Radar target threat level evaluation method and device

Abstract

The invention discloses a radar target threat level evaluation method and device, and relates to the technical field of radar air defense. The method comprises the steps of obtaining motion characteristic parameters of a target scanned by a search radar; according to the obtained motion characteristic parameters of the target, combining the battlefield environment, utilizing a threat evaluation model, setting different weights for the motion characteristic parameters, and calculating the threat level of the target to the position according to the principle that the shorter the defense time is, the larger the threat is; and after the threat levels of the targets are obtained, the threat targets with the same threat level are sorted according to the principle that the threat of the time when the targets arrive at the position in the air is lower. According to the invention, a decision maker can quickly determine a preferential coping target and reasonably allocate preferential resources. Excessive investment on low-threat targets is avoided, it is ensured that key targets get enough attention, and the overall defense or monitoring efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of radar air defense technology, and more specifically to a radar target threat level assessment method and device. Background Art

[0002] In today's complex and ever-changing environment, whether in the military or civilian fields, it is of vital importance to accurately assess and rank the threat level of radar detection targets.

[0003] In a military battlefield environment, search radar is responsible for detecting many targets in a specific airspace. It can output key information such as target track and target distance, azimuth, pitch, speed, etc. With the continuous development of military technology, the battlefield situation has become more and more complex, and various types of targets have emerged in an endless stream. When facing a large number of incoming targets, if the threat level of each target cannot be quickly and accurately determined and reasonably ranked, decision makers will find it difficult to make effective response decisions in a short period of time and will not be able to reasonably allocate limited defense resources. This may lead to excessive investment of defense forces in low-threat targets, while high-threat-level targets have not received timely and sufficient attention, thus posing a major threat to the life safety of battlefield personnel and the property safety of important military facilities and equipment, seriously affecting the overall defense efficiency and combat effectiveness, and may even change the direction of the war.

[0004] In the field of aviation control, with the increasing air traffic, there are a large number of aircraft in different flight states. Radar needs to monitor these targets. If the threat level of each aircraft target cannot be effectively assessed, for example, when an aircraft has an abnormal flight trajectory or approaches sensitive airspace, its potential threat cannot be quickly determined and prioritized, it will be difficult to ensure the safe and orderly operation of air traffic, increasing the risk of mid-air collisions and other accidents, posing a serious challenge to aviation safety.

[0005] Similarly, in the field of maritime transportation, there are a large number of ships and complex routes, and targets detected by radar also need to be assessed and ranked for threats. For example, near busy ports or on important sea lanes, when unidentified ships or ships moving abnormally appear, if their threat level to other ships and offshore facilities cannot be determined in time, effective traffic control measures cannot be taken, which can easily lead to safety accidents such as maritime collisions and damage to offshore facilities, greatly affecting the safety and stability of maritime transportation and operations.

[0006] In summary, the existing technology lacks an effective target threat level assessment and ranking method when facing numerous radar detection targets in complex environments, which makes it difficult to meet the urgent needs of military, aviation control, and maritime traffic to ensure safety and efficient operation. Therefore, a new radar target threat level assessment method is urgently needed to solve these problems. Summary of the invention

[0007] In order to solve the defects and shortcomings in the above-mentioned prior art, the present invention provides a radar target threat level assessment method and device. The invention aims to provide a radar target threat level assessment method for ranking the threat level of radar detection target outputs. It can be used in military battlefield environments, aviation control fields, and maritime traffic fields.

[0008] The radar target threat level assessment method of the present invention first obtains the target information sent back by the radar, the speed, distance, target navigation, altitude, azimuth and pitch angle of the main target, assesses the RCS parameter value of the target radar, and the target residence time; sets the threat assessment model parameters, sets the target characteristic weight value according to the battlefield use environment, and calculates the threat level of the target to the position. According to the threat level value, the same level threat targets are sorted according to the principle that the longer the target arrives at the position, the lower the threat. Threat level sorting allows decision makers to quickly determine the priority targets and reasonably allocate priority resources. Avoid excessive investment in low-threat targets. High-threat level targets pose a major threat to the safety of life and property of battlefield personnel, ensure that key targets receive sufficient attention, improve the overall defense or monitoring efficiency, and play a key role in ensuring safety and stability.

[0009] In order to solve the above problems existing in the prior art, the present invention is implemented through the following technical solutions.

[0010] A first aspect of the present invention provides a radar target threat level assessment method, the method comprising the following steps: S1. Obtain motion characteristic parameters of the target scanned by the search radar; the motion characteristic parameters include target speed, target distance, target heading, target altitude, target azimuth and pitch angle, target RCS parameter value and target dwell time; S2. According to the acquired motion characteristic parameters of the target, combined with the battlefield environment, and using the threat assessment model, different weights are set for the motion characteristic parameters, and the threat level of the target to the position is calculated based on the principle that the less defense time, the greater the threat; S3. After obtaining the threat level of the target, for threat targets with the same threat level, they are sorted according to the principle that the lower the threat, the longer the target is in the air when it arrives at the position.

[0011] Further preferably, in step S2, the threat assessment model adopts a neural network-based architecture, wherein the input layer receives the motion feature parameters of the target, the hidden layer performs feature extraction and nonlinear transformation on the input data through multiple neurons, and the output layer outputs the threat level value of the target.

[0012] Further preferably, in the initialization stage of the threat assessment model, an initial weight range is assigned to each motion feature parameter based on historical battlefield data and expert experience; the threat assessment model is accessed to real-time battlefield environment data; Collect a large amount of historical actual combat or simulation exercise data, including target motion characteristic parameters, corresponding battlefield environment information, and the final target threat level assessment structure as a training data set; The neural network model is trained using the back propagation algorithm, and the error between the threat level output by the model and the actual threat level assessment result is minimized by continuously adjusting the weights of each motion feature parameter. During the training process, an adaptive learning rate strategy is adopted to dynamically adjust the learning rate according to the progress of training to speed up the training convergence and avoid falling into the local optimal solution; The introduction of online learning mechanism enables the model to update weight parameters in real time as new target data and battlefield environment data continue to flow in, so as to adapt to the dynamic changes of the battlefield environment; The trained and optimized model, based on the input target motion feature parameters and real-time battlefield environment information, calculates and distributes weights in the hidden layer, and finally outputs the threat level value of the target to the position in the output layer.

[0013] More preferably, the threat assessment model is obtained based on MLP neural network model training.

[0014] More preferably, the threat assessment model is obtained based on RBFNN neural network model training.

[0015] More preferably, the threat assessment model is trained based on an LSTM neural network model or a GRU neural network model.

[0016] Further preferably, in step S2, the threat assessment model is constructed based on the analytic hierarchy process, and the specific construction steps are as follows: S201, constructing a hierarchical model, taking the threat level of the target to the position as the target layer, taking the motion characteristic parameters of the target as the criterion layer, and taking each target individual scanned by the search radar as the solution layer; S202, constructing a judgment matrix, inviting experts to compare each motion feature parameter in the criterion layer in pairs, constructing a judgment matrix, and performing consistency test on the judgment matrix; Calculate the maximum eigenvalue of the judgment matrix , through the formula Calculate the consistency index CI, where n is the order of the judgment matrix; determine the average random consistency index RI based on the order of the judgment matrix, and calculate the random consistency ratio ; If CR < 0.1, the judgment matrix passes the consistency test, otherwise the judgment matrix is ​​readjusted until the consistency requirement is met; S203, calculating the weight vector, calculating the eigenvector corresponding to the maximum eigenvalue of the judgment matrix by the eigenvector method, and obtaining the weight vector of each motion feature parameter after normalizing the eigenvector; S204, adjusting the weights in combination with the battlefield environment, and further optimizing and adjusting the weight vectors calculated in step S203 according to the battlefield environment information actually obtained; S205. Calculate the target threat level. For each scanned target, multiply the corresponding motion feature parameter value by the weight adjusted in step S204 and add the results to obtain the threat level score of the target. According to the preset threat level classification standard, map the threat level score to the corresponding threat level to determine the threat level of the target to the position.

[0017] Further preferably, in step S3, the calculation method of the time when the target reaches the position in the air is: the time when the target reaches the position in the air = the distance from the current position of the target to the shortcut point of the route / the target speed; the shortcut point of the route refers to the shortest distance between the position and the target's forward trajectory; When calculating the distance from the target's current position to the shortcut point on the route, you need to first calculate the angle between the target azimuth and the target heading. The angle between the target azimuth and the target heading is calculated as follows: Calculate the absolute value θ of the difference between the target azimuth and the target heading. If θ>180°, the angle between the target azimuth and the target heading is 360°-θ. At this time, the distance from the current position of the target to the shortcut point of the route is calculated as A=C*sin(270-θ). If θ≤180°, the angle between the target azimuth and the target heading is θ; at this time, the distance from the current position of the target to the shortcut point on the route is calculated as A=C*sin(θ-90); Where A represents the distance from the target's current position to the shortcut point on the route, and C represents the distance between the target's current position and the position.

[0018] Further preferably, in step S3, the targets are sorted according to the principle that the longer the target is in the air when it arrives at the position, the lower the threat, specifically, S301, according to the motion characteristic parameters of the target, first determine whether the motion characteristic parameters of the target are valid, if the motion characteristic parameters of the target are invalid, assign a value of Dj=0; if the motion characteristic parameters of the target are valid, proceed to the next step of determination; S302, judging whether the target is heading towards the position according to the target heading, if the target heading is away from the position, assigning the target a value of Dj=-1; if the target heading is heading towards the position, proceeding to the next step of judging; S303, judging whether the target's airborne duration is less than a preset first time threshold, if less than the first preset time threshold, assigning the target a value of Dj=1, if greater than the first preset threshold, proceeding to the next step of judgment; S304, determining whether the target's airborne duration is less than a preset second time threshold, if less than the second preset time threshold, assigning the target a value of Dj=2, if greater than the second preset threshold, assigning the target a value of Dj=3; S305. Sort targets at the same threat level according to the Dj value of the final target.

[0019] A second aspect of the present invention provides a radar target threat level assessment device, the device comprising: The first module is used to obtain the motion characteristic parameters and battlefield environment of the target scanned by the search radar; the motion characteristic parameters include target speed, target distance, target heading, target altitude, target azimuth and pitch angle, target RCS parameter value and target dwell time; The second module is used to set different weights for the motion characteristic parameters of the acquired target in combination with the battlefield environment and the threat assessment model, and calculate the threat level of the target to the position based on the principle that the less defense time, the greater the threat; The third module is used to sort the threat targets with the same threat level according to the principle that the lower the threat, the longer the target is in the air when it arrives at the position.

[0020] Further preferably, the device also includes a radar display and control interface for displaying targets sorted by threat level values ​​output by the third module.

[0021] The third aspect of the present invention provides a computer device, comprising a processor, an input device, an output device and a memory, wherein the processor, the input device, the output device and the memory are interconnected, wherein the memory is used to store a computer program, the computer program comprises program instructions, and the processor is configured to call the program instructions to execute some or all of the steps described in the first aspect of the present invention.

[0022] The fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program for electronic data exchange, wherein the computer program enables a computer to execute some or all of the steps described in the first aspect of the present invention.

[0023] Compared with the prior art, the beneficial technical effects brought by the present invention are as follows: 1. The radar target threat level assessment method of the present invention obtains multiple motion characteristic parameters of the target, such as speed, distance, heading, altitude, azimuth, pitch angle, RCS parameter value and dwell time, and uses advanced threat assessment models (such as models based on neural networks or hierarchical analysis methods) to set reasonable weights for these parameters in combination with the battlefield environment. It can accurately calculate the threat level of the target to the position, accurately distinguish different levels such as no threat (level -1), unknown threat (level 0), low threat (level 1), medium threat (level 2) and high threat (level 3), avoids the ambiguity and inaccuracy of the target threat level judgment in the traditional method, and provides clear and reliable target threat information for decision makers, so that they can quickly determine the priority targets and reasonably allocate limited defense resources, effectively improving the ability to cope with complex battlefield environments.

[0024] 2. The method of the present invention fully considers battlefield environmental factors, such as terrain height, electromagnetic interference intensity, meteorological conditions, and the distribution of one's own defense forces, and can dynamically adjust the weights of various motion characteristic parameters in threat assessment according to different environmental conditions. For example, when electromagnetic interference is strong, the weights of parameters related to radar signal strength are reduced, and the weights of relevant target parameters are specifically increased in areas with complex terrain or weak defense, thereby ensuring that the target threat level can be accurately and stably assessed in various complex and changeable battlefield environments, greatly improving the adaptability and practicality of the assessment method, reducing misjudgments and missed judgments caused by environmental factors, and enhancing the reliability and effectiveness of the defense system.

[0025] 3. After determining the threat level of the target, the present invention sorts the targets with the same threat level according to the principle that the longer the target arrives at the position, the lower the threat. When facing multiple threat targets of the same level, the urgency and processing order of the targets can be further clarified, and high-urgency targets that are about to arrive at the position are given priority. This helps to optimize the allocation of defense resources, avoid waste and unreasonable allocation of resources, improve the overall defense efficiency, ensure that key targets receive sufficient attention, and protect the safety of personnel and property. It has important application value in the fields of military defense, air traffic control, and maritime transportation, and can effectively maintain the security and stability of related areas.

[0026] 4. The present invention adopts a neural network architecture to construct a threat assessment model, which can effectively handle the complex nonlinear relationship between target motion characteristic parameters and threat level. The neural network extracts features and nonlinear transformations of input data through neurons in the hidden layer, and can accurately fit these complex relationships, thereby more accurately assessing the threat level. The present invention uses a large amount of collected historical actual combat or simulation exercise data for training, so that the model can automatically learn the degree of influence of each parameter on the threat level under different circumstances. By continuously adjusting the weights through the back-propagation algorithm, the model can continuously optimize itself and reduce the error between the actual threat level assessment results. For example, with the addition of new threat scenarios and target type data, the model can gradually learn and adapt, and continuously improve the accuracy of threat assessment.

[0027] 5. When the present invention adopts a neural network architecture to construct a threat assessment model, an adaptive learning rate strategy is introduced to further improve the training efficiency of the model. It can dynamically adjust the learning rate according to the progress of training, speed up the training convergence speed, avoid falling into the local optimal solution, and ensure that the model can quickly and stably reach a better performance state. An online learning mechanism is also introduced, so that the model can access new target data and battlefield environment data in real time and update weight parameters in time. In a battlefield environment, the situation changes rapidly. For example, when new weapons and equipment appear, sudden changes in electromagnetic interference, or changes in target behavior patterns, the model can respond quickly and dynamically adjust the threat assessment strategy to ensure the timeliness and accuracy of the assessment results. This enables the defense system to respond to new threats in a timely manner and allocate resources reasonably.

[0028] 6. The present invention uses a neural network architecture to construct a threat assessment model. The threat level output by the model is accompanied by a confidence index, which provides additional reference information for decision makers. When the confidence level is low, it prompts that further manual analysis or additional information is needed, which to a certain extent makes up for the uncertainty that may exist in the model and enhances the reliability of decision-making. For example, for some special cases or targets with incomplete data, a lower confidence level can remind operators to be cautious about the assessment results.

[0029] 7. The present invention adopts the hierarchical analysis method to construct a threat assessment model. By constructing a hierarchical model, the threat assessment problem of the target to the position is decomposed into the target layer, the criterion layer and the scheme layer, and the hierarchical relationship between the various factors is clearly sorted out. This systematic structure makes the threat assessment process more organized, in line with human thinking logic, and easy to understand and explain. For example, decision makers can intuitively see how each motion feature parameter affects the final threat level in a progressive manner.

[0030] 8. When the present invention adopts the hierarchical analysis method to construct the threat assessment model, in the process of constructing the judgment matrix, military experts and radar technology experts are invited to make pairwise comparisons, making full use of their professional knowledge and experience. These experts can accurately judge the relative importance of each motion feature parameter to the threat level based on the actual situation and past experience. For example, experts can reasonably determine the relative weights of target speed and target distance in threat assessment according to different battlefield environments and target types, so that the model is more in line with the actual situation. The judgment matrix is ​​checked for consistency by consistency index (CI) and random consistency ratio (CR), ensuring the rationality and reliability of expert judgment. This process avoids logical contradictions that may occur in the expert judgment process and ensures the scientific nature of weight allocation. Only when CR is less than 0.1, the judgment matrix is ​​considered to have satisfactory consistency, otherwise it needs to be readjusted. This strict inspection mechanism provides a strong guarantee for the accuracy of the model. After calculating the weight vectors of each motion feature parameter, the weights can be flexibly adjusted according to real-time battlefield environment information, such as terrain, electromagnetic interference, meteorological conditions, and the distribution of one's own defense forces. This flexibility enables the model to adapt to different battlefield environments and mission requirements. For example, in areas with complex terrain, the weights of parameters such as target altitude and target heading are adjusted based on terrain data to more accurately assess the potential threat of the target. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 This is a flowchart of threat level assessment of the present invention; Figure 2 Threat ranking flow chart for targets with the same threat level; Figure 3 This is a schematic diagram for calculating the time the target spends in the air when it reaches the position. DETAILED DESCRIPTION

[0032] The technical solution of the present invention is further described in detail below in conjunction with specific embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] Example 1 As a preferred embodiment of the present invention, refer to the attached specification Figure 1 As shown, this embodiment discloses a radar target threat level assessment method, which includes the following steps: S1. Obtain motion characteristic parameters of the target scanned by the search radar; the motion characteristic parameters include target speed, target distance, target heading, target altitude, target azimuth and pitch angle, target RCS parameter value and target dwell time; In this embodiment, the high-precision detection system of the search radar is used to accurately measure the basic motion parameters of the target, such as speed, distance, heading, altitude, azimuth and pitch angle. By optimizing the radar signal processing algorithm, the ability to analyze the target echo is improved, so as to obtain these parameter values ​​more accurately and reduce the measurement error.

[0034] To obtain the target RCS (radar cross-sectional area) parameter value, we can use advanced multi-band radar measurement technology, combined with the target's shape characteristics, material properties and the incident angle of the radar wave, and estimate and correct it through a complex electromagnetic calculation model to improve the accuracy and reliability of the RCS parameter value. At the same time, we can use the long-term tracking capability of the radar to count the target's residence time, ensure the measurement accuracy of the residence time, and avoid statistical errors caused by clutter interference or signal loss.

[0035] S2. According to the acquired motion characteristic parameters of the target, combined with the battlefield environment, and using the threat assessment model, different weights are set for the motion characteristic parameters, and the threat level of the target to the position is calculated based on the principle that the less defense time, the greater the threat; In this embodiment, an intelligent threat assessment model is constructed, which has adaptive learning capabilities and can automatically adjust the weight distribution of each motion feature parameter according to different battlefield environment characteristics, such as topography, distribution of key defense areas, and surrounding electromagnetic interference conditions. For example, in battlefield environments with complex terrain such as mountainous areas, the weight of the target height parameter may be appropriately increased, because low-flying targets may be more likely to evade radar detection and cause sudden attacks on ground targets; while in areas close to important military facilities, the weight of target distance and speed will be more critical.

[0036] When calculating the threat level, the concept of dynamic time window can be introduced to consider not only the current motion state of the target, but also its motion trend in the past period of time to make a more comprehensive assessment of the target's potential threat. For example, if a target suddenly accelerates and changes its course in a short period of time and flies towards the position, even if it is currently far away, its threat level can be increased in time by analyzing its motion trend through the dynamic time window, thereby more accurately reflecting the actual threat level of the target.

[0037] S3. After obtaining the threat level of the target, for threat targets with the same threat level, they are sorted according to the principle that the lower the threat, the longer the target is in the air when it arrives at the position.

[0038] Based on the current motion state and historical trajectory data of the target, combined with the meteorological conditions (such as wind speed and direction) and geographical information (such as mountains, rivers and other factors that may affect the target's flight path) in the battlefield environment, the time when the target arrives in the air is accurately predicted. For targets with the same threat level, they are sorted in ascending order according to the predicted time in the air, ensuring that when dealing with multiple targets of the same threat level, those high-urgency targets that are about to arrive at the position can be dealt with first.

[0039] As the radar continues to track the target, the target's motion characteristic parameters and environmental information are constantly updated, and the target's threat level and ranking results are dynamically adjusted. For example, when the target's speed or heading changes, its threat level and time in the air are recalculated in a timely manner, and the ranking is updated accordingly to ensure the timeliness and accuracy of threat assessment, and provide decision makers with the most timely and accurate target threat information, so as to reasonably allocate defense resources and effectively respond to various potential threats.

[0040] As an example of this embodiment, five threat levels are used for introduction. Among them, a target that is moving away from the target can be considered as a target with no attack intention, that is, no threat (set to -1 level); when the target is approaching the position at a relatively high speed, it can be considered as a target with attack intention, that is, the highest threat (set to 3 levels); when the information is incomplete or the parameters are invalid, the target lacks a basis for judgment, and an unknown threat is set (set to 0 level); the remaining targets are given a threat level according to the threat assessment model (set to 1 to 2 levels respectively).

[0041] After obtaining the threat level of the target, the target is sorted according to the threat level. Among them, the threat level of -1 and 0 are not sorted. For the threat level of 1 to 3, the same level of threat targets are sorted according to the principle that the longer the target arrives at the position, the lower the threat.

[0042] Example 2 As another preferred embodiment of the present invention, this embodiment further supplements and elaborates on the technical solution of the present invention on the basis of the above embodiment 1. This embodiment focuses on elaborating and supplementing the threat assessment model of step S2 in the above embodiment 1.

[0043] In this embodiment, in step S2, the threat assessment model adopts a neural network-based architecture, whose input layer receives the motion feature parameters of the target, the hidden layer performs feature extraction and nonlinear transformation on the input data through multiple neurons, and the output layer outputs the threat level value of the target.

[0044] The construction process of the threat assessment model is as follows: in the initialization stage of the threat assessment model, the initial weight range is assigned to each movement feature parameter based on historical battlefield data and expert experience; the threat assessment model is connected to real-time battlefield environment data; Collect a large amount of historical actual combat or simulation exercise data, including target motion characteristic parameters, corresponding battlefield environment information, and the final target threat level assessment structure as a training data set; The neural network model is trained using the back propagation algorithm, and the error between the threat level output by the model and the actual threat level assessment result is minimized by continuously adjusting the weights of each motion feature parameter. During the training process, an adaptive learning rate strategy is adopted to dynamically adjust the learning rate according to the progress of training to speed up the training convergence and avoid falling into the local optimal solution; The introduction of online learning mechanism enables the model to update weight parameters in real time as new target data and battlefield environment data continue to flow in, so as to adapt to the dynamic changes of the battlefield environment; The trained and optimized model, based on the input target motion feature parameters and real-time battlefield environment information, calculates and distributes weights in the hidden layer, and finally outputs the threat level value of the target to the position in the output layer.

[0045] As an example of this embodiment, the threat assessment model is obtained by training based on the MLP neural network model. MLP is a simple and commonly used feedforward neural network, which consists of an input layer, one or more hidden layers and an output layer. For the above-mentioned threat assessment model, the number of neurons in the input layer corresponds to the number of motion feature parameters. For example, if there are 8 motion feature parameters (target speed, target distance, target heading, target altitude, target azimuth and pitch angle, target RCS parameter value and target dwell time), the input layer has 8 neurons. The hidden layer can be set to 1-3 layers as needed, and the number of neurons in each layer can be determined by experiments and adjustments, for example, 10-50 neurons are used. The output layer has one neuron for outputting the threat level value. In this example, MLP can handle the nonlinear relationship between the input features well, and the input data is nonlinearly transformed through the neurons in the hidden layer, which is suitable for feature extraction and processing of data with complex relationships such as target motion feature parameters. Its structure is relatively simple, easy to understand and implement, and performs well in processing small and medium-sized data and less complex classification problems (such as threat level classification).

[0046] As another example of this embodiment, the threat assessment model is obtained by training based on the RBFNN neural network model. RBFNN is composed of an input layer, a hidden layer and an output layer. The input layer also receives target motion feature parameters. The neurons of the hidden layer use radial basis functions (such as Gaussian functions) as activation functions, and parameters such as the center and width need to be determined through training. The neurons of the output layer obtain the final result by linearly combining the outputs of the neurons of the hidden layer. The number of hidden layer neurons in this network structure is usually large, which is related to the distribution and complexity of the input data. RBFNN is also very effective in dealing with nonlinear problems, especially when the distribution of the input data is more complex or there are local features. In the threat assessment model, the relationship between the target motion feature parameters and the threat level may have certain local characteristics. For example, under certain specific speed-distance combinations, the threat level will have a specific change pattern. RBFNN can better capture this local feature, thereby improving the accuracy of threat level assessment.

[0047] As another example of this embodiment, the threat assessment model is trained based on an LSTM neural network model or a GRU neural network model. LSTM and GRU are variants of recurrent neural networks (RNNs) that have memory cells and can process sequence data. Structurally, they have input gates, forget gates, and output gates (LSTM) or update gates and reset gates (GRU) to control the flow and storage of information. If the time series of target motion feature parameters is considered in threat assessment, such as the speed change, distance change, etc. of the target over a period of time, these networks can process the parameters of each time step as input. When it is desired to consider the dynamic characteristics and historical information of the target motion, LSTM or GRU is a good choice. For example, if a target has been away from the position before, but suddenly changes its course and accelerates to approach, this dynamic change can be better captured by LSTM or GRU, and its threat level can be more accurately assessed based on historical information. However, this network structure is relatively complex, the computational cost is high, and more training data and computing resources are required.

[0048] Example 3 As another preferred embodiment of the present invention, this embodiment further supplements and elaborates on the technical solution of the present invention on the basis of the above embodiment 1. This embodiment focuses on elaborating and supplementing the threat assessment model of step S2 in the above embodiment 1.

[0049] In step S2, the threat assessment model is constructed based on the hierarchical analysis method, and the specific construction steps are as follows: S201, constructing a hierarchical model, taking the threat level of the target to the position as the target layer, taking the motion characteristic parameters of the target as the criterion layer, and taking each target individual scanned by the search radar as the solution layer; S202, constructing a judgment matrix, inviting experts to compare each motion feature parameter in the criterion layer in pairs, constructing a judgment matrix, and performing consistency test on the judgment matrix; Calculate the maximum eigenvalue of the judgment matrix , through the formula Calculate the consistency index CI, where n is the order of the judgment matrix; determine the average random consistency index RI based on the order of the judgment matrix, and calculate the random consistency ratio ; If CR < 0.1, the judgment matrix passes the consistency test, otherwise the judgment matrix is ​​readjusted until the consistency requirements are met.

[0050] S203, calculating the weight vector, calculating the eigenvector corresponding to the maximum eigenvalue of the judgment matrix by the eigenvector method, and obtaining the weight vector of each motion feature parameter after normalizing the eigenvector; S204, adjusting the weights in combination with the battlefield environment, and further optimizing and adjusting the weight vectors calculated in step S203 according to the battlefield environment information actually obtained; S205. Calculate the target threat level. For each scanned target, multiply the corresponding motion feature parameter value by the weight adjusted in step S204 and add the results to obtain the threat level score of the target. According to the preset threat level classification standard, map the threat level score to the corresponding threat level to determine the threat level of the target to the position.

[0051] This embodiment uses the analytic hierarchy process to construct a threat assessment model. By constructing a hierarchical model, the threat assessment problem of the target position is decomposed into the target layer, the criterion layer, and the solution layer, and the hierarchical relationship between the various factors is clearly sorted out. This systematic structure makes the threat assessment process more organized, in line with human thinking logic, and easy to understand and explain. For example, decision makers can intuitively see how each motion feature parameter affects the final threat level in a progressive manner.

[0052] In the process of constructing the judgment matrix, military experts and radar technology experts were invited to conduct pairwise comparisons, making full use of their professional knowledge and experience. These experts can accurately judge the relative importance of each motion feature parameter to the threat level based on actual conditions and past experience. For example, experts can reasonably determine the relative weights of target speed and target distance in threat assessment according to different battlefield environments and target types, making the model more in line with actual conditions. The judgment matrix is ​​checked for consistency through consistency index (CI) and random consistency ratio (CR), ensuring the rationality and reliability of expert judgment. This process avoids possible logical contradictions in the expert judgment process and ensures the scientific nature of weight allocation. Only when CR is less than 0.1, the judgment matrix is ​​considered to have satisfactory consistency, otherwise it needs to be readjusted. This strict inspection mechanism provides a strong guarantee for the accuracy of the model. After calculating the weight vector of each motion feature parameter, the weight can be flexibly adjusted according to real-time battlefield environment information, such as terrain, electromagnetic interference, meteorological conditions, and the distribution of one's own defense forces. This flexibility enables the model to adapt to different battlefield environments and mission requirements. For example, in areas with complex terrain, the weights of parameters such as target altitude and target heading are adjusted based on terrain data to more accurately assess the potential threat of the target.

[0053] Example 4 As another preferred embodiment of the present invention, this embodiment further supplements and elaborates on the technical solution of the present invention on the basis of the above-mentioned embodiment 1, embodiment 2 or embodiment 3. This embodiment focuses on the explanation and supplement of step S3 of the radar target threat level assessment method.

[0054] Refer to the instruction manual Figure 2 As shown in the figure, in step S3, the targets are sorted according to the principle that the longer the target is in the air, the lower the threat. Specifically, S301, according to the motion characteristic parameters of the target, first determine whether the motion characteristic parameters of the target are valid, if the motion characteristic parameters of the target are invalid, assign a value of Dj=0; if the motion characteristic parameters of the target are valid, proceed to the next step of determination; S302, judging whether the target is heading towards the position according to the target heading, if the target heading is away from the position, assigning the target a value of Dj=-1; if the target heading is heading towards the position, proceeding to the next step of judging; S303, judging whether the target's airborne duration is less than a preset first time threshold, if less than the first preset time threshold, assigning the target a value of Dj=1, if greater than the first preset threshold, proceeding to the next step of judgment; S304, determining whether the target's airborne duration is less than a preset second time threshold, if less than the second preset time threshold, assigning the target a value of Dj=2, if greater than the second preset threshold, assigning the target a value of Dj=3; S305. Sort targets at the same threat level according to the Dj value of the final target.

[0055] In this embodiment, refer to the attached manual. Figure 3 As shown, the calculation method of the target's airborne time when it arrives at the position is: the target's airborne time when it arrives at the position = the distance from the target's current position to the shortcut point on the route / the target speed; the shortcut point on the route refers to the shortest distance between the position and the target's forward trajectory; When calculating the distance from the target's current position to the shortcut point on the route, you need to first calculate the angle between the target azimuth and the target heading. The angle between the target azimuth and the target heading is calculated as follows: Calculate the absolute value θ of the difference between the target azimuth and the target heading. If θ>180°, the angle between the target azimuth and the target heading is 360°-θ. At this time, the distance from the current position of the target to the shortcut point of the route is calculated as A=C*sin(270-θ). If θ≤180°, the angle between the target azimuth and the target heading is θ; at this time, the distance from the current position of the target to the shortcut point on the route is calculated as A=C*sin(θ-90); Where A represents the distance from the target's current position to the shortcut point on the route, and C represents the distance between the target's current position and the position.

[0056] Example 5 As another preferred embodiment of the present invention, this embodiment discloses a device applicable to the radar target threat level assessment method described in any one of the above embodiments 1, 2, 3 and 4, the device comprising The first module is used to obtain the motion characteristic parameters and battlefield environment of the target scanned by the search radar; the motion characteristic parameters include target speed, target distance, target heading, target altitude, target azimuth and pitch angle, target RCS parameter value and target dwell time; The second module is used to set different weights for the motion characteristic parameters of the acquired target in combination with the battlefield environment and the threat assessment model, and calculate the threat level of the target to the position based on the principle that the less defense time, the greater the threat; The third module is used to sort the threat targets with the same threat level according to the principle that the lower the threat, the longer the target is in the air when it arrives at the position.

[0057] Example 6 As another preferred embodiment of the present invention, in order to achieve the above-mentioned purpose, according to another aspect of the present application, a computer device is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the above-mentioned radar target threat level assessment method when executing the computer program.

[0058] In this embodiment, the processor may be a central processing unit (CPU). The processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or a combination of the above chips.

[0059] The memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer executable programs and units, such as the corresponding program module units in the above method embodiments of the present invention. The processor executes various functional applications of the processor and works data processing by running the non-transitory software programs, instructions and modules stored in the memory, that is, implementing the method in the above method embodiments.

[0060] The memory may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created by the processor, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory may optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0061] The one or more units are stored in the memory, and when executed by the processor, perform the steps of the above-mentioned embodiment 1, embodiment 2, embodiment 3 or embodiment 4.

[0062] Example 7 As another preferred embodiment of the present invention, this embodiment discloses a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps in the above-mentioned embodiment 1, embodiment 2, embodiment 3 or embodiment 4 are implemented.

[0063] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A radar target threat level assessment method, characterized in that: The method comprises the following steps, S1. Obtain motion characteristic parameters of the target scanned by the search radar; the motion characteristic parameters include target speed, target distance, target heading, target altitude, target azimuth and pitch angle, target RCS parameter value and target dwell time; S2. According to the acquired motion characteristic parameters of the target, combined with the battlefield environment, and using the threat assessment model, different weights are set for the motion characteristic parameters, and the threat level of the target to the position is calculated based on the principle that the less defense time, the greater the threat; S3. After obtaining the threat level of the target, for threat targets with the same threat level, they are sorted according to the principle that the lower the threat, the longer the target is in the air when it arrives at the position.

2. A radar target threat level assessment method as claimed in claim 1, characterized in that: In step S2, the threat assessment model adopts a neural network-based architecture, in which the input layer receives the motion feature parameters of the target, the hidden layer extracts features and performs nonlinear transformation on the input data through multiple neurons, and the output layer outputs the threat level value of the target.

3. A radar target threat level assessment method as claimed in claim 2, characterized in that: In the initialization phase of the threat assessment model, initial weight ranges are assigned to each motion feature parameter based on historical battlefield data and expert experience; the threat assessment model is connected to real-time battlefield environment data; Collect a large amount of historical actual combat or simulation exercise data, including target motion characteristic parameters, corresponding battlefield environment information, and the final target threat level assessment structure as a training data set; The neural network model is trained using the back propagation algorithm, and the error between the threat level output by the model and the actual threat level assessment result is minimized by continuously adjusting the weights of each motion feature parameter. During the training process, an adaptive learning rate strategy is adopted to dynamically adjust the learning rate according to the progress of training to speed up the training convergence and avoid falling into the local optimal solution; The introduction of online learning mechanism enables the model to update weight parameters in real time as new target data and battlefield environment data continue to flow in, so as to adapt to the dynamic changes of the battlefield environment; The trained and optimized model, based on the input target motion feature parameters and real-time battlefield environment information, calculates and distributes weights in the hidden layer, and finally outputs the threat level value of the target to the position in the output layer.

4. A radar target threat level assessment method as claimed in claim 2 or 3, characterized in that: The threat assessment model is obtained based on the training of the MLP neural network model.

5. A radar target threat level assessment method as claimed in claim 2 or 3, characterized in that: The threat assessment model is obtained based on the training of the RBFNN neural network model.

6. A radar target threat level assessment method as claimed in claim 2 or 3, characterized in that: The threat assessment model is obtained by training based on an LSTM neural network model or a GRU neural network model.

7. A radar target threat level assessment method as claimed in claim 1, characterized in that: In step S2, the threat assessment model is constructed based on the hierarchical analysis method, and the specific construction steps are as follows: S201, constructing a hierarchical model, taking the threat level of the target to the position as the target layer, taking the motion characteristic parameters of the target as the criterion layer, and taking each target individual scanned by the search radar as the solution layer; S202, constructing a judgment matrix, inviting experts to compare each motion feature parameter in the criterion layer in pairs, constructing a judgment matrix, and performing consistency test on the judgment matrix; Calculate the maximum eigenvalue of the judgment matrix , through the formula Calculate the consistency index CI, where n is the order of the judgment matrix; determine the average random consistency index RI based on the order of the judgment matrix, and calculate the random consistency ratio ; If CR < 0.1, the judgment matrix passes the consistency test, otherwise the judgment matrix is ​​readjusted until the consistency requirement is met; S203, calculating the weight vector, calculating the eigenvector corresponding to the maximum eigenvalue of the judgment matrix by the eigenvector method, and obtaining the weight vector of each motion feature parameter after normalizing the eigenvector; S204, adjusting the weights in combination with the battlefield environment, and further optimizing and adjusting the weight vectors calculated in step S203 according to the battlefield environment information actually obtained; S205. Calculate the target threat level. For each scanned target, multiply the corresponding motion feature parameter value by the weight adjusted in step S204 and add the results to obtain the threat level score of the target. According to the preset threat level classification standard, map the threat level score to the corresponding threat level to determine the threat level of the target to the position.

8. A radar target threat level assessment method as claimed in any one of claims 1 to 3 or 7, characterized in that: In step S3, the targets are ranked according to the principle that the longer the target is in the air, the lower the threat. Specifically, S301, according to the motion characteristic parameters of the target, first determine whether the motion characteristic parameters of the target are valid, if the motion characteristic parameters of the target are invalid, assign a value of Dj=0; if the motion characteristic parameters of the target are valid, proceed to the next step of determination; S302, judging whether the target is heading towards the position according to the target heading, if the target heading is away from the position, assigning the target a value of Dj=-1; if the target heading is heading towards the position, proceeding to the next step of judging; S303, judging whether the target's airborne duration is less than a preset first time threshold, if less than the first preset time threshold, assigning the target a value of Dj=1, if greater than the first preset threshold, proceeding to the next step of judgment; S304, determining whether the target's airborne duration is less than a preset second time threshold, if less than the second preset time threshold, assigning the target a value of Dj=2, if greater than the second preset threshold, assigning the target a value of Dj=3; S305. Sort targets at the same threat level according to the Dj value of the final target.

9. A radar target threat level assessment method as claimed in claim 8, characterized in that: In step S3, the calculation method of the target's airborne time when it arrives at the position is: the target's airborne time when it arrives at the position = the distance from the target's current position to the shortcut point on the route / the target speed; The route shortcut point refers to the shortest distance between the position and the target's forward trajectory; When calculating the distance from the target's current position to the shortcut point on the route, you need to first calculate the angle between the target azimuth and the target heading. The angle between the target azimuth and the target heading is calculated as follows: Calculate the absolute value θ of the difference between the target azimuth and the target heading. If θ>180°, the angle between the target azimuth and the target heading is 360°-θ. At this time, the distance from the current position of the target to the shortcut point of the route is calculated as A=C*sin(270-θ). If θ≤180°, the angle between the target azimuth and the target heading is θ; at this time, the distance from the current position of the target to the shortcut point on the route is calculated as A=C*sin(θ-90); Where A represents the distance from the target's current position to the shortcut point on the route, and C represents the distance between the target's current position and the position.

10. A radar target threat level assessment device for implementing the radar target threat level assessment method according to any one of claims 1 to 9, the device comprising: The first module is used to obtain the motion characteristic parameters and battlefield environment of the target scanned by the search radar; the motion characteristic parameters include target speed, target distance, target heading, target altitude, target azimuth and pitch angle, target RCS parameter value and target dwell time; The second module is used to set different weights for the motion characteristic parameters of the acquired target in combination with the battlefield environment and the threat assessment model, and calculate the threat level of the target to the position based on the principle that the less defense time, the greater the threat; The third module is used to sort the threat targets with the same threat level according to the principle that the lower the threat, the longer the target is in the air when it arrives at the position.

11. A computer device, characterized in that: The method comprises a processor, an input device, an output device and a memory, wherein the processor, the input device, the output device and the memory are interconnected, wherein the memory is used to store a computer program, the computer program comprises program instructions, and the processor is configured to call the program instructions to execute the radar target threat level assessment method according to any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the processor executes the radar target threat level assessment method according to any one of claims 1 to 9.

Images