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
The modern warfare has turned into a confrontation between the systems and the evaluation of operational system is an important method to strengthen the construction of the systems. This paper discusses the connotation and the challenges of operational system evaluation, analyzes the application and limitation of the traditional methods. Then new methods such as complex network, war-gaming and deep learning are studied and analyzed. Finally, thoughts and suggestions on the research are proposed.
As an efficient and intelligent means of knowledge organization, knowledge graphs can help users quickly and accurately obtain the information they care about, and it has developed rapidly in recent years. Knowledge graphs, deep learning, big data and other technologies have become the core force to promote the development of artificial intelligence. Starting from the basic concepts of knowledge graphs, the article systematically analyzes the logical structure, technical architecture and construction methods of knowledge graphs. Afterwards, it focuses on the full life cycle technology of knowledge graphs, from knowledge modeling, acquisition, fusion, processing, storage, reasoning and the typical applications of knowledge graphs and other aspects explain the domestic and foreign research progress of key technologies used in the construction of knowledge graphs. Then, it introduces the use of knowledge graphs in intelligent semantic search, knowledge question and answer systems, and vertical industries such as public security, medical treatment, and industrial production. Finally, the future development trend of the knowledge map and the many challenges that exist at present are discussed.
In order to solve such problems in the field of weapon equipment as scattered data sources, lack of good association between data, inconsistent representation, and a large number of redundancy, this paper proposes a kind of ontology-based method to construct knowledge graph to realize effective management and intelligent information searching. Using knowledge graph, we can excavate the potential value of weapon equipment data further. Firstly, the domain ontology was constructed based on the knowledge of weapon equipment domain. Then, the entity, relationship and attributes were extracted from a large number of collected original data. The redundant data was fused. The result data was represented by triple. Finally, the knowledge graph was constructed with Neo4j Database. At the same time, the constructed knowledge graph can direct the extraction or fusion of the original data. So, the all above formed a virtuous circle of knowledge graph optimization.
With the rapid growth of consumer drone market, the low altitude security threat caused by "black flight" of low-slow-small target represented by quadrotor UAV is more and more serious. Different from the existing defense system, in this paper, a new low-slow-small target defense system with physical damage interception as the main counter measure is presented. The details of system components and its key principle is introduced, and some simulation results are also given. The construction of the system has a certain reference significance for guiding the design of the new anti-UAV defense system.
Aiming at the analysis and application requirements of equipment data, the paper tries to apply Apriori algorithm to make association rules analysis of equipment data. It first introduces the principles and implementation process of Apriori algorithm, then establishes an association rule analysis model of equipment data based on the process of data mining, and carries out case simulation based on real cases to obtain the valuable correlation between influencing factors and efficiency in equipment application, and finally proves that the proposed method is effective and feasible.
In order to meet the military needs of aerospace commanders, customize the situation products they care about, and improve the timeliness of decision-making, this article focuses on the extraction of situation elements in the space mission situation integration link. Firstly, the principle of situation element extraction and the relationship between the space mission and situation elements are studied and analyzed; Secondly, there is a set of situation elements of common concern according to different space missions, and in the process of element extraction introduce the concept of modules, it builds situation modules around different functional themes according to the connections between elements, and determine the hierarchical structure between situation modules within the mission; Finally, a concept lattice-based method is proposed to extract the situation elements of aerospace missions, combined with aerospace an emergency launch example verifies the effectiveness of this method.
To understand the impact of marine environment on laser guided missile, based on the semi-active guidance system and its working principle of laser guided missile, this paper analyzes the target characteristics in marine environment, sea surface anti/scattering in background environment, sea situation, wave, wind speed in climate environment and the factors affecting visibility, and obtains the influence of the above factors on laser transmission and reception, based on the semi-active guidance system and its working principle of laser guided missile, target search and recognition. The range and hit accuracy of the missile are affected to a certain extent. In view of the above effects and the differences between the sea and land environment, the sea suitability test requirements are proposed for several laser seeker performance indexes, including the operating range, blind range, anti sun angle, anti target loss ability, etc., which have a greater impact, so as to provide a theoretical basis for correctly guiding the sea operation of laser guided missile.
In order to solve the problem of improving the accuracy of projectile miss-distance prediction, a recursive prediction model of projectile miss-distance based on innovation adaptive Kalman Filter algorithm is established. Firstly, it uses the characteristics of each error source to establish an error source model, combines its characteristics to superimpose a reasonable miss distance sequence, and establishes innovation adaptive Kalman Filter state equation. Then, the innovation sequence is used to calculate the system noise matrix and the measured noise covariance matrix to construct innovation adaptive Kalman Filter prediction model. Finally, the improved prediction results are compared and analyzed with the traditional Kalman Filter prediction. Matlab simulation results show that the performance and estimation accuracy of Kalman Filter depend on the system model and noise statistical characteristics, and the improved Innovation Adaptive Kalman Filter model is more accurate in predicting results. At the same time, the algorithm has universal applicability in random miss distance sequence.
In order to efficiently evaluate the target value of each military target in the complex combat network, and to support the combat form of rapid decision and rapid attack under the condition of information war, a military target value evaluation method based on the complex network theory is proposed. In this method, the complex network theory is used to process the network topology of combat target system. On the basis of fully considering the importance of node to edge and the characteristics of military target, the support value of node to edge is introduced. The idea of grey relation is used to quantify the target characteristic value, combined with node degree value, node betweenness and edge betweenness, the algorithm model is constructed by entropy method to evaluate the value of each military target in the combat network, and an example is analyzed. The results show that the military target value evaluation model based on complex network theory and entropy method can be applied to the general military target system, and can provide data support and certain reference for decision-makers to make operational plans.
Under joint operations, rockets pose a great threat to land combat forces, and air defense force are in urgent need of anti-rocket capabilities. This paper analyzes the ballistic characteristics, speed characteristics, radar scattering characteristics and radar tracking stability of rockets, and puts forward the main tactical capability requirements such as field air defense reconnaissance radar pattern of detection and tracking rocket, resolution, measurement accuracy, intelligence data rate and tracking target capacity etc.,to provide technical support for intercepting and combating rockets.
Currently, traditional methods such as empirical estimation and mathematical formula calculations are widely used to predict the consumption of ammunition in the army. The feasibility of the prediction results are difficult to guarantee, resulting in mismatch between operations and support and inaccurate planning. In order to perform the accurate prediction of the ammunition consumption of combined forces, this paper constructs a combined forecasting model of ammunition consumption based on an adaptive deep neural network, selecting and quantifying data related to ammunition combat consumption in activities such as domestic and foreign typical battle cases, exercise training, and simulation experiments. And it uses multiple evaluation indicators to train and test the model. By comparing and analyzing the accuracy of the results with neural networks such as BP and Elman, it proves the reliability and accuracy of the model used to predict the combat consumption of various types of ammunition.
The types of armored equipment in-service evaluation data are complex and large, traditional evaluation methods are difficult to effectively mine the hidden information of the data. By analyzing the content and principles of armored equipment in-service evaluation, analyzing the advantages of deep learning, researching the deep belief network model, and proposing an armored equipment in-service evaluation model based on deep belief network. Taking the service economy of armored equipment as an example, sample data is used to train and verify the model. The example shows that the model can realize the objective evaluation of armored equipment in-service.
To test and evaluate the equipment support capability of the synthesis brigade, an evaluation index system of the equipment support capability of the synthetic brigade was constructed based on the method of “mission-capability” mapping through literature review and troop investigation, and the weight of the evaluation index was determined through the analytic hierarchy process, and the evaluation results were calculated by fuzzy comprehensive evaluation method. Finally, it provides reference for the evaluation of equipment support capability of synthetic brigade.
Simulation training is an important function of warship combat system. Simulation training integration design of warship combat system is proposed. Simulation training software is deployed in the TSCE of warship combat system. The function demand of warship combat system simulation training is analyzed. The architecture, deployment, interface and flow of simulation training software are designed. The architecture of simulation training software model is designed. The application example is given. The feasibility of design plan is verified. The integration design plan of warship combat system simulation training is applicable to different kinds of warship combat system simulation training development, and it has great maneuverability.
The cycle difference and data transmission delay between simulation nodes will cause the mismatch of collision detection and collision response, lead to the penetration phenomenon between entities in the virtual scene, and bring unreal virtual experience. Therefore, the consistency of collision detection and response in distributed simulation environment is studied. A state updating method of simulation object based on motion lock is proposed, which separates entity dynamic simulation and collision detection. The simulation object is divided into two parts: simulation entity and detection entity. The detection entity receives the dynamic simulation results. When collision is detected, the state of the simulation entity is locked and corrected by loop feedback, the state of the simulation entity is unlocked and the state of the simulation entity coincides with the state of the detection entity until the detection entity in the visual simulation unit is no longer collided. It can effectively avoid the crossing problem caused by asynchronous detection and response.
To solve the problem that the recoil force of a small caliber high rate of fire gun does not match the unmanned vehicle, the truncated barrel is adopted and the corresponding muzzle brake structure is designed. The optimized muzzle brake shows that the muzzle brake with an angle of 15 degrees between baffle and y axis has the best performance. In order to study the influence of overpressure and noise generated by the muzzle brake on the shooter, numerical simulation of overpressure and noise in the area around the muzzle brake was carried out by using computational fluid dynamics software Fluent. The simulation results show that the overpressure and noise intensity of the receiving point decrease with the distance between the receiving point and muzzle increasing, and the shooter should wear the corresponding protective equipment within the area of 1m around the muzzle brake. The simulation results are helpful to study the characteristics of muzzle shock wave and provide a reference for the safety protection of marksman near muzzle.
In order to alleviate the sub-optimal problem of the scheme, caused by the relative independence of task assignment and route planning, in the process of static task planning of multi-UUV, a hybrid optimization algorithm is proposed by the means of combining non dominated sorting genetic algorithm Ⅱ (NSGA-Ⅱ) with dynamic programming algorithm. NSGA-Ⅱ is used as the overall optimization framework to assign tasks to each UUV. Then, the dynamic programming algorithm is applied to each UUV task subset to obtain the task sequence based on the shortest path criterion, which is used evaluate and optimize the task allocation scheme. The simulation results based on typical scenarios show that by embedding dynamic programming algorithm into NSGA-Ⅱ optimization framework and explicitly considering the influence of task execution order on the performance of the scheme in the process of task assignment, the quality of the optimization scheme can be improved and the convergence of NSGA-Ⅱ optimization process can be accelerated.
Over the modern army’s battlefield, army helicopters are the most complex airspace coordination object due to the diversity of its aerial activities and the multiple cross-links with multiple firepower in airspace and ground. This paper draws on the airspace control measures of the United States Army, integrates multiple design dimensions that including the operational requirements of the army aviation, air combat functions, weapon performance and others to draw up an airspace coordination method. The introduction is mainly divided into two aspects: general airspace coordination method and dedicated airspace coordination method, including the division of airspace space, the digitization of airspace, the advancement of airspace planning, and the conflict and resolution of airspace. It draws up an airspace coordination method that meets the operational mission requirements and conforms to the flight characteristics of helicopters, ensures that army helicopters can safely and flexibly use the limited airspace to avoid conflicts and accidental injuries.
Aiming at the problem of multi-dimensional parameter estimation of a single vector hydrophone, a joint estimation method of frequency-azimuth-elevation based on sparse decomposition theory is proposed. According to the time-space transformation technology (TST), the channel data received by the single-vector hydrophone is stacked according to the time-domain sampling points, and then converted into spatial snapshot data, and the time-space receiving data model of the single-vector hydrophone is established. With reference to the data model of the single vector hydrophone, an over-complete atomic library of the same form is constructed, so as to realize the joint estimation of the signal's frequency, azimuth and pitch angle based on the sparse decomposition theory. Numerical simulation results show that the proposed method has good estimation performance.
In order to improve the attack ability of terminal guided projectile on maritime over the horizon maneuvering target, it is necessary to quickly identify the turning maneuvering behavior of the target. Based on the image information obtained by UAV continuous observation of ships, the ship pixel width in the observation image is converted into the observation width of the ship vertically projected on the sea surface by the projection conversion method, and the rolling angle of the ship is calculated by the observation width. On this basis, a method to quickly judge the turning maneuver state of ships based on the change of ship roll is proposed. This method can be used to predict the turning rate, assist in selecting the turning maneuver model of fire control targets, and predict the future turning maneuver point of maritime maneuvering targets. However, this paper mainly studies the method of quickly judging target turning maneuver, and does not involve the research on the initial value prediction and filtering of target turning rate.
An integrated 2-3D display architecture for the tactical mission planning system of the air force is constructed from the perspective of system design according to the usage requirements, and the implementation of 2-3D integrated display technology is analysed in terms of data drawing principles and map linkage. Based on the functional characteristics of the system, three integrated display algorithms, namely the rectangular area jump method, the scale jump method and the perspective projection jump method, are proposed, together with their characteristics and applicability, to realise the simultaneous transformation of the 2D flat map and the 3D earth view during the interaction process in a message-driven form.
Event extraction method can help users quickly and accurately identify the interesting events from the massive, disordered and unstructured information, which is widely used in the field of natural language processing. On the basis of sorting out the concept, the knowledge representation of event and the development of event extraction, the extraction methods of meta-event and topic event are summarized and analyzed respectively. The research status of event extraction methods in military field is summarized, and the application trend of event extraction methods in military field is discussed.
The multifunction phased array radar has multiple subarrays, and the multi-task parallel scheduling can be realized by dynamic aperture reconstruction. In this paper, the multi-task parallel scheduling technology for multifunction phased array radar is studied, and a two-dimensional aperture dynamic reconstruction method based on heuristic algorithm is proposed. On this basis, the joint management method of time resource and two-dimensional aperture resource is studied and a hybrid multi-task parallel EDF (Earliest Deadline First) algorithm and a hybrid genetic algorithm are proposed respectively. The two algorithms are verified and are compared from task execution time offset penalty value and task scheduling success rate by simulation.
In order to realize automatic equipment matching, use equipment resources as efficiently as possible, and give full play to the overall operational effectiveness, the problem of equipment combination operation is put forward and analyzed. This paper discusses the principle of deep reinforcement learning, and establishes the concept, model and framework of equipment combination application method based on deep reinforcement learning. The feasibility of equipment combination application method based on deep reinforcement learning is verified by the experiment of aircraft anti-ship penetration, which shows that this method can effectively solve the problem of “combination explosion” in equipment application field and provide theoretical and technical support for equipment command and decision-making departments to formulate equipment application plans.
