A drone identification system and method

Abstract

The application discloses an unmanned aerial vehicle (UAV) identification system and method, and relates to the technical field of UAV identification.The method comprises the following steps: collecting basic information of a target area, dividing dangerous time periods of the target area based on attributes of the target area, and obtaining a theoretical limit identification boundary distance of the dangerous time periods and regular time periods; collecting types of UAV identification means and historical identification records of the target area; establishing a UAV identification basic information database; recording dangerous time periods of the target area and basic data of the UAV identification means; and according to the historical identification records of the UAV identification means of the target area, aiming at different single identification means and combined identification means.The application selects a UAV identification mode according to the types of different target areas, the time periods, and the UAV identification means and interception means of the different target areas, so as to comprehensively identify effects and use costs.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) identification technology, and more specifically, to a UAV identification system and method. Background Technology

[0002] With the rapid popularization of drone technology, its applications in aerial surveying, logistics transportation, and emergency rescue are becoming increasingly widespread, but it also brings many risks. Existing drone identification technologies mostly adopt a single identification scheme, lacking consideration for the actual conditions in different regions. The geographical and electromagnetic environments of different regions vary significantly, and the appropriate drone identification and interception methods differ. However, existing methods do not systematically collect and analyze identification and interception data for different regions, often adopting a "one-size-fits-all" identification mode. Furthermore, they lack a scientific time-segmentation mechanism and fail to differentiate between dangerous and normal time periods based on the target area's timetable. Resource allocation lacks specificity, making it difficult to reasonably streamline resources and reduce costs during normal time periods. In practical applications, the combined effects and costs of different identification methods vary greatly. Existing technologies fail to combine the theoretical limit of identification boundary distance with the current identification distance of suspicious targets to construct a scientific utility evaluation system, and cannot output the optimal combination of identification distance and identification methods. This easily leads to insufficient scenario adaptability of the identification scheme and low practicality. Summary of the Invention

[0003] In view of the problems in related technologies, the present invention proposes an unmanned aerial vehicle (UAV) identification system and method to overcome the above-mentioned technical problems existing in the existing related technologies.

[0004] Therefore, the specific technical solution adopted by the present invention is as follows: A method for identifying unmanned aerial vehicles (UAVs), the method comprising the following steps: S1. Collect basic information about the target area, divide the dangerous time periods of the target area based on the attributes of the target area, and obtain the theoretical limit recognition boundary distance between the dangerous time periods and the normal time periods. Collect the types of UAV recognition methods and historical recognition records in the target area, establish a UAV recognition basic information database, and record the dangerous time periods and basic data of UAV recognition methods in the target area. S2. Based on the historical identification records of UAV identification methods in the target area, construct a model of the relationship between the distance of suspicious targets and the identification success rate under different weather conditions for different single identification methods and combined identification methods. Obtain the relationship between the target distance and the identification success rate of single identification methods and combined identification methods under different weather conditions, and calculate the current usage cost of different methods. S3. Based on the initial detection distance, theoretical limit recognition boundary distance, and weather conditions of the current suspicious target, and based on the target distance, success rate, and cost of different single and combined recognition methods, under the current initial detection distance and theoretical limit recognition boundary distance of the suspicious target, the UAV identification of the suspicious target is performed based on the utility function to output the current optimal judgment distance and recognition method.

[0005] In a preferred embodiment, S1 includes the following steps: S11. Collect basic information about the target area through the GIS system, including the type and area of ​​the target area. Divide the dangerous time periods based on the work schedule of the target area for different types of target areas. Combine the working range of the drone interception and expulsion equipment in the target area to divide the theoretical limit identification boundary distance between the dangerous time period and the normal time period. S12. Collect the types of UAV identification methods in the target area and historical identification records under different weather conditions, including identification records of single identification methods and identification records of combined identification methods. Establish a UAV identification basic information database through MySQL, create files based on the target area number, and create sub-files in the files. Record the UAV identification records of the target area and the theoretical limit identification boundary distances for dangerous periods, normal periods, and corresponding periods.

[0006] In a preferred embodiment, S11 includes the following steps: S111, Types based on target region Based on the current standard working day schedule and special event schedule of the target area, threat levels are assigned to different time periods. calculate: ; in, These represent the base threat level, the time-weighted function, and the special event weight, respectively. Represents the permission coefficient, and Threat levels based on different time periods t in the output The dangerous and normal time periods are divided into different target areas: ; Where 1 represents a dangerous period and 0 represents a normal period. Represents a danger threshold, based on Divide the standard workday schedule and the danger period and regular period under the special event schedule for different target areas; S112. Statistically analyze historical UAV interception data for the target area. Based on the interception distances under the historical interception data, divide the theoretical limit recognition boundary distances between dangerous periods and normal periods. Add three standard deviations to the mean of historical interception distances as the theoretical limit recognition boundary distances for dangerous periods, and use the mean of historical interception distances as the theoretical limit recognition boundary distances for normal periods.

[0007] In a preferred embodiment, S2 includes the following steps: S21. Based on the UAV identification basic information database, obtain the historical identification records of UAV identification methods in the target area, and construct a model of the relationship between the distance of suspicious targets and the identification success rate of different single identification methods and combined identification methods under different weather conditions for the current target area. S22. Calculate the usage costs of different single identification methods and combined identification methods in historical identification records, and comprehensively calculate the usage costs of different single identification methods and combined identification methods in a single identification.

[0008] In a preferred embodiment, S21 includes the following steps: S211. For historical records identified by a single identification method in the basic information database for UAV identification, descriptive weather is encoded as discrete categorical variables. Where k represents different weather state numbers in historical records, all single means in the current target area are encoded as discrete categorical variables. s represents the unique identifier of a single drone in the current target area, and a dataset is constructed for the current target area under different weather conditions. ,in and represent the initial detection distance and final identification result of the suspicious target in the j-th record, respectively. In the identification result, 0 represents identification failure and 1 represents identification success. Logistic regression is used to construct a model of the relationship between the suspicious target distance and the identification success rate under a single identification method. The specific steps are as follows: For specific weather and methods The relationship between its success rate and the suspicious target identification distance d is as follows: ; in, Represents the situation under given weather conditions. and methods The predicted probability of successfully identifying a suspicious target at a distance d. These represent the intercept, the coefficient of the first-order distance term, and the coefficient of the second-order distance term, respectively. S212. Based on the historical records of combined identification methods in the UAV identification basic information database, assign an identifier to each unique combination of methods. Where m represents the number of different combinations of methods, to construct the dataset. The theoretical success rate of combined methods is calculated based on the independent parallel model. ,in For a single means Under the same conditions, the recognition success rate The probability that all methods fail simultaneously under the probability formula representing independent events; Based on dataset The success rate of combinations in historical data is calculated based on distance intervals, and parameterized fitting is performed on the success rates of combinations in different distance intervals of historical data to obtain empirical success rates. Calculate the empirical correction factor function Using calculations from historical data The correction function is obtained based on polynomial fitting of the data points. ; A model combining the correction function and theoretical success rate was obtained to demonstrate the relationship between the distance to suspicious targets and the success rate of identification. ,in This represents the correction factor function obtained by fitting a specific weather-combination pair.

[0009] In a preferred embodiment, S22 includes the following steps: S221. Based on the energy cost, equipment depreciation cost, and additional costs of different devices, calculate the single-use cost of different single identification methods and combined identification methods. The energy cost is the power consumption per unit time of various sensors and devices in typical working modes, and the energy consumption cost per unit time is calculated in combination with the local industrial electricity price. The equipment depreciation cost is (equipment purchase price - residual value) / expected full life cycle working time, which is the depreciation cost per unit working hour. The additional cost is the extra cost of using different devices. The energy cost, equipment depreciation cost, and additional cost are added together to obtain the single-use cost of single identification methods and combined identification methods. S222. In the target area file corresponding to the UAV identification basic information database, record the cost of a single identification method and the cost of a combination of identification methods per use.

[0010] In a preferred embodiment, step S3 includes the following steps: S31. Based on the initial detection distance, time, and weather conditions of the current suspicious target, and according to the detection time of the current suspicious target and the type of target area, obtain the theoretical limit recognition boundary distance, output the success rate and cost of different single and combined recognition methods at the current initial detection distance and theoretical limit recognition boundary distance, and make multi-dimensional judgments to output the current best judgment distance and the type of recognition method.

[0011] In a preferred embodiment, S31 includes the following steps: S311. Based on the initial discovery time of the current suspicious target Obtain the current initial discovery time Determine the type of the current initial discovery time, including dangerous periods and normal periods, in order to obtain the theoretical limit of the identification boundary distance at the current time; S312. When the initial detection distance of a suspicious target is less than or equal to the theoretical limit of the identification boundary distance, calculate the identification success rate of the single identification method and the combined identification method at the initial detection distance of the suspicious target in the current area based on the current weather conditions, and select the method with the highest identification success rate as the identification method of the suspicious target for UAV identification of the suspicious target. When the initial detection distance of a suspicious target is greater than the theoretical limit of identification boundary distance, for distance data points within the range between the initial detection distance and the theoretical limit of identification boundary distance, calculate the identification success rate and single-use cost of the current single identification method and the combined identification method for different distance data points in the current area, and determine the current optimal identification method through a utility function: ; in, These represent the distance data points and the recognition method, respectively. include and , Represents the current weather condition. Representative identification methods The standard cost of a single identification, Represents weight, and is based on the lowest recognition success rate threshold. right Apply constraints to satisfy Iterate through all elements that satisfy the constraints. Yes, the combination that maximizes the utility function U is selected as the current optimal identification method for identifying suspicious targets by drones.

[0012] A drone identification system includes a data acquisition module, a relationship model establishment module, and a situation determination and identification method selection module; The data acquisition module collects basic information about the target area, divides the dangerous time periods of the target area based on the attributes of the target area, obtains the theoretical limit recognition boundary distance between the dangerous time periods and the normal time periods, collects the types of UAV recognition methods and historical recognition records of the target area, establishes a UAV recognition basic information database, and records the dangerous time periods and basic data of UAV recognition methods in the target area. The relationship model building module, based on the historical identification records of UAV identification methods in the target area, constructs a relationship model between the distance of suspicious targets and the identification success rate under different weather conditions for different single identification methods and combined identification methods. It obtains the relationship between the target distance and the identification success rate of single identification methods and combined identification methods under different weather conditions, and calculates the current usage cost of different methods. The situation determination and identification method selection module compares and determines the initial detection distance and theoretical limit identification boundary distance of the current suspicious target based on the initial detection distance, theoretical limit identification boundary distance, and weather conditions, as well as the target distance, success rate, and cost of different single and combined identification methods. Based on the determination result, it selects a utility function to output the current optimal determination distance and identification method for UAV identification of the suspicious target.

[0013] The beneficial effects of this invention are as follows: 1. This invention collects data on drone identification and interception methods in different regions, divides dangerous and normal time periods into time periods based on the timetable of different target areas, calculates the theoretical limit of identification boundary distance, and combines the identification distance of the current suspicious target with the success rate and cost of different single and combined identification methods to select drone identification methods. It selects the most suitable drone identification scheme according to local conditions, optimizes drone identification costs, and enhances practicality. 2. This invention comprehensively considers success rate and cost, and combines the detection distance and time period of suspicious targets. Based on the actual situation, it rationally allocates resources and prioritizes the identification method with the most reasonable utility function while meeting the identification requirements. It outputs the comprehensive optimal combination of identification distance and identification method, thereby enhancing functionality. 3. This invention calculates threat levels based on different regional types and divides dangerous periods into normal times. It can make targeted dangerous period divisions according to the characteristics of different regional types to meet the needs of different use scenarios. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a flowchart of a drone identification method according to an embodiment of the present invention. Detailed Implementation

[0016] To further illustrate the various embodiments, the present invention provides accompanying drawings, which are part of the disclosure of the present invention. These drawings are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these drawings, those skilled in the art should be able to understand other possible implementation methods and the advantages of the present invention. The components in the drawings are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0017] According to an embodiment of the present invention, an unmanned aerial vehicle (UAV) identification system and method are provided.

[0018] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments: Example 1

[0019] like Figure 1 As shown, according to an embodiment of the present invention, a method for identifying unmanned aerial vehicles (UAVs) includes the following steps: S1. Collect basic information about the target area, divide the dangerous time periods of the target area based on the attributes of the target area, and obtain the theoretical limit recognition boundary distance between the dangerous time periods and the normal time periods. Collect the types of UAV recognition methods and historical recognition records in the target area, establish a UAV recognition basic information database, and record the dangerous time periods and basic data of UAV recognition methods in the target area. S11. Collect basic information about the target area through the GIS system, including the type and area of ​​the target area. Divide the dangerous time periods based on the work schedule of the target area for different types of target areas. Combine the working range of the drone interception and expulsion equipment in the target area to divide the theoretical limit identification boundary distance between the dangerous time period and the normal time period. S111, Types based on target region Based on the current standard working day schedule and special event schedule of the target area, threat levels are assigned to different time periods. calculate: ; in, These represent the base threat level, the time-weighted function, and the special event weight, respectively. Represents the permission coefficient, and Threat levels based on different time periods t in the output The dangerous and normal time periods are divided into different target areas: ; Where 1 represents a dangerous period and 0 represents a normal period. Represents a danger threshold, based on Divide the standard workday schedule and the danger period and regular period under the special event schedule for different target areas; It should be noted that, These represent the basic threat level, time weighting function, and special event weight, respectively. The basic threat level and special event weight need to be adjusted based on the target area type i. Target area types include military / government-managed areas, large event areas, and residential areas. For example, in military / government-managed areas, the danger period needs to be 24 hours a day, and the basic threat level is usually set to 1 with a weighting coefficient of 1 to ensure the security of the target area. In large event areas, since the population density is not usually high, the basic threat level can be set lower, and the special event weight can be increased during the special event start time. Specific settings require consultation with relevant experts and empirical methods. The time weighting function needs to be adjusted based on the different time periods of different types of areas each day, such as the weighting during morning and evening rush hours and the weighting during nighttime. It needs to be set based on the area type and empirical methods. At the same time, the corresponding permission coefficients should be adjusted based on the area type. It is usually set to 0.55, based on the current local time arrangement under the standard workday schedule and special event schedule. Perform the calculation.

[0020] S112. Statistically analyze historical UAV interception data for the target area. Based on the interception distances under the historical interception data, divide the theoretical limit recognition boundary distances between dangerous periods and normal periods. Add three standard deviations to the mean of historical interception distances as the theoretical limit recognition boundary distances for dangerous periods, and use the mean of historical interception distances as the theoretical limit recognition boundary distances for normal periods.

[0021] It should be noted that the theoretical limit of recognition boundary distance is set higher for dangerous periods to ensure that drones can be effectively identified and intercepted in advance during dangerous periods. Drones need to be detected earlier during dangerous periods to ensure sufficient time for processing. Therefore, the theoretical limit of recognition boundary distance will be farther than during normal periods.

[0022] S12. Collect the types of UAV identification methods in the target area and historical identification records under different weather conditions, including identification records of single identification methods and identification records of combined identification methods. Establish a UAV identification basic information database through MySQL, create files based on the target area number, and create sub-files in the files. Record the UAV identification records of the target area and the theoretical limit identification boundary distances for dangerous periods, normal periods, and corresponding periods.

[0023] It should be noted that the single identification method and the combined identification method need to be determined based on the actual situation of the current area. The single identification method includes image recognition, acoustic recognition, radar recognition, radio frequency identification, etc. The combined identification method is the drone identification method under the combination of the single identification methods in the current area.

[0024] Example 2 S2. Based on the historical identification records of UAV identification methods in the target area, construct a model of the relationship between the distance of suspicious targets and the identification success rate under different weather conditions for different single identification methods and combined identification methods. Obtain the relationship between the target distance and the identification success rate of single identification methods and combined identification methods under different weather conditions, and calculate the current usage cost of different methods. S21. Based on the UAV identification basic information database, obtain the historical identification records of UAV identification methods in the target area, and construct a model of the relationship between the distance of suspicious targets and the identification success rate of different single identification methods and combined identification methods under different weather conditions for the current target area. To identify historical records using a single identification method in the basic information database for drone identification, descriptive weather is encoded as discrete categorical variables. Where k represents different weather state numbers in historical records, all single means in the current target area are encoded as discrete categorical variables. s represents the unique identifier of a single drone in the current target area, and a dataset is constructed for the current target area under different weather conditions. ,in and represent the initial detection distance and final identification result of the suspicious target in the j-th record, respectively. In the identification result, 0 represents identification failure and 1 represents identification success. Logistic regression is used to construct a model of the relationship between the suspicious target distance and the identification success rate under a single identification method. The specific steps are as follows: For specific weather and methods The relationship between its success rate and the suspicious target identification distance d is as follows: ; in, Represents the situation under given weather conditions. and methods The predicted probability of successfully identifying a suspicious target at a distance d. These represent the intercept, the coefficient of the first-order distance term, and the coefficient of the second-order distance term, respectively. S212. Based on the historical records of combined identification methods in the UAV identification basic information database, assign an identifier to each unique combination of methods. Where m represents the number of different combinations of methods, to construct the dataset. The theoretical success rate of combined methods is calculated based on the independent parallel model. ,in For a single means Under the same conditions, the recognition success rate The probability of all means failing simultaneously under the probability formula representing independent events; Based on dataset The success rate of combinations in historical data is calculated based on distance intervals, and parameterized fitting is performed on the success rates of combinations in different distance intervals of historical data to obtain empirical success rates. Calculate the empirical correction factor function Using calculations from historical data The correction function is obtained based on polynomial fitting of the data points. ; A model combining the correction function and theoretical success rate was obtained to demonstrate the relationship between the distance to suspicious targets and the success rate of identification. ,in This represents the correction factor function obtained by fitting a specific weather-combination pair.

[0025] It should be noted that for each pair The above model can output a smooth success rate-distance relationship curve from the minimum detection distance to the maximum effective distance. By recording the success rate-distance relationship curve in the corresponding file, it is convenient to call it later to output the success rate. At the same time, when the system accumulates new combination recognition records, it is merged into the historical database, and the above model building steps are periodically re-executed to update the correction factor function.

[0026] S22. Calculate the usage cost of different single identification methods and combined identification methods in historical identification records, and comprehensively calculate the usage cost of different single identification methods and combined identification methods in a single identification. S221. Based on the energy cost, equipment depreciation cost, and additional costs of different devices, calculate the single-use cost of different single identification methods and combined identification methods. The energy cost is the power consumption per unit time of various sensors and devices in typical working modes, and the energy consumption cost per unit time is calculated in combination with the local industrial electricity price. The equipment depreciation cost is (equipment purchase price - residual value) / expected total life cycle working time, which is the depreciation cost per working hour. The additional cost is the extra cost of using different devices. The energy cost, equipment depreciation cost, and additional costs are added together to obtain the single-use cost of single identification methods and combined identification methods.

[0027] It should be noted that for consumables with a limited lifespan, such as optical lenses and filters, the cost of equipment wear and tear is amortized based on the duration or number of uses to obtain the cost of equipment wear and tear per use. Additional costs need to be set in advance based on the specific identification methods of different regions and situations, such as additional spectrum usage fees.

[0028] S3. Based on the initial detection distance, theoretical limit recognition boundary distance, and weather conditions of the current suspicious target, and based on the target distance, success rate, and cost of different single and combined recognition methods, under the current initial detection distance and theoretical limit recognition boundary distance of the suspicious target, the drone identification of the suspicious target is performed based on the utility function to output the current optimal judgment distance and recognition method. S31. Based on the initial detection distance, time, and weather conditions of the current suspicious target, and according to the detection time of the current suspicious target and the type of target area, obtain the theoretical limit recognition boundary distance, output the success rate and cost of different single and combined recognition methods at the current initial detection distance and theoretical limit recognition boundary distance, and make multi-dimensional judgments to output the current best judgment distance and the type of recognition method. S311. Based on the initial discovery time of the current suspicious target Obtain the current initial discovery time Determine the type of the current initial discovery time, including dangerous periods and normal periods, in order to obtain the theoretical limit of the identification boundary distance at the current time; S312. When the initial detection distance of a suspicious target is less than or equal to the theoretical limit of the identification boundary distance, calculate the identification success rate of the single identification method and the combined identification method at the initial detection distance of the suspicious target in the current area based on the current weather conditions, and select the method with the highest identification success rate as the identification method of the suspicious target for UAV identification of the suspicious target. When the initial detection distance of a suspicious target is greater than the theoretical limit of identification boundary distance, for distance data points within the range between the initial detection distance and the theoretical limit of identification boundary distance, calculate the identification success rate and single-use cost of the current single identification method and the combined identification method for different distance data points in the current area, and determine the current optimal identification method through a utility function: ; in, These represent the distance data points and the recognition method, respectively. include and , Represents the current weather condition. Representative identification methods The standard cost of a single identification, Represents weight, and is based on the lowest recognition success rate threshold. right Apply constraints to satisfy Iterate through all elements that satisfy the constraints. Yes, the combination that maximizes the utility function U is selected as the current optimal identification method for identifying suspicious targets by drones.

[0029] It should be noted that when the initial detection distance of a suspicious target is greater than the theoretical limit of the identification boundary distance, weights are set for the dangerous period and the normal period to measure the utility. Specifically, during the dangerous period... The values ​​were set to 0.8 and 0.2 respectively to prioritize success rate, and then set to 0.5 and 0.5 respectively during normal periods to balance success rate and cost. It is usually set to 0.85.

[0030] Example 3 A drone identification system includes a data acquisition module, a relationship model establishment module, and a situation determination and identification method selection module; The data acquisition module collects basic information about the target area, divides the dangerous periods of the target area based on the attributes of the target area, obtains the theoretical limit recognition boundary distance between the dangerous periods and the normal periods, collects the types of UAV recognition methods and historical recognition records in the target area, establishes a database of basic information on UAV recognition, and records the dangerous periods of the target area and the basic data of UAV recognition methods. The relationship model building module, based on the historical identification records of UAV identification methods in the target area, constructs a relationship model between the distance of suspicious targets and the identification success rate under different weather conditions for different single identification methods and combined identification methods. It obtains the relationship between the target distance and the identification success rate under different weather conditions for single identification methods and combined identification methods, and calculates the current usage cost of different methods. The situation determination and identification method selection module compares and determines the initial detection distance and theoretical limit identification boundary distance of the current suspicious target based on the initial detection distance, theoretical limit identification boundary distance, and weather conditions, as well as the target distance, success rate, and cost of different single and combined identification methods. Based on the determination result, a utility function is selected to output the current optimal determination distance and identification method for UAV identification of the suspicious target.

[0031] In summary, this invention collects data on drone identification and interception methods for different regions, divides dangerous and normal time periods based on timetables for different target areas, calculates the theoretical limit of identification boundary distance, and selects drone identification methods by considering the current identification distance of suspicious targets and the success rate and cost of different single and combined identification methods. It chooses the most suitable drone identification scheme according to local conditions, optimizes drone identification costs, and rationally allocates resources based on actual circumstances by comprehensively considering success rate and cost, combined with the detection distance and time period of suspicious targets. While meeting identification requirements, it prioritizes the identification method with the most reasonable utility function, outputting the optimal combination of identification distance and identification method, thus enhancing functionality.

[0032] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for identifying unmanned aerial vehicles (UAVs), characterized in that, The method includes the following steps: S1. Collect basic information about the target area, divide the dangerous time periods of the target area based on the attributes of the target area, and obtain the theoretical limit recognition boundary distance between the dangerous time periods and the normal time periods. Collect the types of UAV recognition methods and historical recognition records in the target area, establish a UAV recognition basic information database, and record the dangerous time periods and basic data of UAV recognition methods in the target area. S2. Based on the historical identification records of UAV identification methods in the target area, construct a model of the relationship between the distance of suspicious targets and the identification success rate under different weather conditions for different single identification methods and combined identification methods. Obtain the relationship between the target distance and the identification success rate of single identification methods and combined identification methods under different weather conditions, and calculate the current usage cost of different methods. S3. Based on the initial detection distance, theoretical limit recognition boundary distance, and weather conditions of the current suspicious target, and based on the target distance, success rate, and cost of different single and combined recognition methods, under the current initial detection distance and theoretical limit recognition boundary distance of the suspicious target, the UAV identification of the suspicious target is performed based on the utility function to output the current optimal judgment distance and recognition method.

2. The UAV identification method according to claim 1, characterized in that, S1 includes the following steps: S11. Collect basic information about the target area through the GIS system, including the type and area of ​​the target area. Divide the dangerous time periods based on the work schedule of the target area for different types of target areas. Combine the working range of the drone interception and expulsion equipment in the target area to divide the theoretical limit identification boundary distance between the dangerous time period and the normal time period. S12. Collect the types of UAV identification methods in the target area and historical identification records under different weather conditions, including identification records of single identification methods and identification records of combined identification methods. Establish a UAV identification basic information database through MySQL, create files based on the target area number, and create sub-files in the files. Record the UAV identification records of the target area and the theoretical limit identification boundary distances for dangerous periods, normal periods, and corresponding periods.

3. The method for identifying unmanned aerial vehicles (UAVs) according to claim 2, characterized in that, S11 includes the following steps: S111, Types based on target region Based on the current standard working day schedule and special event schedule of the target area, threat levels are assigned to different time periods. calculate: ; in, These represent the base threat level, the time-weighted function, and the special event weight, respectively. Represents the permission coefficient, and Threat levels based on different time periods t in the output The dangerous and normal time periods are divided into different target areas: ; Where 1 represents a dangerous period and 0 represents a normal period. Represents a danger threshold, based on Divide the standard workday schedule and the danger period and regular period under the special event schedule for different target areas; S112. Statistically analyze historical UAV interception data for the target area. Based on the interception distances under the historical interception data, divide the theoretical limit recognition boundary distances between dangerous periods and normal periods. Add three standard deviations to the mean of historical interception distances as the theoretical limit recognition boundary distances for dangerous periods, and use the mean of historical interception distances as the theoretical limit recognition boundary distances for normal periods.

4. The UAV identification method according to claim 3, characterized in that, S2 includes the following steps: S21. Based on the UAV identification basic information database, obtain the historical identification records of UAV identification methods in the target area, and construct a model of the relationship between the distance of suspicious targets and the identification success rate of different single identification methods and combined identification methods under different weather conditions for the current target area. S22. Calculate the usage costs of different single identification methods and combined identification methods in historical identification records, and comprehensively calculate the usage costs of different single identification methods and combined identification methods in a single identification.

5. The UAV identification method according to claim 4, characterized in that, S21 includes the following steps: S211. For historical records identified by a single identification method in the basic information database for UAV identification, descriptive weather is encoded as discrete categorical variables. Where k represents different weather state numbers in historical records, all single means in the current target area are encoded as discrete categorical variables. s represents the unique identifier of a single drone in the current target area, and a dataset is constructed for the current target area under different weather conditions. ,in and represent the initial detection distance and final identification result of the suspicious target in the j-th record, respectively. In the identification result, 0 represents identification failure and 1 represents identification success. Logistic regression is used to construct a model of the relationship between the suspicious target distance and the identification success rate under a single identification method. The specific steps are as follows: For specific weather and methods The relationship between its success rate and the suspicious target identification distance d is as follows: ; in, Represents the situation under given weather conditions. and methods The predicted probability of successfully identifying a suspicious target at a distance d. These represent the intercept, the coefficient of the first-order distance term, and the coefficient of the second-order distance term, respectively. S212. Based on the historical records of combined identification methods in the UAV identification basic information database, assign an identifier to each unique combination of methods. Where m represents the number of different combinations of methods, to construct the dataset. The theoretical success rate of combined methods is calculated based on the independent parallel model. ,in For a single means Under the same conditions, the recognition success rate The probability of all means failing simultaneously under the probability formula representing independent events; Based on dataset The success rate of combinations in historical data is calculated based on distance intervals, and parameterized fitting is performed on the success rates of combinations in different distance intervals of historical data to obtain empirical success rates. Calculate the empirical correction factor function Using calculations from historical data The correction function is obtained based on polynomial fitting of the data points. ; A model combining the correction function and theoretical success rate was obtained to demonstrate the relationship between the distance to suspicious targets and the success rate of identification. ,in This represents the correction factor function obtained by fitting a specific weather-combination pair.

6. The unmanned aerial vehicle (UAV) identification method according to claim 5, characterized in that, S22 includes the following steps: S221. Based on the energy cost, equipment depreciation cost, and additional costs of different devices, calculate the single-use cost of different single identification methods and combined identification methods. The energy cost is the power consumption per unit time of various sensors and devices in typical working modes, and the energy consumption cost per unit time is calculated in combination with the local industrial electricity price. The equipment depreciation cost is (equipment purchase price - residual value) / expected full life cycle working time, which is the depreciation cost per unit working hour. The additional cost is the extra cost of using different devices. The energy cost, equipment depreciation cost, and additional cost are added together to obtain the single-use cost of single identification methods and combined identification methods. S222. In the target area file corresponding to the UAV identification basic information database, record the cost of a single identification method and the cost of a combination of identification methods per use.

7. The UAV identification method according to claim 6, characterized in that, S3 includes the following steps: S31. Based on the initial detection distance, time, and weather conditions of the current suspicious target, and according to the detection time of the current suspicious target and the type of target area, obtain the theoretical limit recognition boundary distance, output the success rate and cost of different single and combined recognition methods at the current initial detection distance and theoretical limit recognition boundary distance, and make multi-dimensional judgments to output the current best judgment distance and the type of recognition method.

8. The method for identifying unmanned aerial vehicles (UAVs) according to claim 7, characterized in that, S31 includes the following steps: S311. Based on the initial discovery time of the current suspicious target Obtain the current initial discovery time Determine the type of the current initial discovery time, including dangerous periods and normal periods, in order to obtain the theoretical limit of the identification boundary distance at the current time; S312. When the initial detection distance of a suspicious target is less than or equal to the theoretical limit of the identification boundary distance, calculate the identification success rate of the single identification method and the combined identification method at the initial detection distance of the suspicious target in the current area based on the current weather conditions, and select the method with the highest identification success rate as the identification method of the suspicious target for UAV identification of the suspicious target. When the initial detection distance of a suspicious target is greater than the theoretical limit of identification boundary distance, for distance data points within the range between the initial detection distance and the theoretical limit of identification boundary distance, calculate the identification success rate and single-use cost of the current single identification method and the combined identification method for different distance data points in the current area, and determine the current optimal identification method through a utility function: ; in, These represent the distance data points and the recognition method, respectively. include and , Represents the current weather condition. Representative identification methods The standard cost of a single identification, Represents weight, and is based on the lowest recognition success rate threshold. right Apply constraints to satisfy Iterate through all elements that satisfy the constraints. Yes, the combination that maximizes the utility function U is selected as the current optimal identification method for identifying suspicious targets by drones.

9. A drone identification system, characterized in that, The system employs a drone identification method as described in any one of claims 1-8, comprising a data acquisition module, a relationship model establishment module, and a situation determination and identification method selection module; The data acquisition module collects basic information about the target area, divides the dangerous time periods of the target area based on the attributes of the target area, obtains the theoretical limit recognition boundary distance between the dangerous time periods and the normal time periods, collects the types of UAV recognition methods and historical recognition records of the target area, establishes a UAV recognition basic information database, and records the dangerous time periods and basic data of UAV recognition methods in the target area. The relationship model building module, based on the historical identification records of UAV identification methods in the target area, constructs a relationship model between the distance of suspicious targets and the identification success rate under different weather conditions for different single identification methods and combined identification methods. It obtains the relationship between the target distance and the identification success rate of single identification methods and combined identification methods under different weather conditions, and calculates the current usage cost of different methods. The situation determination and identification method selection module compares and determines the initial detection distance and theoretical limit identification boundary distance of the current suspicious target based on the initial detection distance, theoretical limit identification boundary distance, and weather conditions, as well as the target distance, success rate, and cost of different single and combined identification methods. Based on the determination result, it selects a utility function to output the current optimal determination distance and identification method for UAV identification of the suspicious target.

Images