中国科技核心期刊
中国指挥与控制学会会刊
军事装备类重点期刊
中国指挥与控制学会会刊
军事装备类重点期刊
中国指挥与控制学会会刊
军事装备类重点期刊
This paper summarizes the origin, development and interrelation of system science systems, such as system engineering, system engineering and digital engineering. Based on the analysis of the development of underwater attack and defense systems at home and abroad, and in view of the main problems faced by underwater attack and defense systems in the development process, such as complex system operation mechanism and weak technical foundation, this paper expounds the necessity of applying the relevant theories of system science system to the construction and development of underwater attack and defense systems. From the perspective of system engineering, system engineering and digital engineering, the paper puts forward relevant development suggestions of underwater attack and defense system, which has certain military significance and theoretical value for the subsequent research of underwater attack and defense system.
On the basis of analyzing the development of underwater multi-platform cluster variable granularity situational awareness, this paper studies the system of systems and the synchronization of spatiotemporal evolution for multi granularity underwater situation awareness. Virtual physical artificial intelligence theory and method for multi granularity underwater situation awareness are proposed to establish the dynamic virtual physical neural network embedded with interpretation models of context and knowledge. Driven by spatiotemporal data and context information, multi granularity underwater situation awareness and synchronous situation evolution analysis across domains are realized to make collaborative intelligent decisions for underwater offensive and defensive confrontation tasks. Theoretical analysis and numerical results show that multi granularity underwater situation awareness and synchronous situation evolution analysis can enhance information superiority by integrating cluster resources, with optimal involved functional domains and space-time scales. Our approach can provide a theoretical basis and technical reference for the construction and development of underwater all-around attack and defense systems.
To get the best strategy of submarine underwater acoustic countermeasure against active/passive airborne sonar in normal conditions, this paper constructs the underwater acoustic environment model, sonar detection model and submarine underwater acoustic countermeasure model, proposes an airborne sonar information processing model considering acoustic contact, tracking, positioning and recognition processing, and simulates and analyzes the effectiveness of different submarine underwater acoustic countermeasure in the conditions of sonar contacting submarines. The results show that it is more effective to reduce the radiated noise than to increase the speed against the passive sonar. Submarines should adopt greater depth, minimum noise speed and shortest distance separation. If the submarine recognizes the position of the passive sonar buoy based on electronic reconnaissance or listening to the sound of sonobuoy impacting water, it can insert the bubble shell to worsen the common view condition of the passive buoy array, which avoids tracking and positioning and improves the effectiveness of countermeasures. It is the best strategy for the submarine to counter the active sonar, change the course, and release the wideband jammers and decoys at a high speed.The bubble shell can be released to worsen the sonar tracking, positioning and recognition environment while submarine fleeing, which providing conditions for the submarine to extend the distance from the sonar. The research in this paper can provide a simulation environment and decision-making reference for submarine underwater acoustic countermeasure active/passive airborne sonar.
Task allocation is one of the basic and key issues in the research of agent clusters. UUV clusters are limited by underwater detection and communication capabilities in task allocation. UUV individuals can only obtain partial information around them, and conventional global algorithms cannot be applied well. A task allocation algorithm based on deep reinforcement learning and distributed UUV cluster organizational structure is proposed. The algorithm first realizes the partial task allocation of each UUV individual, and then the information between adjacent individuals is consistent and coordinated, thus realizing the optimal task allocation of UUV cluster. Through the simulation experiment, the algorithm in this paper converges faster than the genetic algorithm, has less traffic and higher task allocation efficiency than the contract network algorithm, and the distributed architecture does not rely on the "command center". The UUV cluster system has higher robustness and higher task allocation reliability.
With the development of range to the integration of test and training, its function has changed from evaluating single weapon equipment to the whole combat system. In order to construct a realistic system of system-level test and training environment, the concept of digital twin is introduced into the range, and the overall framework of digital twin range to construct a complete digital image of combat system is proposed. The key technologies such as intelligent measurement and control, computational test, intelligent decision-making and combat cloud are analyzed, and application cases of underwater attack and defense digital twin range are briefly offered. The study is hoped to provide a reference for the realization of system of system-level test and training under the condition of large-scale and high-frequency real combat.
For the underwater dynamic target searching problem, an AUV dynamic target search algorithm based on improved genetic algorithm is proposed. The Monte Carlo method is used to generate a large number of target motion trajectories as the basis for calculating the fitness; a novel way to calculate the cumulative detection probability is proposed in combination with the hydroacoustic model ; the population selection adopts a combination of catastrophe idea and elitism method to ensure the non-inferiority and diversity of the population and accelerate the jump out of local extremes; chaotic sequence is used to select the crossover points and variation points to increase the population randomness; adopting dynamic adaptive crossover probability and variation probability reduce empirical dependence and ensure population diversity in the later stage. The simulation experimental results show that the improved genetic algorithm can effectively avoid falling into local extremes and improve the search probability compared with the traditional algorithm and classical genetic algorithm.
Aiming at the problem of mounting adaptability evaluation of shipboard gun weapon system, a mounting adaptability evaluation model based on analytic hierarchy process is put forward. The mounting adaptability of guns, tracking sensors and fire control equipment are chosen as fuzzy variables to evaluate the mounting adaptability of shipboard gun weapon system in the model. And the evaluation ruler of them is given. And the calculation method of mounting adaptability based on analytic hierarchy process is put forward. The example analysis shows the model can realize the quantitative evaluation for the adaptability of shipboard gun weapon system. The model has good operability and provides a powerful support for type selection decision-making.
Maintenance equipment consumption forecast usually adopts a single prediction model, which has a narrow scope of application and is not accurate in predicting the consumption of complete maintenance equipment. In this paper, the stability consumption characteristics of equipment and maintenance equipment are considered. The comprehensive prediction method of the classification of equipment consumption forecast is carried out. The combination prediction method combines the advantages of the exponential smoothing method, Croston method and TSB method to predict the materials with sTab.consumption and unsTab.consumption respectively. The results verify that the combination forecasting method which avoids the limitation of single forecasting model, has more extensive applicability and improves the accuracy of maintenance material consumption forecasting.
Aiming at the problem that the PID control algorithm can not meet the requirement of tracking and docking of moving flange due to its slow response speed when controlling the LNG unloading arm, which has strong nonlinearity and time-varying system, the fuzzy control theory is introduced, and the fuzzy control algorithm based on experience and the fuzzy PID control algorithm which is combined with PID control are proposed and implemented respectively, which does not depend on the dynamic model of the LNG unloading arm. The two intelligent control algorithms are respectively applied to the LNG unloading arm of the test site. Taking a single joint as the test object, the performance test and the tracking test of the simulated moving target flange are carried out. The test show that the algorithm designed by introducing the fuzzy control theory has significantly improved the comprehensive response speed compared with the PID control algorithm, and the fuzzy PID control algorithm can fully meet the docking requirements of moving target flange in practical application because of its strong resistance to time delay.
The pole assignment method is commonly used in the design of three-loop autopilot, and the constraints for the key parameter that the open-loop crossover frequency was blurred. By concluding the analytical relationship between control gains, the close-loop poles and the open-loop crossover frequency, the control system design problem is transformed into a judicious choice of the open-loop crossover frequency. Combined with engineering experience of the Canard layout missile, the constraint criteria and modified methods for the selection of the open-loop crossover frequency of the rudder bandwidth, the fin rate, the close-loop zeros and the gain polarity are discussed. The applicability of the criteria is verified in typical conditions.
The capability assessment of the joint reconnaissance and early warning system provides the fundamental information for the joint combat commanders to execute combat planning and command and control. According to the urgent needs of the joint operations commander to analyze and evaluate the capability of the joint reconnaissance and early warning system, based on the basic scenarios and real-time deduction data of the joint combat wargame, this paper evaluates the joint reconnaissance and early warning system by combining the static capability boundary evaluation and the real-time combat effectiveness evaluation. According to the evaluation requirements, fifteen specific evaluation indicators from two categories of basic combat capability evaluation and real-time combat effectiveness evaluation are designed and constructed. And the connotation, extension and calculation method of each evaluation index are elaborated. Finally, based on the basic scenario data of the joint combat wargame system and the output data of the game model, this paper constructs the capability evaluation software tool for the joint reconnaissance and early warning system, which provides new ideas and new methods for the capability evaluation of the joint reconnaissance and early warning system to solve the key information requirements of the commander in the wargame.
In view of the case that the fleet of ships threatens the safety of the sea area, the scenarios of red side penetration attack, blue side air defense and antimissile, as well as three kinds of anti aircraft carrier combat strategies of land, sea and air are designed. Based on the Mozi system, the multi scheme combat simulation is carried out, and the battle damage data of both sides are obtained. On this basis, an index system based on the performance parameters of decoy and attack is established, and the operational effectiveness of the three strategies is analyzed with the analytic hierarchy process (AHP). Furthermore, a Deep Reinforcement Learning (DRL) algorithm based on AHP weight is proposed, which optimizes the sea based strategy and improves the combat effectiveness by 5.36%. The research results show that the method of scenario design, combat strategy simulation, and AHP-DQN for operational efficiency optimization based on such combat simulation software as Mozi system can provide reference for anti-ship warfare.
Aiming at the problem of operational effectiveness evaluation of anti-tank minefield, a method of operational effectiveness evaluation of anti-tank minefield based on Harris Hawks Optimization (HHO) optimized Monte Carlo neural network (MCNN) is proposed. Firstly, the basic principles and algorithm flow of IHHO algorithm and Monte Carlo neural network are introduced. Then the main influencing factors of the effectiveness evaluation of anti-tank minefield are analyzed, and the index system of the effectiveness evaluation of anti-tank minefield is summarized. Finally, the IHHO-MCNN neural network evaluation model is constructed. The simulation results show that the model can effectively evaluate the effectiveness of anti-tank minefield.
The assessment of airfield runway destruction effectiveness and functional recovery time is of great importance in the field of operational assessment. In this paper, the criteria and methods of damage assessment of airfield runways are studied in the context of the sub-bomb against airfield runways and the benefits of striking airfield runways as a starting point. Firstly, the main purpose of attacking airport runway is blockade, and the damage assessment criteria of runway should focus on its time benefit. Then, it is concluded that calculating the minimum lifting window can only achieve the purpose of blockade of airport, and the optimal lifting window is needed to calculate the airport function recovery time. A direction search algorithm is proposed to search the optimal lifting window on the airport runway after striking, and an example is given.
In recent years, the breakthrough of machine learning based on deep reinforcement learning provides a new development direction for intelligent game confrontation. In order to solve the problems of slow convergence speed and great difference in training effect of heterogeneous multi-agent reinforcement learning algorithm in intelligent confrontation, this paper proposes a priori knowledge-driven multi-agent reinforcement learning game antagonism algorithm PK-MADDPG, and constructs a MADDPG model under the framework of double Critic. The model uses the experience first replay technique to optimize the prior knowledge extraction, thus achieving remarkable results in the training of game confrontation. In the national competition of MaCA heterogeneous multi-agent game confrontation, the paper compares the game confrontation results of PK-MADDPG algorithm with classical rule algorithm, and verifies the effectiveness of the algorithm proposed in this paper.
In order to accurately link the entity references in the commander’s demand statement to the standardized entity nodes in the knowledge graph, an entity linking method in the military domain based on improved edit distance is proposed. By summarizing the non-standard forms of entity referencing to establish an index, and using the BM25 model to integrate the improved edit distance algorithm that considers the character position exchange, inclusion similarity and other similarities, the entities to be linked are sorted to achieve the link. The experimental results show that the entity linking method in the military field can effectively improve the matching accuracy in similarity calculation.
According to the lack of sea detection research and the high complexity of genetic algorithm (GA) based on actual terrain data in radar network deployment optimization, this paper carries out the research on seaside radar deployment optimization with the digital elevation model (DEM) data and genetic algorithm. Based on the given positions and seaside radar detection model, the proposed algorithm can greatly reduce the searching space of GA and improve the solving speed. Simulation results show that, our algorithm can obtain almost the same optimal results compared with the exhausted searching method, while enjoying obvious advantage in speeding with the increase of deployment scale.
Radar networking is an effective method to improve detection accuracy and fault tolerance in complex electromagnetism environment. It is necessary to study data fusion schemes which can address the challenges from interference and signal-to-noise ratio reduction. In this paper, a data fusion method for multiple-radar point fusion based on bayesian statistical theory is proposed. The multi-source data fusion method based on bayesian theory is combined with kalman filtering, with the prediction of kalman filter and its covariance as the prior knowledge for bayesian theory. The points of multiple-radar are regarded as the observation value of bayesian theory. A real-time estimation method for the standard deviations of radar points is also proposed based on signal-to-noise ratio. The simulation results show that the filtering accuracy of the proposed data fusion method is better than that of the individual radar track and track fusion, and it can adapt to changing standard deviations caused by target distance changing and RCS (Radar Cross-Section) fluctuating. The proposed method is of great value to area air defense.
Aiming at the problem of bearing-only target steering maneuver detection, a multi-platform bearing-only target maneuver detection algorithm based on heading estimation is proposed. The algorithm selects the sequence of hypothetical maneuvering points, solves the two target motion elements before and after the hypothetical maneuvering point, and determines whether the target maneuvers according to the change of the sequence of the heading difference of the adjacent segments solved. Based on the Taylor series element solution model, a joint solution model of two motion elements and an independent solution model of two motion elements were established. The false alarm rate, target maneuver detection rate, maneuver detection delay time and maneuver time estimation accuracy of the maneuver detection algorithm under these two solution models are statistically analyzed. Simulation results show that the maneuver detection algorithms under the two solution models can effectively detect maneuvering targets.
Facing the requirements of dynamic planning for aviation communication under the multi-domain cooperative system-level war in the future, according to the objectives of virtual-real mapping, real-time synchronization, symbiotic evolution and closed-loop optimization between the simulation system and the actual system, the paper proposes some application cases that parallel simulation works on improving the combat effectiveness of aviation communication equipment, in-field incremental integration and verification, intelligent health management, virtual and real combined flight training etc. Establish a parallel simulation system architecture for aviation communication and put forward the system composition, then analyze the key technologies such as real-time data acquisition, multi-branch parallel simulation deduction and artificial intelligence-based situation prediction and intelligent decision-making. Lay the theoretical foundation for the application of parallel simulation technology in aviation communication field.
To improve the positioning accuracy of the followers in leader-follower cooperative navigation algorithm for multi-USVs based on distance measurement, this algorithm integrates the distance measurements between the followers into the traditional leader-follow structure, reconstructs the equations of state and measurements in this system, calculating positions through extended Kalman filter. Simulation results shows that the algorithm added distance measurements between the followers compared with the traditional algorithm can restrain the errors diffuse effectively and improve positioning accuracy of the followers.
In view of the sinking of the Cruiser Moscow during towing, the large probability that the Cruiser Moscow was hit by a Ukrainian missile is simulated based on scenario deduction. Subsequently, a mathematical model is constructed to quantitatively assess the impact of the marine environment on the sinking of the Cruiser Moscow. The simulation results show that the marine environment has a great influence on the detection efficiency of air defense radar of the Cruiser Moscow and the capsizing of the ship during towing. Finally, by analyzing the level of equipment technology and the use of tactics demonstrated by Russia and Ukraine in this naval warfare, it provides reference for the development of our naval equipment and provides a scientific basis for the improvement of our naval warfare tactics.
In view of the requirements of air defense equipment system construction and equipment development, take the air attack/anti air attack operation in the Russia-Ukraine military conflict in February 2022 as an example to analyze the key factors that affect the countermeasure effect, such as the attacker’s electronic jamming effect, the cruise missile penetration height, and the defender’s air defense and anti-missile weapon performance. Three exploratory experimental schemes are designed by using exploratory analysis combined with simulation. The influence of the above factors on the hit quantity of tactical ballistic missiles, cruise missiles and other munitions of the attacking side is quantitatively analyzed through simulation, and the development suggestions of the defense’s air defense and anti-missile equipment system, such as networked operations, integrated air defense and anti-missile are put forward, providing reference for the development, system construction and application of air defense and missile defense equipment.
