A method for improving the positioning accuracy of unmanned aerial vehicle for observation targets

Abstract

The present invention provides a method for improving the positioning accuracy of an unmanned aerial vehicle for an observation target, monitoring a receiving end, obtaining first data, verifying the obtained first data, and then obtaining a real-time data set, filtering the data in the real-time data set, and obtaining first result data. By filtering the obtained first result data, a measurement point with the smallest error can be selected, and data with higher positioning accuracy can be obtained, thereby improving the accuracy of the positioning data, avoiding the inefficient process of analyzing a large amount of data one by one, and improving the measurement efficiency and positioning accuracy.

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial vehicle (UAV) positioning, and in particular to a method for improving the positioning accuracy of an unmanned aerial vehicle (UAV) for an observation target. Background Art

[0002] At present, in the prior art, referring to the principle of GPS single-point positioning, the observed target can be solved based on the distance measurement values ​​of the drone to the observed target at at least three different positions and the longitude, latitude and altitude of the pod at the time of measurement. Among them, the accuracy of the solution depends on two factors: the distance measurement value of the observed target and the longitude, latitude and altitude of the pod at the time of measurement. The smaller the measurement error, the higher the accuracy of positioning; the geometric configuration of the three points. The geometric configuration is approximately close to an equilateral triangle, and the higher the accuracy of positioning. However, each parameter of the data sent back by the drone in the prior art has errors, which leads to low positioning accuracy. Summary of the invention

[0003] The present invention aims to solve at least one of the technical problems existing in the prior art or related art.

[0004] To this end, a first aspect of the present invention proposes a method for improving the positioning accuracy of a drone for an observation target.

[0005] In view of this, a first aspect of the present invention provides a method for improving the positioning accuracy of a drone for an observation target, comprising: obtaining first data; verifying the first data and obtaining a first verification result; based on the first verification result, obtaining a real-time data set; filtering the data in the real-time data set to obtain first result data; archiving and processing the first result data.

[0006] The present invention provides a method for improving the positioning accuracy of an unmanned aerial vehicle for an observation target, monitoring a receiving end, obtaining first data, verifying the obtained first data, and then obtaining a real-time data set, filtering the data in the real-time data set, and obtaining first result data. By filtering the obtained first result data, a measurement point with the smallest error can be selected, and data with higher positioning accuracy can be obtained, thereby improving the accuracy of the positioning data, avoiding the inefficient process of analyzing a large amount of data one by one, and improving the measurement efficiency and positioning accuracy.

[0007] In addition, a method for improving the positioning accuracy of a drone for an observation target in the above technical solution provided by the present invention may also have the following additional technical features:

[0008] In some technical solutions of the present invention, optionally, verifying the first data and obtaining a first verification result includes: filtering outliers in the first data; obtaining a first task identifier that matches the first data; and determining whether the first task identifier is consistent with the current task identifier to obtain the first verification result.

[0009] In this technical solution, outliers in the first data are filtered, and a first task identifier matching the first data is obtained. It is determined whether the first task identifier is consistent with the current task identifier, and then a first verification result is obtained. By filtering outliers, invalid or missing information in the first data can be removed, the overall quality of the data can be improved, and errors and deviations in subsequent data processing and analysis can be reduced. In addition, by determining whether the first task identifier is consistent with the current task identifier, task confusion or execution errors can be prevented, the consistency of tasks can be ensured, and the stability and reliability of the entire system can be improved.

[0010] In some technical solutions of the present invention, optionally, based on the first verification result, a real-time data set is obtained, including: if the first verification result is that the first task identifier is consistent with the current task identifier, the first data is added to the real-time data set; if the first verification result is that the first task identifier is inconsistent with the current task identifier, the current task identifier is updated to the first task identifier, the first variable and the timer are reset, and the first data is added to the real-time data set.

[0011] In this technical solution, different processing is performed based on the first verification result. If the first verification result is that the first task identifier is consistent with the current task identifier, the first data is added to the real-time data set; if the first verification result is that the first task identifier is inconsistent with the current task identifier, the current task identifier is updated to the first task identifier, the first variable and the timer are reset, and the first data is added to the real-time data set. By updating the current task identifier to the first task identifier, task confusion or execution errors are prevented, the consistency and continuity of tasks are guaranteed, errors and conflicts are reduced, and the stability and reliability of the system are improved.

[0012] In some technical schemes of the present invention, optionally, data in a real-time data set is filtered to obtain first result data, including: dividing the real-time data set into a pod point set and a pod identification target point set; filtering the pod point set and the pod identification target point set to obtain a pod point optimization set and a pod identification target point optimization set; filtering the pod point optimization set and the pod identification target point optimization set to obtain a pod point advanced optimization set and a pod identification target point advanced optimization set; based on the pod point advanced optimization set and the pod identification target point advanced optimization set, obtaining an alternative combination set; based on the pod point advanced optimization set and the pod identification target point advanced optimization set, obtaining first statistical information of the pod identification target point advanced optimization set; based on the first statistical information, positioning and solving the alternative combination set to obtain an alternative target point set; filtering the alternative target point set to obtain the coordinates of the real target point.

[0013] In this technical solution, the real-time data set is divided into a pod point set and a pod identification target point set, and the pod identification target point set is used as the main body for a first screening, and then the pod point optimization set and the pod identification target point optimization set are obtained, and the pod point optimization set is used as the main body for a second screening, and the pod point advanced optimization set and the pod identification target point advanced optimization set are obtained. By screening the real-time data set twice, the data that does not meet the conditions or is of low value is initially eliminated, so that the data set that is screened twice is more accurate and reliable, and the complexity of data processing is reduced. On this basis, based on the pod point advanced optimization set and the pod identification target point advanced optimization set, that is, the data after the second screening, the value range is further narrowed to obtain the alternative combination set, and the first statistical information based on the pod identification target point advanced optimization set is located and settled to obtain the alternative target point set, and finally the alternative target point set is filtered to obtain the coordinates of the real target point. By screening the data multiple times, the range of data values ​​can be gradually narrowed, thereby achieving accurate positioning of the real target point, improving the speed and efficiency of data processing, reducing errors and uncertainties, improving the accuracy and efficiency of task execution, and ensuring the accuracy and reliability of the final data.

[0014] In some technical schemes of the present invention, optionally, a pod point set and a pod identification target point set are screened to obtain a pod point optimization set and a pod identification target point optimization set, including: obtaining the centroid of the pod identification target point set; obtaining a first distance set based on each pod identification target point of the pod identification target point set and the centroid of the pod identification target point set; obtaining one times the standard deviation of the first distance set based on the first distance set; eliminating pod identification target points in the first distance set that exceed one times the standard deviation of the first distance set to obtain an ordered pod identification target point set; based on the task identification set of the ordered pod identification target point set, screening pod points and pod identification target points that are the same as the task identification set of the ordered pod identification target point set to obtain a pod point optimization set and a pod identification target point optimization set.

[0015] In this technical solution, the real-time data set is initially screened with the pod identification target point set as the main body, the centroid of the pod identification target point set is obtained, and the first distance set is obtained based on each pod identification target point in the pod identification target point set and the centroid of the pod identification target point set. On this basis, based on the first distance set, one standard deviation of the first distance set is obtained, and the pod identification target points in the first distance set that exceed one standard deviation of the first distance set are eliminated to obtain an ordered pod identification target point set. In addition, based on the point identifier of the ordered pod identification target point set, the pod points and pod identification target points with the same point identifier as the ordered pod identification target points in the ordered pod identification target point set are screened, thereby obtaining a pod point optimization set and a pod identification target point optimization set.

[0016] By using the pod to identify the target point set as the main body and conducting an initial screening of the real-time data set, it is possible to effectively identify and eliminate data points caused by errors, significantly improving the accuracy and reliability of data in subsequent processing, enhancing the positioning accuracy of the target point, and improving the intelligence level of the system.

[0017] In some technical solutions of the present invention, optionally, a pod point optimization set and a pod identification target point optimization set are screened to obtain a pod point advanced optimization set and a pod identification target point advanced optimization set, including: obtaining the centroid of the pod point optimization set; obtaining a second distance set based on each pod point in the pod point optimization set and the centroid of the pod point optimization set; obtaining one standard deviation of the second distance set based on the second distance set; eliminating pod points in the second distance set that exceed one standard deviation of the second distance set to obtain an ordered pod point set; based on the point identifier of the ordered pod point set, screening pod points and pod identification target points that have the same point identifier as the ordered pod point set to obtain an advanced optimization set of pod points and an advanced optimization set of pod identification target points.

[0018] In this technical solution, the real-time data set is secondary screened with the pod point set as the main body to obtain the centroid of the pod point optimization set. Based on each pod point in the pod point optimization set and the centroid of the pod point optimization set, a second distance set is obtained. On this basis, based on the second distance set, one times the standard deviation of the second distance set is obtained, and the pod points in the second distance set that exceed one times the standard deviation of the second distance set are eliminated to obtain an ordered pod point set. In addition, based on the point identifier of the ordered pod point set, that is, the point identifier of the ordered pod points in the ordered pod point set, the pod points and pod identification target points with the same point identifier as the ordered pod points in the ordered pod point set are screened, thereby obtaining an advanced optimization set of pod points and an advanced optimization set of pod identification target points.

[0019] By taking the pod point set as the main body and conducting a secondary screening of the real-time data set, we can effectively eliminate possible contradictions or inconsistencies in the data set, improve the consistency and credibility of the data, significantly improve the accuracy and reliability of the data in subsequent processing, enhance the positioning accuracy of the target point, and improve the intelligence level of the system.

[0020] In some technical solutions of the present invention, optionally, based on the pod point advanced optimization set and the pod identification target point advanced optimization set, selecting an alternative combination set includes: based on the pod point advanced optimization set, obtaining the convex hull of the pod point advanced optimization set; randomly selecting three different vertices from the vertices of the convex hull of the pod point advanced optimization set to form a triangle, and constructing a triangle set; eliminating the triangles that meet the first condition or the triangles that meet the second condition in the first triangle set, and obtaining the optimized triangle set; Based on the pod points in the optimized triangle set and the pod points in the optimized triangle set, the task identification set of the pod points is the same as the pod identification target point, and the third distance is obtained; the triangles whose difference between the third distance and the pod ranging value exceeds the error threshold are eliminated to obtain the initial combination set; based on the initial combination set, the second statistical information is obtained; the second statistical information is sorted in ascending order, and the initial combination set is sorted based on the sorting order; it is determined whether the number of triangles in the initial combination set is greater than the default number, and the set number is determined based on the judgment result; based on the sorting order, the same number of triangles as the set number in the initial combination set are selected to construct an alternative combination set; wherein, based on the point identification of the first triangle set, a second triangle set having a mapping relationship with the first triangle set is obtained; the triangle of the first condition is the first triangle in the first triangle set that has the same point identification as the second triangle in the second triangle set that has a side length greater than the maximum threshold; the triangle of the second condition is the first triangle in the first triangle set that has a side length less than the minimum threshold.

[0021] In this technical solution, based on the pod point advanced optimization set, the convex hull of the pod point advanced optimization set is obtained; based on the convex hull of the pod point advanced optimization set, a triangle set is constructed. First, even if the real-time data set has been screened twice, there may still be a large number of pod point advanced optimization sets. If any combination forms a triangle, the data processing time may increase. Therefore, the pod point advanced optimization set and the pod identification target point advanced optimization set select the smallest convex polygon that contains all, that is, the convex hull.

[0022] On this basis, every three points are formed into a group by permutation and combination in the convex hull to construct a triangle group. By constructing a convex hull and constructing a triangle set in the convex hull, the data processing time is reduced, and the calculation amount and time complexity of the algorithm are reduced. In addition, the first triangle whose side length is greater than the maximum threshold in the second triangle formed by the pod identification target point with the same point identifier as the first triangle in the first triangle set is eliminated, or the first triangle whose side length is less than the minimum threshold in the first triangle set is eliminated, and the optimized triangle set is obtained. The pod identification target point with the same point identifier as the pod point in the optimized triangle set is obtained based on these two points, and the triangle whose difference between the third distance and the pod ranging value exceeds the error threshold is eliminated to obtain the initial combination set. By screening the triangles in the triangle set, the points hovering near the same point of the pod, the points whose measured pod identification target points are not in the same position, and the points with ranging errors are eliminated, thereby improving the accuracy and reliability of the data and optimizing the data in the triangle set.

[0023] Furthermore, based on the initial combination set, the second statistical information is obtained, the second statistical information is sorted in ascending order, and the initial combination set is sorted based on the sorting order, it is determined whether the number of triangles in the initial combination set is greater than the default number, and the set number is determined based on the judgment result, and the same number of triangles as the set number in the initial combination set is selected based on the sorting order to construct an alternative combination set. By constructing the alternative combination set, triangles with a lower degree of internal angle dispersion are selected, and the amount of data processed is limited, unnecessary calculations and resource consumption are avoided, and data processing efficiency is improved.

[0024] In some technical solutions of the present invention, optionally, based on the first statistical information, the alternative combination set is positioned and solved to obtain a set of alternative target points, including: based on the alternative combination set, obtaining third statistical information; using the third statistical information as a parameter and the first statistical information as an iteration initial value, and performing positioning and solving in the alternative combination set according to the Newton iteration method and the least squares method to obtain a set of alternative target points.

[0025] In this technical solution, based on the candidate combination set, the third statistical information is obtained; the third statistical information is used as a parameter, and the first statistical information is used as the initial value of the iteration, and the positioning solution is performed in the candidate combination set according to the Newton iteration method and the least squares method to obtain the candidate target point set. The positioning solution is performed on the candidate combination set obtained after multiple screening by the Newton iteration method and the least squares method, which further improves the positioning accuracy, optimizes the calculation process, improves the calculation efficiency, and ensures the stability and reliability of the positioning settlement results.

[0026] In some technical solutions of the present invention, optionally, a set of alternative target points is screened to obtain the coordinates of the real target point, including: obtaining the center of mass of the set of alternative target points; obtaining a third distance set based on each solved target point of the set of alternative target points and the center of mass of the set of alternative target points; obtaining the mean of the third distance set based on the third distance set; selecting final data in the third distance set that is closest to the mean of the third distance set, and using the final data as the coordinates of the real target point.

[0027] In this technical solution, the centroid of the candidate target point set is obtained, and the third distance set is obtained based on each solution target point of the candidate target point set and the centroid of the candidate target point set. Based on the third distance set, the mean of the third distance set is obtained, and the final data closest to the mean of the third distance set is selected from the third distance set, and the final data is used as the coordinates of the real target point. After multiple screenings, the data with the smallest difference from the mean of the third distance set is used as the final data, thereby optimizing the positioning accuracy and improving the positioning accuracy and reliability.

[0028] In some technical solutions of the present invention, optionally, archiving and processing the first result data includes: storing the first result data; sending the first result data based on a communication protocol; and obtaining final calculation points based on a real-time data set.

[0029] In this technical solution, the final calculation points are obtained based on the real-time data set, and then when executing subsequent tasks, it can be determined whether to perform algorithm processing based on the final calculation points, avoiding unnecessary calculations, thereby optimizing resource utilization and improving the response speed of the system.

[0030] In some technical solutions of the present invention, optionally, before filtering the data in the real-time data set to obtain the first result data, it includes: determining whether the current final calculation points are consistent with the final calculation points at the end of the last task, and if they are consistent, skipping the step of filtering the data in the real-time data set; if they are inconsistent, filtering the data in the real-time data set.

[0031] In this technical solution, by judging and selecting to process a real-time data set or skipping the processing of a real-time data set, computing resources are optimized, the system is made more flexible, and the scalability of the system is improved.

[0032] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0034] Figure 1 A first flow chart of a method for improving positioning accuracy of a drone for an observation target according to an embodiment of the present invention;

[0035] Figure 2 A second flow chart of a method for improving positioning accuracy of a drone for an observation target according to an embodiment of the present invention;

[0036] Figure 3 A third flow chart of a method for improving positioning accuracy of a drone for an observation target according to an embodiment of the present invention;

[0037] Figure 4 A fourth flow chart of a method for improving positioning accuracy of a drone for an observation target according to an embodiment of the present invention;

[0038] Figure 5 A fifth flow chart of a method for improving positioning accuracy of a drone for an observation target according to an embodiment of the present invention;

[0039] Figure 6 The sixth flow chart of a method for improving the positioning accuracy of a drone for an observation target according to an embodiment of the present invention. DETAILED DESCRIPTION

[0040] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0041] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0042] Refer to the following Figures 1 to 6 A method for improving the positioning accuracy of a drone for an observation target according to some embodiments of the present invention is described.

[0043] like Figure 1 and Figure 6 As shown, the first aspect of the present invention provides a method for improving the positioning accuracy of a drone for an observation target, comprising:

[0044] Step 102, obtaining first data;

[0045] Step 104, verifying the first data and obtaining a first verification result;

[0046] Step 106, acquiring a real-time data set based on the first verification result;

[0047] Step 108, filtering the data in the real-time data set to obtain first result data;

[0048] Step 110: archive and process the first result data.

[0049] The present invention provides a method for improving the positioning accuracy of an unmanned aerial vehicle for an observation target, monitoring a receiving end of first data, and obtaining the first data based on the receiving end of the first data, verifying the obtained first data and obtaining the verification result of this verification, i.e., the first verification result, and further processing according to the first verification result to obtain a real-time data set, which is a set of data collected by the unmanned aerial vehicle when performing an observation mission, and filtering the data in the real-time data set to obtain the positioning coordinates of the unmanned aerial vehicle for the observation target, i.e., the first result data. By filtering the first result data obtained, it is possible to select a measurement point with the smallest error, and then obtain data with higher positioning accuracy, thereby improving the accuracy of the positioning data, avoiding the inefficient process of analyzing one by one in a large amount of data, and improving the measurement efficiency and positioning accuracy.

[0050] Specifically, the first data includes the longitude of the pod point, the latitude of the pod point, the altitude of the pod point, the distance value of the pod point to the target point, the time of pod point measurement, the longitude of the pod identification target point, the latitude of the pod identification target point and the altitude of the pod identification target point.

[0051] Specifically, the real-time data set includes the longitudes of multiple pod points, the latitudes of multiple pod points, the altitudes of multiple pod points, the ranging values ​​of multiple pod points to target points, the measurement time of multiple pod points, the longitudes of multiple pod identified target points, the latitudes of multiple pod identified target points and the altitudes of multiple pod identified target points in the same mission.

[0052] Specifically, the first result data is the coordinates of the real target point.

[0053] Optionally, before acquiring the first data, the process includes initializing a communication module, an initial point selection algorithm module, an initialization file operation module, resetting variables and a timer, etc.

[0054] Optionally, for the real-time data set, an algorithm based on point selection is called for processing, three appropriate points are selected, and positioning and solving are performed.

[0055] Optionally, after obtaining the first result data, the first result data is output in a format of the latitude of the target location, the longitude of the target location, the altitude of the target location and the time of the last state.

[0056] Furthermore, in some embodiments of the present invention, verifying the first data and obtaining a first verification result includes: filtering outliers in the first data; obtaining a first task identifier that matches the first data; and determining whether the first task identifier is consistent with a current task identifier to obtain a first verification result.

[0057] In this embodiment, since there may be abnormal values ​​in the first data obtained, the abnormal values ​​in the first data are first filtered, and after the filtering operation, the first task identifier matching the first data is obtained. The first data is obtained based on the acquisition when the drone performs the flight positioning task, and each time the drone performs a task, it has its own unique task identifier. Acquiring the first task identifier matching the first data can understand the data in which task the first data is obtained. Determine whether the first task identifier is consistent with the current task identifier, and then obtain the first verification result, wherein the current task identifier is the task identifier currently retained in the system. Determining whether the first task identifier is consistent with the current task identifier is to determine whether the flight mission of the drone is consistent with the task being processed by the system. By filtering abnormal values, invalid or missing information in the first data can be removed, the overall quality of the data can be improved, and errors and deviations in subsequent data processing and analysis can be reduced. In addition, by determining whether the first task identifier is consistent with the current task identifier, task confusion or execution errors can be prevented, the consistency of the task is guaranteed, and the stability and reliability of the entire system are improved.

[0058] It is understandable that a complete flight is considered as a mission, and each mission has a unique mission identifier. The present invention refers to a mission.

[0059] Specifically, the abnormal values ​​include the null value of the longitude of the pod point, the null value of the latitude of the pod point, the null value of the longitude of the pod identification target point, the null value of the latitude of the pod identification target point, the null value of the pod ranging value, the zero value of the longitude of the pod point, the zero value of the latitude of the pod point, the zero value of the longitude of the pod identification target point, the zero value of the latitude of the pod identification target point and the zero value of the pod ranging value. Any of the above values ​​appearing in the first data will be filtered.

[0060] Optionally, the first task identifier and the current task identifier may be selected from ID (abbreviation of Identifier or Identification, meaning identifier or identification code).

[0061] Further, in some embodiments of the present invention, based on the first verification result, a real-time data set is obtained, including: if the first verification result is that the first task identifier is consistent with the current task identifier, the first data is added to the real-time data set; if the first verification result is that the first task identifier is inconsistent with the current task identifier, the current task identifier is updated to the first task identifier, the first variable and the timer are reset, and the first data is added to the real-time data set.

[0062] In this embodiment, different processing is performed based on the first verification result. If the first verification result is that the first task identifier is consistent with the current task identifier, that is, the flight mission of the drone is consistent with the task being processed by the system, the system can directly perform operation processing on the first data, that is, add the first data to the real-time data set; if the first verification result is that the first task identifier is inconsistent with the current task identifier, that is, the flight mission of the drone is inconsistent with the task being processed by the system, the system obtains the data collected by the new flight mission of the drone, then updates the current task identifier to the first task identifier, resets the first variable and timer, and adds the first data to the real-time data set. By updating the current task identifier to the first task identifier, task confusion or execution errors are prevented, the consistency and continuity of the tasks are guaranteed, errors and conflicts are reduced, and the stability and reliability of the system are improved.

[0063] Specifically, the first variable includes a real-time data set and a final calculation point number. Resetting the first variable includes clearing the real-time data set and resetting the final calculation point number to zero.

[0064] Furthermore, if Figure 2 As shown, in some embodiments of the present invention, filtering data in a real-time data set to obtain first result data includes:

[0065] Step 202, dividing the real-time data set into a pod point set and a pod identification target point set;

[0066] Step 204, filtering the pod point set and the pod identification target point set to obtain the pod point optimization set and the pod identification target point optimization set;

[0067] Step 206, filtering the pod point optimization set and the pod identification target point optimization set, and obtaining the pod point advanced optimization set and the pod identification target point advanced optimization set;

[0068] Step 208, obtaining a candidate combination set based on the pod point advanced optimization set and the pod identification target point advanced optimization set;

[0069] Step 210, based on the advanced optimization set of pod recognition target points, obtaining first statistical information of the advanced optimization set of pod recognition target points;

[0070] Step 212: Based on the first statistical information, perform positioning calculation on the candidate combination set to obtain a candidate target point set;

[0071] Step 214, screening the candidate target point set to obtain the coordinates of the real target point.

[0072] In this embodiment, the real-time data set is a set of data collected by the drone when performing an observation mission. These data naturally include the data of the drone's pod point and the data of the pod identification target point. Then, the real-time data set can be classified based on the data of the pod point and the data of the pod identification target point, that is, the real-time data set is divided into a pod point set and a pod identification target point set. The pod point set is a set constructed by the data of the pod point, and the pod identification target point set is a set constructed by the data of the pod identification target point. First, the real-time data set is filtered for the first time. The first filtering is based on the pod identification target point set as the main body, and based on the one-to-one correspondence between the point identifier of the pod point and the point identifier of the pod identification, after filtering the pod identification target point set, the pod point set can also be screened based on the point identifier, that is, a screening is performed based on the pod identification target point set as the main body, and then the pod point optimization set and the pod identification target point optimization set are obtained. The second filtering is based on the filtered pod point set, that is, the pod point optimization set is used as the main body for filtering. Based on the one-to-one correspondence between the point identifier of the pod point and the point identifier of the pod identification, after filtering the pod point optimization set, the pod identification target point optimization set can also be filtered based on the point identifier, that is, the pod point optimization set is used as the main body for secondary filtering to obtain the pod point advanced optimization set and the pod identification target point advanced optimization set. By filtering the real-time data set twice, the data that does not meet the conditions or is of low value is initially eliminated, making the twice filtered data set more accurate and reliable, and reducing the complexity of data processing.

[0073] On this basis, based on the pod point advanced optimization set and the pod identification target point advanced optimization set, that is, the data set after the second screening, the data range is further narrowed to obtain the data set after the third screening, that is, the alternative combination set, and the first statistical information based on the pod identification target point advanced optimization set is used to locate and settle the alternative combination set to obtain the alternative target point set. After multiple screenings, the remaining data is finally screened to obtain the coordinates of the real target point. By screening the data multiple times, the range of data values ​​can be gradually narrowed, thereby achieving accurate positioning of the real target point, improving the speed and efficiency of data processing, reducing errors and uncertainties, improving the accuracy and efficiency of task execution, and ensuring the accuracy and reliability of the final data.

[0074] Specifically, the pod point set includes: the longitude of the pod point, the latitude of the pod point, the altitude of the pod point, the distance value of the pod point to the target point, and the time of measurement of the pod point.

[0075] Specifically, the pod identification target point set includes: the longitude of the pod identification target point, the latitude of the pod identification target point and the altitude of the pod identification target point. The pod is equipped with an identification module, and there is an error in the target identified when the identification module is working. The pod identification target point is the target point identified by the pod's identification module.

[0076] Specifically, the first statistical information includes: the mean in the pod identification target point advanced optimization set, the maximum in the pod identification target point advanced optimization set, the variance in the pod identification target point advanced optimization set, the standard deviation in the pod identification target point advanced optimization set and the mean of the three-dimensional coordinates in the pod identification target point advanced optimization set (i.e., the center of mass).

[0077] Specifically, each time a drone performs a flight mission, it has its own unique mission identifier, that is, a unique mission identifier. Based on the mission identifier, the mission currently performed by the drone can be confirmed to facilitate further understanding of the situation.

[0078] It is understandable that the data obtained by the drone when performing a flight mission, that is, the real-time data set, can be classified into pod points and pod identification target points, wherein both the pod points and the pod identification target points are provided with point identifiers, and the point identifiers of the pod points and the point identifiers of the pod identification target points are in a one-to-one mapping relationship. For example, if the first point identifier in the pod point set is 1, then the first point identifier in the pod target point set is also 1.

[0079] Furthermore, if Figure 3 As shown, in some embodiments of the present invention, screening the pod point set and the pod identification target point set to obtain the pod point optimization set and the pod identification target point optimization set includes:

[0080] Step 302, obtaining the centroid of the pod recognition target point set;

[0081] Step 304, acquiring a first distance set based on each pod identification target point of the pod identification target point set and the centroid of the pod identification target point set;

[0082] Step 306, based on the first distance set, obtaining one standard deviation of the first distance set;

[0083] Step 308, removing the pod identification target points in the first distance set that are more than one standard deviation of the first distance set, and obtaining an ordered pod identification target point set;

[0084] Step 310, based on the point identifier of the ordered pod identification target point set, pod points and pod identification target points having the same point identifier as the ordered pod identification target point set are screened to obtain a pod point optimization set and a pod identification target point optimization set.

[0085] In this embodiment, the real-time data set is initially screened with the pod identification target point set as the main body to obtain the centroid of the pod identification target point set (the centroid is obtained by the mean of the three-dimensional coordinates). On this basis, the first distance from each pod identification target point to the centroid is calculated, and the obtained first distances are integrated into a set, that is, based on each pod identification target point of the pod identification target point set and the centroid of the pod identification target point set, a first distance set is obtained.

[0086] On this basis, according to the data in the set, the mean, maximum and variance of the set can be calculated in turn, and finally one standard deviation of the set can be obtained, that is, based on the first distance set, one standard deviation of the first distance set can be obtained. Then, the pod identification target points whose first distance from the pod identification target point to the centroid exceeds one standard deviation of the first distance set are eliminated, that is, the pod identification target points in the first distance set that exceed one standard deviation of the first distance set are eliminated, and an ordered pod identification target point set is obtained, wherein the ordered pod identification target point set is the ID of the remaining points after the pod identification target point set is eliminated from the pod identification target point set from which the first distance from the pod identification target point to the centroid exceeds one standard deviation of the first distance set. In other words, the ordered pod identification target point set is the point identifier after the pod identification target point set is screened. Furthermore, based on the one-to-one correspondence between the point identifier of the pod point and the point identifier of the pod identification, based on the point identifier of the ordered pod identification target point set, the pod points and pod identification target points with the same point identifier as those in the ordered pod identification target point set are screened, and then the optimized set of pod points and the optimized set of pod identification target points are obtained, that is, the set of pod points after one screening and the set of pod identification target points after one screening are confirmed based on the point identifier.

[0087] By taking the pod identification target point set as the main body and performing the initial screening of the real-time data set, it is possible to effectively identify and eliminate the data points caused by errors, significantly improving the accuracy and reliability of the data in subsequent processing, enhancing the positioning accuracy of the target point, and improving the intelligence level of the system. It should be noted that in addition to each flight mission having a unique mission identifier, the points detected multiple times in a flight mission also have unique point identifiers. This application screens out the point identifiers and further screens the data corresponding to the point identifiers based on the point identifiers.

[0088] Specifically, the first distance set includes the distance from each pod identification target point to the mass center, which is obtained via the distance between each pod identification target point in the pod identification target point set and the mass center of the pod identification target point set.

[0089] Specifically, one times the standard deviation of the first distance set is gradually acquired through the mean of the first distance set, the maximum of the first distance set, and the variance of the first distance set.

[0090] Furthermore, if Figure 4 As shown, in some embodiments of the present invention, screening the pod point optimization set and the pod identification target point optimization set, obtaining the pod point advanced optimization set and the pod identification target point advanced optimization set, includes:

[0091] Step 402, obtaining the centroid of the optimized set of pod points;

[0092] Step 404, obtaining a second distance set based on each pod point of the pod point optimization set and the centroid of the pod point optimization set;

[0093] Step 406, based on the second distance set, obtaining one standard deviation of the second distance set;

[0094] Step 408, removing the pod points in the second distance set that exceed one standard deviation of the second distance set, and obtaining an ordered pod point set;

[0095] Step 410, based on the point identifier of the ordered pod point set, filter the pod points and pod identification target points with the same point identifier as the ordered pod point set, and obtain the pod point advanced optimization set and the pod identification target point advanced optimization set.

[0096] In this embodiment, the real-time data set is screened twice based on the pod point set to obtain the centroid of the pod point optimization set. On this basis, the first distance from each pod point to the centroid is calculated, and the obtained second distances are integrated into a set, that is, based on each pod point of the pod point optimization set and the centroid of the pod point optimization set, a second distance set is obtained.

[0097] On this basis, according to the data in the set, the mean, the maximum and the variance of the set can be calculated in turn, and finally one standard deviation of the set can be obtained, that is, based on the second distance set, one standard deviation of the second distance set can be obtained. Then, the pod points whose second distance from the pod point to the centroid exceeds one standard deviation of the second distance set are eliminated, that is, the pod points in the second distance set that exceed one standard deviation of the second distance set are eliminated, and an ordered pod point set is obtained, wherein the ordered pod point set is the point identification after the pod point optimization set is screened. And, based on the one-to-one correspondence between the point identification of the pod point and the point identification of the pod identification, based on the point identification of the ordered pod point set, that is, the point identification of the ordered pod points in the ordered pod point set, the pod points and the pod identification target points with the same point identification as the ordered pod points in the ordered pod point set are screened, and then the pod point advanced optimization set and the pod identification target point advanced optimization set are obtained, that is, the set of pod points after secondary screening based on the point identification and the set of pod identification target points after secondary screening.

[0098] By taking the pod point set as the main body and conducting a secondary screening of the real-time data set, we can effectively eliminate possible contradictions or inconsistencies in the data set, improve the consistency and credibility of the data, significantly improve the accuracy and reliability of the data in subsequent processing, enhance the positioning accuracy of the target point, and improve the intelligence level of the system.

[0099] Specifically, the second distance set includes the distance from each pod point to the centroid of the pod point optimization set, and is obtained via the distance between each pod point of the pod point set and the centroid.

[0100] Specifically, one times the standard deviation of the second distance set is gradually acquired through the mean of the second distance set, the maximum of the second distance set, and the variance of the second distance set.

[0101] Furthermore, if Figure 5 As shown, in some embodiments of the present invention, based on the pod point advanced optimization set and the pod identification target point advanced optimization set, an alternative combination set is selected, including:

[0102] Step 502, based on the advanced optimization set of pod points, obtaining the convex hull of the advanced optimization set of pod points;

[0103] Step 504, randomly select three different vertices from the vertices of the convex hull of the advanced optimization set of pod points and combine them into a triangle to construct a triangle set;

[0104] Step 506, removing triangles that meet the first condition or the second condition from the first triangle set, to obtain an optimized triangle set;

[0105] Step 508, obtaining a third distance based on a pod identification target point having the same point identifier as the pod point in the optimized triangle set and the pod point in the optimized triangle set;

[0106] Step 510, eliminating triangles whose difference between the third distance and the pod ranging value exceeds the error threshold, and obtaining an initial combination set;

[0107] Step 512, obtaining second statistical information based on the initial combination set;

[0108] Step 514, sorting the second statistical information in ascending order, and sorting the initial combination set based on the sorting order;

[0109] Step 516, determining whether the number of triangles in the initial combination set is greater than a default number, and determining a set number based on the determination result;

[0110] Step 518, based on the sorting order, select the triangles in the initial combination set that are equal to the set number to construct an alternative combination set;

[0111] Among them, based on the point identification of the first triangle set, a second triangle set having a mapping relationship with the first triangle set is obtained; the triangle of the first condition is the first triangle in the first triangle set that has the same point identification as the second triangle in the second triangle set whose side length is greater than the maximum threshold; the triangle of the second condition is the first triangle in the first triangle set whose side length is less than the minimum threshold.

[0112] In this embodiment, based on the pod point advanced optimization set, the convex hull of the pod point advanced optimization set is obtained; based on the convex hull of the pod point advanced optimization set, a triangle set is constructed. First, even if the real-time data set has been screened twice, there may still be a large number of pod point advanced optimization sets. If any combination forms a triangle, the data processing time may increase. Therefore, the pod point advanced optimization set and the pod identification target point advanced optimization set select the smallest convex polygon that contains all, that is, the convex hull.

[0113] When a drone performs a flight mission, it may measure the same target point at multiple different locations. In this case, the convex hull, which is already the smallest convex polygon, is further reduced, that is, every three points are grouped together in the convex hull using permutations and combinations to construct a triangle group. By constructing a convex hull and constructing a first triangle set in the convex hull, the data processing time is reduced, and the amount of calculation and time complexity of the algorithm are reduced. In addition, judging the distance between the two vertices of the triangle in the triangle set is the judgment of the screened pod points. If the distance between the two pod points is too close, it means that the drone's pod is hovering near the same point when performing a flight mission, and such points should be eliminated; and if the distance between the two pod recognition target points is too far, it means that the pod recognition target points are not the same point, so this obviously cannot be counted as a measurement of the same target point. Therefore, the triangles satisfying the first condition or the triangles satisfying the second condition in the first triangle set are eliminated to obtain an optimized triangle set, wherein, based on the point identification of the first triangle set, a second triangle set having a mapping relationship with the first triangle set is obtained; the triangle of the first condition is the first triangle in the first triangle set that has the same point identification as the second triangle in the second triangle set whose side length is greater than the maximum threshold; the triangle of the second condition is the first triangle in the first triangle set whose side length is less than the minimum threshold, to obtain an optimized triangle set, it should be noted that the first triangle set is composed of the first triangles constructed on the convex hull of the pod points after two screenings, and the second triangle set is mapped on the basis of the first triangle in the first triangle set based on the one-to-one correspondence between the point identification of the pod point and the point identification of the pod recognition.

[0114] Moreover, based on the one-to-one correspondence between the point identifier of the pod point and the point identifier of the pod identification, the pod identification target point corresponding to it can be obtained according to the optimized triangle set, and the distance between the two points can be obtained based on the coordinates of the two points, that is, the pod identification target point with the same point identifier as the pod point in the optimized triangle set and the pod point in the optimized triangle set, and the third distance is obtained based on these two points. If the difference between the distance between the two points and the ranging value of the pod point is too large, then there is a ranging gap at this point, and the triangle formed by this point should be eliminated, that is, the triangle whose difference between the third distance and the pod ranging value exceeds the error threshold is eliminated to obtain the initial combination set. By screening the triangles in the triangle set, points that wander around the same point of the pod, points whose pod identification target points are not in the same position, and points with ranging errors are eliminated, thereby improving the accuracy and reliability of the data and optimizing the data in the triangle set.

[0115] And, finally, based on the situation, how many values ​​are retained, based on the initial combination set, the second statistical information is obtained, the second statistical information is sorted in ascending order, and the initial combination set is sorted based on the sorting order, and it is determined whether the number of triangles in the initial combination set is greater than the default number, and the set number is determined based on the judgment result, and the same number of triangles as the set number in the initial combination set is selected based on the sorting order to construct an alternative combination set. By constructing the alternative combination set, triangles with lower internal angle discreteness are selected, and the amount of data processed is limited, unnecessary calculations and resource consumption are avoided, and data processing efficiency is improved.

[0116] It should be noted that there is no need to use the pod identification target point to construct the convex hull and then construct the triangle. The second triangle set is a set of second triangles obtained by mapping the vertices of the first triangle in the first triangle set, that is, the second triangles formed by the pod identification target points with the same point identifier as the secondary screened pod point, and the construction relationship between the second triangles is based on the first triangle constructed with the same point identifier as the pod point. If there is a first triangle composed of three different point identifiers in the first triangle set, then there will be a second triangle in the second triangle set with the point identifiers of the three vertices of the first triangle that are exactly the same.

[0117] It can be understood that, based on the fact that all pods are shooting at the target, the average value is considered to be closer to the real target, and the smaller the distance from the pod's identified target point to the center of mass, the closer it is to the real target and the more reliable the data; the farther the distance from the pod's identified target point to the center of mass, the farther it is from the real target and the less reliable the data.

[0118] It should be understood that, during a flight mission, the drone can obtain the distance measurement values ​​of the same observation target at at least three different positions, as well as the longitude, latitude and altitude of the pod at the time of measurement, to solve the observation target. Among them, when the geometric configuration of these three positions is closer to an equilateral triangle, the positioning accuracy is higher. The reason why the present invention selects data based on triangles is based on this principle.

[0119] Specifically, the second statistical information includes the three interior angles of the triangle, the mean, variance, standard deviation, etc. of the three interior angles. Among them, the initial combination set is sorted in ascending order according to the standard deviation. It can be understood that the standard deviation describes the degree of dispersion of the interior angles of the triangle. After the sorting set, the standard deviation is smaller the closer to the front, indicating that the difference between the interior angles of the triangle is smaller and the triangle is closer to a regular triangle. Based on the sorting order, selecting the same number of triangles as the set number in the initial combination set means selecting the set number of triangles with a lower degree of dispersion of the interior angles.

[0120] Furthermore, in some embodiments of the present invention, based on the first statistical information, the alternative combination set is positioned and solved to obtain a set of alternative target points, including: based on the alternative combination set, obtaining third statistical information; using the third statistical information as a parameter and the first statistical information as an iteration initial value, and according to the Newton iteration method and the least squares method, performing positioning and solving in the alternative combination set to obtain a set of alternative target points.

[0121] In this embodiment, after multiple screenings, the burden of the system is greatly reduced, the algorithm is optimized, and the remaining data is positioned and solved, that is, based on the alternative combination set, the third statistical information is obtained, the third statistical information is used as a parameter, the first statistical information is used as the initial value of the iteration, and the alternative combination set is positioned and solved according to the Newton iteration method and the least squares method to obtain the alternative target point set. The alternative combination set obtained after multiple screenings is positioned and solved by the Newton iteration method and the least squares method, which further improves the positioning accuracy, optimizes the calculation process, improves the calculation efficiency, and ensures the stability and reliability of the positioning settlement results.

[0122] Specifically, the third statistical information includes: the longitudes of the three vertices of the first triangle, the latitudes of the three vertices of the first triangle, the altitudes of the three vertices of the first triangle, and the ranging value of the laser rangefinder carried by the pod to the target.

[0123] Specifically, the first statistical information includes: the centroid of the advanced optimization set of pod identification target points, that is, summing up and averaging the three-dimensional coordinates of all pod identification target points that have been screened twice

[0124] Furthermore, in some embodiments of the present invention, screening a set of alternative target points to obtain the coordinates of the real target point includes: obtaining the centroid of the set of alternative target points; obtaining a third distance set based on each solved target point of the set of alternative target points and the centroid of the set of alternative target points; obtaining a mean of the third distance set based on the third distance set; selecting final data in the third distance set that is closest to the mean of the third distance set, and using the final data as the coordinates of the real target point.

[0125] In this embodiment, the centroid of the candidate target point set is obtained, and the third distance set is obtained based on each solution target point of the candidate target point set and the centroid of the candidate target point set. Based on the third distance set, the mean of the third distance set is obtained, and the final data closest to the mean of the third distance set is selected from the third distance set, that is, the coordinates of the solution target point in the candidate target point set that satisfies the distance to the centroid closest to the mean of the set are obtained, and this point is used as the final data, and the final data is used as the coordinates of the real target point. After multiple screenings, the data with the smallest difference from the mean of the third distance set is used as the final data, thereby optimizing the positioning accuracy, and improving the positioning accuracy and reliability.

[0126] Optionally, given the third distance set, the mean, maximum, variance, standard deviation, one times the standard deviation, etc. can also be obtained.

[0127] Furthermore, in some embodiments of the present invention, archiving and processing the first result data includes: storing the first result data; sending the first result data based on a communication protocol; and obtaining final calculation points based on a real-time data set.

[0128] In this embodiment, by obtaining the final calculation points based on the real-time data set, it is possible to determine whether to perform algorithm processing based on the final calculation points when executing subsequent tasks, thereby avoiding unnecessary calculations, thereby optimizing resource utilization and improving the response speed of the system.

[0129] Furthermore, in some embodiments of the present invention, before filtering the data in the real-time data set to obtain the first result data, it includes: determining whether the current final calculation points are consistent with the final calculation points at the end of the last task, if they are consistent, skipping the step of filtering the data in the real-time data set; if they are inconsistent, filtering the data in the real-time data set.

[0130] In this embodiment, by judging and selecting to process the real-time data set or skip the processing of the real-time data set, the computing resources are optimized, the system is made more flexible, and the scalability of the system is improved.

[0131] In the claims, specification and drawings of the present invention, the term "multiple" refers to two or more than two. Unless otherwise clearly defined, the orientation or positional relationship indicated by the terms "upper" and "lower" is based on the orientation or positional relationship shown in the drawings, which is only for the purpose of more conveniently describing the present invention and making the description process simpler, rather than indicating or implying that the device or element referred to must have the specific orientation described, be constructed and operated in a specific orientation, so these descriptions cannot be understood as limiting the present invention; the terms "connect", "install", "fix" and the like should be understood in a broad sense. For example, "connection" can be a fixed connection between multiple objects, or a detachable connection between multiple objects, or an integral connection; it can be a direct connection between multiple objects, or an indirect connection between multiple objects through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood based on the specific circumstances of the above data.

[0132] In the claims, specification and drawings of the present invention, the description of the terms "one embodiment", "some embodiments", "specific embodiments" and the like means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In the claims, specification and drawings of the present invention, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0133] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for improving the positioning accuracy of an unmanned aerial vehicle for an observation target, characterized in that: include: Acquire first data, wherein the first data is based on pod point information and pod identification target point information acquired when the UAV performs a flight positioning task; Verifying the first data and obtaining a first verification result; Based on the first verification result, a real-time data set is acquired, wherein the real-time data set includes a pod point set corresponding to the pod point information and a pod identification target point set corresponding to the pod identification target point information; Filtering the data in the real-time data set to obtain first result data, where the first result data is the coordinates of the real target point; Archiving and processing the first result data; The step of screening the data in the real-time data set to obtain first result data includes: The real-time data set is divided into a pod point set and a pod identification target point set; Filtering the pod point set and the pod identification target point set to obtain a pod point optimization set and a pod identification target point optimization set; Filter the pod point optimization set and the pod identification target point optimization set to obtain a pod point advanced optimization set and a pod identification target point advanced optimization set; Based on the pod point advanced optimization set and the pod identification target point advanced optimization set, obtaining an alternative combination set; Based on the pod identification target point advanced optimization set, obtaining first statistical information of the pod identification target point advanced optimization set; Based on the first statistical information, positioning and solving the candidate combination set are performed to obtain a candidate target point set; Filtering the candidate target point set to obtain the coordinates of the real target point; The screening of the pod point set and the pod identification target point set to obtain the pod point optimization set and the pod identification target point optimization set includes: Obtaining the centroid of the pod identification target point set; Acquire a first distance set based on each pod identification target point of the pod identification target point set and the centroid of the pod identification target point set; Based on the first distance set, obtaining one times the standard deviation of the first distance set; Eliminate pod identification target points in the first distance set that exceed one standard deviation of the first distance set, and obtain an ordered pod identification target point set; Based on the point identifier of the ordered pod identification target point set, pod points and pod identification target points having the same point identifier as the ordered pod identification target point set are screened to obtain the pod point optimization set and the pod identification target point optimization set.

2. The method for improving the positioning accuracy of an unmanned aerial vehicle for an observation target according to claim 1, characterized in that: The verifying the first data and obtaining a first verification result includes: Filtering abnormal values ​​in the first data, wherein the abnormal values ​​are null values ​​and zero values ​​of pod point information and pod identification target points; Acquire a first task identifier that matches the first data, wherein the first task identifier is a task identifier retained when the drone performs a task; It is determined whether the first task identifier is consistent with the current task identifier, and the first verification result is obtained.

3. The method for improving the positioning accuracy of an unmanned aerial vehicle for an observation target according to claim 2, characterized in that: The acquiring of a real-time data set based on the first verification result includes: If the first verification result is that the first task identifier is consistent with the current task identifier, then adding the first data to the real-time data set; If the first verification result is that the first task identifier is inconsistent with the current task identifier, updating the current task identifier to the first task identifier, resetting the first variable and the timer, and adding the first data to the real-time data set; Among them, the first variable includes the real-time data set and the final calculation points, and the current task identifier is the task identifier currently retained in the system.

4. The method for improving the positioning accuracy of an unmanned aerial vehicle for an observation target according to claim 1, characterized in that: The step of screening the pod point optimization set and the pod identification target point optimization set to obtain the pod point advanced optimization set and the pod identification target point advanced optimization set includes: Obtaining the centroid of the optimized set of pod points; Acquire a second distance set based on each pod point of the pod point optimization set and the centroid of the pod point optimization set; Based on the second distance set, obtaining one times the standard deviation of the second distance set; Eliminate the pod points in the second distance set that are greater than one standard deviation of the second distance set, and obtain an ordered pod point set; Based on the point identifier of the ordered pod point set, pod points and pod identification target points having the same point identifier as the ordered pod point set are screened to obtain the pod point advanced optimization set and the pod identification target point advanced optimization set.

5. The method for improving the positioning accuracy of an unmanned aerial vehicle for an observation target according to claim 1, characterized in that: The acquiring of a candidate combination set based on the pod point advanced optimization set and the pod identification target point advanced optimization set includes: Based on the advanced optimization set of pod points, obtaining the convex hull of the advanced optimization set of pod points; Select three different vertices from the vertices of the convex hull of the advanced optimization set of pod points and combine them into triangles to construct a first triangle set; Eliminate triangles that meet the first condition or the second condition in the first triangle set to obtain an optimized triangle set; Acquire a third distance based on a pod identification target point having the same point identifier as the pod point in the optimized triangle set and the pod point in the optimized triangle set; Eliminate triangles whose difference between the third distance and the pod ranging value exceeds an error threshold, and obtain an initial combination set; Based on the initial combination set, obtaining second statistical information; sorting the second statistical information in ascending order, and sorting the initial combination set based on the sorting order; Determining whether the number of triangles in the initial combination set is greater than a default number, and determining a set number based on the determination result; Based on the sorting order, selecting triangles in the initial combination set whose number is the same as the set number to construct the candidate combination set; Wherein, based on the point identifiers of the first triangle set, a second triangle set having a mapping relationship with the first triangle set is acquired; The triangle of the first condition is a first triangle in the first triangle set that has the same point identifier as a second triangle in the second triangle set whose side length is greater than the maximum threshold; The triangle of the second condition is a first triangle in the first triangle set whose side length is smaller than a minimum threshold.

6. The method for improving the positioning accuracy of an unmanned aerial vehicle for an observation target according to claim 1, characterized in that: The performing positioning and solving on the candidate combination set based on the first statistical information to obtain a candidate target point set includes: Based on the candidate combination set, obtaining third statistical information; The third statistical information is used as a parameter and the first statistical information is used as an iteration initial value. According to Newton iteration method and least square method, positioning solution is performed on the candidate combination set to obtain the candidate target point set.

7. The method for improving the positioning accuracy of an unmanned aerial vehicle for an observation target according to claim 1, characterized in that: The step of screening the candidate target point set and obtaining the coordinates of the real target point includes: Obtaining the centroid of the candidate target point set; Acquire a third distance set based on each calculated target point of the candidate target point set and the centroid of the candidate target point set; Based on the third distance set, obtaining a mean value of the third distance set; Select the final data in the third distance set that is closest to the mean of the third distance set, and use the final data as the coordinates of the real target point.

8. The method for improving the positioning accuracy of an unmanned aerial vehicle for an observation target according to claim 1, characterized in that: The archiving and processing of the first result data includes: storing the first result data; Based on the communication protocol, sending the first result data; Based on the real-time data set, final calculation points are obtained.

Images