A data visualization method and system for simulating flight environment

By processing flight data using the rotating door algorithm and the similarity weighting method, the problem that linear interpolation cannot capture the changing characteristics of missing data points is solved, achieving more accurate data filling and more accurate simulation of the flight environment.

CN118626777BActive Publication Date: 2024-11-08JINAN ANXUN TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411087379.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-09
Publication Date
2024-11-08
Estimated Expiration
2044-08-09

AI Technical Summary

Technical Problem

In existing technologies, when processing multiple consecutive missing data points using conventional linear interpolation, it is impossible to effectively capture the characteristics of numerical changes at the missing data points, causing the processing results to deviate from the true values ​​and affecting the accuracy of the simulated flight environment.

Method used

The rotating door algorithm is used to divide the flight data into multiple subsequences, and the missing data points are re-divided according to the changing trend of the subsequences. The filling value of the missing data points is determined by combining the similarity weighting method. The data is filled by the weighted result to generate processed flight data that is closer to the true value.

Benefits of technology

It improves the accuracy and reliability of the fill values ​​for missing data points, ensures the accuracy and practicality of the simulated flight environment, reduces the jumps in fill values, and improves the overall continuity and consistency of the processed flight data.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118626777B_ABST
    Figure CN118626777B_ABST
Patent Text Reader

Abstract

The present invention belongs to the field of data processing technology, and specifically relates to a data visualization method and system for simulating a flight environment, the method comprising: collecting real flight data, dividing the flight data into multiple subsequences through a revolving door algorithm and consecutive missing data points, re-dividing the multiple consecutive missing data points into subsequences, calculating the similarity between the subsequence with missing data points and other subsequences without missing data points, weighting the other subsequences without missing data points according to the similarity, determining the filling value of each missing data point in the subsequence, filling the value of the missing data point according to the filling value, obtaining processed flight data, simulating a flight environment according to the processed flight data, and performing data visualization. The present invention improves the accuracy and reliability of predicting the filling value of missing data points.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and more specifically, to a data visualization method and system for simulating a flight environment. Background Art

[0002] The simulated flight environment not only provides a safe and controllable learning platform for training and education, but also provides important support for aircraft design, development and system verification.

[0003] In the related technology, for example, the patent application document with application announcement number CN116923724A discloses a ground simulation flight test technology for aircraft maneuverable variable trajectory, which belongs to the technical field of aircraft ground simulation, including ground simulation system and digital twin system; the ground simulation system mainly includes a six-degree-of-freedom platform and an electronic control subsystem; the digital twin system mainly includes a modeling and simulation subsystem and a data management subsystem, based on the ground simulation flight test, giving full play to the advantages of modeling and simulation and digital twin, and strengthening the mutual integration of the above two technologies. Among them, modeling and simulation methods including aircraft geometric modeling, flow field numerical simulation and flight simulation provide basic preparation for ground simulation tests, and ground simulation tests provide aircraft with a semi-physical simulation environment that is closer to real working conditions.

[0004] In the related technologies, the flight environment is simulated based on the real flight data during the actual flight process and the geographical data of the flight area, ensuring the accuracy and practicality of the simulated flight environment, and effectively improving various applications and work efficiency in the aviation field.

[0005] Flight data is collected through sensors during the actual flight process. The sensors are affected by harsh environments such as vibration, impact, and high and low temperature conversions, which may result in the failure to record values ​​at multiple consecutive moments in a timely manner, causing continuous missing values ​​in the actual flight data. When processing multiple missing data points that appear continuously using conventional linear interpolation methods, it is impossible to capture the numerical change characteristics at the missing data points well, which may cause the missing data point processing results to deviate from the true values. Summary of the invention

[0006] In order to solve the technical problem that when the conventional linear interpolation method is used to process multiple missing data points that appear continuously, the numerical change characteristics at the missing data points cannot be well captured, which will cause the missing data point processing results to deviate from the true values, the present invention provides solutions in the following aspects.

[0007] In a first aspect, the present invention provides a data visualization method for simulating a flight environment, comprising: collecting real flight data, each data point in the flight data having a time and a value; dividing the flight data into a plurality of subsequences by a revolving door algorithm and consecutive missing data points; re-dividing the first A missing data points among the consecutive D missing data points into a previous subsequence, re-dividing the remaining missing data points into a later subsequence, and , B represents the number of data points in the previous subsequence, , , The angles between the two reference data points representing the 1st data point of the previous subsequence and the lines connecting the Bth, b+1th, and bth data points of the previous subsequence are respectively represented, and the difference between the values ​​of the two reference data points and the 1st data point at the same time and the 1st data point is equal to the preset parameter; for any subsequence with missing data points, the similarity between the subsequence and other subsequences without missing data points is calculated, and the other subsequences without missing data points are weighted according to the similarity, and the filling value of each missing data point in the subsequence is determined according to the weighted result; the values ​​of the missing data points are filled by the filling values ​​of the missing data points in the subsequence with missing data points to obtain the processed flight data; the flight environment is simulated according to the processed flight data, and the data is visualized.

[0008] The present invention divides the flight data, and re-divides the missing data points into each subsequence according to the change trend of each divided subsequence. By clarifying the subsequence to which the missing data point belongs, the numerical change characteristics at the missing data point can be better captured, so that the estimation of the subsequent filling value is more accurate and localized, the jump of the filling value is reduced, and the overall continuity and consistency of the processed flight data are improved; according to the similarity between the subsequence with missing data points and other subsequences without missing data points, the other subsequences without missing data points are weighted, and the filling value of each missing data point is determined according to the weighted result. The reference degree of the subsequence without missing data points to the filling value of the missing data point can be quantified through weighting, and then combined with the numerical change characteristics at the missing data point, the internal change characteristics and external similarity characteristics of the missing data point are integrated to determine the filling value of the missing data point, thereby improving the accuracy and reliability of predicting the filling value of the missing data point, and making the processed flight data closer to the real value, thereby ensuring the accuracy and practicality of simulating the flight environment according to the processed flight data.

[0009] Preferably, the method of dividing the flight data into multiple subsequences through a revolving door algorithm and consecutively occurring missing data points includes: obtaining a first straight line and a second straight line corresponding to each data point in the flight data and multiple starting data points in the flight data through a revolving door algorithm, during which, in response to the occurrence of consecutively occurring multiple missing data points, the first missing data point in the consecutively occurring multiple missing data points and the data point after the last missing data point are respectively used as starting data points; dividing the flight data into multiple subsequences according to the multiple starting data points, wherein the i-th subsequence includes the i-th starting data point and all data points between the i-th starting data point and the i+1-th starting data point.

[0010] The present invention obtains multiple starting data points in the flight data through a revolving door algorithm, divides the data points reflecting the change trend of a local area of ​​the flight data into a subsequence, and adds a condition for judging the starting data point, divides multiple missing data points that appear continuously into a subsequence, avoids the interference of the missing data points on the division of other normal data points, and makes the obtained subsequences more concentrated on the change trend of the local area.

[0011] Preferably, calculating the similarity between the subsequence and other subsequences without missing data points comprises: ; In the formula, represents the similarity between the subsequence and the rth subsequence without missing data points, represents the number of all data points in the subsequence, represents the number of all data points in the rth subsequence that does not have missing data points, represents an exponential function with a natural constant as base, It represents the first DTW distance between this subsequence and the rth subsequence without missing data points.

[0012] Preferably, the step of weighting other subsequences without missing data points according to similarity and determining the filling value of each missing data point in the subsequence according to the weighted result includes: in response to the ath data point of the subsequence being a missing data point, the filling value of the missing data point is Satisfies the expression: ; Where N represents the number of all subsequences without missing data points, , Respectively represent the similarity between the subsequence and the rth and nth subsequences without missing data points, Indicates the preset threshold value, Represents the value of the a-th data point in the r-th subsequence that has no missing data points.

[0013] The consistency of the changing trends of other subsequences without missing data points and this subsequence is reflected through the similarity between other subsequences without missing data points and this subsequence, and then the filling value of each missing data point in the subsequence is determined through the subsequence without missing data points that is relatively consistent with the changing trend of the subsequence, thereby improving the accuracy and reliability of predicting the filling value of the missing data point, and can make the processed flight data closer to the true value, thereby ensuring the accuracy and practicality of simulating the flight environment according to the processed flight data.

[0014] Preferably, calculating the similarity between the subsequence and other subsequences without missing data points comprises: ; In the formula, represents the similarity between the subsequence and the rth subsequence without missing data points, represents the number of all data points in the subsequence, represents the number of all data points in the rth subsequence that does not have missing data points, represents an exponential function with a natural constant as base, , , They respectively represent the first DTW distance, the second DTW distance and the third DTW distance between the subsequence and the rth subsequence without missing data points.

[0015] In addition to the numerical values ​​of all data points in the subsequence being able to reflect the change trend of the subsequence, the slope sequence of the first straight line of the subsequence and the slope sequence of the second straight line of the subsequence can also reflect the change trend of the subsequence. Therefore, the similarity between the subsequence and other subsequences without missing data points is calculated through the first DTW distance, the second DTW distance and the third DTW distance between the subsequence and the r-th subsequence without missing data points. The obtained similarity can more accurately reflect the consistency of the change trend between the subsequence and the r-th subsequence without missing data points.

[0016] Preferably, the step of weighting other subsequences without missing data points according to similarity and determining the filling value of each missing data point in the subsequence according to the weighted result includes: in response to the ath data point of the subsequence being a missing data point, the filling value of the missing data point is Equal to the first predicted value for the missing data point and the second predicted value The mean of .

[0017] Preferably, the first predicted value of the missing data point and the second predicted value Satisfies the expression: ; ; Where N represents the number of all subsequences without missing data points, , Respectively represent the similarity between the subsequence and the rth and nth subsequences without missing data points, Indicates the preset threshold value, , They represent the slopes of the first and second straight lines corresponding to the a-th data point of the r-th subsequence without missing data points, represents the moment of the missing data point, , , , Respectively represent the time and value of the first reference data point and the second reference data point of the starting data point of the subsequence, wherein the first reference data point and the second reference data have the same time as the starting data point and the difference with the value of the starting data point is equal to the preset parameter, and the value of the first reference point is larger than that of the second reference point.

[0018] Represents the predicted slope of the first straight line corresponding to the missing data point, using the predicted slope and the coordinates of the first reference data point , to calculate the first predicted value of the missing data point; Represents the predicted slope of the second straight line corresponding to the missing data point, using the predicted slope and the coordinates of the second reference data point , to calculate the second predicted value of the missing data point; fully consider the various change characteristics at the missing data point to determine the filling value of the missing data point, improve the accuracy and reliability of predicting the filling value of the missing data point, and make the processed flight data closer to the true value, thereby ensuring the accuracy and practicality of simulating the flight environment according to the processed flight data.

[0019] Preferably, it also includes: in response to the first data point of a subsequence with missing data points being a missing data point, calculating by a dynamic time warping algorithm the DTW distance between the reverse result of the subsequence and the reverse result of the r-th subsequence without missing data points, the DTW distance between the reverse result of the slope sequence of the first straight line of the subsequence and the reverse result of the slope sequence of the first straight line of the r-th subsequence without missing data points, and the DTW distance between the reverse result of the slope sequence of the second straight line of the subsequence and the reverse result of the slope sequence of the second straight line of the r-th subsequence without missing data points, as the first DTW distance, the second DTW distance and the third DTW distance between the subsequence and the r-th subsequence without missing data points, respectively.

[0020] Preferably, the method of re-dividing the first A missing data points among the D consecutive missing data points into the previous subsequence and re-dividing the remaining missing data points into the next subsequence also includes: using the starting data point of the previous subsequence before re-dividing as the starting data point of the previous subsequence after re-dividing; using the starting data point of the next subsequence before re-dividing as the starting data point of the next subsequence after re-dividing.

[0021] In a second aspect, the present invention provides a data visualization system for simulating a flight environment, which adopts the following technical solution:

[0022] A data visualization system for simulating a flight environment comprises: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned data visualization method for simulating a flight environment is implemented.

[0023] By adopting the above technical solution, the above-mentioned data visualization method for simulating a flight environment is generated into a computer program and stored in a memory so as to be loaded and executed by a processor, thereby making a terminal device based on the memory and the processor for easy use.

[0024] The beneficial effects of the present invention are:

[0025] The present invention divides the flight data, and re-divides the missing data points into each subsequence according to the change trend of each divided subsequence. By clarifying the subsequence to which the missing data point belongs, the numerical change characteristics at the missing data point can be better captured, so that the estimation of the subsequent filling value is more accurate and localized, the jump of the filling value is reduced, and the overall continuity and consistency of the processed flight data are improved; according to the similarity between the subsequence with missing data points and other subsequences without missing data points, the other subsequences without missing data points are weighted, and the filling value of each missing data point is determined according to the weighted result. The reference degree of the subsequence without missing data points to the filling value of the missing data point can be quantified through weighting, and then combined with the numerical change characteristics at the missing data point, the internal change characteristics and external similarity characteristics of the missing data point are integrated to determine the filling value of the missing data point, thereby improving the accuracy and reliability of predicting the filling value of the missing data point, and making the processed flight data closer to the real value, thereby ensuring the accuracy and practicality of simulating the flight environment according to the processed flight data. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The above and other purposes, features and advantages of the exemplary embodiments of the present invention will become readily understood by reading the detailed description below through the accompanying drawings. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0027] Figure 1 is a flow chart schematically illustrating a data visualization method for simulating a flight environment in the present invention;

[0028] Figure 2 Schematically shows the first reference point and the second reference point of the starting data point and the first straight line and the second straight line corresponding to each data point. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0030] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0031] The embodiment of the present invention discloses a data visualization method for simulating a flight environment, referring to Figure 1 , comprising steps S1 to S3:

[0032] S1. Collect real flight data.

[0033] It should be noted that the geographical data of the flight area can be obtained based on the map, while the flight data is collected by sensors during the actual flight.

[0034] Specifically, during the actual flight process, real flight data is collected through sensors and normalized. The collected flight data includes the dynamic characteristics data of the aircraft and the environmental data of the aircraft.

[0035] Among them, the dynamic characteristic data of the aircraft include but are not limited to airspeed, magnetic heading, three-axis angular velocity (referring to the angular changes of the aircraft in the pitch, roll and yaw directions), acceleration and three-dimensional position coordinates (referring to the longitude, latitude and altitude of the aircraft's location), and the environmental data of the aircraft include but are not limited to the atmospheric temperature, air pressure, humidity, wind speed, wind direction, cloud height and visibility; the flight data includes multiple data points, each of which has two dimensions: time and value.

[0036] Among them, real flight data is collected through sensors, including: measuring real-time airspeed, magnetic heading, three-axis angular velocity, acceleration, three-dimensional position coordinates, temperature, air pressure, humidity, cloud height and visibility through airspeed sensor, gyroscope sensor, acceleration sensor, magnetometer sensor, GPS sensor, temperature sensor, air pressure sensor, humidity sensor, lidar altimeter and visibility sensor; measuring wind speed and direction through wind speed and direction sensors.

[0037] S2. Divide the flight data into multiple subsequences using the revolving door algorithm and consecutive missing data points; re-divide the consecutive missing data points into subsequences, calculate the similarity between the subsequence with missing data points and other subsequences without missing data points, weight the other subsequences without missing data points according to the similarity, determine the filling value of each missing data point in the subsequence, fill the missing data points according to the filling value, and obtain the processed flight data.

[0038] It should be noted that flight data is collected through sensors during the actual flight process. The sensors are affected by harsh environments such as vibration, impact, and high and low temperature conversions, which will result in the failure to record the values ​​at multiple consecutive moments in a timely manner, causing continuous missing values ​​in the actual flight data. When processing multiple missing data points that appear continuously through conventional linear interpolation methods, it is impossible to capture the numerical change characteristics at the missing data points well, which will cause the missing data point processing results to deviate from the true values.

[0039] It should be further explained that in order to make the processing result of missing data points of flight data close to the true value, the present invention divides the flight data, and re-divides the missing data points into each subsequence according to the change trend of each divided subsequence, and according to the similarity between the subsequence with missing data points and other subsequences without missing data points, weights the other subsequences without missing data points, and determines the filling value of each missing data point in the subsequence with missing data points according to the weighted result, thereby obtaining the processed flight data.

[0040] Step S2 includes steps S201 to S204:

[0041] S201. Divide the flight data into multiple subsequences by using the revolving door algorithm and consecutive missing data points.

[0042] It should be noted that, when the conventional linear interpolation method is used to process multiple missing data points that appear continuously, it is impossible to capture the numerical change characteristics of the missing data points well. Therefore, the present invention first divides the flight data into multiple sub-sequences through a revolving door algorithm according to the distribution of normal data points, and reflects the change trend of the local area of ​​the flight data through each sub-sequence.

[0043] Specifically, a plurality of starting data points in the flight data are obtained through a spinning door transformation (SDT), wherein the method for obtaining the i+1th starting data point is as follows: for any target data point from the 1st to the m-1th data points after the i-th starting data point, it is required that the slope of the first straight line corresponding to the target data point is greater than the slope of the second straight line. At this time, when the slope of the first straight line corresponding to the m-th data point is not less than the slope of the second straight line, the m-th data point is used as the i+1th starting data point.

[0044] The revolving door algorithm is a well-known technology and will not be described in detail here.

[0045] It should be noted that as long as the slope of the first straight line corresponding to the m-1 data point after the ith starting data point is greater than the slope of the second straight line, it means that the change trends between the ith starting data point to the m-1 data point are relatively consistent. Therefore, continue to judge the change trend at the next m-th data point. When the slope of the first straight line corresponding to the m-th data point is less than or equal to the slope of the second straight line, it means that the previous change trend has changed at the m-th data point. Therefore, take the m-th data point as the i+1-th starting data point to judge the new change trend.

[0046] It should be further explained that the above-mentioned obtaining of multiple starting data points in the flight data through the revolving door algorithm is a well-known technology in the revolving door algorithm. Initially, the revolving door algorithm will use the first data point in the flight data as the first starting data point.

[0047] Among them, the method for obtaining the first straight line corresponding to the mth data point is: in response to the slope of the line connecting the first reference point of the ith starting data point and the mth data point being not less than the first straight line corresponding to the m-1th data point, the line connecting the first reference point of the ith starting data point and the mth data point is used as the first straight line corresponding to the mth data point; otherwise, the first straight line corresponding to the m-1th data point is used as the first straight line corresponding to the mth data point.

[0048] Among them, the method for obtaining the second straight line corresponding to the mth data point is: in response to the slope of the line connecting the second reference point of the ith starting data point and the mth data point being no greater than the second straight line corresponding to the m-1th data point, the line connecting the second reference point of the ith starting data point and the mth data point is used as the second straight line corresponding to the mth data point; otherwise, the second straight line corresponding to the m-1th data point is used as the second straight line corresponding to the mth data point.

[0049] Among them, the method for obtaining the first reference point and the second reference point of the i-th starting data point is: take two data points that are at the same time as the i-th starting data point and whose numerical difference with the starting data point is equal to the preset parameter as the two reference data points of the i-th starting data point, and record them as the first reference point and the second reference point of the i-th starting data point respectively, and the first reference point has a larger numerical value than the second reference point.

[0050] For example, see Figure 2 , which shows the first reference point and the second reference point of the starting data point and the first straight line and the second straight line corresponding to the data point, wherein the starting data point of the first subsequence is the data point with coordinates (t0, z0), and the starting data point of the second subsequence is the data point with coordinates (t6, z6).

[0051] The specific value of the preset parameter can be set according to the actual application scenario and requirements, and the value range of the preset parameter is [0.01, 0.2]. The present invention sets the preset parameter to 0.1.

[0052] It should be noted that, since there are missing data points in the flight data of the present invention, and the missing data points cannot be directly divided by the revolving door algorithm, a condition for determining the starting data point is added.

[0053] Specifically, in the process of obtaining multiple starting data points in the flight data through the revolving door algorithm, in response to the occurrence of multiple missing data points that appear consecutively, the first missing data point among the multiple missing data points that appear consecutively and the next data point after the multiple missing data points that appear consecutively are respectively used as starting data points.

[0054] It should be noted that, by taking the first missing data point among multiple missing data points that appear continuously and the next data point after the multiple missing data points that appear continuously as starting data points, the multiple missing data points that appear continuously can be divided into a subsequence, thus avoiding interference with the division of other normal data points.

[0055] Furthermore, the flight data is divided into a plurality of subsequences through all starting data points, wherein the i-th starting data point and all data points between the i-th starting data point and the i+1-th starting data point constitute the i-th subsequence.

[0056] It should be noted that the present invention obtains multiple starting data points in the flight data through a revolving door algorithm, divides the data points reflecting the changing trend of the local area of ​​the flight data into a subsequence, and adds a condition for judging the starting data point, and divides multiple missing data points that appear continuously into a subsequence, thereby avoiding the interference of missing data points on the division of other normal data points, so that the obtained subsequences are more concentrated on the changing trend of the local area.

[0057] S202. Re-divide multiple missing data points that appear consecutively into subsequences.

[0058] It should be noted that the missing data points are re-divided into each subsequence according to the change trend of each divided subsequence, so that the missing data points belonging to the subsequence can be filled in according to the change trend of each subsequence, so that the missing data point processing result is closer to the true value.

[0059] Specifically, for multiple missing data points that appear consecutively, the first A missing data points among the D missing data points that appear consecutively are re-divided into the previous subsequence, and the remaining missing data points are re-divided into the next subsequence. At the same time, the starting data point of the previous subsequence before the re-division is used as the starting data point of the previous subsequence after the re-division; the starting data point of the next subsequence before the re-division is used as the starting data point of the next subsequence after the re-division.

[0060] Where D represents the number of missing data points that appear consecutively, and A represents the number of missing data points that are divided into the previous subsequence, which satisfies the expression:

[0061] ;

[0062] In the formula, B represents the number of data points in the previous subsequence, The angle between the two reference data points of the first data point of the previous subsequence and the line connecting the Bth data point of the previous subsequence, that is, the angle between the two reference data points of the starting data point of the previous subsequence and the line connecting the last data point of the previous subsequence, The angle between the two reference data points of the first data point of the previous subsequence and the line connecting the b+1th data point of the previous subsequence, that is, the angle between the two reference data points of the starting data point of the previous subsequence and the line connecting the last data point of the previous subsequence. It represents the angle between the two reference data points of the first data point of the previous subsequence and the line connecting the bth data point of the previous subsequence, that is, the angle between the two reference data points of the starting data point of the previous subsequence and the line connecting the last data point of the previous subsequence.

[0063] in, , and Satisfies the expression:

[0064] ;

[0065] ;

[0066] ;

[0067] In the formula, , Respectively represent the slopes of the lines connecting the two reference data points of the 1st data point of the previous subsequence and the Bth data point of the previous subsequence, , Respectively represent the slopes of the lines connecting the two reference data points of the 1st data point of the previous subsequence and the b+1th data point of the previous subsequence, , The slopes of the lines connecting the two reference data points of the 1st data point of the previous subsequence and the bth data point of the previous subsequence are respectively represented.

[0068] It should be noted that when the difference between the slopes of the line connecting the two reference data points of the first data point of the previous subsequence and the Bth data point of the previous subsequence is greater than 0, that is, when the slope of the line connecting the first reference data point of the first data point and the Bth data point is greater than the slope of the line connecting the second reference data point of the first data point and the Bth data point, it means that the change trend between the first data point to the Bth data point of the previous subsequence is relatively consistent, and the difference in slope is The larger is, the longer the change trend can be maintained, and the more data points after the Bth data point conform to the change trend. Therefore, when multiple missing data points that appear consecutively are repartitioned, more missing data points are repartitioned into the previous subsequence.

[0069] It should be further explained that the present invention divides the flight data and re-divides the missing data points into subsequences according to the change trends of the divided subsequences. By clarifying the subsequence to which the missing data points belong, the numerical change characteristics of the missing data points can be better captured, so that the subsequent filling value estimation is more accurate and localized, the jump of the filling value is reduced, and the overall continuity and consistency of the processed flight data are improved.

[0070] S203. Calculate the similarity between the subsequence with missing data points and other subsequences without missing data points, weight the other subsequences without missing data points according to the similarity, and determine the filling value of each missing data point in the subsequence.

[0071] It should be noted that, for a subsequence with missing data points, the consistency of the changing trends of other subsequences without missing data points and the subsequence can be reflected by calculating the similarity between the subsequence and other subsequences without missing data points, and then the filling value of each missing data point in the subsequence can be determined by the subsequence without missing data points whose changing trend is relatively consistent with that of the subsequence.

[0072] Specifically, for any subsequence with missing data points, the similarity between the subsequence and other subsequences without missing data points is calculated, the other subsequences without missing data points are weighted according to the similarity, and the filling value of each missing data point in the subsequence is determined according to the weighted result.

[0073] It should be noted that, according to the similarity between the subsequence with missing data points and other subsequences without missing data points, the other subsequences without missing data points are weighted, and the filling value of each missing data point is determined according to the weighted result. The weighting can quantify the reference degree of the filling value of the missing data point for the subsequence without missing data points. Combined with the numerical change characteristics of the missing data points, the internal change characteristics and external similarity characteristics of the missing data points are integrated to determine the filling value of the missing data points, thereby improving the accuracy and reliability of predicting the filling value of the missing data points, and can make the processed flight data closer to the true value, thereby ensuring the accuracy and practicality of simulating the flight environment according to the processed flight data.

[0074] In one embodiment, the similarity between the subsequence and the rth subsequence without missing data points satisfies the expression:

[0075] ;

[0076] In the formula, represents the similarity between the subsequence and the rth subsequence without missing data points, represents the number of all data points in the subsequence, represents the number of all data points in the rth subsequence that does not have missing data points, represents an exponential function with a natural constant as base, It represents the first DTW distance between this subsequence and the rth subsequence without missing data points.

[0077] It should be noted that when the r-th subsequence without missing data points is shorter than the subsequence, the similarity between the subsequence and the r-th subsequence without missing data points is 0; when the r-th subsequence without missing data points is longer than the subsequence, the closer the length of the r-th subsequence without missing data points is to the subsequence, the more similar the subsequence and the r-th subsequence without missing data points are. In the above formula, ,therefore, ,but The smaller, The larger the , the similarity between the corresponding subsequence and the rth subsequence without missing data points The larger the value, the smaller the first DTW distance between the subsequence and the r-th subsequence without missing data points, and the more similar the r-th subsequence without missing data points is to the subsequence. In the above formula, due to ,therefore, The smaller the value, the similarity between the corresponding subsequence and the rth subsequence without missing data points. The bigger.

[0078] Wherein, when obtaining the first DTW distance between the subsequence and the r-th subsequence without missing data points, in response to the last data point of the subsequence being a missing data point, the DTW distance between the subsequence and the r-th subsequence without missing data points is calculated through a dynamic time warping (Dynamic TimeWarping, DTW) algorithm as the first DTW distance between the subsequence and the r-th subsequence without missing data points; in response to the first data point of the subsequence being a missing data point, the DTW distance between the reverse result of the subsequence and the reverse result of the r-th subsequence without missing data points is calculated through a dynamic time warping algorithm as the first DTW distance between the subsequence and the r-th subsequence without missing data points.

[0079] The subsequences without missing data points are weighted according to the similarity, and the filling value of each missing data point in the subsequence is determined according to the weighted result. Then, the filling value of the a-th missing data point in the subsequence satisfies the expression:

[0080] ;

[0081] In the formula, represents the filling value of the ath missing data point in the subsequence, N represents the number of all subsequences without missing data points, , Respectively represent the similarity between the subsequence and the rth and nth subsequences without missing data points, Indicates the preset threshold value, represents the value of the ath data point in the rth non-missing data point, Represents the maximum value function.

[0082] It should be noted that the greater the similarity between the subsequence and the r-th subsequence without missing data points, the more meaningful the change trend of the r-th subsequence without missing data points is for determining the change trend of the subsequence. As the weight of the ath data point among the rth data points that do not have missing data points, the value of the ath data point among the rth data points that do not have missing data points Weighted, where Used for Normalize.

[0083] The specific value of the preset threshold can be set according to the actual application scenario and requirements, and the value range of the preset threshold is [0.5, 1). The present invention sets the preset threshold to 0.6.

[0084] It should be noted that, for any subsequence, in addition to the numerical values ​​of all data points in the subsequence being able to reflect the changing trend of the subsequence, the slope sequence of the first straight line of the subsequence and the slope sequence of the second straight line of the subsequence can also reflect the changing trend of the subsequence.

[0085] In another embodiment, the similarity between the subsequence and the rth subsequence without missing data points satisfies the expression:

[0086] ;

[0087] In the formula, represents the similarity between the subsequence and the rth subsequence without missing data points, represents the number of all data points in the subsequence, represents the number of all data points in the rth subsequence that does not have missing data points, represents an exponential function with a natural constant as base, represents the first DTW distance between this subsequence and the rth subsequence without missing data points, represents the second DTW distance between this subsequence and the rth subsequence without missing data points, It represents the third DTW distance between the subsequence and the rth subsequence without missing data points.

[0088] It should be noted that when the r-th subsequence without missing data points is shorter than the subsequence, the similarity between the subsequence and the r-th subsequence without missing data points is 0; when the r-th subsequence without missing data points is longer than the subsequence, the closer the length of the r-th subsequence without missing data points is to the subsequence, the more similar the subsequence and the r-th subsequence without missing data points are. In the above formula, ,therefore, ,but The smaller, The larger the , the similarity between the corresponding subsequence and the rth subsequence without missing data points The larger the value, the smaller the first DTW distance, the second DTW distance, and the third DTW distance between the subsequence and the r-th subsequence without missing data points. The more similar the r-th subsequence without missing data points is to the subsequence. In the above formula, due to ,therefore, The smaller the value, the similarity between the corresponding subsequence and the rth subsequence without missing data points. The bigger.

[0089] It should be further explained that, in this embodiment, the first DTW distance, the second DTW distance and the third DTW distance between the subsequence and the r-th subsequence without missing data points are combined to calculate the similarity between the subsequence and the r-th subsequence without missing data points. The obtained similarity can more accurately reflect the consistency of the changing trend between the subsequence and the r-th subsequence without missing data points.

[0090] Wherein, when obtaining the second DTW distance and the third DTW distance between the subsequence and the r-th subsequence without missing data points, in response to the last data point of the subsequence being a missing data point, the DTW distance between the slope sequence of the first straight line of the subsequence and the slope sequence of the first straight line of the r-th subsequence without missing data points and the DTW distance between the slope sequence of the second straight line of the subsequence and the slope sequence of the second straight line of the r-th subsequence without missing data points are calculated by a dynamic time warping algorithm, as the second DTW distance and the third DTW distance between the subsequence and the r-th subsequence without missing data points, respectively; in response to The first data point of the subsequence is a missing data point. The DTW distance between the reverse result of the slope sequence of the first straight line of the subsequence and the reverse result of the slope sequence of the first straight line of the r-th subsequence without missing data points is calculated by the dynamic time warping algorithm as the second DTW distance between the subsequence and the r-th subsequence without missing data points. The DTW distance between the reverse result of the slope sequence of the second straight line of the subsequence and the reverse result of the slope sequence of the second straight line of the r-th subsequence without missing data points is calculated by the dynamic time warping algorithm as the third DTW distance between the subsequence and the r-th subsequence without missing data points.

[0091] The slope sequence of the first straight line of the subsequence refers to a sequence composed of the slopes of the first straight line of all data points in the subsequence, and the slope sequence of the second straight line of the subsequence refers to a sequence composed of the slopes of the second straight line of all data points in the subsequence.

[0092] The method weights other subsequences without missing data points according to similarity, and determines the filling value of each missing data point in the subsequence according to the weighted result, including: in response to the ath data point of the subsequence being a missing data point, calculating the first predicted value and the second predicted value of the missing data point, and and the second predicted value The mean of the missing data points is used as the filling value of .

[0093] Wherein, the first predicted value of the missing data point and the second predicted value Satisfies the expression:

[0094] ;

[0095] ;

[0096] In the formula, represents the first predicted value of the missing data point, represents the second predicted value of the missing data point, N represents the number of subsequences without missing data points, , Respectively represent the similarity between the subsequence and the rth and nth subsequences without missing data points, Indicates the preset threshold value, , They represent the slopes of the first and second straight lines corresponding to the a-th data point of the r-th subsequence without missing data points, represents the moment of the missing data point, , Respectively represent the time and value of the first reference data point of the starting data point of the subsequence, , Respectively represent the time and value of the second reference data point of the starting data point of the subsequence.

[0097] It should be noted that Represents the predicted slope of the first straight line corresponding to the missing data point, using the predicted slope and the coordinates of the first reference data point , to calculate the first predicted value of the missing data point; Represents the predicted slope of the second straight line corresponding to the missing data point, using the predicted slope and the coordinates of the second reference data point , to calculate the second predicted value of the missing data point; fully consider the various change characteristics at the missing data point to determine the filling value of the missing data point, improve the accuracy and reliability of predicting the filling value of the missing data point, and make the processed flight data closer to the true value, thereby ensuring the accuracy and practicality of simulating the flight environment according to the processed flight data.

[0098] S204. Fill the missing data points with the filling values ​​of the missing data points in the subsequence where the missing data points exist, to obtain processed flight data.

[0099] S3. Simulate the flight environment based on the processed flight data and perform data visualization.

[0100] Specifically, the geographical data of the flight area and the processed flight data are input into the Gazebo simulation software. The Gazebo simulation software outputs a simulated virtual flight environment, and the simulated virtual flight environment is displayed in the form of a picture through the Rviz data visualization software.

[0101] Among them, Gazebo simulation software allows users to customize realistic simulated virtual environments; Rviz data visualization software can visualize navigation, point clouds, motion planning, etc., and is a comprehensive data visualization platform; Gazebo simulation software and Rviz data visualization software are both well-known technologies and will not be described in detail here.

[0102] An embodiment of the present invention further discloses a data visualization system for simulating a flight environment, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a data visualization method for simulating a flight environment according to the present invention is implemented.

[0103] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.

[0104] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0105] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A data visualization method for simulating flight environment, It is characterized in that include: Collecting real flight data, each data point in the flight data has a time and a value; The flight data are divided into multiple subsequences by the revolving door algorithm and the consecutively appearing missing data points; The first A missing data points among the D consecutive missing data points are re-divided into the previous subsequence, and the remaining missing data points are re-divided into the next subsequence, and , B represents the number of data points in the previous subsequence, , , The angles between the two reference data points representing the first data point of the previous subsequence and the lines connecting the B, b+1, and b data points of the previous subsequence, respectively, are the values ​​of the two reference data points at the same time as the first data point and the difference between the values ​​of the two reference data points and the first data point is equal to the preset parameter; For any subsequence with missing data points, calculate the similarity between the subsequence and other subsequences without missing data points, weight the other subsequences without missing data points according to the similarity, and determine the filling value of each missing data point in the subsequence according to the weighted result; fill the missing data point with the filling value of the missing data point in the subsequence with missing data points to obtain the processed flight data; The flight environment is simulated based on the processed flight data, and data visualization is performed.

2. The data visualization method for simulating a flight environment according to claim 1, It is characterized in that The flight data is divided into multiple subsequences by the revolving door algorithm and the consecutive missing data points, including: A first straight line and a second straight line corresponding to each data point in the flight data and a plurality of starting data points in the flight data are obtained by a revolving door algorithm. In this process, in response to the occurrence of a plurality of consecutive missing data points, a first missing data point and a data point after a last missing data point in the plurality of consecutive missing data points are used as starting data points respectively; Dividing the flight data into a plurality of subsequences according to a plurality of starting data points, wherein the i-th subsequence includes the i-th starting data point and all data points between the i-th starting data point and the i+1-th starting data point; The method for obtaining the first straight line corresponding to the mth data point is as follows: in response to the slope of the line connecting the first reference point of the ith starting data point and the mth data point being not less than the first straight line corresponding to the m-1th data point, the line connecting the first reference point of the ith starting data point and the mth data point is used as the first straight line corresponding to the mth data point; otherwise, the first straight line corresponding to the m-1th data point is used as the first straight line corresponding to the mth data point; The method for obtaining the second straight line corresponding to the mth data point is: in response to the slope of the line connecting the second reference point of the ith starting data point and the mth data point being no greater than the second straight line corresponding to the m-1th data point, the line connecting the second reference point of the ith starting data point and the mth data point is used as the second straight line corresponding to the mth data point; otherwise, the second straight line corresponding to the m-1th data point is used as the second straight line corresponding to the mth data point.

3. The data visualization method for simulating a flight environment according to claim 1, It is characterized in that The calculating the similarity between the subsequence and other subsequences without missing data points includes: ; In the formula, represents the similarity between the subsequence and the rth subsequence without missing data points, represents the number of all data points in the subsequence, represents the number of all data points in the rth subsequence that does not have missing data points, represents an exponential function with a natural constant as base, It represents the first DTW distance between this subsequence and the rth subsequence without missing data points.

4. The data visualization method for simulating a flight environment according to claim 3, It is characterized in that The step of weighting other subsequences without missing data points according to similarity, and determining the filling value of each missing data point in the subsequence according to the weighted result, includes: In response to the ath data point of the subsequence being a missing data point, the filling value of the missing data point is Satisfies the expression: ; Where N represents the number of subsequences without missing data points. , Respectively represent the similarity between the subsequence and the rth and nth subsequences without missing data points, Indicates the preset threshold value, Represents the value of the a-th data point in the r-th subsequence that has no missing data points.

5. The data visualization method for simulating a flight environment according to claim 1, It is characterized in that The calculating the similarity between the subsequence and other subsequences without missing data points includes: ; In the formula, represents the similarity between the subsequence and the rth subsequence without missing data points, represents the number of all data points in the subsequence, represents the number of all data points in the rth subsequence that does not have missing data points, represents an exponential function with a natural constant as the base, and the DTW distance between the slope sequence of the first straight line of the subsequence and the slope sequence of the first straight line of the r-th subsequence without missing data points and the DTW distance between the slope sequence of the second straight line of the subsequence and the slope sequence of the second straight line of the r-th subsequence without missing data points are calculated by the dynamic time warping algorithm, as the second DTW distance and the third DTW distance between the subsequence and the r-th subsequence without missing data points, respectively. , , They respectively represent the first DTW distance, the second DTW distance and the third DTW distance between the subsequence and the rth subsequence without missing data points.

6. The data visualization method for simulating a flight environment according to claim 5, It is characterized in that The step of weighting other subsequences without missing data points according to similarity, and determining the filling value of each missing data point in the subsequence according to the weighted result, includes: In response to the ath data point of the subsequence being a missing data point, the filling value of the missing data point is Equal to the first predicted value for the missing data point and the second predicted value The mean of .

7. The data visualization method for simulating a flight environment according to claim 6, It is characterized in that The first predicted value of the missing data point and the second predicted value Satisfies the expression: ; ; Where N represents the number of subsequences without missing data points. , Respectively represent the similarity between the subsequence and the rth and nth subsequences without missing data points, Indicates the preset threshold value, , They represent the slopes of the first and second straight lines corresponding to the a-th data point of the r-th subsequence without missing data points, represents the moment of the missing data point, , , , Respectively represent the time and value of the first reference data point and the second reference data point of the starting data point of the subsequence, wherein the first reference data point and the second reference data have the same time as the starting data point and the difference with the value of the starting data point is equal to the preset parameter, and the value of the first reference point is larger than that of the second reference point.

8. The data visualization method for simulating a flight environment according to claim 3 or 5, It is characterized in that Also includes: In response to the first data point of the subsequence with missing data points being a missing data point, the DTW distance between the reverse result of the subsequence and the reverse result of the r-th subsequence without missing data points, the DTW distance between the reverse result of the slope sequence of the first straight line of the subsequence and the reverse result of the slope sequence of the first straight line of the r-th subsequence without missing data points, and the DTW distance between the reverse result of the slope sequence of the second straight line of the subsequence and the reverse result of the slope sequence of the second straight line of the r-th subsequence without missing data points are calculated through a dynamic time warping algorithm as the first DTW distance, second DTW distance and third DTW distance between the subsequence and the r-th subsequence without missing data points, respectively.

9. The data visualization method for simulating a flight environment according to claim 1, It is characterized in that The method of re-dividing the first A missing data points among the consecutive D missing data points into the previous subsequence and re-dividing the remaining missing data points into the next subsequence further includes: The starting data point of the previous subsequence before repartitioning is used as the starting data point of the previous subsequence after repartitioning; The starting data point of the next subsequence before repartitioning is used as the starting data point of the next subsequence after repartitioning.

10. A data visualization system for simulating flight environment, It is characterized in that include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a data visualization method for simulating a flight environment according to any one of claims 1 to 9 is implemented.