A Priority Quantification Method for Radar Tasks
By receiving and processing a variety of information about radar missions, and combining with the LightGBM model to quantify the radar target's own priority, it solves the problem of incomplete evaluation of radar mission priorities in the existing technology, and realizes dynamic priority calculation and synthesis of radar missions Priority evaluation improves the performance of the perception system.
Patent Information
- Application Number
- CN202210846100.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-07-05
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2042-07-05
AI Technical Summary
The prior art has not yet effectively handled the priority quantization evaluation elements of radar missions, and is unable to adaptively dig the priority quantization information of the radar targets themselves, resulting in a large number of uncontrollable satellites and debris in orbit that threaten the safety of in-orbit satellites.
Provide a radar task priority quantization method, which performs data preprocessing and quantization processing by receiving target task historical information, spatial target real-time information, general task observation requirements and emergency needs, and combines the LightGBM model to quantify the target's own real-time priority, and Conduct comprehensive assessments to obtain overall comprehensive priorities.
Dynamic priority calculation for radar targets/tasks is realized, combined with observation requirements in actual environments, the rationality of comprehensive priority is ensured and the overall performance of the perception system is improved.
Smart Images

Figure CN115146863B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of computer processing technology, and more particularly, to a priority quantification method for radar tasks. Background Art
[0002] Countries around the world have launched over 9,000 satellites. Currently, over 5,000 intact satellites are tracked, and approximately 2,000 are in orbit. This means a large number of uncontrollable satellites are in Earth's orbit, becoming space debris after losing their functionality. The exact number of debris in orbit is difficult to estimate. These fragments orbit the Earth in specific orbits at speeds of up to 6-16 km / s, posing a threat to the safety of orbiting satellites.
[0003] Prioritization of radar missions requires consideration of numerous factors. Existing literature lacks a comprehensive methodology for comprehensively addressing these various factors. Furthermore, prior literature typically uses predefined priority tables or scaling methods to quantify radar target priorities, failing to adaptively mine radar target priority information from data. Summary of the Invention
[0004] The purpose of the present invention is to overcome the deficiencies in the prior art and provide a method for quantifying the priority of radar tasks.
[0005] In a first aspect, a radar task priority quantization method is provided, which is performed by a quantization system and includes:
[0006] S1. Receive four pieces of information, including target mission history information, space target real-time information, general mission observation requirements, and emergency requirements;
[0007] S2. performing data preprocessing on the four parts of information;
[0008] S3. quantifying the real-time priority of the target task itself by combining the historical information of the target task and the real-time information of the space target;
[0009] S4. Combining the target mission history information, the space target real-time information, and the general mission observation requirements, performing quantitative processing to obtain a general requirement priority;
[0010] S5. Combining the target mission history information, the space target real-time information, and the emergency demand, performing quantitative processing to obtain the emergency demand priority;
[0011] S6. Comprehensively evaluate the real-time priority, general demand priority, and emergency demand priority of the target task itself to obtain the overall comprehensive priority.
[0012] Preferably, S1 includes:
[0013] S101, receiving target mission history information, wherein the target mission history information includes: mission information previously processed by a quantization system and previous physical information of a space target;
[0014] S102: Receive real-time space target information, where the real-time space target information is provided in real time by radar equipment and includes physical information detected in real time by the radar equipment;
[0015] S103, receiving general task observation requirements, wherein the general task observation requirements mainly include: a general observation expectation period, wherein the general observation expectation period refers to a target observation expectation period according to established planning information;
[0016] S104: Receive emergency demands, including emergency demand instructions from superior equipment, experts, and relevant personnel for specific emergency target situations.
[0017] Preferably, in S2, the data preprocessing includes: processing missing values, noise data and outliers in the data.
[0018] Preferably, S3 includes:
[0019] S301. Target data feature screening: Utilize offline or online monitoring data from past target missions to construct a formatted quantitative data set. This data set includes the parameters contained in the two lines of orbital information, as well as the type, purpose, and country information corresponding to the target name in the searchable TLE information, and the scheduling system's past target mission scheduling history.
[0020] S302, target self-priority quantization model training: normalize and preprocess the data in the data set obtained in S301; use the data in the normalized preprocessed data set to train LightGBM, and during the training process, record the ratio of the total number of splits of a single data in the data set to the maximum number of splits in all data, and select data with a ratio greater than a set threshold to construct a low-dimensional data set; divide the low-dimensional data set into training sample data and test sample data; train the Bayesian optimized LightGBM again with the training sample data to obtain the optimal parameters of LightGBM, and then obtain the trained LightGBM; and use the test sample data to test the trained LightGBM;
[0021] S303, target real-time priority quantitative model prediction: Use the real-time information of the spatial target to construct formatted quantitative data, input the data into the trained LightGBM, and obtain the target real-time priority quantitative evaluation result;
[0022] S304. Generate the quantitative evaluation result of the real-time priority of the target task itself: weight the quantitative evaluation result of the real-time priority of the target task itself in combination with the task type table to obtain the quantitative evaluation result of the real-time priority of the target task itself.
[0023] Preferably, S4 includes:
[0024] S401. Calculate the general observation cutoff period. Based on the target's real-time position, velocity, acceleration information, and historical track information, and taking into account errors due to gravity, the atmosphere, and other factors, calculate the time when the target disappears from the observation range according to the motion model. This time is the general observation cutoff period.
[0025] S402: Quantify the general observation deadline and the general observation expectation period to obtain a general demand priority.
[0026] In a second aspect, a quantization system is provided for performing the priority quantization method of any radar task described in the first aspect, including:
[0027] A receiving module is used to receive four pieces of information, including target mission history information, space target real-time information, general mission observation requirements and emergency requirements;
[0028] A processing module, configured to perform data preprocessing on the four parts of information;
[0029] A first quantification processing module is used to combine the target task history information and the space target real-time information to quantify the real-time priority of the target task itself;
[0030] A second quantification processing module is configured to combine the target mission history information, the space target real-time information, and the general mission observation requirements to perform quantification processing to obtain a general requirement priority;
[0031] A third quantification processing module is configured to combine the target mission history information, the space target real-time information, and the emergency demand to perform quantification processing to obtain the emergency demand priority;
[0032] The evaluation module is used to comprehensively evaluate the real-time priority, general demand priority and emergency demand priority of the target task itself to obtain the overall comprehensive priority.
[0033] In a third aspect, a computer storage medium is provided, wherein a computer program is stored in the computer storage medium; when the computer program is executed on a computer, the computer executes the radar task priority quantification method as described in any one of the first aspects.
[0034] In a fourth aspect, a computer program product is provided, characterized in that when the computer program product is run on a computer, the computer is enabled to execute the radar task priority quantification method as described in any one of the first aspects.
[0035] The beneficial effects of the present invention are as follows: the quantification system in the present invention dynamically calculates and obtains the comprehensive priority of radar targets / tasks based on the historical information of target tasks, real-time information of space targets, general task observation requirements, and emergency requirements, thereby not only realizing the priority quantification of the task targets themselves, but also combining general task observation requirements, emergency requirements, etc., taking into account the observation requirements in the actual environment, ensuring the rationality of the comprehensive priority, and improving the overall performance of the perception system. BRIEF DESCRIPTION OF THE DRAWINGS
[0036] Figure 1 A flowchart of a radar task priority quantification method provided in this application;
[0037] Figure 2 A flowchart of a method for quantifying and generating the real-time priority of a target task based on LightGBM provided in this application;
[0038] Figure 3 A general task observation requirement quantification flow chart provided for this application;
[0039] Figure 4 A quantitative flow chart of emergency mission observation requirements provided for this application;
[0040] Figure 5 An overall comprehensive priority quantification flow chart provided for this application. DETAILED DESCRIPTION
[0041] The present invention will be further described below with reference to the following examples. The following examples are provided only to facilitate understanding of the present invention. It should be noted that, without departing from the principles of the present invention, it is possible for a person skilled in the art to make various modifications to the present invention, and such improvements and modifications fall within the scope of the claims of the present invention.
[0042] Example 1:
[0043] LightGBM is an efficient model based on decision trees and boosting. It supports efficient parallel training and features fast training speed, low memory consumption, high accuracy, and distributed support. It can quickly process massive amounts of data and has been widely used in industrial practice. Currently, no literature has used a LightGBM-based method to quantitatively assess radar target priorities.
[0044] This application provides a priority quantization method for radar tasks based on LightGBM, which is performed by the quantization system, such as Figure 1 As shown, including:
[0045] S1. Receive four pieces of information, including: target mission history information, space target real-time information, general mission observation requirements and emergency requirements.
[0046] S2. Perform data preprocessing on the four parts of information.
[0047] S3. Combine the historical information of the target mission and the real-time information of the space target to quantify the real-time priority of the target mission itself.
[0048] S4. Combine the historical information of the target mission, the real-time information of the space target and the general mission observation requirements to perform quantitative processing to obtain the general requirement priority.
[0049] S5. Combine the historical information of the target mission, the real-time information of the space target and the urgent needs, and perform quantitative processing to obtain the priority of the urgent needs.
[0050] S6. Comprehensively evaluate the real-time priority, general demand priority, and emergency demand priority of the target task itself to obtain the overall comprehensive priority and complete the priority quantification of the radar task.
[0051] Specifically, S1 includes:
[0052] S101. Receive target mission history information, where the target mission history information includes: mission information previously processed by a quantization system and previous physical information of a space target.
[0053] S102: Receive real-time space target information. The real-time space target information is provided by radar equipment in real time. The real-time space target information includes physical information detected by the radar equipment in real time.
[0054] S103: Receive general task observation requirements. These requirements primarily include: a general observation expectation period, which refers to the target observation period expected based on established planning information. For example, the general observation period for target i is Ti, meaning that target i must be observed within time period Ti, with the difference between the next observation time and the previous observation time being less than or equal to Ti.
[0055] S104. Receive emergency demands, which include: emergency demand instructions from superior equipment, experts, and relevant personnel for specific sudden target situations.
[0056] In S2, data preprocessing includes processing missing values, noisy data, and outliers. Data preprocessing here can also be called data pre-cleaning, which aims to achieve the following goals: format standardization, removal of abnormal data, error correction, and removal of duplicate data.
[0057] It should be noted that, in specific applications, the mission information processed by the past quantization system may include: scheduling history information of past missions, arrangement history of past missions, etc. The physical information of the space target may include: the target's past two-row orbital element (TLE) information, as well as other related physical information.
[0058] In S2, after the data is input, data preprocessing is first performed to clean the data and process missing values, noise data, and outliers to achieve the following goals: format standardization, abnormal data removal, error correction, and duplicate data removal.
[0059] It should be noted that data preprocessing can be implemented in various forms. This application focuses on the implementation and process of quantitative methods. For data preprocessing, please refer to other literature.
[0060] Additionally, S3 includes:
[0061] S301. Target data feature screening: Use offline or online monitoring data from previous target missions to construct a formatted quantitative data set; the data set includes the parameters contained in two lines of orbital information, as well as the type, purpose, and country information corresponding to the target name in the searchable TLE information, and the scheduling history of the scheduling system's previous target missions.
[0062] In S301 , the data set is substantially equivalent to the above-mentioned historical information of the target task, and in the process of constructing the formatted quantitative data set, the input data needs to meet certain requirements such as input format and dimension.
[0063] S302. Target self-priority quantization model training: perform normalization preprocessing on the data in the data set obtained in S301; use the data in the normalized preprocessed data set to train LightGBM, and during the training process, record the ratio of the total number of splits of a single data in the data set to the maximum number of splits among all the data, and select data with a ratio greater than the set threshold to construct a low-dimensional data set; divide the low-dimensional data set into training sample data and test sample data; use the training sample data to train the Bayesian optimized LightGBM again to obtain the optimal parameters of LightGBM, and then obtain the trained LightGBM; and use the test sample data to test the trained LightGBM.
[0064] S303: Target real-time priority quantitative model prediction: Using the real-time information data of the space target, the data is input into the trained LightGBM to obtain the quantitative evaluation result of the target real-time priority. The real-time information of the space target used in S303 is the real-time information of the space target processed in S2, and the processed real-time information of the space target needs to be converted into the data format of S302 before being input into the trained LightGBM.
[0065] It should be noted that if Figure 2 As shown, there's a difference between target priority and target task priority. One target corresponds to multiple tasks. For example, radar tasks for target A may include search, confirmation, and tracking. A predefined task type weighting table for various radar tasks is used to perform task weighting in S304 on the target priorities generated in S301, S302, and S303 to obtain the target task priority.
[0066] S304. Generate the quantitative evaluation result of the real-time priority of the target task itself: weight the quantitative evaluation result of the real-time priority of the target task itself in combination with the task type table to obtain the quantitative evaluation result of the real-time priority of the target task itself.
[0067] Additionally, the S4 includes:
[0068] S401. The general observation deadline is the time when target i disappears from the observation range according to the motion model based on the real-time position, speed, acceleration and other information of target i, as well as the historical track information of target i, taking into account the errors due to gravity, atmosphere, etc., and is recorded as t i , which is the general observation cut-off period.
[0069] S402: Quantify the general observation deadline and the general observation expectation period to obtain a general demand priority.
[0070] It should be noted that if Figure 3 As shown, when the general observation deadline t i Compared with the general observation expectation period T i The function f quantified as the general demand priority needs to meet the following conditions (here it is assumed that the larger the priority number, the higher the priority): f and t i -t is inversely proportional to T i -(tt i k-1 ) is inversely proportional to the observation deadline. That is, the smaller the observation deadline, the greater the priority; the smaller the observation expectation period, the greater the priority. Where t is the current time, t i k-1 is the last moment of observing target i.
[0071] In addition, if Figure 4As shown, the quantification system in S5 quantifies the emergency demand to obtain the priority of the emergency demand, which can be obtained through a priority table of various emergency demands defined in advance.
[0072] The S6 quantification system conducts a comprehensive evaluation of the target task's own real-time priority, general demand priority, and emergency demand priority generated by the above steps to obtain the overall comprehensive priority and complete the priority quantification of the radar task.
[0073] It should be noted that if Figure 5 As shown in Figure 1, the overall priority can be obtained by taking the weighted sum of the target task's real-time priority, general demand priority, and emergency demand priority. The weights should be selected based not only on the dimensionality but also on the actual application environment. Each weight corresponds to the degree of influence of each priority on the overall priority.
[0074] In summary, the quantification system in the present invention dynamically calculates and obtains the comprehensive priority of radar targets / tasks based on the historical information of target tasks, real-time information of space targets, general task observation requirements, and emergency requirements. It not only quantifies the priority of the task target itself, but also takes into account the observation requirements in the actual environment, combines general task observation requirements, emergency requirements, etc., ensures the rationality of the comprehensive priority, and improves the overall performance of the perception system.
Claims
1. A radar task priority quantification method, characterized in that: Performed by quantitative systems, including: S1. Receive four pieces of information, including: historical information of target missions, real-time information of space targets, general mission observation requirements, and emergency requirements; S1 includes: S101, receiving target mission history information, wherein the target mission history information includes: mission information processed by a previous quantization system and previous physical information of a space target; S102, receiving real-time space target information, where the real-time space target information is provided in real time by radar equipment, and the real-time space target information includes physical information detected in real time by the radar equipment; S103, receiving a general task observation requirement, wherein the general task observation requirement includes: a general observation expectation period, wherein the general observation expectation period is a target observation expectation period according to established planning information; S104, receiving emergency demands, wherein the emergency demands include: emergency demand instructions from superior equipment, experts, and relevant personnel for specific emergency target situations; S2, performing data preprocessing on the four parts of information; S3, combining the target task history information and the space target real-time information, quantifying the real-time priority of the target task itself; S3 includes: S301, target data feature screening: using offline or online monitoring data of previous target tasks to construct a formatted quantitative data set; the data set includes the parameters contained in the two-line orbit information, as well as the type, purpose and country information corresponding to the target name in the searchable TLE information, and the previous target task scheduling history information of the scheduling system; S302, target self-priority quantization model training: normalize and preprocess the data in the data set obtained in S301; use the data in the normalized preprocessed data set to train LightGBM, and during the training process, record the ratio of the total number of splits of a single data in the data set to the number of splits with the largest number of splits among all the data, and select the data with a ratio greater than the set threshold to construct a low-dimensional data set; divide the low-dimensional data set into training sample data and test sample data; train the Bayesian optimized LightGBM again with the training sample data to obtain the optimal parameters of LightGBM, and then obtain the trained LightGBM; and use the test sample data to test the trained LightGBM; S303, target real-time priority quantitative model prediction: Use the real-time information of the space target to construct formatted quantitative data, input the data into the trained LightGBM, and obtain the target real-time priority quantitative evaluation result; S304, generating the quantitative evaluation result of the real-time priority of the target task itself: weighting the quantitative evaluation result of the real-time priority of the target task itself in combination with the task type table to obtain the quantitative evaluation result of the real-time priority of the target task itself; S4, combining the target mission history information, the space target real-time information and the general mission observation requirements, performing quantitative processing to obtain the general requirement priority; S4 includes: S401, calculating the general observation deadline, based on the real-time position, speed, acceleration information of the target, and the historical track information of the target, taking into account the gravity and atmospheric errors, and obtaining the time when the target disappears from the observation range according to the motion model, which is the general observation deadline; S402, quantifying the general observation deadline and the general observation expectation period to obtain a general demand priority; S5, combining the target mission history information, the space target real-time information and the emergency demand, and performing quantitative processing to obtain the emergency demand priority; S6. Comprehensively evaluate the real-time priority, general demand priority and emergency demand priority of the target task itself to obtain the overall comprehensive priority.
2. The radar task priority quantification method according to claim 1, characterized in that: In S2, the data preprocessing includes: processing missing values, noise data and outliers in the data.
3. A quantization system, characterized in that: A priority quantification method for executing a radar task as described in any one of claims 1 or 2, comprising: A receiving module, used to receive four pieces of information, the four pieces of information including: target mission history information, space target real-time information, general mission observation requirements and emergency requirements; A processing module, used for performing data preprocessing on the four parts of information; A first quantification processing module is used to combine the target task history information and the space target real-time information to quantify the real-time priority of the target task itself; A second quantification processing module is used to combine the target mission history information, the space target real-time information and the general mission observation requirements to perform quantification processing to obtain the general requirement priority; A third quantification processing module is used to combine the target mission history information, the space target real-time information and the emergency demand to perform quantification processing to obtain the emergency demand priority; The evaluation module is used to comprehensively evaluate the real-time priority, general demand priority and emergency demand priority of the target task itself to obtain the overall comprehensive priority.
4. A computer storage medium, characterized in that: The computer storage medium stores a computer program; when the computer program is executed on a computer, the computer executes the radar task priority quantification method as claimed in any one of claims 1 or 2.
5. A computer program product, characterized in that When the computer program product is run on a computer, the computer is enabled to execute the radar task priority quantification method as claimed in any one of claims 1 or 2.
