A motion pattern mining method based on moving target trajectory and related equipment

Abstract

The present application relates to the technical field of motion pattern recognition, and provides a motion pattern mining method based on mobile target trajectory and related equipment. The method includes: obtaining the distance change sequence of each mobile target pair based on all trajectory data; calculating the convergence rate and separation rate according to all distance change sequences, and determining multiple convergent mobile targets and multiple separated mobile targets from all mobile targets according to the convergence rate and separation rate; clustering the convergent mobile targets according to the trajectory data at the current moment to obtain multiple convergent clustering clusters, and clustering the separated mobile targets according to the trajectory data at the first moment to obtain multiple separated clustering clusters; taking the movement of all mobile targets in each convergent clustering cluster within a time period as a convergent pattern, and taking the movement of all mobile targets in each separated clustering cluster within a time period as a separated pattern. The method of the present application can improve the versatility and reliability of motion pattern mining.

Description

Technical Field

[0001] The present application relates to the technical field of motion pattern recognition, and in particular to a motion pattern mining method based on moving target trajectories and related equipment. Background Art

[0002] With the rapid development of positioning services, mobile Internet and Internet of Things technologies, a large number of mobile targets (such as vehicles, pedestrians, ships, animals, etc.) have accumulated massive spatiotemporal trajectory data, which are also continuously generated in real-time streams. These mobile target trajectory data contain rich information, which not only reflects the motion information of individual targets, but also reflects the spatial distribution characteristics, motion behavior laws and dynamic changes of group targets. Therefore, it is of great significance to mine the motion patterns of mobile targets, which has important application value in many fields.

[0003] The movement patterns of mobile targets are diverse, and there are many types of movement patterns, such as the accompanying pattern of multiple mobile targets moving together, the convergence pattern of multiple mobile targets gathering in space, and the periodic pattern with time regularity. Different movement patterns can reflect different behaviors of mobile targets and model different types of geographical events. Among them, the convergence pattern refers to the phenomenon that multiple mobile targets move in the same direction or position within a certain period of time, while the separation pattern refers to the movement of these objects in different directions or positions within a certain period of time. On the one hand, the convergence-separation pattern can reflect the collective behavior of mobile targets. For example, the mining of the convergence-separation pattern of traffic flow can model the congestion and dispersion of urban traffic networks; the mining of the convergence-separation pattern of animal communities can reveal the group behavior of biological populations and changes in the ecological environment. On the other hand, the identification and analysis of the convergence-separation pattern can also provide a reference for modeling and predicting geographical spatiotemporal events. The occurrence of certain natural disasters and emergencies is usually accompanied by specific convergence-separation patterns. Therefore, mining the convergence-separation pattern of mobile targets is of great significance.

[0004] Existing methods for mining convergence-separation patterns mainly fall into two categories. The first category is mining methods based on clustering-query. This type of method first clusters the trajectory points of mobile targets at different timestamps to obtain different mobile target clusters; then queries the mobile target cluster groups that meet the given density change, time persistence, continuity, number of mobile targets and other constraints in different timestamps. This type of method continues the definition of the adjoint pattern, aims to discover mobile targets that move together, and mines convergence-separation patterns by constraining the density changes in the timestamps before and after the mobile targets. It can identify mobile targets that are clustered in space, but it cannot model the aggregation pattern formed by mobile targets from different directions. The second category is methods based on group motion morphology, which can be further divided into two methods. One is to determine whether the mobile target has centripetal motion by judging whether the straight lines of the moving directions of the mobile targets at different timestamps intersect in an area of ​​a specified radius. This method can effectively identify which objects are moving in the same direction and position within a specific time period, thereby inferring possible convergence behavior. However, this method is not suitable for the convergence or separation patterns of mobile targets under network space constraints. The other method is to first identify the area where mobile targets are densely concentrated as the candidate convergence center area through density clustering (such as density peak clustering), and then identify the convergence target groups at different timestamps in turn according to the motion patterns of the mobile targets in the convergence center area. However, this type of method cannot effectively identify potential convergence behavior when the convergence pattern has not yet been formed, because in this case, the distribution of mobile targets may still be in a relatively scattered state and fail to show obvious aggregation characteristics.

[0005] In general, the current convergence-divergence pattern mining algorithm for mobile targets still needs to be further improved in terms of the ability to identify the convergence pattern of mobile targets, the ability to model the early motion before the formation of the convergence center, and the applicability of convergence pattern mining in different geographical scenarios. This is mainly reflected in the following aspects: (1) Limited convergence pattern recognition ability: The early motion stage of the convergence pattern includes initial convergence, gradual convergence and other processes. The existing algorithms focus on the recognition of the convergence motion pattern after the formation of the convergence pattern. There is little research on the convergence pattern before the formation of the convergence pattern, and it is difficult to find convergence patterns with different motion forms (such as different directions and different speeds); (2) Insufficient ability to model the early motion characteristics of the convergence pattern: The existing convergence pattern mining algorithm does not accurately model the early motion characteristics of the mobile target before the formation of the convergence center, and it is difficult to capture the potential convergence trend; (3) Poor adaptability to geographical scenarios: The complexity of the convergence-dispersion model recognition of mobile targets in different geographical scenarios (such as urban road networks, oceans, and outdoors) varies, and has a certain degree of scene dependence. It can be seen that the current motion pattern mining method has the problem of poor reliability of motion pattern mining. Summary of the invention

[0006] The present application provides a motion pattern mining method based on moving target trajectories, which can solve the problem of poor reliability of motion pattern mining.

[0007] In a first aspect, an embodiment of the present application provides a motion pattern mining method based on a moving target trajectory, the motion pattern mining method comprising:

[0008] Obtain the trajectory data of each mobile target in the study area at T moments; the Tth moment is the current moment;

[0009] Pair all moving targets in pairs to obtain multiple moving target pairs, and obtain the distance change sequence of each moving target pair based on all trajectory data;

[0010] The convergence rate and separation rate of each mobile target pair are calculated according to all distance change sequences, and a plurality of convergence mobile targets and a plurality of separation mobile targets are determined from all mobile targets according to all convergence rates and separation rates; the convergence mobile target is a mobile target whose corresponding convergence rate satisfies the convergence condition, and the separation mobile target is a mobile target whose corresponding separation rate satisfies the separation condition;

[0011] According to the trajectory data of all converging moving targets at the current moment, all converging moving targets are clustered to obtain multiple converging cluster clusters, and according to the trajectory data of all separating moving targets at the first moment among T moments, all separating moving targets are clustered to obtain multiple separating cluster clusters;

[0012] The movement of all mobile targets in each convergent cluster within the time period covered by T moments is regarded as a convergent pattern, and the movement of all mobile targets in each separate cluster within the time period covered by T moments is regarded as a separate pattern.

[0013] Optionally, a distance change sequence of each moving target pair is obtained based on all trajectory data, including:

[0014] For each moving target pair, perform the following steps:

[0015] According to the trajectory data of two moving targets in the moving target pair, the distance between the two moving targets at each moment is calculated;

[0016] Calculate the distance change between two moving targets between each two consecutive moments based on all distances;

[0017] All distance changes are integrated to obtain the distance change sequence of the moving target pairs.

[0018] Optionally, calculating the distance between the two moving targets at each moment according to the trajectory data of the two moving targets in the moving target pair includes:

[0019] By formula:

[0020] ;

[0021] Calculate in moment, The moving target and The distance between moving targets ;

[0022] in, Indicated in Moment The first coordinate value in the moving target trajectory data, Indicated in Moment The second coordinate value in the moving target trajectory data, Indicated in Moment The first coordinate value in the moving target trajectory data, Indicated in Moment The second coordinate value in the moving target trajectory data, , , A set of numbers representing all moving targets.

[0023] Optionally, the distance change between two moving targets between each two consecutive moments is calculated based on all distances, including:

[0024] By formula:

[0025]

[0026] Calculate in The moment and Between moments, Moving Target With Moving Target The distance change between ;

[0027] in, Indicated in Moment The moving target and The distance between moving targets, Indicated in Moment The moving target and The distance between moving targets, Indicated in Moment The trajectory data of a moving target, Indicated in Moment The trajectory data of a moving target, Indicated in Moment The trajectory data of a moving target, Indicated in Moment The trajectory data of a moving target, .

[0028] Optionally, the convergence rate and separation rate of each moving target pair are calculated based on all distance change sequences, including:

[0029] By formula:

[0030]

[0031] Calculate moving target pairs The convergence rate ;

[0032] in, represents the time step from the initial moment, Indicates A moving target, Indicates A moving target;

[0033] By formula:

[0034]

[0035] Calculate moving target pairs Separation rate .

[0036] Optionally, determining a plurality of converging moving targets and a plurality of separating moving targets from all moving targets according to all converging rates and separating rates includes:

[0037] For each moving target pair, perform the following steps:

[0038] If the distance between the two moving targets of the moving target pair at the current moment is less than the distance threshold, then determining whether the convergence rate of the moving target pair meets the convergence condition;

[0039] If the convergence rate of the mobile target pair meets the convergence condition, the two mobile targets corresponding to the mobile target pair are both used as convergence mobile targets;

[0040] If in If the distance between the two moving targets of the moving target pair at a certain moment is less than the distance threshold, it is determined whether the separation rate of the moving target pair meets the separation condition;

[0041] If the separation rate of the mobile target pair meets the separation condition, the two mobile targets corresponding to the mobile target pair are both used as separated mobile targets.

[0042] Optionally, the convergence condition is:

[0043] The convergence rate is greater than the convergence rate threshold;

[0044] The separation conditions are:

[0045] The separation rate is greater than the separation rate threshold.

[0046] In a second aspect, an embodiment of the present application provides a motion pattern mining device based on a moving target trajectory, comprising:

[0047] The acquisition module obtains the trajectory data of each mobile target in the study area at T moments; the Tth moment is the current moment;

[0048] The pairing module pairs all the moving targets in pairs to obtain multiple moving target pairs, and obtains the distance change sequence of each moving target pair based on all trajectory data;

[0049] A determination module calculates the convergence rate and separation rate of each mobile target pair according to all distance change sequences, and determines multiple convergence mobile targets and multiple separation mobile targets from all mobile targets according to all convergence rates and separation rates; the convergence mobile targets are mobile targets whose corresponding convergence rates meet the convergence conditions, and the separation mobile targets are mobile targets whose corresponding separation rates meet the separation conditions;

[0050] A clustering module clusters all converging moving targets according to their trajectory data at the current moment to obtain a plurality of converging clustering clusters, and clusters all separating moving targets according to their trajectory data at the first moment among T moments to obtain a plurality of separating clustering clusters;

[0051] The marking module regards the movement of all mobile targets in each convergent cluster within the time period covered by T moments as a convergent pattern, and regards the movement of all mobile targets in each separate cluster within the time period covered by T moments as a separate pattern.

[0052] In a third aspect, an embodiment of the present application provides a terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned motion pattern mining method based on mobile target trajectory when executing the above-mentioned computer program.

[0053] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, it implements the above-mentioned motion pattern mining method based on the moving target trajectory.

[0054] The above solution of the present application has the following beneficial effects:

[0055] In an embodiment of the present application, the trajectory data of each mobile target in the study area at T moments are obtained, and then all the mobile targets are paired two by two to obtain multiple mobile target pairs, and the distance change sequence of each mobile target pair is obtained based on all the trajectory data, and then the convergence rate and separation rate of each mobile target pair are calculated according to all the distance change sequences, and according to all the convergence rates and separation rates, multiple converging mobile targets and multiple separating mobile targets are determined from all the mobile targets, and then according to the trajectory data of all the converging mobile targets at the current moment, all the converging mobile targets are clustered to obtain multiple converging clustering clusters, and according to the trajectory data of all the separating mobile targets at the first moment among the T moments, all the separating mobile targets are clustered to obtain multiple separating clustering clusters, and finally, the movement of all the mobile targets in each converging clustering cluster within the time period covered by T moments is taken as a convergence mode, and the movement of all the mobile targets in each separating clustering cluster within the time period covered by T moments is taken as a separation mode. Among them, obtaining the distance change sequence of the moving target pairs can effectively capture and analyze the movement changes between the moving targets. Calculating the convergence rate and separation rate based on the moving target pairs can improve the accuracy and rationality of the convergence rate and separation rate, thereby improving the accuracy of motion pattern mining based on the convergence rate and separation rate. Calculations are performed on trajectory data without restricting the geographical space where the moving targets are located, which effectively improves the versatility and reliability of motion pattern mining.

[0056] Other beneficial effects of the present application will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0058] Figure 1 A flow chart of a motion pattern mining method based on moving target trajectory provided by an embodiment of the present application;

[0059] Figure 2 A schematic diagram of vehicle trajectory data provided by an embodiment of the present application;

[0060] Figure 3 A schematic diagram of ship trajectory data provided by an embodiment of the present application;

[0061] Figure 4 A schematic diagram of the result of a vehicle convergence mode provided in an embodiment of the present application;

[0062] Figure 5 A schematic diagram of the result of a vehicle separation mode provided in an embodiment of the present application;

[0063] Figure 6 A schematic diagram of the result of a ship convergence mode provided in an embodiment of the present application;

[0064] Figure 7 A schematic diagram of the result of a ship separation mode provided in an embodiment of the present application;

[0065] Figure 8 A schematic diagram of the structure of a motion pattern mining device based on moving target trajectory provided by an embodiment of the present application;

[0066] Fig. 9 A schematic diagram of the structure of a terminal device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0067] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0068] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0069] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0070] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0071] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0072] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0073] In view of the problem of poor reliability of motion pattern mining in existing motion pattern mining methods, an embodiment of the present application provides a motion pattern mining method based on mobile target trajectories. The motion pattern mining method obtains trajectory data of each mobile target in a study area at T moments, then pairs all mobile targets in pairs to obtain multiple mobile target pairs, and obtains a distance change sequence of each mobile target pair based on all trajectory data, and then calculates the convergence rate and separation rate of each mobile target pair according to all distance change sequences, and determines multiple converging mobile targets and multiple separating mobile targets from all mobile targets based on all convergence rates and separation rates, and then clusters all converging mobile targets according to the trajectory data of all converging mobile targets at the current moment to obtain multiple converging clustering clusters, and clusters all separating mobile targets according to the trajectory data of all separating mobile targets at the first moment among T moments to obtain multiple separating clustering clusters, and finally takes the movement of all mobile targets in each converging clustering cluster within the time period covered by T moments as a converging pattern, and takes the movement of all mobile targets in each separating clustering cluster within the time period covered by T moments as a separating pattern. Among them, obtaining the distance change sequence of the moving target pairs can effectively capture and analyze the movement changes between the moving targets. Calculating the convergence rate and separation rate based on the moving target pairs can improve the accuracy and rationality of the convergence rate and separation rate, thereby improving the accuracy of motion pattern mining based on the convergence rate and separation rate. Calculations are performed on trajectory data without restricting the geographical space where the moving targets are located, which effectively improves the versatility and reliability of motion pattern mining.

[0074] Next, the motion pattern mining method based on moving target trajectory provided by this application is exemplified.

[0075] like Figure 1 As shown, the motion pattern mining method based on the moving target trajectory provided by the present application includes the following steps:

[0076] Step 11, obtain the trajectory data of each mobile target in the study area at T moments.

[0077] The research area is an area where the motion pattern of moving targets needs to be mined, such as a certain range of sea areas, etc. The Tth moment is the current moment. For example, if the current moment is 8 o'clock and T is 4, then the T moments can be 6:30, 7 o'clock, 7:30, and 8 o'clock. The above-mentioned moving targets can be pedestrians, cars, ships, etc. The above-mentioned trajectory data includes the coordinates of the trajectory points of the moving target at that moment, etc. (the coordinates can be the coordinates in the longitude and latitude coordinate system of the earth).

[0078] In some embodiments of the present application, the trajectory data of the moving target may be acquired using a device such as a global positioning system.

[0079] It should be noted that after obtaining the trajectory data, preprocessing such as time alignment is required, and time alignment can be performed using interpolation methods, dynamic time warping, etc.

[0080] For example, when the moving target is a vehicle, the trajectory data of the vehicle is as follows: Figure 2 As shown in the figure, the solid line represents the trajectory of the vehicle, the dots represent the trajectory points of the vehicle, and the hash lines represent the roads. When the moving target is a ship, the trajectory data of the ship is as follows Figure 3 As shown in the figure, the points represent ports, A, B, C, and D represent the port numbers, the dotted line represents the coastline, and the solid line represents the data of the ship's Automatic Identification System (AIS).

[0081] Step 12: Pair all the moving targets in pairs to obtain multiple moving target pairs, and obtain the distance change sequence of each moving target pair based on all trajectory data.

[0082] In some embodiments of the present application, the above step of pairing all moving targets in pairs to obtain multiple moving target pairs, and obtaining the distance change sequence of each moving target pair based on all trajectory data includes:

[0083] In the first step, all moving targets are paired with each other to obtain multiple moving target pairs.

[0084] Exemplarily, all mobile targets may be randomly paired in pairs to obtain multiple mobile target pairs. If the number of mobile targets is an odd number, that is, after pairing is completed, there are unpaired mobile targets, then the mobile targets are discarded.

[0085] In the second step, for each moving target pair, perform the following steps:

[0086] First, according to the trajectory data of two moving targets in the moving target pair, the distance between the two moving targets at each moment is calculated.

[0087] Specifically, through the formula:

[0088]

[0089] Calculate in moment, The moving target and The distance between moving targets .

[0090] in, Indicated in Moment The first coordinate value in the moving target trajectory data, Indicated in Moment The second coordinate value in the moving target trajectory data, Indicated in Moment The first coordinate value in the moving target trajectory data, Indicated in Moment The second coordinate value in the moving target trajectory data, , , A set of numbers representing all moving targets.

[0091] Then, the distance change between the two moving targets between each two consecutive moments is calculated based on all the distances.

[0092] Specifically, through the formula:

[0093]

[0094] Calculate in The moment and Between moments, Moving Target With Moving Target The distance change between .

[0095] in, Indicated in Moment The moving target and The distance between moving targets, Indicated in Moment The moving target and The distance between moving targets, Indicated in Moment The trajectory data of a moving target, Indicated in Moment The trajectory data of a moving target, Indicated in Moment The trajectory data of a moving target, Indicated in Moment The trajectory data of a moving target, .

[0096] Finally, all distance changes are integrated to obtain the distance change sequence of the moving target pairs.

[0097] Specifically, all distance changes are sequentially integrated into a sequence from early to late in the order of time to obtain a distance change sequence.

[0098] It should be noted that the distance change sequence reflects the movement trend between two moving targets. t i arrive t i+1 The distance change between moving targets at any moment is less than 0, that is, Δ d <0, then o k and o q from t i arrive t i+1 There is a trend of increasing movement at all times; on the contrary, if the moving target o k and o q from t i arrive t i+1 The distance change between moments is greater than 0, that is, Δ d >0, then o k and o q from t i arrive t i+1 The time has a trend of increasing and decreasing; if Δ d =0, then move the target o k and o q There is no obvious movement trend.

[0099] For example, when calculating the distance above, the method of calculating the distance can be adjusted according to the different types of moving targets. For example, when the moving target is a city vehicle or a pedestrian, the distance between two vehicles or pedestrians in the road network space at any time needs to use the network distance to calculate the distance between the moving targets.

[0100] It is worth mentioning that obtaining the distance change sequence of the moving target pair can effectively capture and analyze the movement changes between the moving targets.

[0101] Step 13, calculating the convergence rate and separation rate of each mobile target pair according to all distance change sequences, and determining a plurality of convergent mobile targets and a plurality of separated mobile targets from all mobile targets according to all the convergence rates and separation rates.

[0102] The above-mentioned convergence moving target is a moving target whose corresponding convergence rate satisfies the convergence condition, and the separation moving target is a moving target whose corresponding separation rate satisfies the separation condition.

[0103] In some embodiments of the present application, the steps of calculating the convergence rate and separation rate of each moving target pair according to all distance change sequences, and determining a plurality of converging moving targets and a plurality of separating moving targets from all moving targets according to all the convergence rates and separation rates include:

[0104] In the first step, the convergence rate and separation rate of each moving target pair are calculated based on all distance change sequences.

[0105] By formula:

[0106]

[0107] Calculate moving target pairs The convergence rate .

[0108] in, Indicates the time step from the initial moment, which is used to control the time window size for calculating the aggregation rate or separation rate. Indicates A moving target, Indicates A moving target.

[0109] By formula:

[0110]

[0111] Calculate moving target pairs Separation rate .

[0112] For example, The value of is an integer less than T but not less than 0. When setting it, it is necessary to consider the moments in the time window as meaningful as possible, but not more than T moments.

[0113] In the second step, for each moving target pair, perform the following steps:

[0114] If the distance between the two moving targets of the moving target pair is less than the distance threshold at the current moment, it is determined whether the convergence rate of the moving target pair meets the convergence condition.

[0115] If the convergence rate of the mobile target pair meets the convergence condition, the two mobile targets corresponding to the mobile target pair are both used as converged mobile targets.

[0116] If in If the distance between the two moving targets of the moving target pair at a certain moment is less than the distance threshold, it is determined whether the separation rate of the moving target pair meets the separation condition.

[0117] If the separation rate of the mobile target pair meets the separation condition, the two mobile targets corresponding to the mobile target pair are both used as separated mobile targets.

[0118] It should be noted that the convergence condition is: the convergence rate is greater than the convergence rate threshold; the separation condition is: the separation rate is greater than the separation rate threshold. If the convergence rate of the moving target does not meet the convergence condition and the separation rate does not meet the separation condition, it is considered that there is no potential convergence and separation trend between the two moving targets in the time window and no processing is performed. If the distance between the two moving targets of the moving target pair at the current moment is greater than or equal to the distance threshold, and in the first If the distance between the two moving targets of a moving target pair at a certain moment is greater than or equal to the distance threshold, the moving target pair will not be processed.

[0119] Step 14: cluster all the converging mobile targets according to their trajectory data at the current moment to obtain a plurality of converging clusters, and cluster all the separating mobile targets according to their trajectory data at the first moment among T moments to obtain a plurality of separating clusters.

[0120] Specifically, clustering algorithms such as density-based spatial clustering of applications with noise (DBSCAN) can be used to cluster all converging mobile targets according to their trajectory data at the current moment, and obtain multiple potential converging clusters. If the number of mobile targets in the potential converging cluster is greater than or equal to the minimum number of mobile targets (such as 6), the potential converging cluster is taken as a converging cluster. Similarly, clustering algorithms such as DBSCAN can be used to cluster all separated mobile targets according to their trajectory data at the first moment among T moments, and obtain multiple potential separated clusters. If the number of mobile targets in the potential separated cluster is greater than or equal to the minimum number of mobile targets, the potential separated cluster is taken as a separated cluster.

[0121] It should be noted that the DBSCAN algorithm is based on the clustering idea of ​​density, and can find clusters of any shape and automatically determine the number of clusters. At the same time, it can effectively process noise data points, that is, it can effectively find a set of mobile targets in the same convergence mode, separate mobile target sets in different convergence / separation modes, and filter among a group of mobile target sets that have convergence / separation trends. The DBSCAN algorithm has two key parameters: neighborhood radius Eps and minimum cluster point data MinObjects, which are used to identify the convergence center area (separation core area in separation mode) and mobile targets that constitute the convergence mode, and to distinguish mobile targets interfered by noise. By setting these two parameters reasonably, the convergence / separation mode clusters of mobile targets can be well discovered.

[0122] Taking the moving target as a vehicle as an example, the basic steps of DBSCAN are as follows:

[0123] (1) Calculate the value of each vehicle o p ε-neighborhood, i.e., the distance from vehicle o p The set of all moving targets in the neighborhood that does not exceed Eps (here set to 50m) is N(o p ).

[0124] (2) If |N(o p )|≥MinObjects, then vehicle o p For core objects, MinObjects is set to 3.

[0125] (3) For each boundary point o q , if o q Belongs to a core object o p Neighborhood N(o p ), then o q Join p The cluster to which it belongs.

[0126] (4) Continue scanning the remaining unprocessed vehicles and repeat steps (2)-(3) until all vehicles are processed.

[0127] It is worth mentioning that in this step, for the convergent clustering cluster, it can reflect the clustering status of the trajectory data of multiple converging mobile targets in the convergent clustering cluster at the current moment, that is, multiple converging mobile targets have a convergence trend at the current moment after moving for T moments; for the separation clustering cluster, it can reflect the clustering status of the trajectory data of multiple separation mobile targets in the separation clustering cluster at the first moment, that is, multiple separation mobile targets have a separation trend starting from the first moment.

[0128] Step 15, taking the movement of all mobile targets in each convergent cluster within the time period covered by T moments as a convergent pattern, and taking the movement of all mobile targets in each separate cluster within the time period covered by T moments as a separate pattern.

[0129] The convergence mode refers to the phenomenon that the movements of multiple moving targets tend to converge within the time period covered by T moments, and the separation mode refers to the phenomenon that the movements of multiple moving targets tend to separate within the time period covered by T moments.

[0130] It should be noted that, starting from the current moment, the mobile target set of the current convergence mode can be continuously tracked to detect the time when the mobile targets in different convergence modes gather. At the same time, the mobile target sets in different convergence modes are used as candidate mobile targets for the separation mode in the next process, and the moment when the mobile targets in the convergence mode are formed is used as the reference initial time for setting the time window size of the separation mode.

[0131] For example, the results of the mining of the convergence mode and separation mode can be visualized. Figure 2 Taking the trajectory data of the vehicle in as an example, the results of the convergence mode are as follows Figure 4 As shown in the figure, the solid line represents the trajectory line, the dot represents the trajectory point, the arrow represents the moving direction of the vehicle, the hash line represents the road, and the circle represents the convergence center (that is, the trajectory data of all vehicles marked as convergence mode at the current moment); the result of the separation mode is shown in Figure 5 As shown in the figure, the solid line represents the trajectory line, the dot represents the trajectory point, the arrow represents the moving direction of the vehicle, the hash line represents the road, and the circle represents the dispersion center (that is, the trajectory data of all vehicles marked as separation mode at the first moment).

[0132] by Figure 3 Taking the trajectory data of the ship in the example, the results of the convergence mode are as follows Figure 6 As shown in the figure, the dots represent ports, A, B, C, and D represent the port numbers, the dotted lines represent the coastline, the solid lines represent the AIS data, the arrows represent the moving direction, and the circles represent the convergence center of the ships (i.e., the trajectory data of all ships marked as convergence mode at the current moment). The results of the separation mode are shown in the figure. Figure 7 As shown in the figure, the dots represent ports, the starting point of all trajectories in the figure is port C, the solid line represents AIS data, the arrow represents the moving direction, the dotted line represents the coastline, and the circle represents the dispersion center of the ship (that is, the trajectory data of all ships marked as separation mode at the first moment).

[0133] After identifying the convergence pattern or separation pattern of mobile targets, it can be used for tasks such as analyzing urban traffic flow and group behavior analysis of animal communities. If the mobile target is a vehicle, mining the convergence pattern and separation pattern of traffic flow can model the congestion and dispersion phenomena of urban traffic networks.

[0134] It is worth mentioning that obtaining the distance change sequence of moving target pairs can effectively capture and analyze the movement changes between moving targets. Calculating the convergence rate and separation rate based on the moving target pairs can improve the accuracy and rationality of the convergence rate and separation rate, thereby improving the accuracy of motion pattern mining based on the convergence rate and separation rate. Calculations are performed on trajectory data without restricting the geographical space where the moving targets are located, which effectively improves the versatility and reliability of motion pattern mining.

[0135] In addition, the method of the present application can effectively capture the potential convergence and dispersion characteristics of mobile targets by modeling the movement trends between mobile targets at different times; compared with the existing convergence pattern mining algorithm, it is not necessary to first locate the central convergence area of ​​the mobile target, and can identify the early convergence characteristics of the mobile target, thereby enhancing the accuracy of early warning and decision support for group events and abnormal events formed by the convergence of mobile targets; in addition, the method of the present application can be adapted to the mining of convergence patterns of mobile targets in different geographical spaces (such as urban road network space and free space at sea), and can process real-time inflowing mobile target data, providing support for real-time monitoring of convergence-divergence patterns of mobile targets. The method of the present application shows broad application prospects and practical value in the fields of data mining and pattern recognition.

[0136] The following is an exemplary description of the motion pattern mining device based on the moving target trajectory provided by the present application.

[0137] like Figure 8 As shown, the embodiment of the present application provides a motion pattern mining device based on a moving target trajectory, and the motion pattern mining device based on a moving target trajectory 800 includes:

[0138] Acquisition module 801, acquires the trajectory data of each mobile target in the study area at T moments; the Tth moment is the current moment;

[0139] A pairing module 802 pairs all the moving targets in pairs to obtain a plurality of moving target pairs, and obtains a distance change sequence of each moving target pair based on all trajectory data;

[0140] The determination module 803 calculates the convergence rate and separation rate of each mobile target pair according to all distance change sequences, and determines a plurality of convergent mobile targets and a plurality of separate mobile targets from all mobile targets according to all convergence rates and separation rates; the convergent mobile targets are mobile targets whose corresponding convergence rates satisfy the convergence condition, and the separate mobile targets are mobile targets whose corresponding separation rates satisfy the separation condition;

[0141] The clustering module 804 clusters all the converging moving targets according to the trajectory data of all the converging moving targets at the current moment to obtain a plurality of converging clustering clusters, and clusters all the separating moving targets according to the trajectory data of all the separating moving targets at the first moment among the T moments to obtain a plurality of separating clustering clusters;

[0142] The marking module 805 regards the movement of all mobile targets in each convergent cluster within the time period covered by T moments as a convergent pattern, and regards the movement of all mobile targets in each separate cluster within the time period covered by T moments as a separate pattern.

[0143] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0144] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0145] like Fig. 9 As shown, an embodiment of the present application provides a terminal device. The terminal device D10 of this embodiment includes: at least one processor D100 ( Fig. 9Only one processor is shown in the figure), a memory D101, and a computer program D102 stored in the memory D101 and executable on the at least one processor D100, wherein the processor D100 implements the steps of any of the above-mentioned method embodiments when executing the computer program D102.

[0146] Specifically, when the processor D100 executes the computer program D102, it obtains the trajectory data of each mobile target in the study area at T moments, then pairs all the mobile targets in pairs to obtain multiple mobile target pairs, and obtains the distance change sequence of each mobile target pair based on all the trajectory data, and then calculates the convergence rate and separation rate of each mobile target pair according to all the distance change sequences, and determines multiple converging mobile targets and multiple separating mobile targets from all the mobile targets according to all the convergence rates and separation rates, and then clusters all the converging mobile targets according to the trajectory data of all the converging mobile targets at the current moment to obtain multiple converging clustering clusters, and clusters all the separating mobile targets according to the trajectory data of all the separating mobile targets at the first moment among the T moments to obtain multiple separating clustering clusters, and finally takes the movement of all the mobile targets in each converging clustering cluster within the time period covered by T moments as a convergence mode, and takes the movement of all the mobile targets in each separating clustering cluster within the time period covered by T moments as a separation mode. Among them, obtaining the distance change sequence of the moving target pairs can effectively capture and analyze the movement changes between the moving targets. Calculating the convergence rate and separation rate based on the moving target pairs can improve the accuracy and rationality of the convergence rate and separation rate, thereby improving the accuracy of motion pattern mining based on the convergence rate and separation rate. Calculations are performed on trajectory data without restricting the geographical space where the moving targets are located, which effectively improves the versatility and reliability of motion pattern mining.

[0147] The processor D100 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0148] In some embodiments, the memory D101 may be an internal storage unit of the terminal device D10, such as a hard disk or memory of the terminal device D10. In other embodiments, the memory D101 may also be an external storage device of the terminal device D10, such as a plug-in hard disk, a smart memory card (SMC, SmartMedia Card), a secure digital (SD, Secure Digital) card, a flash card (Flash Card), etc. equipped on the terminal device D10. Further, the memory D101 may also include both an internal storage unit of the terminal device D10 and an external storage device. The memory D101 is used to store an operating system, an application program, a boot loader (BootLoader), data, and other programs, such as the program code of the computer program, etc. The memory D101 may also be used to temporarily store data that has been output or is to be output.

[0149] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0150] An embodiment of the present application provides a computer program product. When the computer program product is run on a terminal device, the terminal device can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0151] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium that can carry the computer program code to the motion pattern mining method device / terminal device based on the moving target trajectory. For example, a USB flash drive, a mobile hard disk, a disk or an optical disk.

[0152] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0153] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0154] The above is a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles described in the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A motion pattern mining method based on moving target trajectory, characterized in that: include: Obtain the trajectory data of each mobile target in the study area at T moments; The Tth moment is the current moment; Pairing all the moving targets in pairs to obtain a plurality of moving target pairs, and acquiring a distance change sequence of each of the moving target pairs based on all the trajectory data; The convergence rate and separation rate of each pair of mobile targets are calculated according to all distance change sequences, and a plurality of converging mobile targets and a plurality of separating mobile targets are determined from all mobile targets according to all the converging rates and separation rates; the converging mobile targets are mobile targets whose corresponding converging rates satisfy the converging conditions, and the separating mobile targets are mobile targets whose corresponding separation rates satisfy the separating conditions; According to the trajectory data of all converging moving targets at the current moment, all converging moving targets are clustered to obtain multiple converging cluster clusters, and according to the trajectory data of all separating moving targets at the first moment among T moments, all separating moving targets are clustered to obtain multiple separating cluster clusters; The movement of all mobile targets in each of the convergent clusters within the time period covered by T moments is regarded as a convergent mode, and the movement of all mobile targets in each of the separate clusters within the time period covered by T moments is regarded as a separate mode; Wherein, the step of acquiring the distance change sequence of each pair of moving targets based on all trajectory data includes: For each pair of moving targets, perform the following steps: Calculating the distance between the two moving targets at each moment according to the trajectory data of the two moving targets in the moving target pair; Calculate the distance change between the two moving targets between each two consecutive moments based on all the distances; Integrate all distance changes to obtain a distance change sequence of the moving target pair; The calculation of the distance change between the two moving targets between each two consecutive moments based on all the distances includes: By formula: Calculate in The moment and Between moments, Moving Target With Moving Target The distance change between ; in, Indicated in The moment described The moving target and the The distance between moving targets, Indicated in The moment described The moving target and the The distance between moving targets, Indicated in The moment described The trajectory data of a moving target, Indicated in The moment described The trajectory data of a moving target, Indicated in The moment described The trajectory data of a moving target, Indicated in The moment described The trajectory data of a moving target, .

2. The motion pattern mining method according to claim 1, characterized in that: The step of calculating the distance between the two moving targets at each moment according to the trajectory data of the two moving targets in the moving target pair comprises: By formula: ； Calculate in moment, The moving target and The distance between moving targets ; in, Indicated in Moment The first coordinate value in the moving target trajectory data, Indicated in Moment The second coordinate value in the moving target trajectory data, Indicated in Moment The first coordinate value in the moving target trajectory data, Indicated in Moment The second coordinate value in the moving target trajectory data, , , A set of numbers representing all moving targets.

3. The motion pattern mining method according to claim 1, characterized in that: The step of calculating the convergence rate and separation rate of each pair of moving targets according to all distance change sequences comprises: By formula: Calculate moving target pairs The convergence rate ; in, represents the time step from the initial moment, Indicates the A moving target, Indicates the A moving target; By formula: Calculate moving target pairs Separation rate .

4. The motion pattern mining method according to claim 3, characterized in that: The step of determining a plurality of converging moving targets and a plurality of separating moving targets from all moving targets according to all converging rates and separating rates comprises: For each moving target pair, perform the following steps: If the distance between the two moving targets of the moving target pair is less than the distance threshold at the current moment, determining whether the convergence rate of the moving target pair meets the convergence condition; If the convergence rate of the mobile target pair meets the convergence condition, both of the two mobile targets corresponding to the mobile target pair are used as convergence mobile targets; If in If the distance between the two moving targets of the moving target pair at a certain moment is less than the distance threshold, then judging whether the separation rate of the moving target pair meets the separation condition; If the separation rate of the mobile target pair meets the separation condition, both of the two mobile targets corresponding to the mobile target pair are used as separated mobile targets.

5. The motion pattern mining method according to claim 4, characterized in that: The convergence condition is: The convergence rate is greater than the convergence rate threshold; The separation conditions are: The separation rate is greater than the separation rate threshold.

6. A motion pattern mining device based on moving target trajectory, characterized in that: include: Acquisition module, obtains the trajectory data of each mobile target in the study area at T moments; The Tth moment is the current moment; A pairing module pairs all the moving targets in pairs to obtain a plurality of moving target pairs, and obtains a distance change sequence of each moving target pair based on all trajectory data; A determination module calculates the convergence rate and separation rate of each of the mobile target pairs according to all distance change sequences, and determines a plurality of convergent mobile targets and a plurality of separate mobile targets from all mobile targets according to all convergence rates and separation rates; the convergent mobile targets are mobile targets whose corresponding convergence rates satisfy the convergence condition, and the separate mobile targets are mobile targets whose corresponding separation rates satisfy the separation condition; A clustering module clusters all converging moving targets according to their trajectory data at the current moment to obtain a plurality of converging clustering clusters, and clusters all separating moving targets according to their trajectory data at the first moment among T moments to obtain a plurality of separating clustering clusters; A marking module, which takes the movement of all mobile targets in each of the convergent clusters within the time period covered by T moments as a convergent mode, and takes the movement of all mobile targets in each of the separate clusters within the time period covered by T moments as a separate mode; The pairing module is specifically used to implement: For each pair of moving targets, perform the following steps: Calculating the distance between the two moving targets at each moment according to the trajectory data of the two moving targets in the moving target pair; Calculate the distance change between the two moving targets between each two consecutive moments based on all the distances; Integrate all distance changes to obtain a distance change sequence of the moving target pair; The calculation of the distance change between the two moving targets between each two consecutive moments based on all the distances includes: By formula: Calculate in The moment and Between moments, Moving Target With Moving Target The distance change between ; in, Indicated in The moment described The moving target and the The distance between moving targets, Indicated in The moment described The moving target and the The distance between moving targets, Indicated in The moment described The trajectory data of a moving target, Indicated in The moment described The trajectory data of a moving target, Indicated in The moment described The trajectory data of a moving target, Indicated in The moment described The trajectory data of a moving target, .

7. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the motion pattern mining method based on the moving target trajectory as described in any one of claims 1 to 5 is implemented.

8. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the motion pattern mining method based on the trajectory of a moving target as described in any one of claims 1 to 5 is implemented.

Images