CN112365743A - Method and device for correcting flight path positioning data offset of aircraft - Google Patents

Abstract

The invention provides a method and a device for correcting aircraft track positioning data offset, wherein the method comprises the steps of emptying historical cache information before correcting the current track positioning data offset of an aircraft; receiving real-time measuring point information of an aircraft sent by a distributed positioning system; determining current parabola and real-time deflection angle information used as a filter cover according to the real-time measuring point information and the current cache information; calculating and generating real-time filtering point information corresponding to the real-time measuring point according to the real-time measuring point information, the current parabola and the real-time deflection angle information; updating current cache information according to the real-time measuring point information, the real-time filtering point information and the real-time deflection angle information; and obtaining the corrected flight path of the current flight path of the aircraft according to all the filtering point information of the current flight path. The invention realizes accurate positioning and high stability, and the track is more in line with the motion rule.

Description

Method and device for correcting flight path positioning data offset of aircraft

Technical Field

The invention relates to the technical field of aircraft positioning, in particular to a method and a device for correcting an aircraft track positioning data offset.

Background

In airport route guidance project application, a software system is required to control runway edge lights and conduct route light guidance on aircrafts. In the guiding process, the aircraft is required to travel according to the optimized path calculated by the software, the road surface light around the relevant path is lightened, and the aircraft travels according to the light indication. And calculating the real-time position data of the aircraft by the distributed positioning system and sending the real-time position data to the application software end so as to draw the flight path information of the aircraft. And the software controls and extinguishes the lamplight along the path of the history track which has been walked in real time through the track information which has been walked by the airplane and keeps the lamplight along the non-driving path. When the aircraft leaves the preset optimized path from the fork temporarily for some reason, the lights on the road surface along which the original optimized path which is not completely traveled is completely extinguished, the optimized path is recalculated according to the real-time position of the aircraft, and the lights on the road surface along which the new path is located are lightened.

Because of the positioning accuracy, there will be great error in the positioning data, so in the route has the guide process, because the error leads to the positioning data drift, will cause the following two aspect's problem above all:

when the aircraft is still at the starting point of the taxiway, the software end receives the positioning data, and can mistakenly find that the aircraft moves forwards and mistakenly control the light behind the advancing direction to be extinguished;

when the aircraft is guided to travel to a certain fork and temporarily changes the direction and deviates from the preset optimized path, the software end can mistakenly continue to keep the light of the path which is not completed for the aircraft to go on according to the preset optimized path after receiving the positioning data, the deviation is not slowly found until the aircraft goes out for a certain distance, and then a route planning suggestion is given again, the light of the previous path is extinguished, and the light for re-planning the path is lightened.

Generally speaking, it is because of the drift of the positioning data, there will be no synchronous coincidence between the light control and the aircraft operation state, and the response speed of the real-time path re-planning will be delayed greatly compared with the change of the aircraft operation plan.

The existing method for solving the problem of positioning data drift and correcting the flight path offset mainly comprises the following three aspects:

and the positioning accuracy of the positioning algorithm is improved. The disadvantages are as follows: in the aspect of hardware, communication equipment, computing equipment and storage equipment with more quantity, higher precision and stronger performance need to be replaced or added, and the cost is high; in the aspect of software, the improvement difficulty of the positioning algorithm is high, the improvement space is small, and the improvement effect is not obvious.

And carrying out path matching through the inherent path information of the road surface. The disadvantages are as follows: data which needs to be subjected to detailed data modeling, transmission or storage on the road surface can become complex in type and structure, and data throughput and calculation overhead can be increased sharply; when the aircraft moves to an area outside the modeling road network due to an unexpected reason, correct matching and positioning cannot be carried out.

And correcting the measured positioning data with errors on the software algorithm level. The method is a mainstream solution mode adopted by considering the factors of development cost, development difficulty, resource consumption, convenience in use and maintenance and the like at present. However, the existing correction methods are all to analyze the motion law of the statistical historical trajectory, including but not limited to position, velocity, acceleration, distribution law, etc., and find the correct position of the current aircraft by a prediction or fitting method. Such a design has the following disadvantages:

(1) the corrected paths of some correction methods are continuous and smooth, can well reflect the advancing trend, but are far away from the actual measurement data, and cannot reflect the real position information.

(2) Some correction methods keep higher precision after correction, can well reflect real position information, but have poor path continuity (for example, zigzag) and can not be matched with a performed pavement;

the above methods cannot well filter out the situation that the trajectory self-intersection caused by the drift of the positioning data and the soft distribution are formed, and the trajectory reciprocates in the globoid distribution, which does not conform to the physical motion law of the large-scale aircraft.

Disclosure of Invention

In view of the above, the present invention provides a method and an apparatus for correcting an aircraft track positioning data offset, so as to achieve accurate positioning and high stability, and the track better conforms to the motion law.

In one aspect, the present invention provides a method for correcting an aircraft track positioning data offset, including:

clearing historical cache information before correcting the current track positioning data offset of the aircraft;

receiving real-time measuring point information of an aircraft sent by a distributed positioning system;

determining current parabola and real-time deflection angle information used as a filter cover according to the real-time measuring point information and the current cache information; the current cache information includes: the cache queue comprises last measurement point cache information, last filtering point cache information, last deflection angle cache information and cache queues of all filtering point information between an initial filtering point and a last filtering point of the current flight path; the last deflection angle cache information is an included angle formed between a unit vector in the positive direction of the x axis and a connecting line from the focus of the current parabola to the vertex of the current parabola along the anticlockwise direction;

calculating and generating real-time filtering point information corresponding to the real-time measuring point according to the real-time measuring point information, the current parabola and the real-time deflection angle information;

updating current cache information according to the real-time measuring point information, the real-time filtering point information and the real-time deflection angle information;

and obtaining the corrected flight path of the current flight path of the aircraft according to all the filtering point information of the current flight path.

Further, determining the real-time deflection angle information according to the real-time measurement point information and the current cache information includes:

determining a common point of a line segment formed by connecting the cache information of the previous filtering point to the information of the real-time measuring point and the boundary of the current parabola;

and forming an included angle between vectors by the unit vector in the positive axial direction and a common point connecting line from the focus of the current parabola to the boundary of the current parabola along the counterclockwise direction to serve as real-time deflection angle cache information.

Further, determining a current parabola for use as a filter mantle based on the real-time measurement point information and current cache information comprises:

and determining the size of the current parabola according to the length of a connecting line between the cache information of the previous measuring point and the real-time measuring point information.

Further, the step of calculating and generating real-time filtering point information corresponding to the real-time measuring point according to the real-time measuring point information, the current parabola and the real-time deflection angle information comprises:

calculating an included angle between a first deflection angle information and a real-time deflection angle information, calculating a distance between a common point of a current parabola boundary and a focus of the current parabola, if the common point exists, determining real-time filtering point information on a line segment formed by the focus of the current parabola and the real-time measuring point information, wherein the real-time filtering point information meets the condition that the distance between a last filtering point information and the common point is equal to the distance between the real-time filtering point information and the real-time measuring point, and if the common point does not exist, the real-time filtering point information is the last filtering point information.

Further, the step of calculating and generating real-time filtering point information corresponding to the real-time measuring point according to the real-time measuring point information, the current parabola and the real-time deflection angle information further comprises the following steps:

and correcting the real-time filtering point information according to the cache queue so as to update the current cache information according to the corrected real-time filtering point information and obtain the corrected flight path of the current flight path of the aircraft.

The invention also provides a device for correcting the flight path positioning data offset of the aircraft, which comprises:

the preprocessing unit is used for emptying historical cache information before the correction operation of the current track positioning data offset of the aircraft is carried out;

the receiving unit is used for receiving the real-time measuring point information of the aircraft sent by the distributed positioning system;

the filtering unit is used for determining the current parabola and the real-time deflection angle information which are used as the filtering cover according to the real-time measuring point information and the current cache information; calculating and generating real-time filtering point information corresponding to the real-time measuring point according to the real-time measuring point information, the current parabola and the real-time deflection angle information; the current cache information includes: the cache queue comprises last measurement point cache information, last filtering point cache information, last deflection angle cache information and cache queues of all filtering point information between an initial filtering point and a last filtering point of the current flight path; the last deflection angle cache information is an included angle formed between a unit vector in the positive direction of the axis and a connecting line from the focus of the current parabola to the vertex of the current parabola along the anticlockwise direction;

the updating unit is used for updating the current cache information according to the real-time measuring point information, the real-time filtering point information and the real-time deflection angle information;

and the track correction unit is used for obtaining the corrected track of the current track of the aircraft according to all the filtering point information of the current track.

Further, the filtration unit is specifically configured to: determining a common point of a line segment formed by connecting the cache information of the previous filtering point to the information of the real-time measuring point and the boundary of the current parabola; and forming an included angle between vectors by the unit vector in the positive axial direction and a common point connecting line from the focus of the current parabola to the boundary of the current parabola along the counterclockwise direction to serve as real-time deflection angle cache information.

Further, the filter unit is specifically further configured to: and determining the size of the current parabola according to the length of a connecting line between the cache information of the previous measuring point and the real-time measuring point information.

Further, the filter unit is specifically further configured to: calculating an included angle between a first deflection angle information and a real-time deflection angle information, calculating a distance between a common point of a current parabola boundary and a focus of the current parabola, if the common point exists, determining real-time filtering point information on a line segment formed by the focus of the current parabola and the real-time measuring point information, wherein the real-time filtering point information meets the condition that the distance between a last filtering point information and the common point is equal to the distance between the real-time filtering point information and the real-time measuring point, and if the common point does not exist, the real-time filtering point information is the last filtering point information.

Further, the device for correcting the aircraft track positioning data offset further includes:

the filtering point information correcting unit is used for: and correcting the real-time filtering point information according to the cache queue, so that the updating unit updates the current cache information according to the corrected real-time filtering point information and the flight path correcting unit obtains the corrected aircraft of the current flight path of the aircraft according to the corrected real-time filtering point information.

According to the method and the device for correcting the flight path positioning data deviation of the aircraft, the position is corrected to the near small-range space field, the situation that the measured data is not excessively deviated is guaranteed, a continuous and smooth advancing track is formed, the positioning is accurate, the stability is high, and the track is more in line with the motion rule.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

FIG. 1 is a flow chart of a method for correcting an aircraft track location data offset according to an exemplary first embodiment of the present invention.

Fig. 2 is a frame overview of a method for correcting an aircraft track location data offset according to an exemplary second embodiment of the invention.

FIG. 3 is a data flow diagram of a method for correcting an aircraft track location data offset according to an exemplary third embodiment of the present invention.

FIG. 4 is a flowchart of a method for correcting an aircraft track location data offset according to an exemplary fourth embodiment of the invention.

Fig. 5 is a spatial relationship between relevant points in the Shell structure.

FIG. 6 is an example of a simple filter trace generation process.

FIG. 7 is a diagram illustrating a modification of a history filter point to a latest filter point.

FIG. 8 is a screenshot of a modification procedure for a historical filter point to a latest filter point.

Fig. 9 is a screenshot of the convergence effect procedure on the path of the long distance round-trip trajectory.

Fig. 10 is a screenshot of the filter effect procedure for drift in a quiescent state.

Fig. 11 is a screenshot of the filter effect procedure for drift in the straight state.

Fig. 12 is a screenshot of a filter effect routine for drift in a stationary launch condition.

Fig. 13 is a screenshot of a filtering effect procedure for a drift in a deceleration stop state.

Fig. 14 is a screenshot of the filter effect procedure for drift in the steered state.

Fig. 15 is a block diagram of an apparatus for correcting an aircraft track positioning data offset according to an exemplary fifth embodiment of the present invention.

Detailed Description

Embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

It should be noted that, in the case of no conflict, the features in the following embodiments and examples may be combined with each other; moreover, all other embodiments that can be derived by one of ordinary skill in the art from the embodiments disclosed herein without making any creative effort fall within the scope of the present disclosure.

It is noted that various aspects of the embodiments are described below within the scope of the appended claims. It should be apparent that the aspects described herein may be embodied in a wide variety of forms and that any specific structure and/or function described herein is merely illustrative. Based on the disclosure, one skilled in the art should appreciate that one aspect described herein may be implemented independently of any other aspects and that two or more of these aspects may be combined in various ways. For example, an apparatus may be implemented and/or a method practiced using any number of the aspects set forth herein. Additionally, such an apparatus may be implemented and/or such a method may be practiced using other structure and/or functionality in addition to one or more of the aspects set forth herein.

The idea of the correction method for the aircraft track positioning data offset is as follows: with a spatial hysteresis design, the unpredictability of the measurement points due to the upcoming front is counteracted. Colloquially, by analogy, are: now there are two people, a and B, starting at the same starting point and going along a definite path from the starting point to the destination. The path A represents a track drawn by a measuring point given by a positioning algorithm, and the path A stops from the starting point to the end point for the determined path in a manner of 'left neglect, right neglect, front neglect and back neglect', and alternates with the path B until the end point after the path A goes further. What B needs to walk is a point-drawn track given by the correction algorithm, and B can step forward according to the position of A, so that the walking path of B is ensured to be a determined path as much as possible, and the walking path continuously and stably walks forward to the terminal point, and the situations of 'left neglect, right neglect, front neglect and back neglect' can not occur. The current practice of B is:

firstly, A walks first, B always keeps a one-step difference with A, and they are all positioned at the same starting point at the beginning;

b always keeps a proper distance from A according to the speed of A, the faster the speed of A, the farther B is away from A, and on the contrary, the closer B is to A;

if the A suddenly stops right ahead of the B or backs towards the B direction, the B station is not moved in situ until the A continuously moves right ahead and exceeds a certain distance, and then the B keeps up with the B;

if A changes direction suddenly and is no longer in the front of B, B will first adjust the advancing direction to the direction opposite to A,

and determining the distance to be followed up or kept in place according to the distance between the A and the B.

The following embodiments of the method for correcting the flight path positioning data offset of the aircraft are implemented by formalizing the game thought into a mathematical method and performing an algorithm.

As shown in fig. 1, a method for correcting an aircraft track positioning data offset includes:

step 101, emptying historical cache information before correcting the current track positioning data offset of the aircraft;

102, receiving real-time measuring point information of an aircraft, which is sent by a distributed positioning system;

103, determining current parabola and real-time deflection angle information used as a filter cover according to the real-time measuring point information and the current cache information; the current cache information includes: the cache queue comprises last measurement point cache information, last filtering point cache information, last deflection angle cache information and cache queues of all filtering point information between an initial filtering point and a last filtering point of the current flight path; the last deflection angle cache information is an included angle formed between a unit vector in the positive direction of the x axis and a connecting line from the focus of the current parabola to the vertex of the current parabola along the anticlockwise direction;

104, calculating and generating real-time filtering point information corresponding to the real-time measuring point according to the real-time measuring point information, the current parabola and the real-time deflection angle information;

105, updating current cache information according to the real-time measuring point information, the real-time filtering point information and the real-time deflection angle information;

and 106, obtaining the corrected flight path of the current flight path of the aircraft according to all the filtering point information of the current flight path.

The position of the embodiment is corrected to the near small-range space field, so that the continuous and smooth advancing track is formed without excessively deviating from the measured data, the positioning is accurate, the stability is high, and the track is more in line with the motion rule.

Specifically, as shown in fig. 2, the schematic frame diagram of the method for correcting the aircraft track positioning data offset includes the following steps:

firstly, a filtering algorithm module comprises two functions of initialization and calculation, and a cache mechanism is arranged in the module;

before the filtering track is generated by using the filtering algorithm module, an initialization function is required to be called to clear the cache;

thirdly, the filtering algorithm module receives a latest measuring point as an input;

invoking a calculation function through input data and cache data to generate a latest filtering point and update the cache data;

the filtering algorithm module returns a latest measuring point as output;

and sixthly, returning to the step III, receiving the next latest measuring point by the filtering algorithm module, or ending.

Further, as shown in fig. 3, the data flow frame diagram (corresponding to the detailed outline frame diagram) of the method for correcting the aircraft track positioning data offset includes the following steps:

the filtering algorithm module comprises two functions: the initialization function and the geometric computation function comprise four buffer variables: a cache queue of a recent measurement point cache, a recent filtering point cache, a recent deflection angle cache and a recent history filtering point;

before the filtering track is generated by using the filtering algorithm module, the cache is emptied by calling the initialization function only once so as to clear temporary data remained in a cache variable in the calculation process of generating the previous complete filtering track;

the filtering algorithm module integrally waits for receiving a latest measuring point as input, the data of the latest measuring point is an instant calculation result from the distributed positioning system, and the point is added into the current original track;

inputting the latest measuring point and all cache data into a geometric calculation function, calculating to generate the latest filtering point, respectively updating the latest measuring point cache and the latest filtering point cache by using the latest measuring point and the latest filtering point, updating the latest deflection angle cache by using the latest deflection angle, and adding the latest filtering point into a ring cache queue of the latest generated history filtering point (if the queue is full, covering the position stored by the earliest filtering point);

the filtering algorithm module returns a latest filtering point as output and adds the latest filtering point into the current filtering track;

and returning to the step III, and waiting for receiving the next latest measuring point by the filtering algorithm module or ending.

To explain in more detail the implementation idea of the method for correcting the aircraft track positioning data offset, the basic variables and the basic function definitions are now described:

p is a trace point for specifying the position of a moving target on the earth surface at a certain moment, P (x, y) is a P data format representation form, and is a two-dimensional rectangular coordinate of a local geodetic reference system converted from longitude and latitude coordinates, and x and y are respectively a horizontal coordinate and a vertical coordinate under a rectangular coordinate system.

② L is a trace, denoted L ═ P₀,P₁,...,P_i-1,P_i}，P₀Referred to as the starting point, P_iCalled the latest point; f and S are track points with a data format of P; LF ═ F₀,F₁,...,F_i-1,F_iIs called the current original trajectory, F_iCalled nearest measurement point, F_i+1Called the latest measurement point; LS ═ S₀,S₁,...,S_i-1,S_iIs called the current filtering track, S_iCalled nearest filter point, S_i+1Called the latest filter point; add is a method for adding new track points in a track in sequence, and a calling mode LF_i+1) Add (S) or LS_i+1)。

If the line segment formed by connecting the P point with the focus of the parabola does not have a common point with the parabola, the P is positioned in the parabola; the parabola and its inner region are referred to herein as the Shell (filter sock); the parabolic focus is a certain filtering point S, called the focus of the Shell; vertex Z of the parabola is called the vertex of Shell; the symmetrical midline of the parabola is called the midline of the Shell; throwingThe object line itself is called the boundary of the Shell; a point on the Shell boundary is called C; unit vector of positive x-axis direction

In the counterclockwise direction and vector

The angle between them is called the angle of deflection of the Shell.

The Shell (s, f, alpha) is a data format representation form of the filter mantle Shell; wherein s represents the focus of the Shell, f represents the distance of the line segment SZ in the Shell, and alpha represents the deflection angle of the Shell; the position, the size and the direction of the Shell are respectively and uniquely determined by s, f and alpha; the minimum allowed value of f is set to Min _ f and the maximum allowed value is set to Max _ f.

For trace LS ═ S₀,S₁,...,S_i-1,S_i}，Shell_iIs the nearest filter point S_iIs a focal point, point Z_iIs a vertex, F_iS_iIs a line segment on the middle line (if F)_i≠S_i) For calculating the latest filtering point S_i+1The Shell of (1); unit vector of positive x-axis direction

In the counterclockwise direction and vector

Angle alpha therebetween_iReferred to as the most recent deflection angle.

⑥Shell_iS focus of_iTo the latest measuring point F_i+1Connected line segment and Shell_iThe common point of the boundary is called C_i(if a common point exists); focal point S_iAnd vertex Z_iFormed vector

And a focus S_iAnd point C_iFormed vector

The angle between them, smaller than the straight angle, is called theta_i(if C is_iPresence); unit vector of positive x-axis direction

In the counterclockwise direction and vector

Angle alpha therebetween_i+1Called the latest deflection angle (if C)_iPresent).

Seventhliter is an example of a filtering algorithm module; init is an initialization function, and is called only once before the module is used to generate a complete filtering track, and the calling mode is as follows: init (); the generation is a calculation function, and is called once every time a measurement point is received to Generate a filter point, and the calling method is as follows: s_i+1＝fliter.Generate(F_i+1)。

⑧F^*Is used in the fliter for storing the most recent measurement point F_iThe cache variable of (2); s^*Is used in the fliter for storing the nearest filtered point S_iThe cache variable of (2); alpha is alpha^*Is used in fliter for storing the most recent deflection angle alpha_iThe cache variable of (2); the RN is a circular cache queue in the fliter and used for storing the history filtering points generated recently in the LS; when the length of LS is less than N, RN equals LS, otherwise RN equals S_i-N+1,S_i-N+2,...,S_i-1,S_iN is the maximum allowed length of the queue.

AddElement is a method for adding elements in RN, and the calling mode is RN_i+1) I.e. the order of the elements S_i+1Is added to S_iThen; clear is a method for clearing all elements in RN, and the calling mode is rn.clear (); IsEmpty () is a method for judging whether the RN is an empty list or not, and is called in a mode RN; size is a method for judging RN length, and a calling mode is RN. Middle Get is a method for obtaining a certain track point in RN and a calling mode

S_i＝RN.Get(RN.Size()-1)。

GeoF in R is the core function in fliter to implement Generate, called the geometric computation functionNumber, satisfying the relation S_i+1＝GeoF(F_i+1,F^*,S^*,α^*RN) for calculating the latest filtering point; create is an internal function of GeoF, is used for generating Shell and satisfies the relation Shell_i＝Create(F_i+1,F^*,S^*,α^*) (ii) a Update is an internal function of GeoF for calculating or updating the latest deflection angle satisfying the relation α_i+1＝UpdateS(α^*,S^*,F_i+1or S_i+1) (ii) a GetRaw is an internal function of GeoF, is used for preliminarily calculating the latest filtering point and satisfies the relation S_i+1＝GetRaw(Shell_i,F_i+1,α_i+1) (ii) a Attach is an internal function of GeoF and is used for correcting the latest filtering point to obtain the latest filtering point as a final result, and the latest filtering point meets the relation S_i+1＝Attach(S_i+1,RN)。

Distance is a tool function used for calculating the Distance between two points and is called in a manner of | P₁P₂|＝Distance(P₁,P₂) In which P is₁、P₂Is any two track points, | P₁P₂L is a distance value of two points, the preset minimum value of the distance is Min _ d, the preset maximum value of the distance is Max _ d, and the preset value can be used in a Create function; atan is a tool function used to calculate the unit vector in the positive x-direction

In the counterclockwise direction to vector

Angle therebetween

The calling mode is

Bettween is a tool function, and calculates the nearest deflection angle alpha_iAnd the latest deflection angleα_i+1Included angle theta less than straight angle_iSatisfies the relation of_i＝Bettween(α_i,α_i+1) (ii) a Linear is a tool function, and the calling mode y is Linear (X)₁,X₂X) when X₁≤x≤X₂When y is more than or equal to 0 and less than or equal to 1, X and y are in a linear relationship, X₁X, y is 0, X > X₂When y is 1; CosP is a tool function, and calling mode l is CosP (Shell)_i.f,θ_i) Wherein theta_iIs Shell_iFocal point S_iAnd vertex Z_iFormed vector

And a focus S_iAnd Shell_iPoint C on the boundary_iFormed vector

The angle between l is the line segment S_iC_iThe distance value of (2).

Specifically, as shown in fig. 4, a computation flow diagram of a method for correcting an aircraft track positioning data offset (mainly, a computation part in a data flow diagram is made into a flow and a form) includes the following steps:

s1: fliter is an example of a filtering algorithm module, calling an initialization function fliter^*、S^*、α^*RN, which is an internal attribute associated with the specific instance fliter, sets the receiving point sequence number i to-1.

And setting the global receiving point sequence index number i to be-1.

S2: receiving a latest measuring point F_i+1And added to the end of the original trajectory LF: add (F) in LF_i+1)。

S3: the step is a process of generating a latest filtering point for the calculation of a geometric computation function and updating cache data:

S_i+1＝GeoF(F_i+1,F^*,S^*,α^*RN), the function GeoF implementation process is detailed later.

The implementation of the GeoF function in step S3 is the main content of the specific implementation, and the sub-step S33 (generating Shell) is discussed with emphasis on_iCorrelation), S34 (calculate S)_i+1Correlation), S35 (correction S)_i+1Related) to the implementation of the process.

S31: judging whether the ring buffer queue RN of the history filtering point is an empty queue: isoempty (); if the cool is true, go to step S32, otherwise go to step S33.

S32: at this time, no filtering point is generated in the LS, a first filtering point needs to be generated, and the first filtering point is set to be a first measuring point: s_i+1＝F_i+1(i.e., S)₀＝F₀) (ii) a Setting the latest deflection angle alpha_i+1═ infinity (i.e., α)₀═ infinity), meaning that the first Shell has not been generated.

The function GeoF is implemented as follows:

calculating the bol (RN. IsEmpty ()), judging whether the bol is true, and if not, turning to the step (c);

secondly, directly taking the latest filtering point as the latest measuring point: s_i+1＝F_i+1And initializing the latest deflection angle alpha_i+1Changing to the step ninthly;

computing Shell_i＝Create(F_i+1,F^*,S^*,α^*) For creating a preliminary calculation S_i+1Time dependent filter mantle Shell_i；

Fourthly, calculating alpha_i+1＝Update(α^*,S^*,F_i+1) For calculating the latest deflection angle α_i+1；

Calculating S_i+1＝GetRaw(Shell_i,F_i+1,α_i+1) For preliminary calculation of the latest filtering point S_i+1Wherein the filter mantle Shell created in the step (c) is utilized_i；

Sixthly, calculate S_i+1＝Attach(S_i+1RN) for preliminary counting in the correction step (c)Calculated latest filtering point S_i+1A value of (d);

seventh judgment S^*≠S_i+1If the result is not true, turning to the step ninthly;

b calculating alpha_i+1＝Update(α^*,S^*,S_i+1) For adjusting the latest deflection angle alpha calculated in step (iv)_i+1A value of (d);

ninthly, updating the value of the cache variable: f^*＝F_i+1、S^*＝S_i+1、α^*＝α_i+1Caching the latest measuring point, the latest filtering point and the latest deflection angle obtained by the calculation respectively to form the latest measuring point, the latest filtering point and the latest deflection angle parameter which are depended by calling the GeoF calculation for the next time;

r is the latest filtering point S_i+1Adding to the RN circular buffer queue: AddElement (S)_i+1)；；

Return to the latest filtering point S_i+1。

S33: call Shell_i＝Create(F_i+1,F^*,S^*,α^*) Generating a calculation S_i+1Dependent filter housing Shell_i(S₀As the first filtering point, it is computationally derived without passing through a filter housing), where S^*Is Shell_iFocus of (Shell)_i.s＝S^*)，α^*Is Shell_iDeflection angle (Shell)_i.α＝α^*) Value of line segment | F_i+1F^*The size of |, determines the Shell_iSize of (Shell)_iF); call alpha_i+1＝Update(α^*,S^*,F_i+1) To calculate the unit vector of the positive direction of the x-axis

In the counterclockwise direction to line segment F_i+1S^*Angle alpha of_i+1When line segment F_i+1S^*Is absent (i.e. F)_i+1＝S^*Of (i) alpha_i+1Is set to a^*。

Wherein the implementation process of the Create function in step S33 is as follows:

creating Shell_iExample (c): shell_iInitially setting Shell () for new Shell ()_iFocal point Shell_i.s＝S^*Initialization setting Shell_iAngle of deflection Shell_i.α＝α^*；

② calculating the latest measuring point F_i+1To the nearest measurement point F^*The distance of (c): i F_i+1F^*|＝Distance(F_i+1,F^*) Calculating the distance value in the interval [ Min _ f, Max _ f]Relative ratio coefficient of (1): phi ═ Linear (Min _ d, Max _ d, | F)_i+1F^*|)；

Third, the Shell generated in the first step of calculation_iMagnitude-dependent f-value: shell_i.f＝Min_f+(Max_f-Min_f)*φ；

Fourthly, the latest filtering point S of the calculation is returned_i+1Dependent filter housing Shell_i。

Wherein the implementation of the tool function Linear is as follows,

firstly, judging that X is less than or equal to X₁If yes, returning to 0;

② judge that X is more than or equal to X₂If yes, returning to 1;

③ return a proportionality coefficient between (0, 1): (X-X)₁)/(X₂-X₁)。

Note: shell_iThe value of f determines the size of the filter housing, which is set to [ Min _ f, Max _ f]Between the ranges, and with the value | F_i+1F^*I is related in size, i.e. the latest measuring point F in the original trajectory_i+1Closest measurement point F to the original trajectory^*Distance between | F_i+1F^*The larger the | is, the larger the absolute value of the drift error can be considered, and the required filter housing Shell_iThe larger the design.

S34: calling S_i+1＝GetRaw(Shell_i,F_i+1,α_i+1) Preliminary calculation of the bestNew filter point S_i+1Here first the Shell is calculated_iAngle of deflection alpha_i(α_i＝Shell_iα) and α)_i+1Angle theta therebetween_i＝Bettween(α_i,α_i+1) Recalculating Shell_iS focus of_iAnd F_i+1Line segment formed with Shell_iCommon point of boundary C_iDistance from focal point | S_iC_i|＝CosP_i(Shell_i,θ_i) If C is present_iIf present, then in line segment S_iF_i+1To find a point S_i+1So that S_iC_i|＝|S_i+1F_i+1I, otherwise (C)_iIn absence of F_i+1Dotted on Shell_iBoundary or interior) S_i+1Take the value of S_i。

The tool function CosP is realized as follows, and is related to a parameter (rho)_i,θ_i) Parabolic polar equation:

|S_iC_i|＝ρ_i＝CosP(Shell_i.f,θ_i)＝2*Shell_i.f/(1+cosθ_i)＝2*S_iZ_i|/(1+cosθ_i)。

note: when alpha is_i< 0 is present when i is 0, alpha_iIn the case of ═ infinity, the Shell that has been generated is explained₀Is that the deflection angle has not been set yet, but α has been generated in step S33₁Is the next filter mantle Shell₁Angle of deflection (alpha)₁Does not change its value after step S35), the Shell is put into service at this time₀Angle of deflection alpha₀Preset to alpha₁I.e. alpha₀And alpha₁Included angle theta₀And (0) keeping the two filter covers from turning. When | S_iF_i+1|＜＝|S_iC_iWhen l, F_i+1Is positioned on the filter housing_iOn the inside or boundary, when F is not_i+1Whether or not to react with S_iCoincident, latest measurement point S_i+1Are all set to be equal to S_iAnd (4) overlapping.

Process description of GetRaw function As shown in the example of FIG. 6, Shell_iIs a line segment S_i-1S_iThe straight line is the symmetric central axis (if S_iIs not modified by the Attach function, then F_iPossibly on the medial axis) with a vertex z_iAt a deflection angle of

In the counterclockwise direction to

Angle alpha of_i，θ_iIs composed of

In the counterclockwise direction to

Angle of (D) F_i+1Is Shell_iOne point outside, calculate | S_iF_i+1|、|S_iC_iAfter | preliminarily calculating the line segment S_iF_i+1Upper is calculated S_i+1Satisfy | S_iC_i|＝|S_{i+ 1}F_i+1|。

The geometry of the Shell will be constructed as a filter shroud because it exhibits symmetry with respect to the direction of motion of the aircraft and is easy to calculate. Shell_iThe direction of (1) is a vector formed by two filtering points of the nearest moving object

Should theoretically be symmetrical, but not symmetrical in the front-to-back direction, i.e. F_i+1Relative to S_iThe positional relationship of (A) can be represented by a vector

A description is given.

And

angle of (theta)_iAnd the nearest measurement point F_iAnd F_i+1Distance d of_i＝|F_iF_i+1Two parameters that are of paramount importance to how the filter housing Shell is designed. Theta_iThe degree to which the currently moving object needs to be steered at the next moment (real-time direction change), d_iThe extent to which the currently moving object needs to be displaced at the next moment (real-time velocity) is illustrated. The method comprises the following steps: theta_iIn addition to reflecting true steering, error-induced directional offset is also reflected, especially at low speeds, which can cause the target to change direction frequently and even back up, so the design of Shell needs to prevent abrupt steering due to error as much as possible. The principle is therefore theta_iThe larger the value, the more it needs to be suppressed, so that the filtered trajectory is more continuously smooth. And when the shape of the Shell is designed as a parabolic region, when theta is_iThe larger the value, S_iF_i+1The intersection C of the straight line and the Shell boundary_iTo the focus S_iThe greater the distance of F_i+1Fall on Shell_iThe higher the probability of the inside, the more effectively F that is severely diverted due to the error_i+1Effectively filtering out. Therefore, θ_iThis index determines the Shell_iThe design of the geometrical shape of (2). The method comprises the following steps: d_iIn addition to the true speed, the error-induced position offset, d_iThe larger the value is, the larger the amount of positional deviation due to the error becomes. If all the shells are_iSet to a fixed small size, the yaw angle is merely changed following the movement of the target, and the size is not changed with the real-time speed. When the target speed is faster, d_i＝|F_iF_i+1The greater the value of | is, the less the velocity will not decrease due to the inertia of the target movement, so the d of the subsequent measurement_i+1＝|F_i+1F_i+2The value of | will also be larger. Because of the Shell_iF is small, so the distance F is obtained after GetRaw function calculation_i+1Very close S_i+1. Because of d_i+1Is larger, so | S_i+1F_i+2The value of | will also be roughly greater, then F_i+2Fall in S_i+1Shell in focus_i+1The probability of insides is low. At the sudden turning of the target due to the error, so that theta_i+1In the case of a relatively large size, F is required to be reduced_i+2Successfully filtered out (i.e. S)_i+2And S_i+1Overlap) is less likely, even if F cannot be adjusted_i+2Filtration, also requiring suppression from S_i+1To F_i+2Distance of directional displacement (selected from S)_i+1Near, far from F_i+2The far point is taken as S_i+2) However, such an effect of suppressing is also weak. If all the shells are_iSet to a fixed large size, thus the distance S will be set_iSufficiently far F_i+1Is also enclosed in the Shell_iInternally, will lead to S_i+1Is superposed on S_i，S_i+1F of true measurement of distance_i+1The deviation is too far away so that the filtering effect does not fit the actual measurement. Thus, in summary, with d_iVariation of (3), Shell_iThe value of f should be reasonably scaled up or down in positive correlation, so that d_iThis index determines the Shell_iThe size of the device is designed.

S35: calling S_i+1＝Attach(S_i+1RN), correcting the latest filtering point S by using the local trace RN constituted by the history filtering points generated most recently_i+1The value of (c): when S is_i+1And removing S in RN_iA point S other than_jSatisfies S_jIs leaving S in RN_i+1Nearest track point, and | S_i+1S_jIf | is less than the minimum allowed value Min _ f of Shell.f, S_i+1Value is S_jOtherwise, when there is S in RN_iTwo points S adjacent to each other_j、S_j+1Line segment S of the composition_jS_j+1So that S_jS_j+1Is leaving S in RN_i+1Nearest line segment and passes through S_i+1As S_jS_j+1The vertical foot S of the straight line_⊥Is located on line segment S_jS_j+1Within, and | S_i+1S_⊥|＜|S_i+1S_iI, and I S_i+1S_⊥If | is less than the minimum allowed value Min _ f of Shell.f, S_i+1Value is S_⊥OtherwiseS_i+1The value of (a) is not changed, i.e. no position correction is required; if S is_i+1Not in contact with S^*In case of coincidence, call α_i+1＝Update(α^*,S^*,S_i+1) To update the unit vector in the positive direction of the x-axis

In the counterclockwise direction to the line segment S_i+1S^*Angle alpha of_i+1。

We demonstrate completely how the function GeoF works without considering the effect of the Attach function modification, i.e. the function GeoF omits step S35. FIG. 6 is a diagram illustrating an example of a simple filtering trace generation process:

obtaining an initial measurement point F₀，S₀Direct value of F₀；

② obtaining F₁According to S₀Determining Shell₀According to S₀F₁Calculate Shell₀And Shell₁The deflection angles (the deflection angles are matched) according to | F₁F₀I determining Shell₀According to S₀F₁Calculate C₀(C₀Is located on Shell₀The vertex) to calculate S₁；

Obtaining F₂According to S₁Determining Shell₁Known as Shell₁Has been calculated, according to S₁F₂Calculate Shell₂According to | F₂F₁I determining Shell₁According to S₁F₂Calculate C₁Thereby calculating S₂；

Obtaining F₃According to S₂Determining Shell₂Known as Shell₂Has been calculated, according to S₂F₃Calculate Shell₃According to | F₃F₂I determining Shell₂According to S₂F₃Calculate C₂Thereby calculating S₃；

Fifthly, F is obtained₄According to S₃Determining Shell₃Known as Shell₃Has been calculated, according to S₃F₄Calculate Shell₄According to | F₄F₃I determining Shell₃According to S₃F₄Calculate C₃Thereby calculating S₄；

Obtaining F₅According to S₄Determining Shell₄Known as Shell₄Has been calculated, according to S₄F₅Calculate Shell₅According to | F₅F₄I determining Shell₄In this case because of F₅In Shell₄Inside, so that S does not exist₄F₅And Shell₄Intersection C of the boundaries₄So that S₅And S₄Overlapping;

obtaining the end point F₆According to S₅Determining Shell₅Known as Shell₅Has been calculated, according to S₅F₆Calculate Shell₆According to | F | (no longer used)₆F₅I determining Shell₅According to S₅F₆Calculate C₅Thereby calculating S₆。

So that the broken line LF is equal to F₀F₁F₂F₃F₄F₅F₆Is converted into a broken line LS ═ S through an algorithm₀S₁S₂S₃S₄S₅S₆(wherein S₄、S₅Overlap, so | S₄S₅0). It can be seen that the direction change between adjacent line segments of the original trajectory LF is severe and the trajectories intersect themselves; the direction change between adjacent line segments of the filtering track LS is relatively smooth, such as ≈ F_k-1F_kF_k+1≤∠S_k-1S_kS_k+1(when k is 1,2,3) and self-intersection is released.

The implementation process of the Update function in step S33 and step S35 is as follows:

given a trace point p, it can be the latest measurement point F_i+1Or the latest filtering point S_i+1；

② judgment of S^*P and α^*Whether more than or equal to 0 is simultaneously established, if so, the latest deflection angle is kept as a cache alpha^*Value of the most recent deflection angle of (1): alpha is alpha_i+1＝α^*Transferring to the step IV;

③ passing through the track point p and buffer S^*The latest deflection angle alpha is calculated from the relative position of the nearest filtering point in (b)_i+1＝Atan(S^*,p)；

Fourthly, the latest deflection angle alpha is returned_i+1。

Note: latest measurement point F_i+1Or the latest filtering point S_i+1Relative to the nearest filter point S^*In the case of no movement (coincidence), the latest deflection angle α_i+1The nearest deflection angle alpha to be stored before^*And the filter housing does not turn. However, there is a precondition that the deflection angle has been successfully set recently, i.e., α^*Not ≠ - ∞. If the latest deflection angle is not set in case of coincidence, then α in the else statement block is called₁＝Atan(S^*,S^*) Set at 0.

The implementation procedure of the GetRaw function in step S34 is as follows:

firstly, from the filter mantle Shell_iTo obtain the nearest deflection angle alpha_i＝Shell_i.α；

② judgment of alpha_iIf < 0, if yes, initializing included angle theta_iIf the value is 0, the step IV is carried out;

calculating the nearest deflection angle alpha_iAnd the latest deflection angle alpha_i+1The included angle between: theta_i＝Bettween(α_i,α_i+1)；

Fourthly, from the filter mantle Shell_iTo obtain the nearest filtering point S_i＝Shell_i.s；

Fifthly, calculating the nearest filtration point S_iTo the latest measuring point F_i+1The distance of (c):|S_iF_i+1|＝Distance(S_i,F_i+1)；

sixthly, calculating the nearest filtering point S_i(focus) on the filter mantle and the vertex of the filter mantle form an included angle theta relative to the focus_iPoint C of_iThe distance between: i S_iC_i|＝CosP(Shell_i.f,θ_i)；

Judgment of | S_iF_i+1|≤|S_iC_iIf |, if true, the latest filtering point S_i+1And the nearest filtering point S_iSuperposing: s_i+1＝S_iTurning to step ninthly;

calculating distance ratio coefficient lambda ═ S [ [ lambda ] ]_iC_i|/|S_iF_i+1L, calculation is located on line segment S_iF_i+1Last filter point above: s_i+1＝λ*S_i+(1-λ)*F_i+1；

Ninthly, returning to the latest filtering point S_i+1。

Wherein the implementation process of the Attach function in step S35 is as follows:

judging whether RN.Size () > 0 is true, if not, turning to step ninthly;

acquiring the nearest trace point S in the ring cache queue RN_iGet (rn. size () -1), calculate the distance to the latest filter point S_i+1Distance | S of_iS_i+1|＝Distance(S_i,S_i+1)；

Calculating every point in RN to S_i+1Finding out the point min _ p with the minimum distance and the minimum distance value min _ d;

fourthly, judging that Min _ d is less than Min _ f and Min _ d is less than | S_iS_i+1If the l is simultaneously established, wherein Min _ f is the allowed minimum filter cover f value, and if not, the step is executed to the step (c);

the latest filtration point S_i+1The assignment is min _ p: s_i+1Min _ p, go to step ninj;

sixthly, calculating the line segment formed by every two adjacent points in the RN to the S_i+1Find the two end points min _ pA and min _ pB of the line segment with the smallest distance, and find the most distantSmall distance values min _ l and S_i+1A proportionality coefficient min _ λ of the position of the foot on the line segment with respect to the entire line segment;

seventhly, judging Min _ l < Min _ f and Min _ l < | S_iS_i+1If the | is simultaneously established, if not, turning to the step ninthly;

according to the proportional coefficient obtained in the step (c), calculating the latest filtering point of the latest filtering point on the line segment min _ pA min _ pB by using min _ lambda: s_i+1＝(1-min_λ)*min_pA+min_λ*min_pB；

Ninthly, returning to the latest filtering point S_i+1。

Note: substituting the parameter S of the function in three cases_i+1Conditional screening of the Attach function, i.e. S, can be circumvented_i+1Is not modified, but remains as it is, here called the "escape Attach function": firstly, when the RN has no track point or only one track point, the S is generated₀、S₁The Attach function can be escaped; when S_i+1The value of (D) and the last trace point S in RN_iThe Attach function can be escaped when the overlapping is carried out; ③ Dang | S_iS_i+1| < S >_i+1Set of all points (S) contained in the polyline formed by RN_iExcept for dots) "separated by S_i+1The distance to the nearest point, and the distance value from the point is less than Min _ f, the Attach function can be escaped.

Description of the procedure of the Attach function as shown in the example of fig. 7, two different modifications S are shown separately_i+1Value of S 'before correction'_i+1Denotes S_i+1. A is the case: when calculating S'_i+1Distance RN-S_iMiddle closest point S_uIts distance value | S_uS′_i+1| satisfy | S_uS′_i+1|＜|S_iS′_i+1I and I S_uS′_i+1If Min _ f is | < S'_i+1Is corrected to S_uSo that S after correction_i+1＝S_u(ii) a The Situation of B is: is S'_i+1And RN-S_iUnder the condition that the position relation of the left half graph is not satisfied, a line segment S is found_uS_u+1Is satisfied with S'_i+1To S_uS_u+1The vertical foot S of the straight line_⊥On line segment S_uS_u+1Inner (two points at the boundary are not included), and S_uS_u+1Is a distance S 'in a broken line composed of RN'_i+1Nearest line segment and distance value | S_⊥S′_i+1| satisfy | S_⊥S′_i+1|＜|S_iS′_i+1I and I S_⊥S′_i+1If Min _ f is | < S'_i+1Is corrected to S_⊥So that S after correction_i+1＝S_⊥。

The reason why the subfunction Attach is designed is that the latest filtering point S preliminarily generated by the subfunction GetRaw in the function GeoF_i+1If step S35 is skipped and added directly to LS, the trajectory formed when the moving object traverses a certain path will have such a feature: in the LS, a plurality of sub-tracks running approximately parallel to the path and not related to each other are formed in the portion passing through the path, and the dispersion width sometimes exceeds the path width, resulting in the overall drift of the track. Because the function GetRaw only reacts with Shell_i、F_i+1、α_i+1Correlation; shell_iOnly with F_i+1、F^*(F_i)、S^*(S_i)、α^*(α_i) Correlation; alpha is alpha_i+1And also with alpha^*(α_i)、S^*(S_i)、F_i+1And (4) correlating. So S_i+1And S_i、α_i、F_i、F_i+1These four parameters are directly related. Wherein the nearest measuring point F_iLatest measuring point F_i+1Is independent data, nearest filtering point S_iMost recently the deflection angle alpha_iIs iteratively calculated based on historical trajectories. The more first occurring or generated historical track points (e.g., S)₀、F₁) For calculating the latest filtering point S_i+1The smaller the impact of (c) is, because in an iterative process, historical data impact weights will diminish rapidly, and recent and up-to-date data impacts will be very large. Therefore, the GeoF function not including step S35 is considered to have no ability to track the history data, and to be on the same routeGenerate the path of the path (ls)₁) Then, a return sub-track (denoted as ls) is generated₂) Usually ls₁And ls₂There is a longer sub-track connecting the two, so ls₁Containing pairs of track points ls₂That is, the long history data, vs. ls₂The generation of each filtering point in the filter has little influence, and is considered to be ls₁And ls₂Are almost irrelevant. In this case, therefore, it is ensured that LS generated by GeoF is smooth, but the sub-trajectories passing through the same road segment will be shifted as a whole. As shown in fig. 9, in the case that the original trajectory LF has multiple round trips in the same link, there is a wider (about 10m) offset width (the surrounding mark is identified by a white thick line in the figure); when the filtering trajectory LS (the label "Without S35") generated by the function GeoF not including the step S35 is used, if the same link has multiple round trips, the offset width is also wide, and is equivalent to the width of the LF on the same link; in the case where multiple round trips occur in the same link (the recent round trip data is stored in RN), the filtered trajectory LS generated by the GeoF function including step S35 (identified as "Include S35") has only a narrow (less than 3m) offset width between sub-trajectories, and all sub-trajectories going to and from the same link are highly overlapped. The strip-shaped area where the white bounding area is located in the case including step S35 is almost located in the middle of the strip-shaped area corresponding to LF, and the distribution of the latter is usually a certain symmetric probability distribution about the center line of the link, so that the function Attach can converge the sub-trajectories reciprocating on the same link into the area where the link is located with high probability, and reduce the situation that a certain segment of LS drifts out of the link boundary.

The actual program run screenshot in which the Attach function works is shown in fig. 8, which shows a scenario involving the collection object illustrated by the two schematic diagrams of fig. 5 and 7. In location A and location B, the trace points are corrected to correspond to location A and location B in FIG. 7, respectively. The small dots marked with numbers in the figure represent track points and positions thereof, and after a complete filtering track (a track record called from a database in advance) is generated by using the numbers with the upper row connected by vertical connecting lines to represent the track points and the positions thereof, the track points overlapped at the corresponding positions of the dots areNumber of the cells. The adjacent trace points are connected by line segments to form the complete LS (because each dot may be a plurality of overlapped trace points, a certain line segment forming the broken line LS may be the result of multiple overlapped line segments). Shell_iIs a dark color area with a parabola as a boundary, and the corresponding track point of a small circle with a star-shaped edge circle is Shell_iS focus of_iSince here F_i+1Is located on Shell_iExternal, so S_iF_i+1Cross Shell_iIs bounded by C_iUsing function GetRaw to preliminarily calculate the latest filtering point S before correction_i'₊₁. In the configuration A, a certain historical filtering point corresponding to the circle center position of the circular area is a modified point S found by the Attach function_i+1，S′_i+1At the boundary of the circular region, vector

That is, the correction amount, the moving distance is the radius of the circle and is S 'in RN'_i+1The nearest trace point, notably the one existing closer in the graph, is the historical trace point earlier or S later than the point in RN _i+1_{In-line with the above}The later track points do not participate in the calculation of the track filtering algorithm; in situ B, the center of the capsule-shaped region is the line segment connecting two adjacent points in RN, which is S 'in RN'_i+1The nearest line segment and the vertical foot between the line segments are the corrected points S found by the Attach function_i+1It is most likely not co-located with any historical track point, S'_i+1A vector at the boundary of the capsule-shaped region and orthogonal to the line segment

I.e. the correction amount.

Wherein in the Update function in step S35, the calling condition is S calculated by the Attach function_i+1Not with the nearest filtering point S^*Coincidence, i.e. S^*≠S_i+1. The logic so designed is that if after the Attach function calculation there is S^*And S_i+1Coincidence, then as a parameter inputS before Attach_i+1Has already reacted with S^*(i.e. the last trace point S in RN_i) Coincidently, having been previously analyzed, S_i+1Is the escape Attach function. So in the process of GetRaw function, F_i+1Is located on Shell_iThe latest deflection angle alpha calculated at that time, inside or on the boundary_i+1Is directed to F_i+1More specifically, if the Update function is called in step S35 without setting the conditional filtering, then α appears after the call is forced_i+1In the case of 0, the next filter housing Shell_i+1In relation to Shell_iThe direction of (2) can change suddenly, which is not in accordance with the movement rule and affects the filtering effect.

S36: at this time, the latest filtering point S_i+1It has been calculated that the cache data needs to be updated here: f^*＝F_i+1(original value is F)_i)，S^*＝S_i+1(original value is S)_i)，α^*＝α_i+1(original value is alpha)_i) Adding S_i+1To the ring buffer queue of the most recently generated history filtering point: AddElement (S)_i+1)。

S4: outputting a latest filtering point S_i+1And added to the end of the filter trace LS: add (S) to LS_i+1) At this time, the sequence number i may be set to be self-increasing (i + +), and the process returns to step S2 (if there are measurement points of a certain trajectory that need to be received), otherwise, the algorithm is ended, and a complete filtering trajectory LS is generated.

The embodiment has the advantages that firstly, the requirement on data complexity is low, only the spatial position information of the measuring points based on time sequence is needed, the time stamp of each measuring point is not needed to calculate the motion rule based on the time difference, and the path matching for improving the precision is not needed to be carried out depending on the road information; secondly, the computing resource consumption is low, the result is obtained by computing simple space geometric relationship, the latest filtering point is rapidly computed after the latest measuring point is input, the motion trend is predicted without statistical analysis such as fitting on historical data, and the computing overhead is greatly reduced; and thirdly, the positioning is accurate, the stability is high, the track is more consistent with the motion rule, the positions are corrected to the near small-range space field, the situation that the measured data are not excessively deviated, a continuous and smooth advancing track is formed, the situations of self-intersection, front-back reciprocating, left-right zigzag swing and the like do not occur, and the motion trend which is consistent with the physical inertia characteristics, such as stable rest, uniform acceleration and deceleration, smooth turning and the like can be reflected.

In this embodiment, when the positioning accuracy error of the positioning data is 5m to 10m, the following effects can be achieved under the parameter settings of Min _ f being 5, Max _ f being 100, Min _ d being 30, Max _ d being 150, and N being 1000:

as shown in fig. 10, when the moving object is actually stationary, the received positioning data from the red marked measurement points appears to be distributed in a lump, the distribution range is in a planar space with a span of about 8m, and random drift with an amplitude of about 1m to 3m appears between the points, forming a self-intersecting trajectory (upper half graph); the filtering algorithm maps the point set of the globoid distribution to a single filtering point marked by blue, the position of the point is a certain point in the globoid distribution space, and the filtering point keeps moving along with the drift of the measuring point in the space, so as to accord with the static actual situation (the lower half picture).

As shown in fig. 11, in the case of a moving object actually moving straight, the received positioning data given by the red marked measuring point drifts along the left and right direction of the straight line and reciprocates back and forth to form a zigzag track (upper half graph) with a local self-intersecting small lump with a size of about 4 m; the filtering algorithm corrects the drawn track of the measuring points to a smooth track (lower half picture) which is formed by a filtering point marked by blue and continuously advances forwards without reciprocating.

As shown in fig. 12, in the process from actual rest to acceleration of the moving object, the received positioning data given by the red marked measuring points appear to be distributed like a lump when starting to gradually accelerate, and a trajectory (upper half diagram) with a self-intersecting head part of a small lump with a size of about 6m and then winding is formed; the filtering algorithm corrects the locus drawn by the measuring points to a smooth locus (lower half picture) which is formed by filtering points marked by blue and gradually accelerates and continuously advances forwards.

As shown in fig. 13, in the process of decelerating the moving object to be stationary from practice, the received positioning data given by the red marked measuring points appears in a lump distribution when gradually decelerating to be stationary, so as to form a track (upper half graph) with a zigzag travel and a self-intersected small lump with a size of about 5m at the tail; the filtering algorithm corrects the locus drawn by the measuring points to a smooth locus (lower half picture) which is formed by filtering points marked by blue and gradually decelerates and continuously advances forwards.

As shown in fig. 14, in the case of a real speed-change turning, the received positioning data from the red marked measurement points shows the distribution of the lumps in the turning processes of "turning after deceleration" and "turning after acceleration", the distribution range is in a plane space with a span of about 12m, and the random drift with an amplitude of 1 m-2 m occurs between the internal points of the lumps, so as to form a self-intersecting trajectory (upper half graph) during the zigzag travel and turning; the filtering algorithm corrects the track drawn by the measuring points to a smooth track (lower half picture) which is formed by filtering points marked by blue and continuously advances forwards, is gradually decelerated, turns at the break and gradually accelerated, and does not reciprocate in the process.

As shown in fig. 15, a structure diagram of an apparatus for correcting an aircraft track positioning data offset according to a fifth exemplary embodiment of the present invention includes:

the preprocessing unit 1501 is configured to empty the history cache information before performing the operation of correcting the current track positioning data offset of the aircraft;

a receiving unit 1502, configured to receive real-time measurement point information of an aircraft sent by a distributed positioning system;

the filtering unit 1503 is used for determining the current parabola and the real-time deflection angle information which are used as the filter cover according to the real-time measuring point information and the current cache information; calculating and generating real-time filtering point information corresponding to the real-time measuring point according to the real-time measuring point information, the current parabola and the real-time deflection angle information; the current cache information includes: the cache queue comprises last measurement point cache information, last filtering point cache information, last deflection angle cache information and cache queues of all filtering point information between an initial filtering point and a last filtering point of the current flight path; the last deflection angle cache information is an included angle formed between a unit vector in the positive direction of the axis and a connecting line from the focus of the current parabola to the vertex of the current parabola along the anticlockwise direction;

an updating unit 1504, configured to update current cache information according to the real-time measurement point information, the real-time filtering point information, and the real-time deflection angle information;

the track correction unit 1505 is used for obtaining a corrected track of the current track of the aircraft according to all the filtering point information of the current track.

Preferably: the filtering unit 1503 is specifically configured to: determining a common point of a line segment formed by connecting the cache information of the previous filtering point to the information of the real-time measuring point and the boundary of the current parabola; and forming an included angle between vectors by the unit vector in the positive axial direction and a common point connecting line from the focus of the current parabola to the boundary of the current parabola along the counterclockwise direction to serve as real-time deflection angle cache information.

Preferably: the filtering unit 1503 is specifically further configured to: and determining the size of the current parabola according to the length of a connecting line between the cache information of the previous measuring point and the real-time measuring point information.

Preferably: the filtering unit 1503 is specifically further configured to: calculating an included angle between a first deflection angle information and a real-time deflection angle information, calculating a distance between a common point of a current parabola boundary and a focus of the current parabola, if the common point exists, determining real-time filtering point information on a line segment formed by the focus of the current parabola and the real-time measuring point information, wherein the real-time filtering point information meets the condition that the distance between a last filtering point information and the common point is equal to the distance between the real-time filtering point information and the real-time measuring point, and if the common point does not exist, the real-time filtering point information is the last filtering point information.

Preferably: the device for correcting the aircraft track positioning data offset further comprises a filtering point information correcting unit 1506, which is used for: correcting the real-time filtering point information according to the cache queue, so that the updating unit updates the current cache information according to the corrected real-time filtering point information and the track correcting unit obtains the corrected track of the current track of the aircraft according to the corrected real-time filtering point information

The corrected path is continuous and smooth, the conditions of sudden change of the running direction and the displacement degree can not occur, and the continuous advancing trend of the aircraft can be well reflected; each point in the corrected path has no large deviation with the corresponding actual measurement data, the error is controlled in a small range, and the error is matched with the road surface in space as much as possible; the corrected path is to filter out the self-intersection of the track caused by the drift of the positioning data error, eliminate the conglomeration distribution, height folding and other conditions of the track in a smaller space range, prevent the reciprocating motion along the advancing direction of the road surface, and enable the form of the track to accord with the physical law of the motion of the aircraft, such as the motion characteristics of large-inertia objects, such as continuous advancing, slow acceleration and deceleration, large turning amplitude, incapability of easily changing the direction and the like. The dependence of the correction device on input data is simple enough, only a measuring point sequence generated according to the time sequence and coming to positioning data needs to be input, and reasonable corrected positioning points are calculated through analysis of the space geometric relationship of the points without depending on an additional data set, so that the increase of use, maintenance and calculation difficulty is avoided.

The above description is only for the specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims (10)

1. A method for correcting an aircraft track positioning data offset, comprising:

clearing historical cache information before correcting the current track positioning data offset of the aircraft;

receiving real-time measuring point information of an aircraft sent by a distributed positioning system;

determining current parabola and real-time deflection angle information used as a filter cover according to the real-time measuring point information and the current cache information; the current cache information includes: the cache queue comprises last measurement point cache information, last filtering point cache information, last deflection angle cache information and cache queues of all filtering point information between an initial filtering point and a last filtering point of the current flight path; the last deflection angle cache information is an included angle formed between a unit vector in the positive direction of the x axis and a connecting line from the focus of the current parabola to the vertex of the current parabola along the anticlockwise direction;

calculating and generating real-time filtering point information corresponding to the real-time measuring point according to the real-time measuring point information, the current parabola and the real-time deflection angle information;

updating current cache information according to the real-time measuring point information, the real-time filtering point information and the real-time deflection angle information;

and obtaining the corrected flight path of the current flight path of the aircraft according to all the filtering point information of the current flight path.

2. The method of correcting for aircraft track positioning data drift of claim 1, wherein: determining real-time deflection angle information according to the real-time measuring point information and the current cache information comprises:

determining a common point of a line segment formed by connecting the cache information of the previous filtering point to the information of the real-time measuring point and the boundary of the current parabola;

and forming an included angle between vectors by the unit vector in the positive axial direction and a common point connecting line from the focus of the current parabola to the boundary of the current parabola along the counterclockwise direction to serve as real-time deflection angle cache information.

3. The method of correcting for aircraft track positioning data drift of claim 2, wherein: determining a current parabola for use as a filter sock from the real-time measurement point information and the current cache information comprises:

and determining the size of the current parabola according to the length of a connecting line between the cache information of the previous measuring point and the real-time measuring point information.

4. A method of correcting for aircraft track positioning data drift as defined in claim 3, wherein: the step of calculating and generating real-time filtering point information corresponding to the real-time measuring point according to the real-time measuring point information, the current parabola and the real-time deflection angle information comprises the following steps:

calculating an included angle between a first deflection angle information and a real-time deflection angle information, calculating a distance between a common point of a current parabola boundary and a focus of the current parabola, if the common point exists, determining real-time filtering point information on a line segment formed by the focus of the current parabola and the real-time measuring point information, wherein the real-time filtering point information meets the condition that the distance between a last filtering point information and the common point is equal to the distance between the real-time filtering point information and the real-time measuring point, and if the common point does not exist, the real-time filtering point information is the last filtering point information.

5. Method for the correction of an aircraft track positioning data offset according to any of claims 1 to 4, characterized in that: the step of calculating and generating real-time filtering point information corresponding to the real-time measuring point according to the real-time measuring point information, the current parabola and the real-time deflection angle information further comprises the following steps:

and correcting the real-time filtering point information according to the cache queue so as to update the current cache information according to the corrected real-time filtering point information and obtain the corrected flight path of the current flight path of the aircraft.

6. An apparatus for correcting an aircraft track positioning data offset, comprising:

the preprocessing unit is used for emptying historical cache information before the correction operation of the current track positioning data offset of the aircraft is carried out;

the receiving unit is used for receiving the real-time measuring point information of the aircraft sent by the distributed positioning system;

the filtering unit is used for determining the current parabola and the real-time deflection angle information which are used as the filtering cover according to the real-time measuring point information and the current cache information; calculating and generating real-time filtering point information corresponding to the real-time measuring point according to the real-time measuring point information, the current parabola and the real-time deflection angle information; the current cache information includes: the cache queue comprises last measurement point cache information, last filtering point cache information, last deflection angle cache information and cache queues of all filtering point information between an initial filtering point and a last filtering point of the current flight path; the last deflection angle cache information is an included angle formed between a unit vector in the positive direction of the axis and a connecting line from the focus of the current parabola to the vertex of the current parabola along the anticlockwise direction;

the updating unit is used for updating the current cache information according to the real-time measuring point information, the real-time filtering point information and the real-time deflection angle information;

and the track correction unit is used for obtaining the corrected track of the current track of the aircraft according to all the filtering point information of the current track.

7. The apparatus for correcting for aircraft track positioning data drift of claim 6, wherein: the filter unit is specifically configured to: determining a common point of a line segment formed by connecting the cache information of the previous filtering point to the information of the real-time measuring point and the boundary of the current parabola; and forming an included angle between vectors by the unit vector in the positive axial direction and a common point connecting line from the focus of the current parabola to the boundary of the current parabola along the counterclockwise direction to serve as real-time deflection angle cache information.

8. The apparatus for correcting for aircraft track positioning data drift of claim 7, wherein: the filter unit is specifically further configured to: and determining the size of the current parabola according to the length of a connecting line between the cache information of the previous measuring point and the real-time measuring point information.

9. The apparatus for correcting for aircraft track positioning data drift of claim 8, wherein: the filter unit is specifically further configured to: calculating an included angle between a first deflection angle information and a real-time deflection angle information, calculating a distance between a common point of a current parabola boundary and a focus of the current parabola, if the common point exists, determining real-time filtering point information on a line segment formed by the focus of the current parabola and the real-time measuring point information, wherein the real-time filtering point information meets the condition that the distance between a last filtering point information and the common point is equal to the distance between the real-time filtering point information and the real-time measuring point, and if the common point does not exist, the real-time filtering point information is the last filtering point information.

10. An aircraft track positioning data offset correction device according to any one of claims 6-9, characterized in that: further comprising:

the filtering point information correcting unit is used for: and correcting the real-time filtering point information according to the cache queue, so that the updating unit updates the current cache information according to the corrected real-time filtering point information and the track correcting unit obtains the corrected track of the current track of the aircraft according to the corrected real-time filtering point information.

Images