A mobile target trajectory data generation method

Abstract

The present application relates to big data analysis technical field, specifically to a kind of mobile target trajectory data generation method, the method includes obtaining point track sampling interval time, point track six-dimensional data sequence obtained by target point track three-dimensional coordinate data sequence and target point track speed sequence combination;Using sliding window extraction method with the first data window as window with pre-set, with pre-set first data interval as step, in point track six-dimensional data sequence, first fitting data sequence to the N fitted data sequence is extracted;Using kinematics equation to construct first predicted point track three-dimensional coordinate data sequence to the N predicted point track three-dimensional coordinate data sequence, the optimal acceleration variable matrix is calculated using particle swarm optimization algorithm, and then the mobile target trajectory data is calculated.The present application establishes optimization model according to discrete data, and then realizes the calculation of mobile target trajectory data.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of big data analysis, and particularly relates to a mobile target trajectory data generation method. BACKGROUND

[0002] In a modern mobile target tracking scenario, in order to accurately position and predict the motion trajectory of a mobile target, it is usually necessary to collect the position and speed information of the target at consecutive time points. In an actual application scenario, a monitoring system collects three-dimensional coordinate data and speed data of a mobile target point trajectory at a certain sampling time interval, and displays them in the form of discrete points. However, it is difficult to accurately restore the continuous motion trajectory of the target by relying on discrete point data alone, and therefore it is necessary to process the discrete data by means of a state estimation algorithm to generate continuous trajectory data.

[0003] In order to obtain the continuous motion trajectory of a mobile target, two types of key information are required: one is the initial condition for kinematic calculation, that is, the initial position and speed data of the target; and the other is the acceleration variation of the target during the motion process. In a mobile target monitoring application, the data collected at the initial time is usually accurate and can be regarded as error-free or less error. However, as time goes by, due to factors such as measurement device precision and environmental interference, the collected position and speed data often introduces errors, leading to deviations in subsequent trajectory calculation.

[0004] In the prior art, a Kalman filter algorithm is often used to smooth and estimate the state of the discrete data with errors. However, this method requires that the data errors conform to a normal distribution, and is sensitive to the system model and noise statistical characteristics. In actual applications, the error distribution often does not conform to the ideal assumption, leading to a decline in algorithm performance. Therefore, there is an urgent need for a trajectory data generation method that has lower requirements for error distribution assumptions and is more adaptable. SUMMARY

[0005] (1) Technical problem to be solved

[0006] The present application aims to provide a mobile target trajectory data generation method to calculate continuous mobile target trajectory data according to target point trajectory three-dimensional coordinates and speed with errors.

[0007] (2) Technical solution

[0008] To achieve the above-mentioned purpose, the present application provides a mobile target trajectory data generation method, which uses a state estimation algorithm to calculate continuous mobile target trajectory data from a target point trajectory three-dimensional coordinate data sequence and a target point trajectory speed sequence with errors. The method comprises the following steps:

[0009] S1, obtaining first data; the first data comprises tracklet sampling interval time, tracklet six-dimensional data sequence composed of target tracklet three-dimensional coordinate data sequence and target tracklet velocity sequence.

[0010] S2, adopting sliding window extraction method to extract first fitting data sequence to the (N-1)th fitting data sequence from the tracklet six-dimensional data sequence by taking the first data window as a window and taking the first data interval as a step. N fitting data sequence; is a first constant.

[0011] S3, constructing first predicted tracklet three-dimensional coordinate data sequence to the (N-1)th predicted tracklet three-dimensional coordinate data sequence according to the first fitting data sequence to the (N-1)th fitting data sequence and the acceleration variable matrix initialized in advance. N predicted tracklet three-dimensional coordinate data sequence; respectively taking the first predicted tracklet three-dimensional coordinate data sequence to the (N-1)th predicted tracklet three-dimensional coordinate data sequence and the first fitting data sequence first three rows to the (N-1)th fitting data sequence first three rows as input variables of the kinematic equation. N predicted tracklet three-dimensional coordinate data sequence; respectively taking the first predicted tracklet three-dimensional coordinate data sequence to the (N-1)th predicted tracklet three-dimensional coordinate data sequence and the first fitting data sequence first three rows to the (N-1)th fitting data sequence first three rows as input variables of the kinematic equation. N predicted tracklet three-dimensional coordinate data sequence; respectively taking the first predicted tracklet three-dimensional coordinate data sequence to the (N-1)th predicted tracklet three-dimensional coordinate data sequence and the first fitting data sequence first three rows to the (N-1)th fitting data sequence first three rows as input variables of the kinematic equation. N predicted tracklet three-dimensional coordinate data sequence; respectively taking the first predicted tracklet three-dimensional coordinate data sequence to the (N-1)th predicted tracklet three-dimensional coordinate data sequence and the first fitting data sequence first three rows to the (N-1)th fitting data sequence first three rows as input variables of the kinematic equation.

[0012] S4, calculating the moving target trajectory data according to the optimal acceleration variable matrix and the tracklet six-dimensional data sequence.

[0013] Further, the method for extracting the first fitting data sequence to the (N-1)th fitting data sequence from the tracklet six-dimensional data sequence by taking the first data window as a window and taking the first data interval as a step comprises: N

[0014] placing the starting point of the first data window at the starting point of the tracklet six-dimensional data sequence; and taking the overlapping part of the first data window and the tracklet six-dimensional data sequence as the first fitting data sequence.

[0015] moving the first data window along the tracklet six-dimensional data sequence by the first data interval; times; taking the overlapping part of the first data window and the tracklet six-dimensional data sequence after the (n-1)th movement as the nth fitting data sequence. times; taking the overlapping part of the first data window and the tracklet six-dimensional data sequence after the (n-1)th movement as the nth fitting data sequence. fitting data sequence, obtaining the first fitting data sequence to the (N-1)th fitting data sequence. N fitting data sequence; is an integer variable with a value of 1 to N; the value of N is greater than 3.

[0016] moving the first data window along the tracklet six-dimensional data sequence by the first data interval; N ​​The fitted data sequence is represented as to ; wherein the first fitted data sequence is represented as .

[0017] .

[0018] wherein, is an integer variable taking values from 1 to ; is an integer variable taking values from 1 to ; represents the window length of the first data window; represents the X-axis first measured point trace coordinate of ; represents the Y-axis first measured point trace coordinate of ; represents the Z-axis first measured point trace coordinate of ; represents the X-axis first measured velocity of ; represents the Y-axis first measured velocity of ; represents the Z-axis first measured velocity of .

[0019] Further, the method for constructing the first predicted point trace three-dimensional coordinate data sequence to the N predicted point trace three-dimensional coordinate data sequence according to the first fitted data sequence to the N fitted data sequence and the pre-initialized acceleration variable matrix comprises:

[0020] The acceleration variable matrix is denoted as ; wherein is represented as:

[0021] .

[0022] wherein, represents the X-axis first acceleration; represents the Y-axis first acceleration; represents the Z-axis first acceleration.

[0023] According to the first fitted data sequence to the NThe first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. N The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix.

[0024] .

[0025] wherein, The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix.

[0026] Further, the method for constructing the first predicted trajectory three-dimensional coordinate data sequence to the nth predicted trajectory three-dimensional coordinate data sequence from the first fitted data sequence to the nth fitted data sequence and the acceleration variable matrix comprises: N The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. N The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix.

[0027] The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. N The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. N The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix.

[0028] The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. .

[0029] wherein, The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix.

[0030] Further, the method for calculating the optimal value of the acceleration variable matrix by taking the minimum value of the Euclidean distance between the first predicted trajectory three-dimensional coordinate data sequence to the nth predicted trajectory three-dimensional coordinate data sequence and the first three rows of the first fitted data sequence to the first three rows of the nth fitted data sequence as the target, and using the particle swarm optimization algorithm comprises: N The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. N The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix.

[0031] The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. N The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. N The first predicted trajectory three-dimensional coordinate data sequence is constructed from the fitted data sequence and an acceleration variable matrix. NThe objective function; the Euclidean distance calculation formula is:

[0032] .

[0033] in, Indicates the first Objective function.

[0034] Based on the first fitted data sequence to the... N The first constraint condition is constructed by fitting the data sequence and the acceleration variable matrix. N Constraints.

[0035] Using the first objective function to the... N The objective function is to minimize its value, and the first constraint is used to determine the next constraint. N Given the constraints, the optimal values ​​of the acceleration variable matrix are calculated using the particle swarm optimization algorithm, denoted as the optimal acceleration variable matrix. .

[0036] .

[0037] Furthermore, the step of adapting the first fitted data sequence to the... N The first constraint condition is constructed by fitting the data sequence and the acceleration variable matrix. N Methods for setting constraints include:

[0038] Based on the first fitted data sequence to the... N The first constraint condition is constructed by fitting the data sequence and the acceleration variable matrix. N Constraints; where the first The constraints are:

[0039] .

[0040] in, This indicates the preset upper limit of the X-axis speed measurement error; This indicates the pre-set upper limit of the Y-axis speed measurement error; This indicates the pre-set upper limit of the Z-axis speed measurement error; This indicates the pre-set upper limit of the X-axis dot measurement error; This indicates the preset upper limit of the Y-axis dot measurement error; This indicates the pre-set upper limit of the Z-axis dot measurement error.

[0041] Furthermore, the method for calculating the trajectory data of the moving target based on the optimal acceleration variable matrix and the six-dimensional data sequence of the point trace includes:

[0042] The three-dimensional coordinate data sequence of the target point is divided into columns to obtain the first data column to the second column.R data columns; wherein ; are first data column to the R data column is assigned an acceleration set initialized as an empty set in advance, obtaining first acceleration set to the R acceleration set; the optimal acceleration variable matrix is divided by column to obtain first optimal acceleration data column to the N optimal acceleration data column.

[0043] traverse to ; the first optimal acceleration data column is cloned to the first acceleration set to the acceleration set; wherein represents the first data interval.

[0044] respectively, the first acceleration set to the R acceleration set in all optimal acceleration data columns is averaged to obtain X-axis first average acceleration to X-axis R average acceleration; respectively, the first acceleration set to the R acceleration set in all optimal acceleration data columns is averaged to obtain Y-axis first average acceleration to Y-axis R average acceleration; respectively, the first acceleration set to the R acceleration set in all optimal acceleration data columns is averaged to obtain Z-axis first average acceleration to Z-axis R average acceleration.

[0045] The first column data in the point trajectory six-dimensional data sequence is represented as:

[0046] ;

[0047] wherein, represents the target initial X-axis position coordinate; represents the target initial Y-axis position coordinate; represents the target initial Z-axis position coordinate; represents the target initial X-axis speed; represents the target initial Y-axis speed; represents the target initial Z-axis speed.

[0048] According to the X-axis first average acceleration to the X-axis R average acceleration, the Y-axis first average acceleration to the Y-axis R average acceleration, the Z-axis first average acceleration to the Z-axis RThe first column of the six-dimensional data sequence of average acceleration and point trace The trajectory data of the moving target is calculated.

[0049] Furthermore, the step of calculating the first average acceleration along the X-axis to the second average acceleration along the X-axis... R Average acceleration, first average acceleration along the Y-axis to the second average acceleration along the Y-axis R Average acceleration, first average acceleration along the Z-axis to the second average acceleration along the Z-axis R The first column of the six-dimensional data sequence of average acceleration and trace Methods for calculating the trajectory data of a moving target include:

[0050] Based on the first average acceleration along the X-axis to the X-axis... R The average acceleration is used to calculate the actual acceleration curve along the X-axis; the actual acceleration curve along the X-axis is as follows:

[0051] .

[0052] in, This represents the actual acceleration curve along the X-axis. Indicates and The corresponding time interval, The value range is 0 to ; Indicates the rounding up symbol; Indicates the floor function; Indicates the X-axis number Average acceleration; Indicates the X-axis number Average acceleration; Indicates the X-axis number Average acceleration.

[0053] Based on the first average acceleration along the Y-axis to the second... R The average acceleration is used to calculate the actual acceleration curve along the Y-axis; the actual acceleration curve along the Y-axis is as follows:

[0054] .

[0055] in, This represents the actual acceleration curve along the Y-axis. Indicates the Y-axis number Average acceleration; Indicates the Y-axis number Average acceleration; Indicates the Y-axis number Average acceleration.

[0056] Based on the first average acceleration along the Z-axis to the Z-axis... RThe average acceleration calculation yields the actual acceleration curve along the Z-axis; the actual acceleration curve along the Z-axis is as follows:

[0057] .

[0058] in, This represents the actual acceleration curve along the Z-axis; Indicates the Z-axis number Average acceleration; Indicates the Z-axis number Average acceleration; Indicates the Z-axis number Average acceleration.

[0059] The first column of data in the six-dimensional data sequence of the point trace is used as the initial boundary condition for kinematic calculation. Based on the actual acceleration curves of the X-axis, Y-axis, and Z-axis, the trajectory data of the moving target is calculated using numerical integration and fitting algorithms.

[0060] (3) Beneficial effects

[0061] Compared with the prior art, the beneficial effects of the present invention are:

[0062] 1. Construct the first predicted point trace three-dimensional coordinate data sequence based on the kinematic equations up to the next... N Predict the three-dimensional coordinate data sequence of the point trace, and then use the first predicted three-dimensional coordinate data sequence of the point trace to the second predicted three-dimensional coordinate data sequence. N The predicted point trace 3D coordinate data sequence and the first three rows to the first fitted data sequence N The goal is to minimize the Euclidean distance of the first three rows of the fitted data sequence. The optimal value of the acceleration variable matrix is ​​calculated using the particle swarm optimization algorithm. This method relies on physical laws rather than statistical assumptions, which makes the algorithm robust under various error distribution conditions.

[0063] 2. Using the sliding window extraction method for data extraction preserves the temporal characteristics of the data while increasing data redundancy, which helps to offset the influence of random errors, reduces dependence on the error characteristics of a single measurement point, and improves robustness to non-normally distributed errors. Attached Figure Description

[0064] Figure 1 This is a flowchart of a method for generating trajectory data of a moving target according to Embodiment 1 of the present invention. Detailed Implementation

[0065] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0066] Before providing examples, it is necessary to describe the application scenario of this invention. In a moving target monitoring scenario under complex conditions, the monitoring center needs to track a high-speed moving target. An external sensing device acquires the target's three-dimensional coordinate data sequence and velocity sequence at a sampling interval of 0.5 seconds. Because the target gradually moves away from the sensing device during its movement, all measurement data except the initial sampling data contain errors and are discrete. This embodiment uses a moving target trajectory data generation method to process the data and calculate the continuous motion trajectory of the moving target.

[0067] Example 1: As Figure 1 As shown, this embodiment provides a method for generating trajectory data of a moving target. It employs a state estimation algorithm to calculate continuous trajectory data of a moving target based on a sequence of three-dimensional coordinate data of target points with errors and a sequence of velocity data of target points. The method includes the following steps:

[0068] S1, acquire the first data; the first data includes the sampling interval time of the target trace and the six-dimensional data sequence of the target trace obtained by combining the three-dimensional coordinate data sequence of the target trace and the velocity sequence of the target trace.

[0069] S2, using the sliding window extraction method with a pre-set first data window as the window and a pre-set first data interval as the step size, extracts the first fitted data sequence up to the sixth dimension of the point trace six-dimensional data sequence. N Fitting the data sequence; This is a pre-defined first constant.

[0070] S3, based on the first fitted data sequence to the... N The fitted data sequence and the pre-initialized acceleration variable matrix are used to construct the first predicted point trace three-dimensional coordinate data sequence using kinematic equations, up to the second... N Predicted point trace 3D coordinate data sequence; respectively from the first predicted point trace 3D coordinate data sequence to the second... N The predicted point trace 3D coordinate data sequence and the first three rows to the first fitted data sequence N The goal is to minimize the Euclidean distance of the first three rows of the fitted data sequence. The optimal value of the acceleration variable matrix is ​​calculated using the particle swarm optimization algorithm and is denoted as the optimal acceleration variable matrix.

[0071] S4, calculating the moving target trajectory data according to the optimal acceleration variable matrix and the point trajectory six-dimensional data sequence.

[0072] Exemplarily, the first data includes point trajectory sampling interval time and point trajectory six-dimensional data sequence. The point trajectory sampling interval time is 0.5 seconds. The monitoring center collected 29 groups of data in the tracking process at intervals of 0.5 seconds to obtain target point trajectory three-dimensional coordinate data sequence and target point trajectory speed sequence, and then combined to obtain point trajectory six-dimensional data sequence. The point trajectory six-dimensional data sequence contains target point trajectory three-dimensional coordinate data and target point trajectory speed of the moving target at different times. For example, at the initial time of 0 seconds, the collected target point trajectory three-dimensional coordinate data is (998, 501, 299) meters, and the target point trajectory speed is (201, 50, 29) meters / second. Among them, the three variables in the target point trajectory three-dimensional coordinate data respectively represent the position coordinates of the target point trajectory on the X axis, the Y axis and the Z axis. The three variables in the target point trajectory speed respectively represent the speed components of the target point trajectory on the X axis, the Y axis and the Z axis. In this way, the point trajectory six-dimensional data sequence is obtained.

[0073] The first data window is set to 5, the first data interval is set to 3, and the sliding window extraction method is used with a window of 5 and a step of 3 to extract the first fitting data sequence to the ninth fitting data sequence in the point trajectory six-dimensional data sequence. The specific method is as follows: the starting point of the first data window is placed at the starting point of the point trajectory six-dimensional data sequence, and the first fitting data sequence is formed by intercepting the first 5 point trajectory data. Then move the data window 3 points backward, intercept the 4th to 8th point trajectory data in the point trajectory six-dimensional data sequence to form the second fitting data sequence. In this way, 9 groups of fitting data sequences, i.e. the first fitting data sequence to the ninth fitting data sequence, are obtained. The setting of the first data window and the first data interval is related to the expected acceleration change of the target point trajectory. The more frequent the expected acceleration change of the target point trajectory is, the smaller the first data window and the first data interval are. The purpose of setting the first data window and the first data interval is to reduce the data processing amount as much as possible while accurately simulating the target point trajectory.

[0074] The acceleration variable matrix is initialized in advance, and the initial values of the accelerations in the X-axis, Y-axis and Z-axis directions are all set to 0 m / s2. According to the first fitting data sequence to the ninth fitting data sequence and the initialized acceleration variable matrix, the first predicted point trail three-dimensional coordinate data sequence to the ninth predicted point trail three-dimensional coordinate data sequence are constructed by applying kinematic equations. The specific method is as follows: for each fitting data sequence, the predicted positions of the subsequent points are calculated based on the position and velocity of the initial point and the preset acceleration value. For example, for the first fitting data sequence, the position and velocity data of the first point and the initialized acceleration values in the X-axis, Y-axis and Z-axis directions are taken to calculate the predicted position coordinates of the subsequent four points. The predicted point trail coordinates are compared with the actually measured point trail coordinates to calculate the Euclidean distance. Taking the first fitting data sequence as an example, the Euclidean distance between the predicted point trail coordinate sequence and the first three rows of the actual point trail coordinate sequence, i.e., the three-dimensional position coordinates, is calculated. The particle swarm optimization algorithm is adopted to adjust the values of the acceleration variable matrix through iterative optimization until the optimal value of the acceleration variable matrix that can minimize the Euclidean distance between the predicted point trail coordinates and the actual point trail coordinates, i.e., the optimal acceleration variable matrix, is obtained. In the particle swarm optimization algorithm, the number of particles is set to 50, and the maximum number of iterations is set to 200.

[0075] Finally, the moving target trajectory data is calculated according to the optimal acceleration variable matrix and the point trail six-dimensional data sequence. The moving target trajectory data is a continuous curve of the three-dimensional coordinates of the moving target changing with time.

[0076] Further, the first fitting data sequence to the ninth fitting data sequence are extracted from the point trail six-dimensional data sequence by the sliding window extraction method with the first data window preset as the window and the first data interval preset as the step. N The method for extracting the fitting data sequence comprises the following steps:

[0077] The starting point of the first data window is placed at the starting point of the point trail six-dimensional data sequence; and the overlapping part of the first data window and the point trail six-dimensional data sequence is taken as the first fitting data sequence.

[0078] The first data window is moved along the point trail six-dimensional data sequence by the first data interval as the step the first data window and the point trail six-dimensional data sequence after the first movement the overlapping part of the first data window and the point trail six-dimensional data sequence after the second movement is recorded as the second fitting data sequence. The first fitting data sequence to the ninth fitting data sequence are extracted from the point trail six-dimensional data sequence by the sliding window extraction method with the first data window preset as the window and the first data interval preset as the step. N The first fitting data sequence to the ninth fitting data sequence are extracted from the point trail six-dimensional data sequence by the sliding window extraction method with the first data window preset as the window and the first data interval preset as the step. is an integer variable with a value of 1 to The value of n is greater than 3. The value of n is greater than 3.

[0079] The first fitting data sequence to the ninth fitting data sequenceN The fitted data sequence is represented as to ; wherein the first The fitted data sequence is represented as

[0080] .

[0081] wherein, is an integer variable taking values from 1 to ; is an integer variable taking values from 1 to ; represents the window length of the first data window; represents the X-axis first measured trajectory coordinate of ; represents the Y-axis first measured trajectory coordinate of ; represents the Z-axis first measured trajectory coordinate of ; represents the X-axis first measured velocity of ; represents the Y-axis first measured velocity of ; represents the Z-axis first measured velocity of .

[0082] Exemplarily, the starting point of the first data window is set at the starting point of the trajectory six-dimensional data sequence, and the data points at time 0s, 0.5s, 1.0s, 1.5s and 2.0s are cut from the trajectory six-dimensional data sequence to obtain the first fitted data sequence. The first fitted data sequence contains the target trajectory three-dimensional coordinate data and the target trajectory velocity information measured during 0~2s. For example, at 0s, the target trajectory three-dimensional coordinate data is (998, 501, 299) meters, and the target trajectory velocity is (201, 50, 29) meters / s. The first data window is moved along the trajectory six-dimensional data sequence with a step of 3. After the first movement, the first data window covers the fourth to eighth data points in the trajectory six-dimensional data sequence, i.e., the data at time 1.5s, 2.0s, 2.5s, 3.0s and 3.5s, and the second fitted data sequence is extracted. In this way, the first data window is moved along the trajectory six-dimensional data sequence for 8 times to obtain the first fitted data sequence to the ninth fitted data sequence. The first fitted data sequence to the ninth fitted data sequence are represented as to ​By using the sliding window extraction method, it is ensured that each fitting data sequence contains sufficient data points, which provides a reliable data basis for subsequent acceleration calculation. At the same time, since the first data interval is set to 3, it means that there is partial overlapping data between adjacent fitting data sequences, which is beneficial to maintain the continuity and smoothness of the acceleration calculation.

[0083] Further, the first predicted point trail three-dimensional coordinate data sequence to the N predicted point trail three-dimensional coordinate data sequence are constructed according to the first fitting data sequence to the N predicted point trail three-dimensional coordinate data sequence and the acceleration variable matrix obtained by pre-initialization.

[0084] The acceleration variable matrix is denoted as ; wherein is expressed as:

[0085] .

[0086] wherein, represents the X-axis first acceleration; represents the Y-axis first acceleration; represents the Z-axis first acceleration.

[0087] The first predicted point trail three-dimensional coordinate data sequence to the N predicted point trail three-dimensional coordinate data sequence are constructed according to the first fitting data sequence to the N predicted point trail three-dimensional coordinate data sequence and the acceleration variable matrix. The first predicted point trail three-dimensional coordinate data sequence to the predicted point trail three-dimensional coordinate data sequence are denoted as:

[0088] .

[0089] wherein, represents the first predicted point trail three-dimensional coordinate data sequence; represents the X-axis first predicted point trail coordinate of ; represents the Y-axis first predicted point trail coordinate of ; represents the Z-axis first predicted point trail coordinate of .

[0090] Exemplarily, the acceleration variable matrix is set as a 3-row 9-column matrix, representing the values of the X-axis, Y-axis and Z-axis acceleration corresponding to the first fitting data sequence to the ninth fitting data sequence. All elements in the acceleration variable matrix are initialized to 0 m / s2. The first predicted trajectory three-dimensional coordinate data sequence to the ninth predicted trajectory three-dimensional coordinate data sequence are constructed according to the pre-initialized acceleration variable matrix and the first fitting data sequence to the ninth fitting data sequence.

[0091] Further, the method of constructing the first predicted trajectory three-dimensional coordinate data sequence to the ninth predicted trajectory three-dimensional coordinate data sequence according to the first fitting data sequence to the ninth fitting data sequence and the acceleration variable matrix includes: N N The method of constructing the first predicted trajectory three-dimensional coordinate data sequence to the ninth predicted trajectory three-dimensional coordinate data sequence according to the first fitting data sequence to the ninth fitting data sequence and the acceleration variable matrix includes:

[0092] The method of constructing the first predicted trajectory three-dimensional coordinate data sequence to the ninth predicted trajectory three-dimensional coordinate data sequence according to the first fitting data sequence to the ninth fitting data sequence and the acceleration variable matrix includes: N N The method of constructing the first predicted trajectory three-dimensional coordinate data sequence to the ninth predicted trajectory three-dimensional coordinate data sequence according to the first fitting data sequence to the ninth fitting data sequence and the acceleration variable matrix includes:

[0093] .

[0094] Wherein, t represents the trajectory sampling interval time.

[0095] Exemplarily, the first predicted trajectory three-dimensional coordinate data sequence to the ninth predicted trajectory three-dimensional coordinate data sequence are constructed according to the first fitting data sequence to the ninth fitting data sequence and the acceleration variable matrix by using the first prediction formula. The first prediction formula is obtained according to the kinematics equation of uniform acceleration motion, and since the value of t is 0.5 seconds, which is less than the minimum time scale of the kinetic state change of the moving target, the first prediction formula is suitable for the kinematics equation of uniform acceleration motion.

[0096] Further, the method of calculating the optimal value of the acceleration variable matrix, denoted as the optimal acceleration variable matrix, by taking the minimum value of the Euclidean distance between the first predicted trajectory three-dimensional coordinate data sequence to the ninth predicted trajectory three-dimensional coordinate data sequence and the first three rows of the first fitting data sequence to the ninth fitting data sequence as the target, by using the particle swarm optimization algorithm includes: N N The method of calculating the optimal value of the acceleration variable matrix, denoted as the optimal acceleration variable matrix, by taking the minimum value of the Euclidean distance between the first predicted trajectory three-dimensional coordinate data sequence to the ninth predicted trajectory three-dimensional coordinate data sequence and the first three rows of the first fitting data sequence to the ninth fitting data sequence as the target, by using the particle swarm optimization algorithm includes:

[0097] The method of calculating the optimal value of the acceleration variable matrix, denoted as the optimal acceleration variable matrix, by taking the minimum value of the Euclidean distance between the first predicted trajectory three-dimensional coordinate data sequence to the ninth predicted trajectory three-dimensional coordinate data sequence and the first three rows of the first fitting data sequence to the ninth fitting data sequence as the target, by using the particle swarm optimization algorithm includes: N N The method of calculating the optimal value of the acceleration variable matrix, denoted as the optimal acceleration variable matrix, by taking the minimum value of the Euclidean distance between the first predicted trajectory three-dimensional coordinate data sequence to the ninth predicted trajectory three-dimensional coordinate data sequence and the first three rows of the first fitting data sequence to the ninth fitting data sequence as the target, by using the particle swarm optimization algorithm includes: N The method of calculating the optimal value of the acceleration variable matrix, denoted as the optimal acceleration variable matrix, by taking the minimum value of the Euclidean distance between the first predicted trajectory three-dimensional coordinate data sequence to the ninth predicted trajectory three-dimensional coordinate data sequence and the first three rows of the first fitting data sequence to the ninth fitting data sequence as the target, by using the particle swarm optimization algorithm includes:​​​​​​

[0098] .

[0099] wherein, the objective function. According to the first fitting data sequence to the ninth fitting data sequence and the acceleration variable matrix, a first constraint condition to a ninth constraint condition is constructed.

[0100] According to the first fitting data sequence to the ninth fitting data sequence and the acceleration variable matrix, a first constraint condition to a ninth constraint condition is constructed. N According to the first fitting data sequence to the ninth fitting data sequence and the acceleration variable matrix, a first constraint condition to a ninth constraint condition is constructed. N According to the first fitting data sequence to the ninth fitting data sequence and the acceleration variable matrix, a first constraint condition to a ninth constraint condition is constructed.

[0101] The particle swarm optimization algorithm is used to calculate the optimal value of the acceleration variable matrix, denoted as the optimal acceleration variable matrix N . N

[0102] .

[0103] Exemplarily, according to the first predicted point trail three-dimensional coordinate data sequence to the ninth predicted point trail three-dimensional coordinate data sequence and the first fitting data sequence to the ninth fitting data sequence, the first objective function to the ninth objective function is constructed by using the Euclidean distance calculation formula. The particle swarm optimization algorithm is used to calculate the optimal value of the acceleration variable matrix, denoted as the optimal acceleration variable matrix. Among them, the optimal acceleration value corresponding to the first fitting data sequence is X-axis acceleration 0.8 m / s², Y-axis acceleration 0.2 m / s², and Z-axis acceleration 0.1 m / s²; the optimal acceleration value corresponding to the second fitting data sequence is X-axis acceleration 0.7 m / s², Y-axis acceleration 0.25 m / s², and Z-axis acceleration 0.12 m / s². Similarly, the optimal acceleration values corresponding to the third to ninth fitting data sequences are obtained respectively, and the optimal acceleration variable matrix is obtained by combining them.

[0104] Further, the method of constructing the first constraint condition to the ninth constraint condition according to the first fitting data sequence to the ninth fitting data sequence and the acceleration variable matrix comprises: N N Further, the method of constructing the first constraint condition to the ninth constraint condition according to the first fitting data sequence to the ninth fitting data sequence and the acceleration variable matrix comprises:

[0105] According to the first fitting data sequence to the ninth fitting data sequence and the acceleration variable matrix, a first constraint condition to a ninth constraint condition is constructed. N According to the first fitting data sequence to the ninth fitting data sequence and the acceleration variable matrix, a first constraint condition to a ninth constraint condition is constructed. N According to the first fitting data sequence to the ninth fitting data sequence and the acceleration variable matrix, a first constraint condition to a ninth constraint condition is constructed. According to the first fitting data sequence to the ninth fitting data sequence and the acceleration variable matrix, a first constraint condition to a ninth constraint condition is constructed.

[0106] . ​​

[0107] wherein, represents a pre-set upper limit of X-axis velocity measurement error; represents a pre-set upper limit of Y-axis velocity measurement error; represents a pre-set upper limit of Z-axis velocity measurement error; represents a pre-set upper limit of X-axis point trajectory measurement error; represents a pre-set upper limit of Y-axis point trajectory measurement error; represents a pre-set upper limit of Z-axis point trajectory measurement error.

[0108] Exemplarily, the upper limits of X-axis velocity measurement error, Y-axis velocity measurement error, Z-axis velocity measurement error, X-axis point trajectory measurement error, Y-axis point trajectory measurement error, and Z-axis point trajectory measurement error are set according to the performance of the detection device, and the pre-set m / s, m / s, m / s, m, m, m.

[0109] Further, the method of calculating the moving target trajectory data according to the optimal acceleration variable matrix and the point trajectory six-dimensional data sequence comprises:

[0110] segmenting the target point trajectory three-dimensional coordinate data sequence by column to obtain a first data column to an nth data column; wherein n represents an integer greater than or equal to 1; and R the first data column to the nth data column are respectively assigned acceleration sets initialized as empty sets to obtain a first acceleration set to an nth acceleration set; and the optimal acceleration variable matrix is segmented by column to obtain a first optimal acceleration data column to an nth optimal acceleration data column. R R N

[0111] traversing the first data interval to the nth data interval; and ; the first optimal acceleration data column is cloned to the first acceleration set to the nth acceleration set; wherein represents the first data interval.

[0112] the first row of all optimal acceleration data columns in the first acceleration set to the nth acceleration set is averaged to obtain X-axis first average acceleration to X-axis nth average acceleration; and R R R ​​​​​​​​​The first average acceleration along the Y-axis is obtained by averaging the second row of all the optimal acceleration data in the acceleration set. R Average acceleration; the first acceleration is collected into the second... R The first average acceleration along the Z-axis is obtained by averaging the third row of all the optimal acceleration data in the acceleration set. R Average acceleration.

[0113] The first column of data in the six-dimensional data sequence of the dots Represented as:

[0114] ;

[0115] in, Indicates the initial X-axis position coordinates of the target; Indicates the initial Y-axis position coordinates of the target; Indicates the initial Z-axis position coordinates of the target; Indicates the target's initial X-axis velocity; This represents the target's initial Y-axis velocity; This represents the initial Z-axis velocity of the target.

[0116] Based on the first average acceleration along the X-axis to the X-axis... R Average acceleration, first average acceleration along the Y-axis to the second average acceleration along the Y-axis R Average acceleration, first average acceleration along the Z-axis to the second average acceleration along the Z-axis R The first column of the six-dimensional data sequence of average acceleration and trace The trajectory data of the moving target is calculated.

[0117] For example, the three-dimensional coordinate data sequence of the target point is divided by columns to obtain the first to the 29th data columns. An acceleration set, pre-initialized to an empty set, is assigned to each of the first to the 29th data columns to obtain the first to the 29th acceleration sets. The optimal acceleration variable matrix is ​​then divided by columns to obtain the first to the ninth optimal acceleration data columns. For example, the first optimal acceleration data column is (0.8, 0.2, 0.1). T The second optimal acceleration data column is (0.7, 0.25, 0.12) m / s². T meters per second². Traversal to , will the Optimal acceleration data column cloned to the first Acceleration set to the first Acceleration set. Specifically... At that time, the first optimal acceleration data column (0.8, 0.2, 0.1) will be used. Tm / s2 Cloned to the first acceleration set to the fifth acceleration set; When t = 0.7s, the second optimal acceleration data column (0.7, 0.25, 0.12) T m / s2 Cloned to the fourth acceleration set to the eighth acceleration set. In this way, until When t = 9s, the ninth optimal acceleration data column is cloned to the 25th acceleration set to the 29th acceleration set. Since there is overlap between the acceleration sets, for example, the fourth acceleration set contains acceleration values from the first optimal acceleration data column and the second optimal acceleration data column, it is necessary to calculate the average value. The average value of all optimal acceleration data columns in the first acceleration set to the 29th acceleration set is taken respectively, and the first average acceleration to the 29th average acceleration of the X axis, the first average acceleration to the 29th average acceleration of the Y axis, and the first average acceleration to the 29th average acceleration of the Z axis are obtained. The first column data is extracted from the point trajectory six-dimensional data sequence, that is, the data at the initial time 0s. The target point trajectory three-dimensional coordinate data is (998, 501, 299) meters, the target point trajectory speed is (201, 50, 29) meters / second, and the combination is . The precise value is considered to be error-free, which is the initial boundary condition for kinematic calculation. This method extracts the true motion characteristics of the moving target from discrete measurement data with errors, and realizes the accurate generation of the moving target trajectory data. Compared with the traditional Kalman filter algorithm, this method does not require the error to be normally distributed.

[0118] Further, the first average acceleration to the 29th average acceleration of the X axis, the first average acceleration to the 29th average acceleration of the Y axis, and the first average acceleration to the 29th average acceleration of the Z axis are calculated according to the first column data in the point trajectory six-dimensional data sequence. R The average acceleration of the Y axis, and the average acceleration of the Z axis. R The average acceleration of the Y axis, and the average acceleration of the Z axis. R The average acceleration of the Y axis, and the average acceleration of the Z axis. The method for calculating the moving target trajectory data comprises:

[0119] The X axis actual acceleration curve is calculated according to the first average acceleration to the 29th average acceleration of the X axis. R The X axis actual acceleration curve is:

[0120] .

[0121] Wherein, represents the X axis actual acceleration curve; represents the interval time corresponding to the time t , the value range of t is 0 to t ; represents the upward rounding symbol; represents a down rounding symbol; represents the first average acceleration of X axis; average acceleration; represents the first average acceleration of X axis; average acceleration; represents the first average acceleration of X axis; average acceleration.

[0122] The Y axis actual acceleration curve is calculated according to the first average acceleration of Y axis to the Nth average acceleration of Y axis; the Y axis actual acceleration curve is: R

[0123] .

[0124] wherein, represents the Y axis actual acceleration curve; represents the first average acceleration of Y axis; average acceleration; represents the first average acceleration of Y axis; average acceleration; represents the first average acceleration of Y axis; average acceleration.

[0125] The Z axis actual acceleration curve is calculated according to the first average acceleration of Z axis to the Nth average acceleration of Z axis; the Z axis actual acceleration curve is: R

[0126] .

[0127] wherein, represents the Z axis actual acceleration curve; represents the first average acceleration of Z axis; average acceleration; represents the first average acceleration of Z axis; average acceleration; represents the first average acceleration of Z axis; average acceleration.

[0128] The first column data in the point trajectory six-dimensional data sequence is taken as the initial boundary condition for kinematic calculation, and the mobile target trajectory data is calculated according to the X axis actual acceleration curve, the Y axis actual acceleration curve and the Z axis actual acceleration curve by using numerical integration and fitting algorithm.

[0129] ​​The X-axis actual acceleration curve, the Y-axis actual acceleration curve and the Z-axis actual acceleration curve are calculated according to the first average acceleration to the 29th average acceleration of the X-axis, the first average acceleration to the 29th average acceleration of the Y-axis, and the first average acceleration to the 29th average acceleration of the Z-axis. The calculation process is essentially a linear interpolation algorithm, which further ensures the continuity of the acceleration change and makes it comply with the laws of physics. The first column of data in the point trace six-dimensional data sequence is taken as the initial boundary condition for kinematic calculation, i.e. the target point trace three-dimensional coordinate data is (998, 501, 299) meters, and the target point trace velocity is (201, 50, 29) meters / second. Starting from the initial time t =0 seconds, the calculation step is pre-set to be 0.01 seconds, and the numerical integration algorithm is used to calculate the position data of the moving target at each time with a step of 0.01 seconds, so as to fit the moving target trajectory data.

[0130] Finally, it should be noted that although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or equivalently replace part of the technical features, as long as they are within the spirit and principles of the present application. Any modification, equivalent replacement, improvement, etc. shall be included in the protection scope of the present application.​

Claims

1. A method for generating moving target trajectory data, which uses a state estimation algorithm to calculate continuous moving target trajectory data from a sequence of three-dimensional coordinate data of target point trails with errors and a sequence of target point trail velocities, characterized in that, The method comprises the following steps: S1, acquiring first data; the first data comprises a track sampling interval time, and track six-dimensional data sequence obtained by combining target track three-dimensional coordinate data sequence and target track velocity sequence; S2, using a sliding window extraction method, taking a first data window as a window, taking a first data interval as a step, and extracting a first fitting data sequence to an nth fitting data sequence in the track six-dimensional data sequence N the fitting data sequence; is a first constant. S3, according to the first fitting data sequence to the first N The fitting data sequence and the pre-initialized acceleration variable matrix are used to construct the first predicted point trail three-dimensional coordinate data sequence to the first N The predicted point trail three-dimensional coordinate data sequence; the first predicted point trail three-dimensional coordinate data sequence to the first N The predicted point trail three-dimensional coordinate data sequence and the first fitting data sequence to the first N The Euclidean distance of the first three rows of the fitting data sequence is the minimum value, and the optimal value of the acceleration variable matrix is calculated by using the particle swarm optimization algorithm, which is recorded as the optimal acceleration variable matrix; S4, calculating mobile target trajectory data according to the optimal acceleration variable matrix and the track six-dimensional data sequence; The first fitting data sequence is extracted from the point six-dimensional data sequence by using the sliding window extraction method with a preset first data window and a preset first data interval as a step. N The method for fitting the data sequence comprises: The starting point of the first data window is placed at the starting point of the track six-dimensional data sequence; the overlapping part of the first data window and the track six-dimensional data sequence is taken as first fitting data sequence; Using the first data interval as the step size, shift the six-dimensional data sequence of the first data window edge points backward. Next; the first The overlapping portion of the first data window and the six-dimensional data sequence of the point after the second movement is denoted as the first... Fit the data sequence to obtain the first fitted data sequence to the second fitted data sequence. N Fitting the data sequence; For values ​​from 1 to Integer variables; The value is greater than 3; fitting the first data sequence to the second data sequence N fitting the data sequence is represented as to ; wherein the first data sequence fitting the data sequence is represented as: ； wherein, is an integer variable taking values from 1 to ; is an integer variable taking values from 1 to ; denotes the window length of the first data window; denotes the X-axis first measurement trace coordinate of the first data window; denotes the Y-axis first measurement trace coordinate of the first data window; denotes the Z-axis first measurement trace coordinate of the first data window; denotes the X-axis first measurement velocity of the first data window; denotes the Y-axis first measurement velocity of the first data window; denotes the Z-axis first measurement velocity of the first data window; denotes the X-axis first measurement velocity of the second data window; denotes the Y-axis first measurement velocity of the second data window; denotes the Z-axis first measurement velocity of the second data window; The first to the nth predicted point trail three-dimensional coordinate data sequences are constructed by using the kinematic equation with the fitting data sequences and the pre-initialized acceleration variable matrix. N The first to the nth predicted point trail three-dimensional coordinate data sequences are constructed by using the kinematic equation with the fitting data sequences and the pre-initialized acceleration variable matrix. N The method for predicting the point trail three-dimensional coordinate data sequence comprises: The acceleration variable matrix is denoted as ; wherein is expressed as: ； wherein, represents the first acceleration of the X axis acceleration; represents the first acceleration of the Y axis acceleration; represents the first acceleration of the Z axis acceleration; According to the first fitted data sequence to the N Fitted data sequence and acceleration variable matrix construct the first predicted point trail three-dimensional coordinate data sequence to the N Predicted point trail three-dimensional coordinate data sequence; the first The predicted point trail three-dimensional coordinate data sequence is represented as: ； in, Indicates the first Predicted 3D coordinate data sequence of point traces; express The X-axis Predict the coordinates of the point; express Y-axis Predict the coordinates of the point; express Z-axis Predict the coordinates of the point; said first fitting data sequence to a N fitting data sequence and acceleration variable matrix constructs a first predicted point trail three-dimensional coordinate data sequence to a N the method of predicting point trail three-dimensional coordinate data sequence includes: According to the first fitting data sequence to the N fitting data sequence and acceleration variable matrix, a first prediction formula is used to construct a first predicted point trail three-dimensional coordinate data sequence to a N predicted point trail three-dimensional coordinate data sequence; the first prediction formula is: ； wherein, represents the time interval between trace samples; The first to the nth predicted track three-dimensional coordinate data sequences are respectively N The Euclidean distance of the first to the third rows of the fitted data sequence is the minimum value, and the optimal value of the acceleration variable matrix is calculated by using a particle swarm optimization algorithm, which is recorded as the optimal acceleration variable matrix. N The Euclidean distance of the first to the third rows of the fitted data sequence is the minimum value, and the optimal value of the acceleration variable matrix is calculated by using a particle swarm optimization algorithm, which is recorded as the optimal acceleration variable matrix. According to the first predicted point trail three-dimensional coordinate data sequence to the first N The predicted point trail three-dimensional coordinate data sequence and the first fitting data sequence to the first N The fitting data sequence adopts a Euclidean distance calculation formula to construct a first objective function to a first N The objective function; the Euclidean distance calculation formula is: ； wherein, represents the first objective function; According to the first fitting data sequence to the nth N Fitting data sequence and acceleration variable matrix construct the first constraint condition to the nth N Constraint conditions; respectively, and the first constraint condition to the fourth constraint condition are respectively represented as N The first target function to the fourth target function are respectively represented as N The first constraint condition to the fourth constraint condition are respectively represented as The optimal acceleration variable matrix is represented as 。 2. The method of claim 1, wherein, said first fitting data sequence to a first N fitting data sequence and acceleration variable matrix constructs a first constraint condition to a first N the method of constraint condition includes: According to the first fitting data sequence to the nth fitting data sequence N The fitting data sequence and the acceleration variable matrix construct the first constraint condition to the nth constraint condition N The constraint condition; wherein the first constraint condition is The constraint condition is: ； wherein represents a pre-set upper limit of the X-axis velocity measurement error; represents a pre-set upper limit of the Y-axis velocity measurement error; represents a pre-set upper limit of the Z-axis velocity measurement error; represents a pre-set upper limit of the X-axis point trajectory measurement error; represents a pre-set upper limit of the Y-axis point trajectory measurement error; represents a pre-set upper limit of the Z-axis point trajectory measurement error.

3. The method of claim 2, wherein, The method for calculating the mobile target trajectory data according to the optimal acceleration variable matrix and the track six-dimensional data sequence comprises: The three-dimensional coordinate data sequence of the target point is divided into columns to obtain the first data column to the second column. R Data columns; where ; respectively from the first data column to the second R The data column is assigned an acceleration set that is pre-initialized to an empty set, resulting in the first acceleration set to the second. R Acceleration set; the optimal acceleration variable matrix is ​​divided by columns to obtain the first optimal acceleration data column to the second optimal acceleration data column. N Optimal acceleration data column; traverse to ; the first optimal acceleration data series is cloned to the first acceleration set to the first acceleration set; wherein represents the first data interval; all optimal acceleration data columns in the first acceleration set to the first R average acceleration in the X-axis; all optimal acceleration data columns in the first acceleration set to the first R average acceleration in the Y-axis; all optimal acceleration data columns in the first acceleration set to the first R average acceleration in the Z-axis; all optimal acceleration data columns in the first acceleration set to the first R average acceleration in the X-axis; all optimal acceleration data columns in the first acceleration set to the first R average acceleration in the Y-axis; all optimal acceleration data columns in the first acceleration set to the first R average acceleration in the Z-axis; extracting a first column of data from the sequence of six-dimensional data points is represented as: ； wherein, represents the target initial X-axis position coordinate; represents the target initial Y-axis position coordinate; represents the target initial Z-axis position coordinate; represents the target initial X-axis velocity; represents the target initial Y-axis velocity; represents the target initial Z-axis velocity; According to the first average acceleration to the sixth average acceleration of the X axis R average acceleration, the first average acceleration to the sixth average acceleration of the Y axis R average acceleration, the first average acceleration to the sixth average acceleration of the Z axis R the first column data in the point trajectory six-dimensional data sequence The moving target trajectory data is calculated.

4. The method of claim 3, wherein, The first average acceleration along the X-axis to the second... R Average acceleration, first average acceleration along the Y-axis to the second average acceleration along the Y-axis R Average acceleration, first average acceleration along the Z-axis to the second average acceleration along the Z-axis R The first column of the six-dimensional data sequence of average acceleration and trace Methods for calculating the trajectory data of a moving target include: According to the first average acceleration along the X axis to the first average acceleration along the X axis R The average acceleration calculation obtains an actual acceleration curve along the X axis; the actual acceleration curve along the X axis is ； wherein, represents the X-axis actual acceleration curve; represents the X-axis average acceleration curve; represents the interval time of the time point corresponding to the value range of is 0 to ; represents the upward rounding symbol; represents the downward rounding symbol; represents the X-axis first average acceleration; represents the X-axis first average acceleration; represents the X-axis first average acceleration; According to the Y-axis first average acceleration to the Y-axis first R The average acceleration calculation obtains a Y-axis actual acceleration curve; the Y-axis actual acceleration curve is: ； wherein, represents the Y-axis actual acceleration curve; represents the Y-axis first average acceleration; represents the Y-axis first average acceleration; represents the Y-axis first average acceleration; According to the first average acceleration along the Z axis to the seventh average acceleration along the Z axis R The average acceleration calculation obtains a Z axis actual acceleration curve; the Z axis actual acceleration curve is: ； wherein, represents the Z-axis actual acceleration curve; represents the Z-axis first average acceleration; represents the Z-axis first average acceleration; represents the Z-axis first average acceleration; The first column data in the track six-dimensional data sequence is taken as initial boundary conditions for kinematic calculation, and the mobile target trajectory data is calculated by using numerical integration and fitting algorithm according to the X-axis actual acceleration curve, the Y-axis actual acceleration curve and the Z-axis actual acceleration curve.

Images