding method is used to get DOA without the spectral peak searching,which not only reduces the spectral peak search time and complexity of the algorithm,but also has better spatial spectrum estimation accuracy than the conventional algorithm in the case of low signal-to-noise ratio.An alternating iterative algorithm is adopted to adaptively calibrate the gain-phase errors of the channels,making the amplitude error less than 0.5 dB and the phase error less than 1 degree,which ensures the accuracy of the spa-tial spectrum estimation.The algorithm can be applied to the direction-finding equipment on unmanned aerial vehicles.Theoretical analysis and computer simulation show that this algorithm has higher spectrum estimation accuracy compared with conventional algo-rithms such as Capon and MEM.

Abstract:A data-driven ideal landing trajectory construction and evaluation method is proposed,taking the flight training da-ta recorded by the avionics equipment of a small general aviation aircraft as the research object.After extracting the landing phase data,abnormal landing data is filtered out.The ideal landing curve is constructed through the quadratic optimal approximation mod-el.And finally,the evaluation index and model of"area-angle dual-dimensional deviation"are proposed.The results show that 82.4％of the landing trajectory deviations are less than 33.3％,which verifies the rationality of the ideal curve construction and the effectiveness of the evaluation method.It provides a data-driven quantitative basis for flight training quality evaluation,which has practical and reference significance for improving the standardization of carrier-based aircraft operations.