JP6677040B2 - Trajectory data processing method, trajectory data processing program and trajectory data processing device - Google Patents

Description

  The present invention relates to a trajectory data processing method, a trajectory data processing program, and a trajectory data processing device.

  Conventionally, a route analysis using a plurality of trajectory data has been performed. For example, in the path analysis, a common part of the path is obtained by comparing the paths for each pair of two pieces of path data from a set of a plurality of path data, and a pair of path data having a high similarity is extracted. Thereby, for example, it can be used for proposing, as friends, users having a high degree of similarity in the locus data, and for proposing a taxi sharing.

JP 2014-41617 A JP 2011-40064 A JP 2014-229069 A JP 2012-202796 A

  However, as the number of trajectory data increases, the number of pairs of trajectory data rapidly increases, and the route analysis may not be able to be processed in a realistic time. For example, there is a case where the processing of the path analysis of a set of trajectory data obtained daily does not complete in one day.

  In general, in path analysis, a pair having a longer common part in a path is more important. The common part of the path is a part where the path of the pair of the track data to be compared matches. For this reason, the common part of the route is shorter than the length of each route of the track data to be compared. Therefore, it is conceivable that the calculation of the route analysis is performed from the locus data having a long locus, and the calculation of the route analysis is terminated based on a predetermined standard, thereby reducing the amount of calculation. However, since it is difficult to appropriately set the criteria for terminating the calculation in advance, the amount of calculation cannot be reduced appropriately.

  An object of one aspect is to provide a trajectory data processing method, a trajectory data processing program, and a trajectory data processing device that can appropriately reduce the amount of calculation.

  In the first plan, in the trajectory data processing method, the computer executes a process of receiving designation of the number of pairs of a plurality of trajectory data. The trajectory data processing method includes storing a plurality of trajectory data including a plurality of pieces of position information respectively associated with any one of a plurality of nodes associated with identification information different from each other. See section. In the trajectory data processing method, the computer executes a process of acquiring trajectory data of a number of nodes equal to or more than a specific number of nodes from the storage unit. In the trajectory data processing method, the computer executes, for each pair of the acquired trajectory data, a process of calculating the number of common nodes having the same identification information between the trajectory data included in each pair. In the trajectory data processing method, a computer executes a process of generating distribution information of the number of pairs for each number of common nodes based on the calculated number of common nodes. In the trajectory data processing method, the computer executes a process of determining the number of other nodes less than the specific number of nodes based on the generated distribution information and the number of accepted pairs. In the trajectory data processing method, the computer executes a process of outputting a pair of trajectory data of the number of nodes between a specific number of nodes and another number of nodes.

  According to one embodiment of the present invention, there is an effect that the amount of calculation can be appropriately reduced.

FIG. 1 is a diagram illustrating an example of a system configuration. FIG. 2 is a diagram schematically showing a functional configuration of the trajectory data processing device. FIG. 3 is a diagram illustrating an example of a data configuration of the movement trajectory data. FIG. 4 is a diagram illustrating an example of a data configuration of node information. FIG. 5 is a diagram illustrating an example of a data configuration of link information. FIG. 6 is a diagram illustrating an example of a data configuration of the normalized trajectory data. FIG. 7 is a diagram illustrating an example of normalization. FIG. 8 is a diagram illustrating an example of a common part between the normalized trajectories. FIG. 9 is a diagram illustrating an example of the distribution of the number of pairs for each number of common nodes. FIG. 10A is a diagram illustrating an example of the distribution of the number of pairs for each number of common nodes. FIG. 10B is a diagram illustrating an example of the distribution of the number of pairs for each number of common nodes. FIG. 11A is a diagram illustrating an example in which it is difficult to appropriately set a conventional standard in advance. FIG. 11B is a diagram illustrating an example in which it is difficult to appropriately set a conventional standard in advance. FIG. 11C is a diagram illustrating an example in which it is difficult to appropriately set a conventional standard in advance. FIG. 12 is a diagram illustrating the flow of estimating the distribution. FIG. 13 is a diagram illustrating an example of a result of estimating the distribution of the number of pairs by the first method. FIG. 14 is a diagram illustrating the estimation of the second method. FIG. 15 is a diagram illustrating an example of the movement trajectory data. FIG. 16 is a diagram illustrating an example of the configuration of a road node. FIG. 17 is a diagram illustrating an example of the normalized trajectory data. FIG. 18A is a diagram illustrating an example of a processing state. FIG. 18B is a diagram illustrating an example of the processing status. FIG. 18C is a diagram illustrating an example of a processing state. FIG. 18D is a diagram illustrating an example of the processing status. FIG. 18E is a diagram illustrating an example of the processing status. FIG. 19 is a flowchart illustrating an example of the flow of the normalization processing. FIG. 20 is a flowchart illustrating an example of the flow of the trajectory data processing. FIG. 21 is a diagram illustrating an example of a computer that executes a trajectory data processing program.

  Hereinafter, embodiments of a trajectory data processing method, a trajectory data processing program, and a trajectory data processing device according to the present invention will be described in detail with reference to the drawings. The present invention is not limited by the embodiment. The embodiments can be appropriately combined within a range that does not contradict processing contents.

[System configuration]
First, an example of the system will be described. FIG. 1 is a diagram illustrating an example of a system configuration. The system 10 is a system that performs a path analysis of a trajectory. The system 10 includes a trajectory data processing device 11 and a trajectory recording device 12. The trajectory data processing device 11 and the trajectory recording device 12 are communicably connected to a network N. As an aspect of the network N, any type of communication, such as a mobile communication network such as a mobile phone, the Internet (Internet), a LAN (Local Area Network), and a VPN (Virtual Private Network), regardless of whether it is wired or wireless. Net can be adopted. Note that, in the example of FIG. 1, the case where one trajectory recording device 12 is provided is exemplified, but the present invention is not limited to this, and the trajectory recording devices 12 can be any number.

  The locus recording device 12 is a device that records a locus of movement. For example, the trajectory recording device 12 is a portable information processing device such as a smartphone or a tablet terminal, or an on-board device mounted on a vehicle such as a train or a car. The trajectory recording device 12 is provided with a sensor capable of detecting a position such as a global positioning system (GPS) module, detects the current position periodically, and records position information of the detected current position as trajectory data. The locus recording device 12 uploads locus data to the locus data processing device 11 at a predetermined timing. For example, the trajectory recording device 12 uploads trajectory data via the network N. The trajectory recording device 12 may upload the trajectory data to the trajectory data processing device 11 via a storage medium such as a memory card.

  The trajectory data processing device 11 is a device that performs path analysis of trajectory data. The trajectory data processing device 11 is, for example, a computer such as a personal computer or a server computer. The trajectory data processing device 11 may be implemented as one computer, or may be implemented by a plurality of computers. In this embodiment, a case where the locus data processing device 11 is a single computer will be described as an example.

[Configuration of trajectory data processing device]
Next, the configuration of the trajectory data processing device 11 will be described. FIG. 2 is a diagram schematically showing a functional configuration of the trajectory data processing device. The trajectory data processing device 11 has a communication I / F (interface) unit 20, an input unit 21, a display unit 22, a storage unit 23, and a control unit 24. Note that the trajectory data processing device 11 may include devices other than the above devices.

  The communication I / F unit 20 is an interface that controls communication with another device. The communication I / F unit 20 transmits and receives various information to and from other devices via the network N. For example, the communication I / F unit 20 receives trajectory data from the trajectory recording device 12 via the network N. As the communication I / F unit 20, a network interface card such as a LAN card can be adopted.

  The input unit 21 is an input device for inputting various information. Examples of the input unit 21 include an input device that receives an input of an operation such as a mouse or a keyboard. The input unit 21 receives various inputs related to the route analysis.

  The display unit 22 is a display device that displays various information. The display unit 22 includes a display device such as an LCD (Liquid Crystal Display) or a CRT (Cathode Ray Tube). The display unit 22 displays various information. For example, the display unit 22 displays various screens such as an operation screen.

  The storage unit 23 is a storage device that stores various data. For example, the storage unit 23 is a storage device such as a hard disk, a solid state drive (SSD), and an optical disk. The storage unit 23 may be a rewritable semiconductor memory such as a random access memory (RAM), a flash memory, or a non-volatile random access memory (NVSRAM).

  The storage unit 23 stores an OS (Operating System) executed by the control unit 24 and various programs. For example, the storage unit 23 stores various programs including a program for executing a later-described normalization process and a trajectory data process. Further, the storage unit 23 stores various data used in the program executed by the control unit 24. For example, the storage unit 23 stores movement trajectory data 30, node information 31, link information 32, normalized trajectory data 33, pair common portion information 34, and analysis result information 35.

  The movement trajectory data 30 is data that stores trajectory data uploaded from the trajectory recording device 12. FIG. 3 is a diagram illustrating an example of a data configuration of the movement trajectory data. As shown in FIG. 3, the movement trajectory data 30 has items of “trajectory ID”, “X coordinate”, and “Y coordinate”. Each item of the movement trajectory data 30 shown in FIG. 3 is an example, and may have other items. Further, the data configuration of the movement trajectory data 30 is also an example, and is not limited to this.

  The item of the track ID is an area for storing identification information for identifying each track. A unique track ID (identification) is assigned to the track as identification information. The trajectory ID may be assigned, for example, for each uploaded trajectory data, for each trajectory recording device 12, or for each user who operates the trajectory recording device 12. Further, the trajectory ID may be newly provided for each trajectory recording device 12 for each day, or may be newly provided for a predetermined period or a predetermined movement range, for example. The trajectory ID only needs to be given in units that divide one trajectory. The items of X coordinate and Y coordinate are areas for storing position information indicating a position. For example, in the items of X coordinate and Y coordinate, information on the latitude and longitude measured by GPS is stored. In the example of FIG. 3, the locus ID “t1” indicates that the X coordinate is “z1” and the Y coordinate is “w1”.

  Returning to FIG. 2, the node information 31 is data that stores the position of the node when the trajectory is normalized. FIG. 4 is a diagram illustrating an example of a data configuration of node information. As shown in FIG. 4, the node information 31 has items of “node”, “X coordinate”, and “Y coordinate”. Each item of the node information 31 shown in FIG. 4 is an example, and may have other items. Further, the data configuration of the node information 31 is also an example, and is not limited to this.

  The node item is an area for storing identification information for identifying each node. The node is provided with identification information. The items of the X coordinate and the Y coordinate are areas for storing position information indicating the position of the node. For example, in the items of the X coordinate and the Y coordinate, information on latitude and longitude indicating the position of the node is stored. In the example of FIG. 4, the node N1 indicates that the X coordinate is “x1” and the Y coordinate is “y1”.

  Returning to FIG. 2, the link information 32 is data that stores information on links between nodes. FIG. 5 is a diagram illustrating an example of a data configuration of link information. As shown in FIG. 5, the link information 32 has an item of “link”. Each item of the link information 32 shown in FIG. 5 is an example, and may have other items. Further, the data configuration of the link information 32 is also an example, and is not limited to this.

  The item of link is an area for storing identification information of each node in a connection relationship. In the link item, identification information of each node having a connection relationship is stored as a link. In the example of FIG. 5, the node Ni and the node Nj have a connection relationship.

  Returning to FIG. 2, the normalized trajectory data 33 is data obtained by normalizing the position information of the movement trajectory data 30 into nodes. FIG. 6 is a diagram illustrating an example of a data configuration of the normalized trajectory data. As shown in FIG. 6, the normalized trajectory data 33 has items of “ID” and “normalized trajectory”. Each item of the normalized trajectory data 33 shown in FIG. 6 is an example, and may have other items. Further, the data configuration of the normalized trajectory data 33 is also an example, and the present invention is not limited to this.

  The item of ID is an area for storing the track ID of the normalized track. The item of the normalized trajectory is an area for storing the identification information of the node corresponding to each position of the trajectory along the trajectory. In the example of FIG. 6, the locus of the ID “t1” indicates that the locus passes through the nodes N11, N12, N13, and N14 in order.

  Returning to FIG. 2, the pair common part information 34 is data in which a path analysis is performed on the movement trajectory data 30 and information about a common part of each trajectory pair is stored. The analysis result information 35 is data that stores information on the analysis result of the route analysis. An example of a data configuration of the pair common portion information 34 and the analysis result information 35 will be described later.

  The control unit 24 is a device that controls the trajectory data processing device 11. As the control unit 24, an electronic circuit such as a CPU (Central Processing Unit) or an MPU (Micro Processing Unit), or an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array) can be adopted. The control unit 24 has an internal memory for storing programs and control data defining various processing procedures, and executes various processing by using these. The control unit 24 functions as various processing units when various programs operate. For example, the control unit 24 includes a storage unit 40, a normalization unit 41, a reception unit 42, an acquisition unit 43, a calculation unit 44, a generation unit 45, a determination unit 46, and an output unit 47.

  The storage unit 40 stores the locus data uploaded from the locus recording device 12 in the moving locus data 30. For example, the storage unit 40 assigns a trajectory ID each time trajectory data is uploaded, and stores the trajectory data in the movement trajectory data 30.

  The normalization unit 41 normalizes the position information stored in the movement trajectory data 30 to a node. For example, based on the node information 31 and the link information 32, the normalization unit 41 normalizes each piece of position information of the trajectory stored in the movement trajectory data 30 into data indicated by a node corresponding to the position. For example, the normalization unit 41 performs a map matching of each position information of the trajectory stored in the movement trajectory data 30 with the position information of each node of the node information 31, and normalizes the position of the trajectory into data indicating the position of the trajectory by the node. Become The normalization unit 41 stores the normalized trajectory data in the normalized trajectory data 33.

  Here, an example of normalization will be described. FIG. 7 is a diagram illustrating an example of normalization. FIG. 7 shows trajectories L1 and L2. FIG. 7 shows nodes A to H. The normalization unit 41 performs map matching of each position information of the trajectories L1 and L2 with each position information of the nodes A to H, and normalizes the trajectories L1 and L2 to data represented by nodes. In the example of FIG. 7, the trajectory L1 is normalized to <A, E, F, B, C, D>. <A, E, F, B, C, D> indicates that the vehicle passes through the nodes A, E, F, B, C, and D in this order. The trajectory L2 is normalized to <A, F, G, C, D>. <A, F, G, C, D> indicates that the light passes through the nodes A, F, G, C, and D in order.

  By normalizing the trajectory to the nodes in this manner, it becomes easy to specify a common part between the trajectories. FIG. 8 is a diagram illustrating an example of a common part between the normalized trajectories. FIG. 8 shows a normalized locus <A, E, F, B, C, D> of the locus L1 shown in FIG. 7 and a normalized locus <A, F, G, C, D> of the locus L2. Are shown. The intersection between the normalized trajectories can be identified by finding the matching nodes. The common part between the trajectories may include an unmatched node between or before and after the matching node. In the example of FIG. 8, the node E exists between the node A and the node F of the trajectory L1 in <A, E, F,...> Of the trajectory L1 and in <A, F,. However, the nodes A and F are common parts. In the example of FIG. 8, the common sentence of <A, E, F, B, C, D> of the trajectory L1 and <A, F, G, C, D> of the trajectory L2 are <A, F, C, D>. Hereinafter, in the present embodiment, the length of the trajectory is represented by the number of nodes. For example, the length of the trajectory is represented by the number of nodes included in the normalized trajectory data. For example, the length of the trajectory L1 is 6 as the number of nodes since the normalized trajectory is <A, E, F, B, C, D>. The length of the common part of the trajectory data is represented by the number of nodes included in the common part. For example, since the common part of the trajectory L1 and the trajectory L2 is <A, F, C, D>, the length is set to 4 as the number of nodes.

  The receiving unit 42 performs various operations. For example, the receiving unit 42 receives input of various types of information related to route analysis. For example, the receiving unit 42 displays an operation screen and receives designation of the number of pairs of a plurality of locus data as the target number of pairs of locus data to be extracted by the route analysis.

  The acquisition unit 43 refers to the normalized trajectory data 33 in the storage unit 23 and acquires trajectory data of a number of nodes equal to or more than a specific number of nodes. For example, the acquisition unit 43 reduces the number of nodes in order from the maximum number of nodes of the trajectory stored in the normalized trajectory data 33 to the number of nodes determined by the determination unit 46, which will be described later. Get number trajectory data. For example, when the maximum number of nodes of the trajectory stored in the normalized trajectory data 33 is 6, the acquiring unit 43 first reads trajectory data having six nodes from the normalized trajectory data 33. When the maximum number of nodes is 6, by reading the locus data having the number of nodes of 6, the locus data having the number of nodes of 6 or more is obtained. Next, the acquiring unit 43 reads the locus data having five nodes. Since the locus data having the number of nodes of 6 or more is acquired first, the locus data having the number of nodes of 5 or more is acquired by reading the locus data having the number of nodes of 5. Similarly, the acquiring unit 43 sequentially decreases the number of nodes by one until the number of nodes reaches the number of nodes determined by the determining unit 46 described later, and reads the locus data of the number of nodes.

  The calculating unit 44 calculates, for each pair of the locus data acquired by the acquiring unit 43, the number of common nodes in which the node identification information is common between the locus data included in each pair. For example, when the acquisition unit 43 acquires the trajectory data, the calculation unit 44 calculates the number of common nodes between the trajectory data included in each pair for each pair obtained by combining the acquired trajectory data. For example, when the locus data having six nodes is acquired, the calculating unit 44 calculates, for each pair of locus data having six nodes, the number of common nodes between the locus data included in each pair. Next, when the trajectory data with the number of nodes of 5 is acquired, the calculation unit 44 calculates, for each pair of the trajectory data with the number of nodes of 6 and 5, the number of common nodes between the trajectory data included in each pair. . Here, the number of common nodes of a pair of locus data having only six nodes is calculated first. Therefore, the calculation unit 44 calculates the number of common nodes for each pair of the trajectory data, excluding the pair for which the number of common nodes has been calculated. For example, the calculation unit 44 calculates the number of common nodes for each pair obtained by combining locus data having 6 nodes and 5 locus data and for each pair obtained by combining only locus data having 5 nodes. I do.

  The generation unit 45 generates distribution information of the number of pairs for each number of common nodes based on the number of common nodes calculated by the calculation unit 44.

  Here, in the route analysis, in general, the longer the common portion of the route, the more important the pair. The common part of the path is a part where the path of the pair of the track data to be compared matches. For this reason, the common part of the route is shorter than the length of each route of the track data to be compared. For example, when two locus data having three nodes are compared, the common part of the route has three or less nodes and does not have four or more nodes. Therefore, when the number of common nodes, which is a common part of the route, is calculated in order from the locus data having a large number of nodes, the number of pairs is determined from the one having a large number of common nodes.

  FIG. 9 is a diagram illustrating an example of the distribution of the number of pairs for each number of common nodes. The example of FIG. 9 shows the distribution of the number of pairs for each number of common nodes by obtaining the number of common nodes between trajectory data that is equal to or more than the number of nodes to be obtained by sequentially reducing the number of nodes to be obtained from 6 to 1 by one. It is a thing. The vertical axis of each graph in FIG. 9 indicates the number of pairs on a logarithmic scale. The horizontal axis of each graph in FIG. 9 indicates the number of common nodes.

  In the example of FIG. 9, first, the number of common nodes of a pair of locus data having six or more nodes is calculated, and a distribution of the number of pairs based on locus data having six or more nodes is generated. In the distribution of the number of pairs based on the trajectory data having the number of nodes of 6 or more, the number of pairs having the common node number of 5 is 1, and the number of pairs having the number of common nodes of 1 is 2. Next, in the example of FIG. 9, the number of common nodes of a pair of locus data having five or more nodes is calculated, and a distribution of the number of pairs based on locus data having five or more nodes is generated. In the distribution of the number of pairs based on the trajectory data having five or more nodes, the number of pairs having five common nodes is one, and the number of pairs having zero common nodes is two. Here, for a pair of locus data having four or less nodes, the number of nodes in the common part of the route is four or less, and the number of nodes in the common part of the route does not become five or more. Therefore, in the distribution of the number of pairs based on the trajectory data having five or more nodes, the number of pairs having five or more nodes at the common part of the route is determined. In the example of FIG. 9, the number of pairs where the number of common nodes is 5 is determined to be 1.

  Next, in the example of FIG. 9, the number of common nodes of a pair of locus data having four or more nodes is calculated, and a distribution of the number of pairs based on locus data having four or more nodes is generated. As a result, the number of pairs in which the number of nodes in the common part of the route is 4 or more is determined. In the example of FIG. 9, the number of pairs with four common nodes is determined to be zero. Next, in the example of FIG. 9, the number of common nodes of a pair of locus data having three or more nodes is calculated, and a distribution of the number of pairs based on locus data having three or more nodes is generated. As a result, the number of pairs in which the number of nodes in the common part of the route is 3 or more is determined. In the example of FIG. 9, the number of pairs is determined to be 4 while the number of common nodes is 3. Next, in the example of FIG. 9, the number of common nodes of a pair of locus data having two or more nodes is calculated, and a distribution of the number of pairs based on locus data having two or more nodes is generated. As a result, the number of pairs for which the number of nodes in the common part of the route is 2 or more is determined. In the example of FIG. 9, the number of pairs with the number of common nodes being 2 is determined to be 19. Next, in the example of FIG. 9, the number of common nodes of a pair of locus data having one or more nodes is calculated, and a distribution of the number of pairs based on locus data having one or more nodes is generated. As a result, the number of pairs in which the number of nodes in the common part of the route is 1 or more is determined. In the example of FIG. 9, the number of pairs where the number of common nodes is 1 is determined to be 67. In addition, the number of pairs whose paths are not common to any of the nodes where the number of nodes in the common part of the path is 0 is also determined. In the example of FIG. 9, the number of pairs where the number of common nodes is 0 is determined to be 209. As described above, when the number of common nodes, which is a common part of the route, is calculated in order from the locus data having a large number of nodes, the number of pairs is determined from the one having a large number of common nodes.

  In general, when the common part of the route is obtained for each pair obtained by combining a plurality of trajectory data, the distribution of the number of pairs for each number of common nodes shows that the number of pairs decreases rapidly as the number of common nodes increases. Often it has an exponential distribution. 10A and 10B are diagrams illustrating an example of the distribution of the number of pairs for each number of common nodes. As shown in FIG. 10A and FIG. 10B, as the number of common nodes increases, the number of pairs sharply decreases, and the distribution is exponential. However, the maximum value of the distribution and how to reduce the number of pairs with respect to the number of common nodes differ for each distribution. For this reason, it is difficult to appropriately set the criteria for terminating the calculation in advance.

  For example, the target number of pairs to be extracted is set to 60. In such a case, for example, as a criterion for terminating the calculation, the length of the trajectory data is set to three nodes in advance. 11A and 11B are diagrams illustrating an example in which it is difficult to appropriately set a conventional standard in advance. FIGS. 11A and 11B show an example of the distribution of the number of pairs for each number of common nodes, and the distributions are different. In the example of FIG. 11A, if the number of nodes of the locus data is extracted up to three nodes of the common part, the number of pairs becomes 79, which is more than the target of 60, and it is not possible to narrow down the pairs of the locus data. On the other hand, in the example of FIG. 11B, if the number of nodes of the common portion is extracted up to three pairs of trajectory data, the number of pairs becomes 1 and less than the target of 60, and a sufficient number of pairs cannot be obtained.

  Further, for example, it is assumed that calculation is performed until the target number of 60 is obtained. FIG. 11C is a diagram illustrating an example in which it is difficult to appropriately set a conventional standard in advance. FIG. 11C shows an example of the distribution of the number of pairs for each number of common nodes. In the distribution of FIG. 11C, the number of pairs of trajectory data in which the number of nodes in the common part is 1 or more is 54, which is smaller than the target of 60. Therefore, if the number of nodes in the common portion is included up to 0, the number of pairs increases, and it is not possible to narrow down the pairs of the track data.

  Therefore, the determining unit 46 determines the number of other nodes that is less than the specific number of nodes based on the distribution information generated by the generating unit 45 and the specified target number. For example, the determination unit 46 estimates the number of pairs of trajectory data that is equal to or more than the number of common nodes for each number of common nodes, with the plurality of trajectory data stored in the normalized trajectory data 33 as the estimation target from the distribution information. . Then, the determination unit 46 determines the number of common nodes whose estimated number of pairs is closest to the specified target number as the number of other nodes. That is, the determination unit 46 determines the number of common nodes as a reference for terminating the calculation as the number of other nodes.

  The output unit 47 performs various outputs. For example, the output unit 47 outputs a pair of locus data of the number of nodes between a specific number of nodes and another number of nodes. For example, the output unit 47 outputs, to the analysis result information 35, a pair of the trajectory data of the specific number of nodes and the number of nodes between the number of common nodes used as a reference for terminating the calculation as the result of the route analysis. The output unit 47 may output the result of the route analysis to another device or another program. The output unit 47 may output a screen displaying the result of the route analysis to the display unit 22.

  As described above, in general, when a common portion of a route is obtained for each pair obtained by combining a plurality of trajectory data, the distribution of the number of pairs for each number of common nodes is an exponential distribution. For this reason, the entire distribution can be estimated from partial information of the distribution. The deciding unit 46 estimates the entire distribution of all the trajectory data based on the distribution of the distribution information generated by the generating unit 45, determines whether to proceed with the calculation, and determines the number of nodes within the range in which the calculation is performed. . FIG. 12 is a diagram illustrating the flow of estimating the distribution. FIG. 12 shows the distribution of the number of pairs for each number of common nodes for locus data having two or more nodes. The number of pairs with two or more nodes is determined. In the example of FIG. 12, the target number is set to 40. In the example of FIG. 12, the number of pairs having two or more common nodes is 24. For example, the determination unit 46 estimates the entire distribution in all the trajectory data using the distribution of the number of pairs having two or more nodes. In the example of FIG. 12, the number of pairs whose common node number is 1 or more is estimated to be 70 from the estimated distribution. The determination unit 46 determines that the target number is closer to 40 when the number of common nodes is 2 or more than 24 when the estimated number of common nodes is 70 or more than 70 when the estimated number of common nodes is 1 or more. The number of common nodes is determined to be two. Accordingly, the calculation is completed without performing the calculation for the trajectory data having one or more nodes, so that the calculation amount can be reduced.

  Here, a calculation method for estimating the number of pairs of trajectory data will be described. A plurality of methods are illustrated as a calculation method for estimating the number of pairs of the trajectory data.

  First, the first method will be described. In the first method, the values of the determined portion and the undetermined portion of the calculated distribution are used to approximate the geometric distribution. The geometric distribution can be generally expressed as in the following equation (1).

P (X = k) = p · q ^k (1)

Equation (1) of this geometric distribution can be approximated from the following equation (2) as in the following equations (3) and (4).

  By averaging the expressions (3) and (4) with k = 1,..., K, they can be expressed as the following expressions (5) and (6).

here,
K: the maximum value of the calculated number of common nodes N: the total number of combinations of the trajectory data to be estimated N _k : the number of pairs where the calculated number of common nodes is k or more N ~ _k : the number of pairs where the number of common nodes is k or more Number estimates

FIG. 13 is a diagram illustrating an example of a result of estimating the distribution of the number of pairs by the first method. The example of FIG. 13 illustrates a case where the distribution of the number of pairs for each number of common nodes is calculated in the order of three or more nodes, two or more nodes, and one or more nodes (all trajectory data calculations). . In the example of FIG. 13, for each common node number k in each distribution, the number of pairs of total number of pairs or the common node number k becomes N _k. In the example of FIG. 13, an estimated value obtained by estimating the distribution of the number of pairs is shown. The estimated value is calculated as follows. In the example of FIG. 13, the maximum value K of the calculated number of common nodes is 5. In the example of FIG. 13, the total number N of combinations of the trajectory data to be estimated is 300. In the example of FIG. 13, assuming that K = 5 and N = 300, the value of q is obtained from Expression (6) using the number of pairs N _k whose number of calculated common nodes is equal to or more than _k . Then, in the example of FIG. 13, the estimated values N to _k in the case of the number k of nodes of the common part are obtained from the equation (5) using the obtained value of q.

Here, in the equation (5), the estimated values N to _{k of the} number of pairs of the trajectory data that is equal to or more than the number k of the common nodes are calculated. For example, when k = 3, estimated values N to _{3 of the} number of pairs of trajectory data in which the number of common nodes is 3 or more are calculated. Also, when the k = 2, the estimated value N ~ ₂ number of pairs locus data number common node is 2 or more is calculated. The estimated value of the number of pairs of the common node number k is estimated from the estimated number of pairs of trajectory data N to _k equal to or more than the number of common nodes _k to the estimated value N to _{k + 1 of the} number of pairs of trajectory data equal to or more than the number k + 1 of common nodes. Is subtracted. For example, estimates of the number of pairs of the common node number 2, from the estimated value N ~ ₂ number of pairs locus data number common node is 2 or more, the number of pairs of locus data as a common node number is 3 or more It is obtained by subtracting the estimated values N to ₃ .

In the first method, the lower limit of the distribution is estimated. The value of N _k is, for example, a value when all the trajectory data is calculated at a stage where all the trajectory data is not calculated, for example, when the number of nodes is 3 or more or the number of nodes is 2 or more. (True value) or less. Therefore, in the intermediate stage, the value of N ~ _k calculated using the value of N _k becomes the lower limit value of N ~ _k calculated using the true value.

Further, in the first method, the number of pairs N _{k obtained} by accumulating the number of pairs equal to or more than the number k of common nodes is used for the calculation of estimation for all the number k of common nodes. Thus, by calculating the number of pairs N _k be added estimate N ~ _k undetermined portion, than without addition also estimate N ~ _k of the number of pairs N _k was calculated using the true value Becomes closer to the value. That is, in the first method, using the values of the determined portion and the undetermined portion in the calculated distribution improves the accuracy.

  Next, the second method will be described. In the second method, the approximate value is approximated by a geometric distribution using the maximum value of the calculated number of common nodes.

The target number and _{N 0.} If Nq ^k0 ≒ N ₀ with respect to the reference k ₀ , it can be expressed as in Expression (8) from Expression (7) below.

The equation (8), the total number N of combinations of locus data to be estimated, can be calculated from only the target number N _0. This is because, in the first method, Logq is approximated as in Expression (6), whereas in the second method, Logq is approximated as in Expression (9) below. That is, in the second method, the parameter q of the distribution is approximated only by the maximum value K of the calculated number of common nodes and the total number N of the combinations of the trajectory data to be estimated.

FIG. 14 is a diagram illustrating the estimation of the second method. In the example of FIG. 14, the distribution of the number of pairs for each number of common nodes is shown for trajectory data having two or more nodes. In the example of FIG. 14, the maximum value K of the number of common nodes is determined to be 5 from the calculated distribution. In the example of FIG. 14, the designated target number N ₀ is set to 40. In the example of FIG. 14, the total number N of the combinations of the trajectory data to be estimated is 300. Thereby, the value on the right side of Expression (8) can be calculated. For example, the right side is 0.35. This value can be calculated in advance from the total number N and the target number N. In the second method, the lower limit of the distribution is estimated using the maximum value K = 5 of the common part. FIG. 14 shows the distribution of the number of pairs with respect to the estimated number of common nodes. FIG. 14 shows the distribution of the number of pairs obtained by accumulating the number of pairs equal to or more than the number of common nodes for each number of common nodes. In the second method, using the maximum value K of the common node number, to calculate the value of the reference k ₀ from equation (8). In the example of FIG. 14, the reference _{k 0} is calculated as 1.75. This corresponds to determining the value of the reference k ₀ corresponding to 40 is the target number. In the second method, the reference k ₀ is rounded off to 1.75, and the number of common nodes closest to the target number is determined as 2. That is, in the second method, at the stage of definite maximum value K of the intersection, the common node number is determined closest to the target number N _0.

  When the number of common nodes determined by the determination unit 46 is equal to or less than the number of nodes for which the current acquisition has been completed, the acquisition unit 43 reduces the number of nodes in order until the number of common nodes is the closest to the determined number of common nodes. The acquisition of the track data of the number of nodes equal to or more than the number is repeated. As a result, the calculation unit 44, the generation unit 45, and the determination unit 46 perform the same processing again on the locus data acquired again. That is, the acquisition unit 43, the calculation unit 44, the generation unit 45, and the determination unit 46 sequentially reduce the number of nodes until the number of common nodes determined by the determination unit 46 becomes the number of nodes for which the current acquisition has been completed. Repeat the process.

  Next, a description will be given using a specific example. FIG. 15 is a diagram illustrating an example of the movement trajectory data. For example, it is assumed that movement locus data 30 shown in FIG. 15 is stored. FIG. 16 is a diagram illustrating an example of the configuration of a road node. For example, it is assumed that a network such as a road has a connection relation of nodes shown in FIG. FIG. 16 shows data of the node information 31 and the link information 32 corresponding to the network.

  The normalization unit 41 normalizes each position information of the trajectory stored in the movement trajectory data 30 into data indicating the position of the trajectory as a node based on the node information 31 and the link information 32. FIG. 17 is a diagram illustrating an example of the normalized trajectory data. FIG. 17 shows normalized trajectory data 33 in which the position of the trajectory is indicated by a node. In the example of FIG. 17, a normalized trajectory having an ID of 1 to 25 is shown. Also, in the example of FIG. 17, the result of sorting the trajectories in the order of the number of nodes (longest trajectory) is shown.

  The receiving unit 42 receives designation of a target number of pairs to be extracted. Here, it is assumed that the target number is designated as 40.

  The acquisition unit 43 refers to the normalized trajectory data 33 and acquires trajectory data of a number of nodes equal to or more than a specific number of nodes. For example, the acquisition unit 43 sequentially reduces the number of nodes from the maximum number of nodes of the trajectory stored in the normalized trajectory data 33 and acquires trajectory data of the number of nodes equal to or greater than the number of nodes. In the example of FIG. 17, the acquiring unit 43 first reads the locus data in which the maximum number of nodes of the locus stored in the normalized locus data 33 is 6 or more.

  Here, in order to make it easier to understand the status during processing, the processing status of data and parameters will be illustrated and described. FIG. 18A is a diagram illustrating an example of a processing state. The example of FIG. 18A shows the processing status for the locus data having the number of nodes of 6 or more.

  The number of processing nodes x stores the number of nodes of the trajectory from which the reading was performed. In the example of FIG. 17, 6 which is the maximum number of nodes is set as the number of processing nodes x. In the candidate list, the read trajectory is stored. When the number x of processing nodes is 6, the trajectories of IDs 6, 7, and 21 having the number of nodes of 6 or more are stored in the candidate list.

  The calculating unit 44 calculates, for each pair of the acquired trajectory data, the number of common nodes between the trajectory data included in each pair. In the example of FIG. 17, the calculation unit 44 calculates the number of common nodes for each pair in which the trajectories with IDs 6, 7, and 21 are combined. For example, the calculation unit 44 compares the locus <B, A, F, A, B, C> with an ID of 6 with the locus <B, A, F, G, B, C> with an ID of 7, The number of common nodes is calculated as 5.

  FIG. 18A shows a calculation result of the common node in a table format. The calculation result of the common node indicates the number of common nodes calculated between the trajectories, using the IDs of the compared trajectories as vertical and horizontal items. For example, since the trajectories with IDs of 6, 7, and 21 are combined, in the calculation result of the common node in FIG. 18A, IDs 6 and 7 are arranged in horizontal items, and IDs 7 and 21 are arranged in vertical items. I have. In the calculation result of the common node in FIG. 18A, since the value of the area corresponding to the horizontal item is 6 and the vertical item is 7 is 5, the common node calculated between the trajectories of ID 6 and ID 7 is This indicates that the number is 5.

  The generation unit 45 generates distribution information of the number of pairs for each number of common nodes based on the number of common nodes calculated by the calculation unit 44. FIG. 18A shows the number of pairs for each number of common nodes in the distribution information. In FIG. 18A, the number of pairs is stored as 1 for the number of common nodes of 5. Further, the number of pairs is stored as 0 for the number of common nodes of 4 to 1. Further, the number of pairs is stored as 2 for the number of common nodes of 0.

  The deciding unit 46 estimates the number of pairs of trajectory data that is equal to or more than the number of common nodes for each number of common nodes, with the plurality of trajectory data stored in the normalized trajectory data 33 as the estimation target from the distribution information. Then, the determining unit 46 determines the number of common nodes whose estimated number of pairs is closest to the specified target number as the reference number of common nodes to terminate the calculation.

  FIG. 18A shows parameters and estimation results in the case of performing estimation using the first method and the second method, respectively. In the example of FIG. 18A, estimation is performed for a pair of trajectories with IDs of 1 to 25, and the total number N of combinations of trajectory data to be estimated is 300.

In the first method, the number of pairs N _k whose number of calculated common nodes is equal to or greater than k is obtained from distribution information. For example, the number of pairs N ₅ number common node is 5 or more, since the number of the common node number pair 5 is 1, the 1. The number of pairs N ₀ whose number of common nodes is 0 or more is 3 by adding 1 of the number of pairs whose common node number is 5 and 2 of the number of pairs whose common node number is 0. In addition, the maximum value K of the calculated number of common nodes is 5 because the number of common nodes is 5 at maximum. In the first method, the value of q is calculated from Expression (6), and the estimated values N to _k in the case of the number k of nodes of the common portion are calculated from Expression (5) using the calculated value of q. In the example of FIG. 18A, the estimated value N- ₁ is 22.2, and the estimated value N- ₀ is 300.0. Determination section 46, the estimated value N ~ ₁ number common nodes 1 for closest to the 40 target number, to determine a reference to abort the computation 1.

On the other hand, in the second method, the maximum value K of the number of common nodes calculated from the distribution information is 5. The total number N of combinations of the trajectory data to be estimated is 300. The target number N ₀ is 40. In the second method, the value of the reference k ₀ is calculated from Expression (8), and the value obtained by rounding off the value of the reference k ₀ is determined as the reference for terminating the calculation. In the example of FIG. 18A, the reference is determined to be 2. In the second method, as shown in FIG. 18A, at the stage when the maximum value K of the calculated number of common nodes is determined, since each parameter of Expression (8) does not change thereafter, the reference for stopping the calculation is determined.

  The output unit 47 stores, in the pair common portion information 34, each ID of a pair of locus data in which the number of common nodes is equal to or greater than the reference number of nodes, and the number of common nodes. In the example of FIG. 18A, the pair common portion information 34 stores 5 as the number of common nodes of the trajectory in association with ID 6 and ID 7.

  In the example of FIG. 18A, the value “6” of the number of processing nodes x is not a reference value in both the first method and the second method. In this case, the acquisition unit 43 subtracts 1 from the value of the number x of processing nodes to obtain “5”, and acquires trajectory data of the number of nodes having five or more nodes. For example, since the track data having the number of nodes of 6 or more has already been obtained, the obtaining unit 43 reads the track data having the number of nodes of 5.

  Here, the normalized trajectory data 33 shown in FIG. 17 does not include a trajectory having five nodes. If there is no newly read trajectory, the distribution information does not change. Therefore, the reference of the first method remains in the state of FIG. 18A. The acquisition unit 43 has determined that the value “5” of the processing node number x is not the reference value. In this case, the acquisition unit 43 subtracts 1 from the value of the number x of processing nodes to obtain “4”, and acquires trajectory data of the number of nodes having four or more nodes. For example, since the locus data having five or more nodes has already been obtained, the acquiring unit 43 reads the locus data having four nodes.

  FIG. 18B is a diagram illustrating an example of the processing status. In the example of FIG. 18B, trajectories with IDs of 8, 12, and 18 having 4 nodes are added to the candidate list.

  The calculation unit 44 calculates the number of common nodes for each pair of the trajectory data except for the pair for which the number of common nodes has been calculated. For example, the calculation unit 44 calculates the number of common nodes for each pair obtained by combining locus data having 6 nodes and 4 locus data and each pair obtained by combining only locus data having 4 nodes. I do. FIG. 18B shows a calculation result of the common node.

  The generation unit 45 generates distribution information of the number of pairs for each number of common nodes based on the number of common nodes calculated by the calculation unit 44. FIG. 18B shows distribution information.

  The deciding unit 46 estimates the number of pairs of trajectory data that is equal to or more than the number of common nodes for each number of common nodes, with the plurality of trajectory data stored in the normalized trajectory data 33 as the estimation target from the distribution information.

In the first method, the number of pairs N _k whose number of calculated common nodes is equal to or more than _k is obtained from the distribution information, the value of q is calculated again from equation (6), and the number of nodes of the common part is calculated from equation (5). The respective estimated values N to _k for _k are calculated. In the example of FIG. 18B, the estimated value N ~ ₂ is 6.2, the estimated value N ~ ₁ is 43.3, and the estimated value N ~ ₀ is 300.0. Determination section 46, the estimated value N ~ ₁ number common nodes 1 for closest to the 40 target number, to determine a reference to abort the computation 1.

  On the other hand, in the second method, the reference for terminating the calculation is determined to be 2.

  The output unit 47 stores, in the pair common portion information 34, each ID of a pair of locus data in which the number of common nodes is equal to or greater than the reference number of nodes, and the number of common nodes. In the example of FIG. 18B, for example, the number 3 of the common nodes of the trajectory is added to the pair common portion information 34 in association with ID 6 and ID 8.

  In the example of FIG. 18B, the value “4” of the number of processing nodes x is not a reference value in both the first method and the second method. In this case, the acquisition unit 43 subtracts 1 from the value of the number of processing nodes x to “3”, and acquires trajectory data of the number of nodes having three or more nodes. For example, since the track data having four or more nodes has already been obtained, the obtaining unit 43 reads the track data having three nodes.

  FIG. 18C is a diagram illustrating an example of a processing state. In the example of FIG. 18C, trajectories with IDs of 2, 3, 10, 23, and 25 having 3 nodes are added to the candidate list.

  The calculation unit 44 calculates the number of common nodes for each pair of trajectory data, except for the pair for which the number of common nodes has been calculated, as described above. FIG. 18C shows a calculation result of the common node.

  The generation unit 45 generates distribution information of the number of pairs for each number of common nodes based on the number of common nodes calculated by the calculation unit 44. FIG. 18C shows distribution information.

  The deciding unit 46 estimates the number of pairs of trajectory data that is equal to or more than the number of common nodes for each number of common nodes, with the plurality of trajectory data stored in the normalized trajectory data 33 as the estimation target from the distribution information.

In the first method, the number of pairs N _k whose number of calculated common nodes is equal to or more than _k is obtained from the distribution information, the value of q is calculated again from equation (6), and the number of nodes of the common part is calculated from equation (5). The respective estimated values N to _k for _k are calculated. In the example of FIG. 18C, the estimated value N ~ ₂ is 13.3, the estimated value N ~ ₁ is 63.2, and the estimated value N ~ ₀ is 300.0. Determination section 46, the estimated value N ~ ₁ number common nodes 1 for closest to the 40 target number, to determine a reference to abort the computation 1.

  On the other hand, in the second method, the reference for terminating the calculation is determined to be 2.

  The output unit 47 stores, in the pair common portion information 34, each ID of a pair of locus data in which the number of common nodes is equal to or greater than the reference number of nodes, and the number of common nodes. In the example of FIG. 18C, for example, the number 3 of the common nodes of the trajectory is added to the pair common portion information 34 in association with ID 6 and ID 3.

  In the example of FIG. 18C, the value “3” of the number of processing nodes x is not a reference value in both the first method and the second method. In this case, the acquisition unit 43 subtracts 1 from the value of the processing node number x to obtain “2”, and acquires trajectory data of the node number of 2 or more. For example, since the track data having three or more nodes has already been obtained, the obtaining unit 43 reads the track data having two nodes.

  FIG. 18D is a diagram illustrating an example of the processing status. In the example of FIG. 18D, trajectories with IDs of 4, 13, 15, and 19 having 2 nodes are added to the candidate list.

  The calculation unit 44 calculates the number of common nodes for each pair of trajectory data, except for the pair for which the number of common nodes has been calculated, as described above. FIG. 18D shows a calculation result of the common node.

  The generation unit 45 generates distribution information of the number of pairs for each number of common nodes based on the number of common nodes calculated by the calculation unit 44. FIG. 18D shows distribution information.

  The deciding unit 46 estimates the number of pairs of trajectory data that is equal to or more than the number of common nodes for each number of common nodes, with the plurality of trajectory data stored in the normalized trajectory data 33 as the estimation target from the distribution information.

In the first method, the number of pairs N _k whose number of calculated common nodes is equal to or more than _k is obtained from the distribution information, the value of q is calculated again from equation (6), and the number of nodes of the common part is calculated from equation (5). The respective estimated values N to _k for _k are calculated. In the example of FIG. 18D, the estimated value N ~ ₂ is 17.6, the estimated value N ~ ₁ is 72.6, and the estimated value N ~ ₀ is 300.0. Determination section 46, the estimated value N ~ ₂ common node number 2 for closest to the 40 target number, to determine a reference to abort the computation 2.

  On the other hand, in the second method, the reference for terminating the calculation is determined to be 2.

  The output unit 47 stores, in the pair common portion information 34, each ID of a pair of locus data in which the number of common nodes is equal to or greater than the reference number of nodes, and the number of common nodes. In the example of FIG. 18D, for example, the number 2 of the common nodes of the trajectory is added to the pair common portion information 34 in association with ID 6 and ID 4.

  In the example of FIG. 18D, both the first method and the second method are based on the number of processing nodes x of 2. As a result, reading is not performed for the trajectory having the number of nodes of 1, and the calculation of the common portion is terminated.

  The output unit 47 uses the pair common part information 34 shown in FIG. 18D to analyze the pair of the locus data of the number of nodes between the specific number of nodes and the number of common nodes used as the reference for terminating the calculation as the analysis result. Output to information 35. FIG. 18E is a diagram illustrating an example of the processing status. FIG. 18E shows a pair of trajectory data in which the number of common nodes is two or more of the number of reference nodes. The output unit 47 may output the analysis result information 35 in association with a trajectory having the number of common nodes equal to or more than the reference for each trajectory, as shown on the right side of FIG. 18E.

  As described above, the trajectory data processing apparatus 11 according to the present embodiment sets the number of pairs of trajectory data that is equal to or more than the number of common nodes for each common node number, with all trajectory data as targets of estimation from the calculated distribution information. Is estimated. Then, the trajectory data processing device 11 dynamically changes the reference for terminating the calculation to the number of common nodes whose estimated number of pairs is closest to the specified target number. Thereby, the trajectory data processing device 11 can appropriately narrow down the pairs of the trajectory data to the number of pairs close to the target number. Further, the trajectory data processing device 11 acquires the trajectory data of the number of nodes equal to or greater than the number of nodes by sequentially decreasing the number of nodes from the specific number of nodes. Then, the trajectory data processing device 11 sequentially calculates the common part of the acquired trajectory data. At this time, if the trajectory data processing device 11 calculates the common part for the trajectory data with a small number of nodes, if the number of pairs is estimated to depart from the target value, the calculation can be terminated. Can be reduced appropriately.

  Next, the flow of various processes by the trajectory data processing device 11 according to the present embodiment will be described. First, the flow of a normalization process in which the trajectory data processing device 11 according to the present embodiment normalizes a trajectory will be described. FIG. 19 is a flowchart illustrating an example of the flow of the normalization processing. This normalization process is executed at a predetermined timing, for example, at a timing when the locus data is uploaded from the locus recording device 12. The storage of the locus data uploaded from the locus recording device 12 and the normalization of the locus data may be processed at different timings. For example, the normalization of the trajectory data may be executed at a timing when a predetermined operation for instructing the start of the normalization is performed from the input unit 21.

  As shown in FIG. 19, when the trajectory data is uploaded from the trajectory recording device 12, the storage unit 40 assigns a trajectory ID to the uploaded trajectory data and stores it in the movement trajectory data 30 (S10).

  The normalization unit 41 performs map matching of each position information of the trajectory stored in the movement trajectory data 30 with the position information of each node of the node information 31, and normalizes the position of the trajectory to data indicated by the node. (S11).

  The normalizing unit 41 deletes the normalized trajectory data from the movement trajectory data 30 and ends the process (S12).

  Next, a flow of the trajectory data processing in which the trajectory data processing device 11 according to the present embodiment extracts a pair of trajectory data having high similarity from a set of a plurality of trajectory data will be described. FIG. 20 is a flowchart illustrating an example of the flow of the trajectory data processing. The trajectory data processing is executed at a predetermined timing, for example, when a predetermined operation for instructing the start of the trajectory data processing is performed from the input unit 21.

  As illustrated in FIG. 20, the receiving unit 42 displays the operation screen, and receives designation of a plurality of pairs of trajectory data as the target number of trajectory data pairs to be extracted in the route analysis (S20).

  The acquisition unit 43 sorts the trajectory data stored in the normalized trajectory data 33 of the storage unit 23 in descending order of the number of nodes (S21). The acquiring unit 43 refers to the normalized trajectory data 33, obtains the maximum number of nodes of the trajectory data, and sets the obtained maximum number of nodes to the processing node number x (S22). Further, the acquisition unit 43 initializes the candidate list to empty (S23).

  The acquisition unit 43 refers to the normalized trajectory data 33, reads the trajectory data whose number of nodes is the processing node number x, and adds it to the candidate list (S24). The calculation unit 44 calculates the number of common nodes for each pair of trajectory data stored in the candidate list (S25). The generation unit 45 generates distribution information of the number of pairs for each number of common nodes based on the number of common nodes calculated by the calculation unit 44 (S26).

  The determination unit 46 estimates the number of pairs of trajectory data that is equal to or greater than the number of common nodes for each common node number, with the plurality of trajectory data stored in the normalized trajectory data 33 as the estimation target from the distribution information (S27). ). The determining unit 46 determines the number of common nodes whose estimated number of pairs is closest to the specified target number as the reference number of common nodes to terminate the calculation (S28).

  The output unit 47 stores, in the pair common part information 34, each ID of the pair of trajectory data whose number of common nodes is equal to or larger than the reference number of nodes at which the calculation is terminated, and the number of common nodes (S29).

  The acquisition unit 43 determines whether or not the number of processing nodes x has become the reference number of common nodes (S30). If the processing node number x is not the reference common node number (No at S30), the acquiring unit 43 subtracts 1 from the value of the processing node number x (S31), and proceeds to S24.

  When the processing node number x has become the reference common node number (Yes at S30), the output unit 47 performs the following processing. That is, the output unit 47 outputs, from the pair common part information 34, pairs of trajectory data having a number of nodes equal to or greater than the number of common nodes used as a reference for terminating the calculation to the analysis result information 35 as a result of the path analysis (S32). The process ends.

[effect]
The trajectory data processing device 11 receives designation of the number of pairs of a plurality of trajectory data. The trajectory data processing device 11 refers to the normalized trajectory data 33 in the storage unit 23 and acquires the trajectory data of the number of nodes equal to or more than the specific number of nodes. The trajectory data processing device 11 calculates, for each pair of the acquired trajectory data, the number of common nodes whose identification information is common between the trajectory data included in each pair. The trajectory data processing device 11 generates distribution information of the number of pairs for each number of common nodes based on the calculated number of common nodes. The trajectory data processing device 11 determines the number of other nodes less than the specific number of nodes based on the generated distribution information and the number of accepted pairs. The trajectory data processing device 11 outputs a pair of trajectory data of the number of nodes between a specific number of nodes and another number of nodes. Thereby, the trajectory data processing device 11 can appropriately reduce the amount of calculation. In addition, when calculating a locus having a high degree of similarity from a large number of locus data, the locus data processing device 11 can appropriately narrow down based on the distribution, so that it is possible to suppress an increase in processing time.

  Further, the trajectory data processing device 11 uses the plurality of trajectory data stored in the normalized trajectory data 33 of the storage unit 23 from the distribution information as an estimation target, and for each common node number, the trajectory data that is equal to or more than the common node number. Estimate the number of pairs. The trajectory data processing device 11 determines the number of common nodes whose estimated number of pairs is closest to the number of accepted pairs as the number of other nodes. Accordingly, the trajectory data processing device 11 appropriately changes the reference for terminating the calculation of the distribution of the number of pairs of the common nodes in the plurality of trajectory data, regardless of the shape of the exponential distribution. it can.

  Further, the trajectory data processing device 11 calculates the number of common nodes using the geometric distribution based on the number of pairs for each number of common nodes of the distribution information and the total number of pairs obtained by combining a plurality of trajectory data stored in the normalized trajectory data 33. In each case, the number of pairs of trajectory data that is equal to or greater than the number of common nodes is estimated. The trajectory data processing device 11 determines the number of common nodes whose estimated number of pairs is closest to the number of accepted pairs as the number of other nodes. The trajectory data processing device 11 repeats the acquisition of the trajectory data of the number of nodes equal to or more than the specified number of nodes, and sequentially decreasing the number of nodes until reaching the determined other number of nodes. Thus, the trajectory data processing device 11 can accurately estimate the number of pairs of trajectory data that is equal to or greater than the number of common nodes.

  In addition, the trajectory data processing device 11 estimates the number of pairs of trajectory data that is equal to or greater than the number of common nodes for each number of common nodes using the geometric distribution from the maximum number of common nodes. The trajectory data processing device 11 determines the number of common nodes whose estimated number of pairs is closest to the number of accepted pairs as the number of other nodes. The trajectory data processing device 11 acquires trajectory data of the number of nodes equal to or larger than the determined number of other nodes. Thus, the trajectory data processing device 11 can quickly obtain the number of other nodes within the read range.

  [C] Third Embodiment Although the embodiments related to the disclosed apparatus have been described above, the disclosed technology may be implemented in various different forms other than the above-described embodiments. Therefore, another embodiment included in the present invention will be described below.

  For example, in the above embodiment, the case where the number of pairs of trajectory data is estimated by the first method and the second method using the geometric distribution is described as an example, but the disclosed device is not limited to this. For example, the number of pairs of locus data may be estimated using an exponential distribution. A third method using the exponential distribution will be described. The exponential distribution can be generally expressed as in the following equation (10).

The expression (10) of the exponential distribution can be approximated from the following expression (11) as in the following expressions (12) and (13).

  By averaging the equations (12) and (13) with k = 1,..., K, they can be expressed as the following equations (14) and (15).

here,
K: the maximum value of the calculated number of common nodes N: the total number of combinations of the trajectory data to be estimated N _k : the number of pairs where the calculated number of common nodes is k or more N ~ _k : the number of pairs where the number of common nodes is k or more Number estimates

The calculated maximum value K of the number of common nodes can be determined by obtaining the distribution of the number of common nodes in order from the locus data having the largest number of nodes. The total number N of combinations of trajectory data to be estimated determines the number of trajectory data. In the third method, as in the first method, when the distribution of the number of common nodes is obtained in order from the locus data having the largest number of nodes, the calculated number of pairs of the common nodes is equal to or more than _k using the number of pairs _Nk. The value of λ is obtained from (15). Then, also in the third method, the estimated values N to _k in the case of the number k of nodes of the common part are obtained from the equation (14) using the obtained value of λ.

  The third method estimates the lower limit of the distribution, as in the first method. In the third method, as in the first method, the accuracy is improved by using the values of the determined portion and the undetermined portion in the calculated distribution.

  The main difference between the geometric distribution and the exponential distribution is that the exponential distribution is a continuous type distribution in contrast to the discrete type distribution. Since the length of the common portion in this embodiment has a discrete type distribution, it is desirable to approximate the length with a geometric distribution. However, a similar calculation result can be obtained by using an approximation method such as the above equation (11).

  Further, in the above embodiment, the case where the maximum number of nodes of the trajectory is used as the specific number of nodes has been described as an example, but the disclosed device is not limited to this. For example, when the target number of pairs of trajectory data is designated as 40, for example, even if six trajectory data are combined, the number of pairs is 15, so that a sufficient number of pairs cannot be obtained. Therefore, the number of nodes in which the number of combinations obtained by combining a plurality of trajectory data in the descending order of the number of nodes may be greater than or equal to the received target value may be initially set as the specific number of nodes. For example, the combination of six locus data is 45. Therefore, when the target number of trajectory data pairs is specified as 40, the specific number of nodes may be set to 10.

In the above embodiment, even in the second method, even if the number of common nodes closest to the target number N ₀ is determined, the number of nodes is sequentially reduced and the acquisition of the locus data of the number of nodes equal to or more than the number of nodes is repeated. Has been described as an example, but the disclosed device is not limited to this. For example, in the second method, when the number of common nodes closest to the target number N ₀ is determined, the acquiring unit 43 may collectively acquire the locus data of the number of nodes equal to or more than the number of common nodes closest to the target number N _0. Good.

  Each component of each device illustrated is a functional concept, and does not necessarily need to be physically configured as illustrated. In other words, the specific state of the distribution / integration of each device is not limited to the one shown in the figure, and all or a part thereof may be functionally or physically distributed / arbitrarily divided into arbitrary units according to various loads and usage conditions. Can be integrated and configured. For example, the processing units of the storage unit 40, the normalization unit 41, the reception unit 42, the acquisition unit 43, the calculation unit 44, the generation unit 45, the determination unit 46, and the output unit 47 may be appropriately integrated. Further, all or any part of each processing function performed by each processing unit may be realized by a CPU and a program analyzed and executed by the CPU, or may be realized as hardware by wired logic. .

[Track data processing program]
The various processes described in the above embodiments can also be realized by executing a prepared program on a computer system such as a personal computer or a workstation. Therefore, an example of a computer system that executes a program having the same function as the above embodiment will be described below. FIG. 21 is a diagram illustrating an example of a computer that executes a trajectory data processing program.

  As shown in FIG. 21, the computer 300 includes a CPU (Central Processing Unit) 310, a HDD (Hard Disk Drive) 320, and a RAM (Random Access Memory) 340. These units 300 to 340 are connected via a bus 400.

  In the HDD 320, a trajectory data processing program 320A that performs the same functions as the above-described storage unit 40, normalization unit 41, reception unit 42, acquisition unit 43, calculation unit 44, generation unit 45, determination unit 46, and output unit 47 is provided. It is stored in advance. The trajectory data processing program 320A may be appropriately separated.

  The HDD 320 stores various information. For example, the HDD 320 stores an OS and various data used for various processes.

  Then, the CPU 310 reads out the locus data processing program 320A from the HDD 320 and executes the locus data processing program 320A, thereby performing the same operation as each processing unit of the embodiment. That is, the trajectory data processing program 320A performs the same operations as the storage unit 40, the normalization unit 41, the reception unit 42, the acquisition unit 43, the calculation unit 44, the generation unit 45, the determination unit 46, and the output unit 47.

  Note that the above-described trajectory data processing program 320A does not necessarily need to be stored in the HDD 320 from the beginning.

  Further, for example, the trajectory data processing program 320A may store the program in a “portable physical medium” such as a CD-ROM, a DVD disk, a magneto-optical disk, or an IC card inserted into the computer 300. Then, the computer 300 may read out the program from these and execute it.

  Further, the program is stored in “another computer (or server)” connected to the computer 300 via a public line, the Internet, a LAN, a WAN, or the like. Then, the computer 300 may read out the program from these and execute it.

Reference Signs List 10 system 11 trajectory data processing device 12 trajectory recording device 21 input unit 22 display unit 23 storage unit 24 control unit 30 movement trajectory data 31 node information 32 link information 33 normalized trajectory data 35 analysis result information 40 storage unit 41 normalization unit 42 Reception unit 43 Acquisition unit 44 Calculation unit 45 Generation unit 46 Determination unit 47 Output unit

Claims (6)

Accepts the number of pairs of multiple trajectory data,
A specific node is referred to by referring to a storage unit that stores a plurality of trajectory data including a plurality of pieces of position information respectively associated with any one of a plurality of nodes associated with mutually different identification information. Get track data with more than the number of nodes,
For each pair of the acquired trajectory data, calculate the number of common nodes in which the identification information is common between the trajectory data included in each pair,
Based on the calculated number of common nodes, generating distribution information of the number of pairs for each number of common nodes,
From the generated distribution information , the plurality of trajectory data stored in the storage unit is an estimation target, and for each common node, the number of pairs of trajectory data that is equal to or greater than the number of common nodes is estimated, and the estimated pair The number determines the number of common nodes closest to the received number of pairs as the number of other nodes ,
A trajectory data processing method, wherein a computer executes a process of outputting a pair of trajectory data whose number of common nodes is equal to or greater than the determined other node . The determining is performed by using a geometric distribution or an exponential distribution from the number of pairs for each common node number of the distribution information and the total number of pairs obtained by combining the plurality of trajectory data stored in the storage unit. For each, the number of pairs of trajectory data that is equal to or greater than the number of common nodes is estimated, and the estimated number of pairs determines the number of common nodes closest to the received number of pairs as the number of other nodes,
The acquisition process repeats acquisition of trajectory data of the number of nodes equal to or greater than the number of nodes by sequentially decreasing the number of nodes from the specific number of nodes until reaching the determined other number of nodes. 2. The trajectory data processing method according to 1. The process of determining, from the maximum number of common nodes of the distribution information, for each common node number using a geometric distribution, estimate the number of pairs of trajectory data is equal to or more than the number of common nodes, the estimated number of pairs, Determine the number of common nodes closest to the received number of pairs as the number of other nodes,
The trajectory data processing method according to claim 1, wherein the acquiring process acquires trajectory data of a number of nodes equal to or greater than the determined number of other nodes. The maximum number of nodes of the plurality of trajectory data stored in the storage unit, or the number of combinations obtained by combining the plurality of trajectory data in descending order of the number of nodes, to the number of nodes that is equal to or greater than the received number of pairs, 3. The trajectory data processing method according to claim 1, wherein a computer further executes a process of setting a specific number of nodes. Accepts the number of pairs of multiple trajectory data,
A specific node is referred to by referring to a storage unit that stores a plurality of trajectory data including a plurality of pieces of position information respectively associated with any one of a plurality of nodes associated with mutually different identification information. Get track data with more than the number of nodes,
For each pair of the acquired trajectory data, calculate the number of common nodes in which the identification information is common between the trajectory data included in each pair,
Based on the calculated number of common nodes, generating distribution information of the number of pairs for each number of common nodes,
From the generated distribution information , the plurality of trajectory data stored in the storage unit is an estimation target, and for each common node, the number of pairs of trajectory data that is equal to or greater than the number of common nodes is estimated, and the estimated pair The number determines the number of common nodes closest to the received number of pairs as the number of other nodes ,
A trajectory data processing program for causing a computer to execute a process of outputting a pair of trajectory data whose number of common nodes is equal to or greater than the determined other node . A receiving unit that receives designation of the number of pairs of a plurality of track data;
A specific node is referred to by referring to a storage unit that stores a plurality of trajectory data including a plurality of pieces of position information respectively associated with any one of a plurality of nodes associated with mutually different identification information. An acquisition unit that acquires trajectory data of the number of nodes equal to or more than the number of nodes,
For each pair of the trajectory data acquired by the acquisition unit, a calculation unit that calculates the number of common nodes in which identification information is common between the trajectory data included in each pair,
Based on the number of common nodes calculated by the calculating unit, a generating unit that generates distribution information of the number of pairs for each number of common nodes,
From the distribution information generated by the generation unit, as the target of the plurality of trajectory data stored in the storage unit, for each common node number, estimate the number of pairs of trajectory data that is equal to or more than the number of common nodes, Estimated number of pairs, a determination unit that determines the number of common nodes closest to the received number of pairs as the number of other nodes ,
An output unit that outputs a pair of trajectory data that is equal to or greater than the number of common nodes, the number of other nodes determined by the determination unit,
A trajectory data processing device comprising:

Images