Location estimation method, location estimation device, and program

Abstract

This invention provides a position estimation method that can improve the accuracy of position estimation. [Solution] The position estimation method includes an acquisition step of acquiring measured internode distances, a first estimated position calculation step of calculating the estimated position of each node in the first node group (1A) based on the internode distances, a first evaluation value calculation step of calculating an evaluation value, and a first correction step of correcting the internode distance between the anchor node (11) and the first actual node (12A) to optimize the evaluation value.

Description

[Technical Field]

[0001] The present invention relates to a position estimation method and a position estimation device. [Background technology]

[0002] Conventionally, various methods have been proposed for estimating the location of wireless nodes in a wireless sensor network composed of multiple wireless nodes. Patent documents 1 to 3 disclose a method for estimating the location of wireless nodes using wireless transmission parameters in indoor environments where communication with satellites is difficult. In particular, Patent document 1 discloses a wireless node location estimation method (SOL: Self-Organizing Localization) using a self-organizing map (SOM) as a method for estimating the location of wireless nodes. [Prior art documents] [Patent Documents]

[0003] [Patent Document 1] Patent No. 6532606 specification [Patent Document 2] Special Publication No. 2014-509381 [Patent Document 3] Special Publication No. 2018-512571 [Overview of the project] [Problems that the invention aims to solve]

[0004] To estimate the position of a radio node based on SOL, it is necessary to measure the distance between nodes. For example, the distance between nodes is measured from the round-trip time of radio waves in ultra-wideband impulse radio (IR-UWB). However, in distance measurement using IR-UWB, if there are obstacles blocking the radio waves in the positioning environment, distance measurement using direct waves cannot be performed, and distance measurement is performed using the round-trip time of indirect waves. In that case, the distance between nodes input to SOL will be a larger value than the true distance, which can cause a deterioration in the accuracy of position estimation.

[0005] One aspect of the present invention has been made in view of the above-mentioned problems, and its object is to provide a position estimation method that can improve position estimation accuracy. [Means for solving the problem]

[0006] To solve the above problems, a position estimation method according to embodiment 1 of the present invention includes, in a first node group, an acquisition step of acquiring the distance between nodes measured by wireless communication between each node in a plurality of nodes, wherein the first node group includes an anchor node whose true position is known and a first actual node whose true position is unknown but which can communicate with the anchor node; a first estimated position calculation step of calculating the estimated position of each node in the first node group by estimating the geometry of the first node group based on the distance between the anchor node and the first actual node; a first evaluation value calculation step of calculating an evaluation value indicating the deviation of the estimated position of a first reference node whose true position is known in the first node group from its true position; and a first correction step of correcting the distance between the anchor node and the first actual node so as to optimize the evaluation value.

[0007] A position estimation method according to aspect 2 of the present invention further includes an additional step in aspect 1 of adding a virtual node in a virtual space whose true position is known to the first node group, wherein in the first estimated position calculation step, the geometry of the first node group is estimated based on the internode distance between the anchor node and the first real node, as well as the internode distance between the anchor node and the virtual node, and in the first evaluation value calculation step, the virtual node may be used as the first reference node.

[0008] In the position estimation method according to aspect 3 of the present invention, in aspect 1, the anchor node may be the first reference node in the first evaluation value calculation step.

[0009] In the position estimation method according to embodiment 4 of the present invention, the first estimated position calculation step, the first evaluation value calculation step, and the first correction step are repeatedly performed in embodiments 1 to 3, and in the second and subsequent first estimated position calculation steps, the geometry of the first node group may be estimated based on the internode distance between the anchor node and the first actual node corrected in the first correction step.

[0010] In the position estimation method according to aspect 5 of the present invention, in aspect 4 above, the second node group includes the first actual node and a second actual node whose true position is unknown and which is not able to communicate with the anchor node but can communicate with the first actual node, and after repeatedly performing the first estimated position calculation step, the first evaluation value calculation step, and the first correction step, the estimated position of each node in the second node group is calculated by estimating the geometry of the second node group based on the internode distance between the first actual node and the second actual node, the second estimated position calculation step is performed, the second evaluation value calculation step is performed using any node in the first node group as a second reference node and calculating an evaluation value that shows the deviation of the estimated position of the second reference node from its true position, and the second correction step is performed repeatedly to correct the internode distance between the first actual node and the second actual node to optimize the evaluation value.

[0011] In the position estimation method according to embodiment 6 of the present invention, in embodiment 5, the estimated position of the first actual node in the first estimated position calculation step is considered to be the true position after repeatedly performing the first estimated position calculation step, the first evaluation value calculation step, and the first correction step, and in the second evaluation value calculation step, the first actual node is considered to be the second reference node.

[0012] In the position estimation method according to embodiment 7 of the present invention, in embodiments 1 to 6 above, the evaluation value may be a value based on the difference between the estimated position of the first reference node and the true position of the first reference node.

[0013] In the position estimation method according to Aspect 8 of the present invention, in the above Aspect 2, the evaluation value may be a value based on the difference between the node-to-node distance between the anchor node and the virtual node before being estimated in the first estimated position calculation step and the node-to-node distance between the anchor node and the virtual node in the geometry estimated in the first estimated position calculation step.

[0014] In the position estimation method according to Aspect 9 of the present invention, in the above Aspects 1 to 8, in the first correction step, the node-to-node distance between the anchor node and the first real node may be corrected so as to decrease the evaluation value.

[0015] In the position estimation method according to Aspect 10 of the present invention, in the above Aspects 1 to 9, in the first correction step, the node-to-node distance between the anchor node and the first real node may be corrected by the gradient method.

[0016] In order to solve the above problems, a position estimation apparatus according to Aspect 11 of the present invention includes a first node group including an anchor node whose true position is known and a first real node whose true position is unknown and capable of communicating with the anchor node, an acquisition unit that acquires the node-to-node distance measured by wireless communication between each pair of nodes in a plurality of nodes, an estimated position calculation unit that calculates the estimated position of each node in the first node group by estimating the geometry of the first node group based on the node-to-node distance between the anchor node and the first real node, an evaluation value calculation unit that calculates an evaluation value indicating the deviation of the estimated position of a first reference node whose true position is known in the first node group from the true position, and a correction unit that corrects the node-to-node distance between the anchor node and the first real node so as to optimize the evaluation value.

[0017] A program according to Aspect 12 of the present invention is a program for causing a computer to execute the position estimation method according to the above Aspects 1 to 10, and is a program for causing the computer to execute the acquisition step, the first estimated position calculation step, the first evaluation value calculation step, and the first correction step. [Effect of the Invention]

[0018] According to one aspect of the present invention, the position estimation accuracy can be improved. [Brief Description of the Drawings]

[0019] [Figure 1] It is a schematic diagram for explaining the influence of an obstacle on distance measurement using IR-UWB. [Figure 2] It is a block diagram showing an example of the configuration of a position estimation device according to Embodiment 1. [Figure 3] It is a schematic diagram showing a plurality of nodes in the real space and a virtual mesh network in the virtual space. [Figure 4] It is a flowchart showing the processing procedure of the SOL algorithm. [Figure 5] It is a schematic diagram showing the topology of the first node group in the real space and the topology of the first node group with virtual nodes whose true positions are known added in the virtual space. [Figure 6] It is a schematic diagram showing the state of correcting the first real node link length so that the estimated position of the virtual node approaches the true position. [Figure 7] It is a flowchart showing the processing flow of the position estimation method executed by the position estimation device according to Embodiment 1. [Figure 8] It is a schematic diagram showing the first node group and the second node group. [Figure 9] It is a flowchart showing the processing flow of the position estimation method executed by the position estimation device according to Embodiment 2. [Modes for Carrying Out the Invention]

[0020] [Embodiment 1] (Regarding the feature points according to one aspect of the present invention) One known method for estimating the location of wireless nodes in a wireless sensor network is the aggregated self-organizing smart device location estimation method (SmartFinder). SmartFinder is an autonomous indoor positioning technology that uses adjacent distance information between multiple wireless nodes to acquire the locations of numerous smart devices (wireless nodes) using only three fixed points. In SmartFinder, ultra-wideband impulse radio (IR-UWB), which enables high-precision distance measurement, is typically used for communication between smart devices. This makes it possible to estimate locations to an order of a few centimeters to more than ten centimeters.

[0021] However, in distance measurement using IR-UWB, the presence of obstacles that block radio waves in the positioning environment can significantly degrade the distance measurement accuracy. This can also lead to a deterioration in the position estimation accuracy of SmartFinder.

[0022] Figure 1 is a schematic diagram illustrating the effect of obstacles OB on distance measurement using IR-UWB. In Figure 1, reference numeral 1011 indicates communication between smart devices D1 and D2 when there are no obstacles OB, and reference numeral 1012 indicates communication between smart devices D1 and D2 when obstacles OB are present.

[0023] As shown by the reference numeral 1011 in Figure 1, when there is no obstacle OB, smart devices D1 and D2 measure distance by the round trip of direct waves. Therefore, the distance between smart devices D1 and D2 can be measured accurately by the round trip time of the radio waves. On the other hand, as shown by the reference numeral 1012 in Figure 1, when an obstacle OB is present, the direct waves from the smart devices are blocked by the obstacle OB, so the smart devices measure distance by the round trip of indirect waves. Therefore, the distance between smart devices D1 and D2 measured by the round trip time of the radio waves will be larger than the true distance. In other words, the obstacle causes an error (distance measurement error) in the measurement of the distance between smart devices D1 and D2.

[0024] To solve these problems, the inventors propose a position estimation method using the position estimation device 100 shown below. Specifically, the inventors propose a position estimation method that can suppress the decrease in the position estimation accuracy of SmartFinder caused by distance measurement errors.

[0025] (Outline configuration of the position estimation device) Figure 2 is a block diagram showing an example of the configuration of the position estimation device 100 according to Embodiment 1. The position estimation device 100 estimates the position of a target node among a plurality of nodes 1 (smart devices) present in the wireless communication space (real space). In this embodiment, the position estimation device 100 that estimates the position of the target node using SmartFinder with IR-UWB will be described. Note that the method of position estimation is not limited to SmartFinder, and known position estimation methods using IR-UWB can be applied.

[0026] The target node in this embodiment is the first real node 12A in the first node group 1A. The first node group 1A includes, for example, three anchor nodes 11 (anchor nodes 11-1 to 11-3) whose true positions are known, and at least one first real node 12A whose true position is unknown and which can communicate with any of the three anchor nodes 11.

[0027] Multiple nodes 1 measure the inter-node distance (link length) between each node by communicating wirelessly with each other. Each of the multiple nodes 1 measures the link length with an adjacent node with which wireless communication is possible, based on the round-trip time of radio waves. The multiple nodes 1 transmit the measured link length to the position estimation device 100 via Wi-Fi (registered trademark) or the like.

[0028] The position estimation device 100 comprises an acquisition unit 21, a storage unit 22, and a control unit 23. The acquisition unit 21 acquires the link length measured by wireless communication between each node 1 in a plurality of nodes 1 from each node 1.

[0029] The control unit 23 includes a geometry estimation unit 231 (estimated position calculation unit), an evaluation value calculation unit 232, and a link length correction unit 233 (correction unit).

[0030] The geometry estimation unit 231 estimates the geometry of the first node group 1A using the Self-Organizing Localization (SOL) algorithm, based at least on the link length between the anchor node 11 and the first real node 12A (first real node link length). The first real node link length may be the first real node link length (measured data) acquired by the acquisition unit 21, or it may be the first real node link length (corrected data) corrected by the link length correction unit 233 described later. The geometry estimation unit 231 modifies the geometry of the first node group 1A so that the first real node link length in the geometry approaches the first real node link length input to the SOL algorithm. If there are multiple first real nodes 12A, the geometry estimation unit 231 may estimate the geometry of the first node group 1A based on the link length between the first real nodes 12A in addition to the first real node link length.

[0031] Furthermore, the geometry estimation unit 231 estimates the estimated position (absolute position) of each node in the first node group 1A by referring to the relative position of the anchor node 11 in the estimated geometry and the true position (absolute position) of the anchor node 11. Details of the SOL algorithm will be described later with reference to Figures 3 and 4.

[0032] In this case, if the first actual node link length acquired by the acquisition unit 21 includes a distance measurement error, the estimated position of each node in the first node group 1A estimated by the geometry estimation unit 231 will deviate from the true position.

[0033] The evaluation value calculation unit 232 calculates an evaluation value that indicates the deviation of the estimated position of the first reference node R1, whose true position in the first node group 1A is known, from its true position. The evaluation value indicates the magnitude of the error in the estimated position of the first reference node R1.

[0034] The link length correction unit 233 corrects the first actual node link length in the estimated geometry (the first actual node link length output by the SOL algorithm) in order to optimize the above evaluation value. That is, the link length correction unit 233 corrects the first actual node link length in the estimated geometry so that the estimated position of the first reference node R1 approaches the true position.

[0035] With the above configuration, the position estimation device 100 can suppress errors in the first actual node link length caused by distance measurement errors. Furthermore, by reestimating the geometry based on the corrected first actual node link length, the estimated position errors of each node in the first node group 1A caused by distance measurement errors can also be suppressed.

[0036] (SOL algorithm) Prior to explaining the correction of the first actual node link length described above, we will first explain the geometry estimation method using the SOL algorithm. Figure 3 is a schematic diagram showing the first node group 1A (reference numeral 1031) in real space and the virtual mesh network (reference numeral 1032) in virtual space. As shown in Figure 3, the position estimation device 100 constructs a virtual mesh network in virtual space corresponding to the topology of the first node group 1A in real space (hereinafter simply referred to as the actual topology). In the example shown in Figure 3, the actual topology is shown, which is formed by three anchor nodes 11 and one first actual node 12A that can communicate with the three anchor nodes 11. The same applies to Figures 5 and 6.

[0037] The position estimation device 100 estimates the geometry of real space by applying the SOL algorithm to a virtual mesh network. SOL is a method that uses the geometry of a completely random network as the initial state and reproduces the geometry of real space by repeatedly correcting the position of each node in the first node group 1A based on the link length acquired by the acquisition unit 21.

[0038] As mentioned above, the first actual node link length, which is input to the SOL algorithm, is measured by wireless communication between nodes and may therefore contain ranging errors. Similarly, the link length between multiple first actual nodes 12A, obtained when there are multiple first actual nodes 12A, is also measured by wireless communication between nodes and may therefore contain ranging errors. On the other hand, the link length between anchor nodes 11 (anchor node link length) is obtained from the true position of anchor node 11 and is therefore the true distance without ranging errors. In Figure 3, link lengths that may contain ranging errors are shown as solid lines, and link lengths that are true distances are shown as dotted lines. The same applies to Figures 5 and 6.

[0039] Figure 4 is a flowchart showing the processing procedure of the SOL algorithm. As shown in Figure 4, first, the geometry estimation unit 231 randomly generates the estimated position of each node in the first node group 1A (S1). Subsequently, the estimated position of node i after the tth position correction is w i Let (t) be the case.

[0040] Next, the geometry estimation unit 231 randomly selects n neighboring nodes from the set of neighboring nodes S (the set of nodes that can communicate with node i) for node i (S2). The geometry estimation unit 231 may select one neighboring node from the set of neighboring nodes S, or it may select all neighboring nodes.

[0041] Next, the geometry estimation unit 231 calculates the link length d between node i and each of the n adjacent nodes (adjacent node m) obtained by the acquisition unit 21. im Based on this, the position of node i is w i (t) is corrected (S3). Specifically, the geometry estimation unit 231 corrects the vector V based on the following formula (1). im Derive the following.

number

number

number

[0042] The geometry estimation unit 231 may narrow the range of adjacent nodes according to the number of corrections. This allows for the reproduction of the overall mesh network shape (relative to the geometry in real space) in the initial stages, and then local position estimation as position corrections progress.

[0043] Next, the geometry estimation unit 231 determines whether the number of position corrections t has reached a specified number (S4). If the number of position corrections has not reached a specified number, the process returns to S2; if the number of position corrections has reached a specified number, the process proceeds to S5.

[0044] In S5, the geometry estimation unit 231 converts the relative position of each node in the first node group 1A into an absolute position. Specifically, the geometry estimation unit 231 estimates the estimated position (absolute position) of each node in the first node group 1A by referring to the relative position of the anchor node 11 in the estimated geometry and the true position (absolute position) of the anchor node 11.

[0045] Preferably, the geometry estimation unit 231 derives estimated positions for all nodes in the first node group 1A after each position correction. However, the geometry estimation unit 231 may correct the estimated positions for a specific group of nodes in the first node group 1A until the number of position corrections t reaches a predetermined number, and then correct the estimated positions for another group of nodes in the first node group 1A until the number of position corrections t reaches a predetermined number.

[0046] According to the processing procedure shown in Figure 4, the position estimation device 100 calculates the estimated position of each node in the first node group 1A. However, if the first actual node link length input in S3 includes a distance measurement error, an error will occur in the estimated position of each node in the first node group 1A estimated by the geometry estimation unit 231. Therefore, the position estimation device 100 suppresses the estimated position error of each node in the first node group 1A by the following method.

[0047] (Adding a virtual node) Figure 5 is a schematic diagram showing the topology of the first node group 1A in real space (indicated by 1051) and the topology of the first node group 1A in virtual space with the addition of a virtual node 13 whose true position is known (indicated by 1052). As shown in Figure 5, the position estimation device 100 expands the topology by adding a virtual node 13 to the virtual mesh network. In addition to the real topology, the virtual mesh network forms a topology (virtual topology) between the anchor node 11 and the virtual node 13. The position estimation device 100 applies the SOL algorithm to this expanded virtual mesh network. That is, the position estimation device 100 estimates the geometry of the first node group 1A including the virtual node 13.

[0048] The first node group 1A in the virtual space includes a first real node 12A whose true position is unknown, an anchor node 11 whose true position is known, and a virtual node 13 whose true position is known. The link length acquired by the acquisition unit 21 includes the link length of the first real node, the link length of the anchor node, and the link length between the anchor node 11 and the virtual node 13 (virtual node link length). Since the virtual node link length is obtained from the true position of the anchor node 11 and the true position of the virtual node 13, it is a true distance that does not include distance measurement errors, similar to the anchor node link length.

[0049] (Correction of the first actual node link length) Figure 6 is a schematic diagram showing how the link length of the first real node is corrected so that the estimated position of the virtual node 13 approaches its true position. Reference numerals 1061 to 1063 in Figure 6 indicate how the link lengths between anchor nodes 11-1 to 11-3 and the first real node 12A are corrected, respectively. In Figure 6, the estimated position of the virtual node 13 is shown by a solid circle, and the true position of the virtual node 13 is shown by a dotted circle.

[0050] If the first actual node link length input to the SOL algorithm includes a distance measurement error, even if the anchor node link length and virtual node link length are true distances, an estimated position error will occur not only in the first actual node 12A but also in the virtual node 13 whose true position is known (indicated by 1061 in Figure 6).

[0051] Therefore, by correcting the first real node link length input to the SOL algorithm so that the estimated position error of the virtual node 13 output by the SOL algorithm approaches zero, the estimated position error of each node in the first node group 1A can be brought closer to zero. In other words, by using the virtual node 13 as the first reference node R1 whose true position is known, and correcting the first real node link length so that the estimated position of the virtual node 13 approaches its true position, the error of the first real node link length can be brought closer to zero (indicated by symbols 1062 and 1063 in Figure 6). Furthermore, by re-estimating the geometry based on the corrected first real node link length, the estimated position error of each node in the first node group 1A can be brought closer to zero.

[0052] (Example of operation of a position estimation device) Figure 7 is a flowchart showing the processing flow of the position estimation method executed by the position estimation device 100 according to Embodiment 1. An example of the operation of the position estimation device 100 that performs the correction of the first actual node link length described above will be explained below with reference to Figure 7.

[0053] First, the position estimation device 100 acquires the link lengths measured by wireless communication between the plurality of nodes 1 from each node 1 (acquisition step S11). Here, the link lengths include the first actual node link length, the anchor node link length, and the virtual node link length. Next, the position estimation device 100 includes the virtual node 13 in the first node group 1A to form a virtual mesh network (addition step S12).

[0054] Next, the position estimation device 100 estimates the geometry that forms the actual topology and the virtual topology by applying the SOL algorithm to the virtual mesh network. In other words, the position estimation device 100 estimates the geometry of the first node group 1A including the virtual node 13 based on the first actual node link length and the virtual node link length. The position estimation device 100 calculates the estimated positions of each node in the first node group 1A from the estimated geometry (first estimated position calculation step S13). The detailed processing procedure of the SOL algorithm has been described above with reference to FIG. 4.

[0055] Next, the position estimation device 100 calculates an evaluation value indicating the deviation of the estimated position of the virtual node 第[1]3 (first reference node R1) from the true position (first evaluation value calculation step S14). The evaluation value is a value based on the difference between the estimated position of the virtual node 13 and the true position of the virtual node 13. For example, the position estimation device 100 calculates the evaluation value E i (t´) according to the following formula (4).

Equation

[0056] Note: In the original text, there is a "第[1]3" which seems to be an incorrect expression. I translated it as "第[1]3" as it is. If it is a misprint, please correct it in the original text for a more accurate translation.Next, the position estimation device 100 corrects the first actual node link length in the estimated geometry to optimize the evaluation value (first correction step S15). The position estimation device 100 corrects the first actual node link length in the estimated geometry to reduce the above evaluation value. For example, the position estimation device 100 corrects the first actual node link length d according to the following formula (5) using the gradient method. x Derive (t'+1).

number

[0057] Next, the position estimation device 100 determines whether or not the termination condition is met (S16). For example, the position estimation device 100 determines whether or not the number of corrections t' of the first actual node link length is reached a specified number, or whether or not the evaluation value E i If (t') falls below a predetermined value, it is determined that the termination condition has been met. If the termination condition is not met (NO in S16), the process returns to S13. In the t'+1th S13, the geometry of the first node group 1A is estimated based on the first actual node link length corrected in the t'th S15. If the termination condition is not met (NO in S16), the process may return to S12. In the t'+1th S12, the virtual node 13 (let's call it virtual node 13t') added in the t'th S12 may be deleted, and a virtual node 13t'+1 with a different true position from virtual node 13t' may be added. If the termination condition is met (YES in S16), the position estimation device 100 outputs the estimated position of each node of the first node group 1A estimated in the previous S13 as the measurement result, and the process ends.

[0058] In other words, the position estimation device 100 repeatedly performs steps S13 to S15 until the termination condition is met. Furthermore, in the second and subsequent S13 steps, the position estimation device 100 estimates the geometry of the first node group 1A based on the first actual node link length corrected in the preceding (or earlier) S15 step.

[0059] Note that when t'=1, the d on the right side of equation (5) x (0) and E i The value of (0) is set to any value. Preferably, d x (1)-d x (0) The value of d becomes negative x The value (0) is set. This allows the estimated position error of each node in the first node group 1A to approach zero more quickly.

[0060] (Effects and Benefits) With the above configuration, the position estimation device 100 can suppress errors in the first actual node link length caused by distance measurement errors by correcting the first actual node link length so that the estimated position of the first reference node R1, whose true position is known, approaches its true position. This also leads to an improvement in position estimation accuracy.

[0061] Furthermore, the position estimation device 100 includes virtual nodes 13 in the first node group 1A to form a virtual mesh network. The position estimation device 100 then corrects the first actual node link length so that the estimated position of the virtual node 13 is closer to its true position. As a result, the geometry of the first node group 1A is estimated based on the first actual node link length and the virtual node link length, thereby achieving further improvement in position estimation accuracy.

[0062] The position estimation device 100 and position estimation method according to this embodiment can be applied to understanding the movement of workers in manufacturing sites, construction sites, fire scenes, etc., as well as the movement of robots, UAVs (Unmanned Aerial Vehicles), etc. Such effects can contribute to achieving, for example, Goal 11 of the United Nations Sustainable Development Goals (SDGs), "Make cities and human settlements inclusive, safe, resilient and sustainable."

[0063] (modified version) The position estimation device 100 may use the anchor node 11 as the first reference node R1. If the first actual node link length input to the SOL algorithm includes a distance measurement error, an estimated position error will also occur in the anchor node 11. Therefore, by correcting the first actual node link length input to the SOL algorithm so that the estimated position error of the anchor node 11 output by the SOL algorithm approaches 0, the estimated position error of each node in the first node group 1A can be brought closer to 0.

[0064] Furthermore, while Figures 5 to 7 illustrate the case of adding one virtual node 13 to the virtual mesh network, multiple virtual nodes 13 may be added to the virtual mesh network. In this case, the position estimation device 100 uses the E of each virtual node i as an evaluation value. i We just need to take the sum of (t').

[0065] Furthermore, the evaluation value may be based on the difference between the virtual node link length in the geometry before estimation (input into the SOL algorithm) and the virtual node link length in the estimated geometry. By minimizing such an evaluation value, the first actual node link length in the estimated geometry can be corrected so that the virtual node link length approaches the true distance. In other words, errors in the first actual node link length caused by distance measurement errors can be suppressed.

[0066] [Embodiment 2] Other embodiments of the present invention will be described below. For the sake of clarity, components having the same function as those described in the above embodiments will be denoted by the same reference numerals, and their descriptions will not be repeated.

[0067] Figure 8 is a schematic diagram showing the first node group 1A and the second node group 1B. Embodiment 2 describes a method for estimating the position when the second actual node 12B in the second node group 1B is used as the target node. The second node group 1B includes the first actual node 12A and the second actual node 12B. The second node group 1B may also include an anchor node 11.

[0068] The true position of the second real node 12B is unknown, and it cannot communicate with any of the three anchor nodes 11, but it can communicate with any of the first real nodes 12A. The second real node 12B is located approximately further away from the anchor nodes 11 compared to the first real node 12A. That is, it is not possible to obtain a measurement result for the link length between the second real node 12B and the anchor nodes 11.

[0069] The position estimation device 100 according to Embodiment 2 performs position estimation for the second node group 1B after completing position estimation for the first node group 1A. That is, the position estimation device 100 performs position estimation for the three anchor nodes 11 in a stepwise manner, starting from the nearest actual node.

[0070] Figure 9 is a flowchart showing the processing flow of the position estimation method executed by the position estimation device 100 according to Embodiment 2. An example of the operation of the position estimation device 100, which performs the stepwise position estimation described above, will be explained below with reference to Figure 9.

[0071] First, the position estimation device 100 acquires the link length measured by wireless communication between each node in the plurality of nodes 1 from each node 1, similar to Embodiment 1 (S21). Here, the link length includes the first actual node link length, anchor node link length, and virtual node link length described above, as well as the link length between the first actual node 12A and the second actual node 12B (second actual node link length).

[0072] Next, the position estimation device 100 performs position estimation for the first node group 1A (S22). The detailed processing flow for position estimation for the first node group 1A is as described in S12 to S16 of Figure 7.

[0073] Next, the position estimation device 100 estimates the geometry of the second node group 1B based on the second actual node link length using the SOL algorithm. The position estimation device 100 calculates the estimated position of each node in the second node group 1B from the estimated geometry (second estimated position calculation step S23). Here, the second actual node link length input to the SOL algorithm is the link length acquired by the acquisition unit 21 or the second actual node link length corrected in S26, which will be described later.

[0074] Next, the position estimation device 100 designates one of the nodes in the first node group 1A as the second reference node R2 and calculates an evaluation value indicating the deviation of the estimated position of the second reference node R2 from the true position (second evaluation value calculation step S24). Here, the estimated position of the first actual node 12A in the first estimation step after performing S22 may be considered as the true position, and the first actual node 12A may be designated as the second reference node R2. The method for calculating the evaluation value is the same as in S14.

[0075] Next, the position estimation device 100 corrects the second real node link length to optimize the evaluation value (second correction step S25). The method for correcting the second real node link length is the same as in S15. Here, the target link x is the link between the first real node 12A, which can communicate with the second real node 12B, and the second real node 12B.

[0076] Next, the position estimation device 100 determines whether the termination condition is met (S26). If the termination condition is not met (NO in S26), the process returns to S23. If the termination condition is met (YES in S26), the position estimation device 100 outputs the estimated position of each node of the second node group 1B, which was estimated in the previous S23, as the measurement result, and terminates the process.

[0077] Furthermore, after completing position estimation for the second node group 1B, the position estimation device 100 may perform position estimation for the third node group, which includes the second real node 12B and the third real node. The third real node's true position is unknown, and it cannot communicate with any of the three anchor nodes 11 or the first real node 12A, but it can communicate with any of the second real nodes 12B.

[0078] Alternatively, a virtual node may be added to the second node group 1B, and this virtual node may be designated as the second reference node R2.

[0079] With the above configuration, the position estimation device 100 can perform position estimation for the three anchor nodes 11 in a stepwise manner, starting from the nearest actual node. This allows the above position estimation method to be applied to a wider topology of the wireless communication space.

[0080] [Examples of implementation using software] The function of the position estimation device 100 (hereinafter referred to as "the device") is a program that causes the device to function as a computer, and can be realized by a program that causes the computer to function as each control block of the device (particularly each part included in the control unit 23).

[0081] In this case, the device includes a computer having at least one control device (e.g., a processor) and at least one storage device (e.g., memory) as hardware for executing the program. By executing the program using this control device and storage device, the functions described in each of the embodiments are realized.

[0082] The above program may be recorded on one or more computer-readable recording media, not temporary ones. These recording media may or may not be provided by the above device. In the latter case, the program may be supplied to the above device via any wired or wireless transmission medium.

[0083] Furthermore, some or all of the functions of each of the above control blocks can also be realized by logic circuits. For example, an integrated circuit in which logic circuits functioning as each of the above control blocks are formed is also included in the scope of the present invention. In addition, it is also possible to realize the functions of each of the above control blocks by, for example, a quantum computer.

[0084] Furthermore, each process described in the above embodiments may be performed by AI (Artificial Intelligence). In this case, the AI ​​may operate on the control device described above, or it may operate on other devices (for example, an edge computer or a cloud server).

[0085] The present invention is not limited to the embodiments described above, and various modifications are possible within the scope of the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present invention. [Explanation of symbols]

[0086] 1. Multiple nodes 1A First Node Group 1B Second Node Group 11 Anchor Nodes 12A First Actual Node 12B Second Real Node 13 virtual nodes 100 Position estimation device 21 Acquisition Department 22 Memory section 23 Control Unit 231 Geometry Estimation Unit (Estimated Position Calculation Unit) 232 Evaluation Value Calculation Unit 233 Link length correction unit (correction unit)

Claims

1. The first node group includes an anchor node whose true position is known, and a first real node whose true position is unknown but which can communicate with the anchor node. An acquisition step to obtain the inter-node distance measured by wireless communication between each node in multiple nodes, A first estimated position calculation step involves calculating the estimated position of each node in the first node group by estimating the geometry of the first node group based on the internode distance between the anchor node and the first actual node, A first evaluation value calculation step of calculating an evaluation value that shows the deviation of the estimated position of a first reference node whose true position in the first node group is known from its true position, with respect to its true position, A position estimation method comprising: a first correction step of correcting the internode distance between the anchor node and the first actual node so as to optimize the evaluation value.

2. The additional step further includes adding virtual nodes in a virtual space whose true location is known to the first group of nodes, In the first estimated position calculation step, the geometry of the first group of nodes is estimated based on the internode distance between the anchor node and the first actual node, as well as the internode distance between the anchor node and the virtual node. The position estimation method according to claim 1, wherein in the first evaluation value calculation step, the virtual node is used as the first reference node.

3. The position estimation method according to claim 1, wherein in the first evaluation value calculation step, the anchor node is the first reference node.

4. The first estimated position calculation step, the first evaluation value calculation step, and the first correction step are repeatedly performed. The position estimation method according to any one of claims 1 to 3, wherein in the second and subsequent first estimated position calculation steps, the geometry of the first group of nodes is estimated based on the internode distance between the anchor node and the first actual node corrected in the first correction step.

5. The second node group includes the first real node and a second real node whose true position is unknown, which is not able to communicate with the anchor node but is able to communicate with the first real node. After repeatedly performing the first estimated position calculation step, the first evaluation value calculation step, and the first correction step, A second estimated position calculation step involves calculating the estimated position of each node in the second node group by estimating the geometry of the second node group based on the internode distance between the first and second actual nodes, A second evaluation value calculation step involves selecting any node in the first group of nodes as a second reference node and calculating an evaluation value that indicates the deviation of the estimated position of the second reference node from its true position, The position estimation method according to claim 4, further comprising repeatedly performing a second correction step of correcting the internode distance between the first real node and the second real node so as to optimize the evaluation value.

6. After repeatedly performing the first estimated position calculation step, the first evaluation value calculation step, and the first correction step, the estimated position of the first actual node in the first estimated position calculation step is considered to be the true position. The position estimation method according to claim 5, wherein in the second evaluation value calculation step, the first actual node is the second reference node.

7. The position estimation method according to claim 1, wherein the evaluation value is a value based on the difference between the estimated position of the first reference node and the true position of the first reference node.

8. The position estimation method according to claim 2, wherein the evaluation value is a value based on the difference between the internode distance between the anchor node and the virtual node before it is estimated in the first estimated position calculation step and the internode distance between the anchor node and the virtual node in the geometry estimated in the first estimated position calculation step.

9. The position estimation method according to claim 1, wherein the first correction step corrects the internode distance between the anchor node and the first actual node so as to reduce the evaluation value.

10. The position estimation method according to claim 1, wherein the first correction step corrects the node-to-node distance between the anchor node and the first actual node using a gradient method.

11. The first node group includes an anchor node whose true position is known, and a first real node whose true position is unknown but which can communicate with the anchor node. An acquisition unit that acquires the distance between nodes measured by wireless communication between each node in multiple nodes, An estimated position calculation unit calculates the estimated position of each node in the first node group by estimating the geometry of the first node group based on the internode distance between the anchor node and the first actual node, An evaluation value calculation unit calculates an evaluation value that shows the deviation of the estimated position of a first reference node whose true position in the first node group is known from its true position, with respect to its true position. A position estimation device comprising: a correction unit that corrects the internode distance between the anchor node and the first actual node so as to optimize the evaluation value.

12. A program for causing a computer to execute the position estimation method described in claim 1, the program for causing the computer to execute the acquisition step, the first estimated position calculation step, the first evaluation value calculation step, and the first correction step.

Images