Position estimation system and position estimation method

Abstract

To provide a location estimation system for suppressing a decrease in estimation precision of a position caused by a loaded state or luster of a coil.SOLUTION: A location estimation system includes: a distance sensor 6a that is installed in a trolley 4 and scans and measures distance to an object located in a side of a floor surface 100 by using a laser La; and a control operation unit 8 that specifies at least one coil C mounted in a side of the floor surface 100 as a position estimation object, moves the trolley 4 along a direction of a cross section of the position estimation object and allows the distance sensor 6a to scan and measure it, calculates a point in which the laser La passes height H of a coil center Ax of the position estimation object for each measurement point Pm as a virtual passage point Pv, extracts the virtual passage point Pv located in a side closest to the position estimation object as an edge candidate point Pc that can become a coil edge Pe in a direction of a cross section of the position estimation object, and derives a position of the position estimation object on the basis of the coil edge Pe estimated from the edge candidate point Pc.SELECTED DRAWING: Figure 10

Description

[Technical Field]

[0001] This disclosure relates to a location estimation system and a location estimation method. [Background technology]

[0002] Conventionally, there are transport devices such as cranes that suspend and transport coils, such as steel plate coils, within a building. Within the building's premises, multiple coils are stacked in two or three layers, in a so-called horizontal position, with their central axes parallel to the floor. Therefore, when a transport device automatically lifts and removes a specific coil that is already placed, it needs to know at least the position of that coil. On the other hand, when a transport device automatically suspends and transports a specific coil to a target position, it needs to know at least the position of the base coils around the target position.

[0003] Patent Document 1 discloses a technology relating to a position and shape recognition device that measures the position of the outer circumference of a coil using a slit-shaped light source and calculates the center position of the coil by substituting the measured value into the equation of a circle and solving it using the least squares method. Patent Document 2 discloses a technology relating to a method for measuring the center position of a roll body based on the position of positive and negative distance changes that have changed by a predetermined set amount relative to a reference value of the distance profile, while moving a distance sensor to measure the distance to an object directly below it. [Prior art documents] [Patent Documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 9-145315 [Patent Document 2] Japanese Patent Publication No. 2001-335281 [Overview of the Initiative] [Problems that the invention aims to solve]

[0005] For example, when multiple coils are stacked, another coil may already be placed diagonally above the coil whose position you wish to identify. In this case, in the recognition device disclosed in Patent Document 1, the upper coil may obstruct part of the measurement range of the distance sensor. In this situation, if the coil is a glossy steel plate coil, for example, stray light or multiple reflections may cause a large error in the measurement data from the distance sensor, which may result in a decrease in the accuracy of estimating the coil's position. In particular, since there is no measurement data for the other side of the coil in the cross-sectional direction, it is expected that the positional displacement in the cross-sectional direction will be large.

[0006] In the measurement method disclosed in Patent Document 2, the undetected range is wide, making it difficult to accurately calculate the position where the distance change is greatest, which corresponds to the coil end. Furthermore, even in the measurement method disclosed in Patent Document 2, if another roll body is placed diagonally above the roll body whose position is to be determined, only one side of the roll body in the cross-sectional direction will be detected. In particular, if the angle at which the laser emitted from the distance sensor strikes the roll body becomes shallow, measurement becomes impossible even though the laser is hitting the roll body, resulting in sparse information near the boundary. Therefore, under such a loading condition, it may be practically impossible to detect either the position of a positive or negative distance change, and as a result, the center position of the roll body may not be measured accurately. Moreover, if the roll body is a glossy steel plate coil, optical phenomena such as stray light or multiple reflections may cause large errors in the measurement data from the distance sensor.

[0007] Therefore, the present disclosure aims to provide a position estimation system and a position estimation method that suppress the decrease in the accuracy of coil position estimation caused by the coil's loading state or gloss. [Means for solving the problem]

[0008] One aspect of the present disclosure is a position estimation system for estimating the position of a coil placed in a position parallel to the floor, comprising: a distance sensor installed on a trolley that moves while suspending the coil, which scans and measures the distance to an object on the floor side using a laser; and a control calculation unit that identifies at least one coil placed on the floor side as a position estimation target, moves the trolley along the cross-sectional direction of the position estimation target and has the distance sensor scan and measure, calculates the point at which the laser passes the height of the center of the coil of the position estimation target as a virtual passing point for each measurement point, extracts the virtual passing point closest to the position estimation target as a candidate end point that could be the coil end in the cross-sectional direction of the position estimation target, and derives the position of the position estimation target based on the coil end estimated from the candidate end point.

[0009] In the above position estimation system, the control calculation unit may estimate each coil end using measurement data obtained by the distance sensor measuring from multiple positions, and identify the coil end to be used for estimating the position of the position to be estimated by performing statistical processing on the estimated values ​​of the multiple coil ends. The control calculation unit may divide the multiple virtual waypoints into multiple groups based on their proximity to each other, identify the group closest to the position to be estimated among the multiple groups based on the calculation results of the virtual waypoints calculated for each multiple frame, and identify the coil end to be used for estimating the position of the position to be estimated based on the candidate end points extracted from the multiple virtual waypoints included in the identified group. Alternatively, the control calculation unit may estimate the coil end by determining the amount of displacement between the virtual waypoint and the coil end at the height of the coil center, which is caused by the tangent angle when the laser emitted from the distance sensor touches the side surface of the position to be estimated, and reflecting the amount of displacement in the candidate end points.

[0010] Another aspect of the present disclosure is a position estimation method for estimating the position of a coil placed in a posture where the central axis is parallel to the floor surface, including: a position estimation target identification step of identifying at least one coil placed on the side of the floor surface as a position estimation target; a measurement step of moving a trolley that suspends and moves the coil along the cross-sectional direction of the position estimation target, and using a laser of a distance sensor installed on the trolley to scan and measure the distance to an object on the side of the floor surface; a virtual passing point calculation step of calculating, for each measurement point measured in the measurement step, a point where the laser passes through the height of the center of the coil of the position estimation target as a virtual passing point; an end candidate point extraction step of extracting the virtual passing point closest to the position estimation target as an end candidate point that can be the end of the coil in the cross-sectional direction of the position estimation target; and a coil position derivation step of deriving the position of the position estimation target based on the coil end estimated from the end candidate point.

Effect of the Invention

[0011] According to the present disclosure, it is possible to provide a position estimation system and a position estimation method that suppress a decrease in the estimation accuracy of the position of a coil due to the stacking state or gloss of the coil.

Brief Description of the Drawings

[0012] [Figure 1] It is a diagram showing a transport device including a position estimation system according to an embodiment. [Figure 2] It is a control block diagram of a position estimation system according to an embodiment. [Figure 3] It is a diagram showing an arrangement example of a plurality of coils in a site within a building. [[ID=2)3]] [Figure 4A] It is a diagram showing a coil to be a transport target or a position estimation target in the unloading process. [Figure 4B] It is a diagram showing a coil to be a transport target or a position estimation target in the loading process. [Figure 5] It is a flowchart of the transport process. [Figure 6] It is a flowchart of the measurement movement process. [Figure 7]It is a diagram for explaining the basic operation of measurement movement in the measurement process. [Figure 8] It is a flowchart of the vertex height estimation process. [Figure 9] It is a flowchart of the coil end estimation process. [Figure 10] It is a diagram for explaining the virtual passing point. [Figure 11] It is a diagram for explaining the end candidate point and the coil end. [Figure 12A] It is a diagram regarding the deviation when the trolley is located outside the coil end. [Figure 12B] It is a diagram regarding the deviation when the trolley is located inside the coil end.

Embodiments for Carrying Out the Invention

[0013] Hereinafter, several exemplary embodiments will be described with reference to the drawings. Here, the dimensions, materials, and other specific numerical values shown in each embodiment are merely examples, and do not limit the present disclosure unless otherwise specified. Also, for elements having substantially the same function and configuration, duplicate explanations are omitted by assigning the same reference numerals, and illustrations of elements not directly related to the present disclosure are omitted.

[0014] FIG. 1 is a perspective view showing the configuration of a conveying device 1 as an example of an apparatus to which a position estimation system or a position estimation method according to an embodiment is applied. FIG. 2 is a control block diagram schematically showing the control flow in the conveying device 1.

[0015] The conveying device 1 is an overhead crane installed in the upper part of a building such as a factory or warehouse, with a cylindrical coil C as the object to be conveyed. The conveying device 1 may also be referred to as a cargo handling device when the object to be conveyed is considered as a load. The position estimation system or position estimation method according to this embodiment is applied to the conveying device 1. In this case, the object of position estimation is the object to be conveyed in the conveying device 1, i.e., the coil C. Note that the object of position estimation is not limited to what is called a coil, but includes, for example, what is called a roll. In the example of this embodiment, the coil C is a steel plate coil in which a thin steel plate is wound into a cylindrical shape.

[0016] Multiple coils C are temporarily stored in the building's premises. For the following explanation of the arrangement of the multiple coils C and the operation of each part of the conveying device 1, the directions are defined as follows: The Z direction is the direction along the vertical. The XY plane is a horizontal plane perpendicular to the Z direction. The X and Y directions are perpendicular to each other on the XY plane.

[0017] Figure 3 is a perspective view showing an example of the arrangement of multiple coils C on a site inside a building. The site has a horizontal floor surface 100. Each coil C is installed on the side of the floor surface 100 with its central axis parallel to the floor surface 100. In this embodiment, when each coil C is installed on the side of the floor surface 100, the axial direction of each coil C is along the Y direction, and the cross-section perpendicular to the axial direction is parallel to the XZ plane. Also, when each coil C is installed on the side of the floor surface 100, they are aligned along the X direction. Furthermore, each coil C is stacked in one layer or in multiple layers along the Z direction. In the example in Figure 3, multiple coils C are stacked in a maximum of three layers. Of these, the bottom layer of coil C is each placed on the floor surface 100 via two rod-shaped sleepers 110 that form a pair. The extension direction of the sleepers 110 is along the axial direction of the coil C. Furthermore, the two sleepers that make up a pair are adjacent to each other in the X direction. Such pairs of sleepers 110 can position the coil C on which they are placed and can also prevent the coil C from rolling. On the other hand, the middle coils C are each supported by the two bottom coils C that are adjacent to each other in the X direction. Similarly, the top coil C is supported by the two middle coils C that are adjacent to each other in the X direction.

[0018] The conveying device 1 can estimate the location of a specific coil C already placed on the site, and then, while gripping the coil C at the estimated location, lift it up and move it to another location. This operation of lifting a desired coil C and moving it to another location may be referred to as the "lifting operation" below. The conveying device 1 can also estimate a target location on the site where a specific coil C will be placed, and then, after moving the coil C from another location, can be lifted and placed at the estimated target location. This operation of lifting and placing the coil C at the target location may be referred to as the "lowering operation" below.

[0019] The transport device 1 comprises a girder 2, a runway 3, a trolley 4, tongs 5, a first sensor 6a and a second sensor 6b as distance sensors, an operator's cab 7, and a control calculation unit 8.

[0020] The girder 2 is a structure that supports the trolley 4 movably near the ceiling inside the building. The girder 2 includes two linear rigid body sections that extend along the X direction and are spaced apart in the Y direction while facing each other. In this case, the two rigid body sections are integrated with each other at both ends in the X direction, although this is not shown. At least one end of the girder 2 in the X direction is equipped with a girder actuator 25 (see Figure 2), such as a motor.

[0021] Runway 3 is a rail that supports the girder 2 so that it can move freely, located near the ceiling inside the building. Runway 3 extends along the Y direction and is installed at two locations opposite to each other, spaced apart in the X direction. Runway 3 may be installed, for example, on the wall of the building, or on an elevated platform provided on the floor surface 100 inside the building. One end of the girder 2 in the X direction is located on one runway 3, and the other end of the girder 2 in the X direction is located on the other runway 3. The girder 2 can move along the runway 3 along the Y direction by being driven by the girder actuator 25 based on a movement command from the control calculation unit 8. Hereinafter, the direction of movement of the girder 2 relative to the runway 3 along the Y direction will be referred to as the "travel direction".

[0022] The trolley 4 is a structure that is supported by two rigid body parts on the girder 2, is movable along the X direction, and supports the tongs 5 ​​so that they are movable along the Z direction. The trolley 4 is equipped with a trolley actuator 26 (see Figure 2), such as a motor. When the trolley actuator 26 is driven based on a movement command from the control calculation unit 8, the trolley 4 can move along the X direction on the rigid body parts of the girder 2. Since the girder 2 can move along the Y direction, the trolley 4 can move to a target coordinate position on the XY plane. Hereinafter, the direction of movement of the trolley 4 relative to the girder 2 along the X direction will be referred to as the "traverse direction".

[0023] The tongs 5 ​​are gripping parts that grasp the coil C and move up and down. The trolley 4 includes a rotating drum (not shown) for lifting and lowering the tongs 5 ​​via a wire rope 9, and a tongs actuator 27 (see Figure 2), such as a motor, that provides rotational force to the rotating drum. The tongs actuator 27 is driven based on a lifting command from the control calculation unit 8, allowing the tongs 5 ​​to move along the Z direction with respect to the height of the trolley 4.

[0024] The tongs 5 ​​also comprises a base 10 and a pair of arms, a first arm 11a and a second arm 11b. The base 10 is connected to a wire rope 9 and supports the first arm 11a and the second arm 11b. The first arm 11a and the second arm 11b each extend downward from the base 10 along the Z direction and face each other in the Y direction. The first arm 11a is connected to one end of the base 10 in the Y direction, and the second arm 11b is connected to the other end of the base 10 in the Y direction. The first arm 11a and the second arm 11b can each slide along the Y direction while maintaining their orientation along the Z direction.

[0025] Although not shown in the diagram, the lower ends of the first arm 11a and the second arm 11b in the Z direction are each equipped with lifting claws that face each other along the Y direction. The first arm 11a and the second arm 11b can approach each other with their two lifting claws positioned on the extension of the coil hole CH of the coil C to be transported, thereby allowing each lifting claw to enter the coil hole CH. In other words, the first arm 11a and the second arm 11b can grip the coil C to be transported by approaching each other. Hereinafter, the state in which the first arm 11a and the second arm 11b are in the position to grip the coil C will be referred to as the closed state of the arms. On the other hand, the first arm 11a and the second arm 11b can release the grip on the coil C by moving apart from each other, thereby retracting each lifting claw from the coil hole CH. Hereinafter, the state in which the first arm 11a and the second arm 11b are in the position to release the grip on the coil C will be referred to as the open state of the arms. The trolley 4 is equipped with a coil-gripping actuator 28 (see Figure 2), such as a motor. The coil-gripping actuator 28 is driven based on an arm opening / closing command from the control calculation unit 8, allowing the tongs 5 ​​to switch between an open state and a closed state of the arm.

[0026] The first sensor 6a and the second sensor 6b, acting as distance sensors, are installed on the trolley 4 and scan and measure the distance to an object on the floor surface 100. The first sensor 6a and the second sensor 6b are of the same type, but their installation positions on the trolley 4 are different. In this embodiment, the first sensor 6a and the second sensor 6b are LiDAR (Light Detection And Ranging), which is an example of a laser distance sensor that measures distance using a laser La (see Figure 7).

[0027] The scanning direction during measurement for the first sensor 6a and the second sensor 6b is along the cross-sectional direction of the coil C whose position is to be estimated, i.e., the lateral direction. In this case, the first sensor 6a and the second sensor 6b are installed on the trolley 4 with a certain distance between them in the X direction. In this embodiment, the first sensor 6a is installed on the side of the trolley 4 facing the positive side in the X direction, and the second sensor 6b is installed on the side of the trolley 4 facing the negative side in the X direction, so that they are located on opposite sides of the main body of the trolley 4 in the X direction. This installation relationship has the advantage that the detection range is less likely to be obstructed by the tongs 5 ​​supported by the trolley 4, and the detection range is less likely to be obstructed by the coil C even when the tongs 5 ​​are gripping the coil C. Furthermore, the field of view during scanning measurement along the lateral direction of the trolley 4 is set to approximately ±45 degrees.

[0028] The driver's cab 7 is fixed to one end of the girder 2 and houses a driver who operates the transport device 1. The driver's cab 7 is equipped with a control calculation unit 8 or a user interface 20, etc. The user interface 20 includes an input unit operated by the driver and can transmit command information entered by the driver to the control calculation unit 8. The user interface 20 can also display estimated position information transmitted from the control calculation unit 8.

[0029] The control calculation unit 8 controls the operation of each component included in the transport device 1 based on transport instructions from the driver via the user interface 20, or measurement data acquired from distance sensors such as the first sensor 6a. The control calculation unit 8 includes a drive control unit 21 and an estimation calculation unit 22.

[0030] The drive control unit 21 transmits operation signals to the various actuators included in the transport device 1. First, the drive control unit 21 receives transport instructions or coil information from the user interface 20, and based on the transport instructions or coil information, it can determine the approximate position to which the trolley 4 and tongs 5 ​​should be directed. The drive control unit 21 can also receive the placement position or target position of the coil C derived by the estimation calculation unit 22 from the estimation calculation unit 22.

[0031] Furthermore, the drive control unit 21 can instruct the girder actuator 25 to specify a target movement position for the girder 2, and move the girder 2 to that position. The drive control unit 21 can instruct the trolley actuator 26 to specify a target movement position for the trolley 4, and move the trolley 4 to that position. The drive control unit 21 can instruct the tong actuator 27 to specify a target lifting position for the tongs 5, and move the tongs 5 ​​to that lifting position. In addition, the drive control unit 21 can instruct the coil gripping actuator 28 to specify a desired open / closed state for the arms of the tongs 5, and open / close the arms.

[0032] The estimation calculation unit 22 acquires measurement data from distance sensors such as the first sensor 6a, and performs an estimation calculation of the position of coil C based on the operation of the position estimation system according to this embodiment or the position estimation method. The estimation calculation unit 22 can also transmit the calculated mounting position or target position of coil C to the drive control unit 21. Furthermore, the estimation calculation unit 22 can receive coil information necessary for the estimation calculation from the user interface 20. In other words, in the example of the transport device 1, the position estimation system according to this embodiment consists of at least distance sensors such as the first sensor 6a and the estimation calculation unit 22.

[0033] Next, the operation of the position estimation system according to this embodiment and the flow of the position estimation method according to this embodiment will be described.

[0034] As the main operations during automatic transport, the transport device 1 performs a lifting operation and a lowering operation, as described above. Hereafter, the transport process related to the lifting operation will be referred to as the "unloading process," and the transport process related to the lowering operation will be referred to as the "inbound process."

[0035] Figure 4A is a side view of multiple coils C, illustrating coils C that are transported or whose position is estimated during the unloading process. In the example in Figure 4A, seven coils C are placed on the side of the floor surface 100. Of these, the first coil C1, second coil C2, third coil C3, and fourth coil C4 are placed on the floor surface 100 via sleepers 110, in parallel along the X direction, as the lower coils C. The fifth coil C5, sixth coil C6, and seventh coil C7 are placed as the upper coils C, each supported by any two of the lower coils C. Specifically, the fifth coil C5 is placed using the first coil C1 and second coil C2 as a base. The sixth coil C6 is placed using the second coil C2 and third coil C3 as a base. The seventh coil C7 is placed using the third coil C3 and fourth coil C4 as a base.

[0036] In the example shown in Figure 4A, it is assumed that the coil C to be transported, i.e., the coil C that is lifted by the transport device 1 and unloaded from its mounting position, is the upper sixth coil C6. In this case, the objects of position estimation in this embodiment are the sixth coil C6, which is the object to be transported, and the fifth coil C5 and seventh coil C7, which are located on either side of the sixth coil C6 in the X direction, for a total of three coils C. Of these, the estimated position of the sixth coil C6 is referenced as the position to which the ungrasped tongs 5 ​​should be directed during unloading. On the other hand, the estimated positions of the fifth coil C5 and seventh coil C7 are referenced as positions that are determined in advance to avoid contact between the sixth coil C6 and the fifth coil C5 or seventh coil C7 during unloading.

[0037] Figure 4B is a side view of multiple coils C, illustrating the target position where the coil C to be transported is brought in during the loading process, and the coil C whose position is to be estimated. In the example in Figure 4B, six coils C, namely the first coil C1, the second coil C2, the third coil C3, the fourth coil C4, the fifth coil C5, and the seventh coil C7, are each placed on the floor surface 100, similar to the example in Figure 4A. In Figure 4B, the area drawn with a dashed line, corresponding to the sixth coil C6 in Figure 4A, is the target position. Hereafter, the coil C placed at the target position will be referred to as the sixth coil C6 for convenience.

[0038] In this embodiment, the objects of position estimation are a total of four coils C: the second coil C2 and the third coil C3 that support the sixth coil C6 to be placed at the target position, and the fifth coil C5 and the seventh coil C7 located on either side of the sixth coil C6 in the X direction. Of these, the estimated positions of the second coil C2 and the third coil C3 are referenced as target positions during delivery. On the other hand, the estimated positions of the fifth coil C5 and the seventh coil C7 are referenced as positions that are determined in advance to avoid contact between the sixth coil C6 and the fifth coil C5 or the seventh coil C7 during delivery.

[0039] Figure 5 is a flowchart of the transport process. The transport process can be divided into two types: an unloading process that includes lifting operations, and an inloading process that includes lowering operations. Below, each step included in the transport process will be explained separately for cases where the transport process is an unloading process and cases where it is an inloading process.

[0040] Furthermore, the position of coil C estimated in this embodiment is defined in a crane coordinate system based on an origin P0 set on the floor surface 100. In this embodiment, the origin P0 is set at a corner of a part of the movement range of trolley 4, as shown in Figure 1. The crane coordinate system is represented by a traverse direction corresponding to the X direction, a travel direction corresponding to the Y direction, and a height direction corresponding to the Z direction.

[0041] The control calculation unit 8 starts the transport process based on a transport instruction from the driver via the user interface 20.

[0042] First, the control calculation unit 8 executes the position estimation target identification process S100. In the position estimation target identification process S100, the control calculation unit 8 identifies whether the transport process is an outbound process or an inbound process based on the transport instruction. If the transport process is an outbound process, the control calculation unit 8 identifies which coil C will be transported. On the other hand, if the transport process is an inbound process, the control calculation unit 8 identifies which target position the transported coil C will be transported to. Next, the control calculation unit 8 identifies the coil C whose position should be estimated, i.e., the position estimation target, based on the placement position of the coil C to be transported in the outbound process, or the target position on which the transported coil C will be placed in the inbound process.

[0043] Furthermore, in the position estimation target identification process S100, the control calculation unit 8 obtains coil information for each coil C to be targeted for position estimation from a database stored in the higher-level system on the user interface 20 side. Here, coil information refers to information regarding the approximate placement position or target position of coil C, or information regarding the size of coil C, etc.

[0044] Next, the control calculation unit 8 executes the primary planar movement process. In the primary planar movement process, the control calculation unit 8 moves the trolley 4 toward a position above the placement position or target position of the coil C. If the transport process is an unloading process, the trolley 4 that does not suspend any coil C is moved toward a position above the placement position of the coil C to be transported. On the other hand, if the transport process is an in-loading process, the trolley 4 that suspends the coil C to be transported is moved toward a position above the target position preset on the floor surface 100. The primary planar movement process includes a high-speed movement process S200, a measurement movement process S300, and a correction movement process S400.

[0045] The high-speed movement process S200 is a process in which the trolley 4 is moved at a speed faster than the movement speed in the measurement movement process S300 to a position above the approximate placement position or target position specified from the transport instruction. At this time, the drive control unit 21 drives the girder actuator 25 or the trolley actuator 26 using the XY coordinates corresponding to the position above the approximate placement position or target position as the movement position. The XY coordinate position as the movement position is the position from which the subsequent measurement movement process S300 begins. In other words, at this stage, the control calculation unit 8 does not know the specific placement position or target position of the coil C. After the high-speed movement process S200, the control calculation unit 8 moves on to the measurement movement process S300.

[0046] The measurement movement process S300 is a process that includes estimating the position of coil C, which is designated as the position estimation target. This measurement movement process S300 will be explained in detail below. After the measurement movement process S300, the control calculation unit 8 moves on to the correction movement process S400.

[0047] The correction movement process S400 is a process of correcting the movement of the trolley 4 from its position at the end of the measurement movement process S300 to the target coordinates of the trolley 4 determined based on the position of the coil C estimated in the measurement movement process S300. At this time, the drive control unit 21 drives the girder actuator 25 or the trolley actuator 26 with the XY coordinates corresponding to the target coordinates of the trolley 4 as the movement position. Upon completion of this correction movement process S400, the XY coordinate position of the trolley 4 when lifting the coil C, which is the object to be transported, from its mounting position, or when suspending the coil C, which is the object to be transported, to the target position, is determined.

[0048] Next, the control calculation unit 8 executes the tongs lowering process S500. In the tongs lowering process S500, the control calculation unit 8 lowers the tongs 5 ​​to the target coordinates of the tongs 5 ​​determined based on the position of the coil C estimated in the measurement movement process S300. At this time, the drive control unit 21 drives the tongs actuator 27 with the Z coordinate corresponding to the target coordinates of the tongs 5 ​​as the lifting position. Upon completion of this tongs lowering process S500, if the transport process is an unloading process, the tongs 5 ​​are in a position to grip the coil C, which is the object to be transported. On the other hand, if the transport process is an inloading process, the coil C, which is the object to be transported, is placed at the target position.

[0049] Next, the control calculation unit 8 executes the tong operation process S600. If the transport process is an unloading process, the tong operation process S600 is the process of causing the tongs 5 ​​to grip the coil C to be transported. On the other hand, if the transport process is an loading process, the tong operation process S600 is the process of causing the tongs 5 ​​to release their grip on the coil C and to place the coil C completely at the target position. At this time, the drive control unit 21 appropriately drives the coil gripping actuator 28 to change between the open and closed states of the arms of the tongs 5.

[0050] Next, the control calculation unit 8 executes the tong raising process S700. In the tong raising process S700, if the transport process is the discharge process, the tongs 5 ​​gripping the coil C are raised to the highest position. On the other hand, if the transport process is the loading process, the tongs 5 ​​not gripping the coil C are raised to the highest position. In this case as well, the drive control unit 21 can raise the tongs 5 ​​by driving the tong actuator 27.

[0051] The control calculation unit 8 then executes the secondary planar movement process S800. In the secondary planar movement process S800, the drive control unit 21 moves the trolley 4 away from the position above the coil C's mounting position or target position. If the transport process is an unloading process, the secondary planar movement process S800 moves the trolley 4, which is suspending the coil C to be transported, to the desired position to transport the coil C. On the other hand, if the transport process is an loading process, the trolley 4 that is not suspending the coil C is moved to the desired position. This completes the series of transport processes.

[0052] Next, we will explain the measurement and movement process S300 in detail.

[0053] Figure 6 is a flowchart of the measurement movement process S300. In the measurement movement process S300, the first sensor 6a or the second sensor 6b, which are distance sensors, move along the cross-sectional direction of the coil C, which is the target of position estimation, and scan measurements for multiple frames to estimate the position of the coil C. In other words, in the measurement movement process S300, the measurement process S310 using the distance sensor and the process of deriving the position of the coil C based on the measurement data obtained in the measurement process S310 run in parallel. In Figure 6, the flow of the measurement process S310 is shown on the left side in chronological order, and the flow of deriving the position of the coil C is shown on the right side.

[0054] Figure 7 is a perspective view illustrating the basic operation of measurement movement in the measurement process S310. Note that Figure 7 corresponds to the perspective view of the transport device 1 in Figure 1. Figure 7 illustrates the case where the transport process is a loading process and the first sensor 6a performs scan measurement in the measurement movement process S300.

[0055] In measurement step S310, the first sensor 6a scans the already placed coil C over it as it moves horizontally along the trolley 4, obtaining measurement data from multiple viewpoints. In Figure 7, three coils C that can be used for position estimation are placed on the sleeper 110. The scan range defined by a specific field of view is shown as the irradiation range of the laser La. Furthermore, in Figure 7, line-shaped point clouds representing three scans aligned in the Y direction are shown on the three coils C. However, in reality, measurements may be taken on more than four lines. By arranging a large number of measurement lines in this way, if a measurement line has a large variation in the point cloud compared to other measurement lines, the measurement line with the large variation can be excluded in advance from the measurement data used for position estimation of the coil C, thereby improving the accuracy of position estimation.

[0056] Here, we assume that the trolley 4 is positioned closest to the origin P0 in the X direction. In this case, if the coil C to be positioned is placed directly below the trolley 4, the first sensor 6a may have difficulty detecting the coil C directly below because the irradiation range of the laser La is obstructed by the coil C being held by the tongs 5. In such a case, the second sensor 6b, which is installed on the trolley 4 on the opposite side of the first sensor 6a in the X direction, can be used as a distance sensor instead of the first sensor 6a to suitably detect the coil C directly below.

[0057] When the measurement process S310 begins following the start of the measurement movement process S300, the estimation calculation unit 22 then derives the position of each coil C whose position needs to be estimated. Here, as an example, we will refer to the multiple coils C shown in Figure 4B relating to the loading process and explain the case where the target of position estimation is the second coil C2.

[0058] First, the estimation calculation unit 22 executes the coordinate transformation process S320. In the coordinate transformation process S320, the measurement data from numerous viewpoints acquired sequentially in the measurement process S310 is corrected by the position of distance sensors such as the first sensor 6a, and transformed into coordinates fixed to the building. This coordinate transformation allows for the pre-removal of isolated noise or point clouds corresponding to parts of the structure of the transport device 1.

[0059] Next, the estimation calculation unit 22 executes the vertex height estimation process S330. In the vertex height estimation process S330, the height Tz of the vertex Pt of the second coil C2 is estimated. Note that the vertex Pt of coil C is indicated by a white triangle in Figures 4A, 4B, and 11.

[0060] Figure 8 is a flowchart of the vertex height estimation process S330.

[0061] First, the estimation calculation unit 22 executes the point cloud extraction process S331. The point cloud extraction process S331 is a process of extracting a point cloud from measurement data as measurement information. In Figure 4B, the extraction range Ar extracted in this process is shown as a dashed rectangular area. At this time, the extraction range Ar is set to include the assumed coil center Ax of the second coil C2 and a part of the surrounding coil C, based on the coil information acquired in advance in the position estimation target identification process S100.

[0062] Next, the estimation calculation unit 22 executes the specific area extraction step S332. The specific area extraction step S332 is a step in which, from the point cloud Am included in the cutout range Ar, constituent points of a specific area, for example, a gentle part with an angle of 30 degrees or less, are extracted. In Figure 4B, the gentle part extracted in this step is shown as a constituent point cloud included in the dashed elliptical region R.

[0063] Next, the estimation calculation unit 22 executes the moving average calculation step S333. The moving average calculation step S333 is a step of calculating the median value, i.e., the moving average, within a predetermined width W using a moving filter at multiple points in the X direction.

[0064] Then, the estimation calculation unit 22 executes the vertex height derivation process S334. The vertex height derivation process S334 is a process of deriving the height Tz and the X-coordinate Tx of the vertex Pt, which is the point at the highest position, based on the moving average calculated in the moving average calculation process S333. Through this series of steps in the vertex height estimation process S330, the estimation calculation unit 22 can estimate the height Tz of the vertex Pt of the second coil C2.

[0065] Next, returning to the measurement movement process S300, the estimation calculation unit 22 executes the coil radius information acquisition process S340 after the vertex height estimation process S330. In the coil radius information acquisition process S340, the estimation calculation unit 22 acquires information relating to the coil radius r of the second coil C2, whose vertex height Tz was estimated in the vertex height estimation process S330. The information relating to the coil radius r may be included in the coil information received from the user interface 20. Alternatively, the information relating to the coil radius r may be measured for each coil C in advance when it is received into storage, or it may be measured using a measuring device that is pre-installed on the tongs 5, for example.

[0066] Next, the estimation calculation unit 22 executes the coil end estimation process S350. In the coil end estimation process S350, the coil end Pe of the second coil C2 is estimated. The coil end Pe is the part of coil C that is located at the very end in the X direction, that is, in the traverse direction of the trolley 4. Note that the coil end Pe is shown as a black diamond in Figures 4A, 4B, 11, 12A, and 12B.

[0067] Figure 9 is a flowchart of the coil end estimation process S350.

[0068] First, the estimation calculation unit 22 executes the coil center height calculation process S351. The coil center height calculation process S351 is a process of calculating the height H of the coil center Ax by subtracting the coil radius r from the height Tz of the apex Pt of the second coil C2.

[0069] Next, the estimation calculation unit 22 executes the virtual waypoint calculation process S352. The virtual waypoint calculation process S352 is a process for calculating multiple virtual waypoints Pv.

[0070] Figure 10 is a conceptual diagram illustrating the virtual waypoint Pv. Figure 10 illustrates a case where the second coil C2 and the third coil C3 are placed in parallel on the floor surface 100 via sleepers 110, and the first coil C1, which is supported by the second coil C2 and the third coil C3, is the target of position estimation. Here, it is assumed that either the first sensor 6a or the second sensor 6b, which are distance sensors, measured the area including the first coil C1 for one scan using the laser La. In this case, the virtual waypoint Pv corresponds to the point where the height H of the coil center Ax of the first coil C1 is passed for each of the multiple measurement points Pm.

[0071] Next, the estimation calculation unit 22 executes the endpoint candidate point extraction process S353. The endpoint candidate point extraction process S353 is a process of extracting endpoint candidate points Pc from among a plurality of virtual passing points Pv.

[0072] Figure 11 is a diagram illustrating the candidate end point Pc and the coil end point Pe that can be estimated from the candidate end point Pc.

[0073] The upper graph in Figure 11 illustrates the case where an end point candidate Pc is extracted from multiple virtual waypoints Pv contained in a single frame. In the upper graph, with the origin P0 as the reference point, the horizontal axis represents the position in the X direction, and the vertical axis represents the position in the Z direction. Furthermore, the upper graph shows a part of coil Ce and the apex Pt of coil Ce, using the example where the estimated coil center Ax of coil Ce is located at approximately X=6.0[m] and Z=1.9[m]. In addition, multiple virtual waypoints Pv calculated in the virtual waypoint calculation process S352 are plotted on the X axis.

[0074] The lower graph in Figure 11 is a graph plotting the candidate end points Pc extracted for each of the multiple frames, aligned with the horizontal axis of the upper graph in Figure 11. As an example, the lower graph shows the candidate end points Pc extracted for 12 frames, from frame number 1 to 12. Of these, the three candidate end points Pc plotted in the column for frame number 1 correspond to the respective candidate end points Pc exemplified in the upper graph. Note that in the actual measurement movement process S300, candidate end points Pc are extracted for a much larger number of frames.

[0075] First, the estimation calculation unit 22 divides a series of consecutive virtual passing points Pv for a given single frame into several groups based on their proximity to each other. In the upper diagram of Figure 11, three groups are shown as examples: the first group G1, the second group G2, and the third group G3. Here, there are regions between adjacent groups where no virtual passing points Pv have been calculated. Such regions are thought to occur, for example, when the coil to be positioned is in the first stage, due to the unevenness of the floor surface 100 below the coil, making measurement difficult, and thus preventing the distance sensor such as the first sensor 6a from detecting the measurement point Pm. Subsequently, the estimation calculation unit 22 extracts, for each group, the virtual passing point Pv closest to the second coil C2, which is the target of position estimation, as a candidate end point to be referenced when estimating the coil end Pe.

[0076] Next, the estimation calculation unit 22 executes the correction amount calculation step S354. The correction amount calculation step S354 is applied to the candidate end point Pc to derive the final estimated coil end Pe by calculating the correction amount X. mod This is the process of calculating the correction amount X. mod As shown in equation (1), the tangent angle θ t Virtual point shift Se(θ) t It is represented as follows:

[0077]

number

[0078] Figures 12A and 12B are conceptual diagrams for explaining Se(θ t ). In Figures 12A and 12B, a case where the first coil C1 placed on the floor surface 100 via the sleeper 110 is the object of position estimation is illustrated. Figure 12A shows Se(θ t ) when the trolley 4 equipped with a distance sensor such as the first sensor 6a is located outside the coil end Pe of the first coil C1, that is, on the plus side in the X direction from the coil end Pe. Figure 12B shows Se(θ t ) when the trolley 4 equipped with a distance sensor such as the first sensor 6a is located inside the coil end Pe of the first coil C1, that is, on the minus side in the X direction from the coil end Pe.

[0079] In this case, Se(θ t ) is represented by Equation (2). However, r is the coil radius. Also, assuming that the relative position of the i-th measurement point from the first sensor 6a etc. is represented by (X r (i), Z r (i)), the tangent angle θ t is represented by Equation (3).

[0080]

Equation

[0081]

Equation

[0082] Also, in Figures 12A and 12B, as an example, the measurement point deviation that may occur when a reflection ghost occurs is illustrated. When a so-called reflection ghost occurs, the measurement point is detected in such a form that it enters the inside of the first coil C1. Since the virtual passing point at this time is not the virtual passing point in contact with the actual coil, the correction by Se(θ t ) becomes a correction that ignores the actual geometric relationship, and the variation in the position after correction is θ tThe value actually increases as the value changes. Figures 12A and 12B show the corrected coil end Pg affected by reflected ghosting.

[0083] Then, the estimation calculation unit 22 executes the coil end derivation process S355. The coil end derivation process S355 applies a correction amount X to the candidate end point Pc. mod This is a process of deriving the coil end Pe by applying the following. In Figure 11, for each group, the correction amount X is applied to the candidate end point Pc. mod When applied, the coil end Pe, which is derived from the position of the candidate end point Pc to the inside of the coil Ce, is shown along the arrow. In other words, in the example of Figure 4B relating to the loading process, the coil end Pe is estimated from the candidate end point Pc by geometric calculation.

[0084] The above coil end estimation process S350 is performed for the measurement data from each viewpoint during the measurement process S310. The estimation calculation unit 22 then stores the estimated data related to the coil end Pe.

[0085] Then, returning to the measurement movement process S300, after the measurement process is completed (S360), the estimation calculation unit 22 executes the coil position derivation process S370. Here, the measurement process may be considered complete when the trolley 4 reaches the end point of the measurement movement. In the coil position derivation process S370, the estimation calculation unit 22 derives the coil end Pe that will be ultimately adopted by statistically processing the measurement data from multiple viewpoints obtained during the measurement process S310.

[0086] For example, referring to Figure 11, the estimation calculation unit 22 identifies the group closest to the second coil C2, which is the target of position estimation, from among multiple groups, based on the calculation results of virtual waypoints Pv calculated for each frame measured at multiple locations. For example, if many of the virtual waypoints Pv calculated for each frame are clustered within a certain range, it is considered to be the group from which the coil end Pe that should ultimately be adopted is derived.

[0087] More specifically, referring to the lower diagram in Figure 11, it can be seen that, as a trend in the candidate end points Pc extracted for each of the multiple frames, the candidate end points Pc in the second group G2 are extracted more frequently and have the highest density than the candidate end points Pc in the other groups. Here, a high density of candidate end points Pc means that, statistically, the variability of the candidate end points Pc is small. On the other hand, the candidate end points Pc in the first group G1 or the third group G3 have a lower density compared to the candidate end points Pc in the second group G2. Therefore, the estimation calculation unit 22 estimates that the second group G2 is the closest to the second coil C2 among the multiple groups. Next, the estimation calculation unit 22 finally adopts the coil end Pe estimated by averaging or other means from the candidate end points Pc included in the second group G2. Then, the estimation calculation unit 22 derives the position of the second coil C2 from the coil end Pe that was finally decided to adopt and the coil radius r of the second coil C2.

[0088] On the other hand, the virtual waypoints Pv calculated for each frame are clustered within a certain range, but Se(θ) ignores the actual geometric relationships as described above. t It is also thought that reflection ghosting occurs due to the correction caused by (), which results in increased variation in the corrected position. In the example shown in Figure 11, the virtual passing point Pv included in the first group G1 is affected by reflection ghosting. Therefore, the coil end Pe estimated from the candidate end point Pc included in the first group G1, i.e., the coil end Pg as exemplified in Figure 12A, is ultimately rejected.

[0089] Up to this point, the measurement and movement process S300 has been explained by referring to the multiple coils C shown in Figure 4B, which relate to the loading process. On the other hand, even when the transport process is the loading process, the measurement and movement process S300 can be executed in a similar manner by defining each part as shown in Figure 4A.

[0090] Next, the effects of the position estimation system and position estimation method according to this embodiment will be described.

[0091] The position estimation system according to this embodiment estimates the position of a coil C that is placed in a position where its central axis is parallel to the floor. The position estimation system is installed on a trolley 4 that suspends and moves the coil C, and includes a first sensor 6a or a second sensor 6b as a distance sensor that scans and measures the distance to an object on the floor surface 100 side using a laser La. The position estimation system also includes a control calculation unit 8. The control calculation unit 8 identifies at least one coil C placed on the floor surface 100 side as the position estimation target, and causes the distance sensor to scan and measure while moving the trolley 4 along the cross-sectional direction of the position estimation target. For each measurement point Pm, the control calculation unit 8 calculates the point at which the laser La passes through the height H of the coil center Ax of the position estimation target as a virtual passing point Pv. The control calculation unit 8 extracts the virtual passing point Pv closest to the position estimation target as a candidate end point Pc that can become the coil end Pe in the cross-sectional direction of the position estimation target. The control calculation unit 8 then derives the position of the position to be estimated based on the coil end Pe estimated from the candidate end point Pc.

[0092] Furthermore, the position estimation method according to this embodiment estimates the position of a coil C that is placed in a position where its central axis is parallel to the floor surface. The position estimation method includes a position estimation target identification step S100 in which at least one coil C placed on the floor surface 100 is identified as the position estimation target. The position estimation method includes a measurement step S310 in which, while moving a trolley 4 that suspends the coil C along the cross-sectional direction of the position estimation target, a distance sensor installed on the trolley 4 is used to scan and measure the distance to an object on the floor surface 100 side using a laser La. For each measurement point Pm measured in the measurement step S310, the position estimation method includes a virtual passing point calculation step S352 in which the point at which the laser La passes through the height H of the coil center Ax of the position estimation target is calculated as a virtual passing point Pv. The position estimation method includes an end candidate point extraction step S353 in which the virtual passing point Pv on the side closest to the position estimation target is extracted as an end candidate point Pc that can become the coil end Pe in the cross-sectional direction of the position estimation target. Furthermore, the position estimation method includes a coil position derivation step S370 in which the position of the target for position estimation is derived based on the coil end Pe estimated from the candidate end point Pc.

[0093] First, the position estimation system or position estimation method according to this embodiment introduces the concept of a virtual passing point Pv, which is defined as the point where the laser La passes through the height H of the coil center Ax of the position to be estimated. Therefore, when attempting to estimate the position of the position to be estimated based on the coil end Pe, it is not always necessary to specify the positions of both coil end Pes, as in the example above. Therefore, even if there is a situation in which another coil is already placed diagonally above the position to be estimated, the coil position can be derived while suppressing a decrease in the accuracy of coil position estimation by estimating only the coil end Pe on one side.

[0094] Furthermore, generally speaking, when a coil has a glossy surface, it can be difficult to detect the coil end using a laser-based distance sensor, or optical phenomena such as stray light or multiple reflections may cause a position far from the actual location to be detected. In contrast, the position estimation system or method according to this embodiment can estimate the position of the coil end Pe with high accuracy, and as a result, it is possible to suppress a decrease in the accuracy of coil position estimation.

[0095] As described above, this embodiment provides a position estimation system and a position estimation method that suppress the decrease in the accuracy of coil position estimation caused by the coil's stacking state or gloss.

[0096] Furthermore, in the position estimation system according to this embodiment, the control calculation unit 8 estimates the coil end Pe using measurement data obtained by the distance sensor measuring from multiple positions. The control calculation unit 8 may then perform statistical processing on the estimated values ​​of the multiple coil end Pe to identify the coil end Pe to be used for estimating the position of the position to be estimated.

[0097] Generally, when using a laser-based distance sensor for measurement, reflective ghosting occurs, resulting in unstable detection of the coil end position. In contrast, this position estimation system estimates the coil end position Pe individually using measurement data from multiple viewpoints, and then uses statistical processing to select a set of results with low variability or high density within a predetermined range. Therefore, this position estimation system makes it easier to suppress the decrease in the accuracy of coil position estimation, even if reflective ghosting occurs.

[0098] Here, the measurement data obtained by the distance sensor measuring from multiple positions may be data obtained by a single distance sensor measuring while moving, such as the first sensor 6a. Alternatively, the measurement data obtained by the distance sensor measuring from multiple positions may be data obtained by multiple distance sensors that are pre-installed at different positions, such as the first sensor 6a and the second sensor 6b.

[0099] Furthermore, in the position estimation system according to this embodiment, the control calculation unit 8 divides the multiple virtual waypoints Pv into multiple groups based on their proximity to each other. Next, the control calculation unit 8 identifies the group closest to the position estimation target among the multiple groups based on the calculation results of the virtual waypoints Pv calculated for each of the multiple frames. Then, the control calculation unit 8 may identify the coil end Pe used to estimate the position of the position estimation target based on the candidate end point Pc extracted from the multiple virtual waypoints Pv included in the identified group.

[0100] According to this position estimation system, the control calculation unit 8 identifies the coil end Pe to be ultimately adopted based on the calculation results of virtual passing point Pv calculated for each of the multiple frames. Therefore, even if reflection ghosting occurs, for example, the measurement data affected by reflection ghosting can be excluded in advance, improving the estimation accuracy of the coil end Pe.

[0101] Furthermore, in the position estimation system according to this embodiment, the control calculation unit 8 determines the tangent angle θ when the laser La emitted from the distance sensor comes into contact with the side surface of the position estimation target. t The amount of displacement between the virtual passing point Pv at height H of the coil center Ax and the coil end Pe, caused by this displacement, is determined. The control calculation unit 8 may then estimate the coil end Pe by reflecting this displacement amount in the candidate end point Pc.

[0102] Here, the amount of deviation mentioned above is the tangent angle θ in the example above. t Virtual point shift Se(θ) t This corresponds to ).

[0103] This position estimation system allows for more accurate estimation of the coil end Pe, and in particular, even when estimating only one coil end Pe, it can further reduce the decrease in the accuracy of the final coil position estimation.

[0104] Although several embodiments have been described, it is possible to modify or transform the embodiments based on the above disclosure. All components of the above embodiments, and all features described in the claims, may be taken individually and combined, provided that they do not conflict with each other. [Explanation of symbols]

[0105] 4 Trolley 6a First Sensor 6b Second sensor 8 Control Calculation Unit Ax coil center C coil H coil center height La laser Pc edge candidate point Pe coil end Pm measurement point Pv Virtual Passing Point θ t tangent angle Se(θ t ) Virtual point shift

Claims

1. A position estimation system for estimating the position of a coil placed in a position where its central axis is parallel to the floor surface, A distance sensor is installed on a trolley that suspends and moves the aforementioned coil, and uses a laser to scan and measure the distance to an object on the floor surface, The control calculation unit identifies at least one of the coils placed on the floor surface as a position estimation target, moves the trolley along the cross-sectional direction of the position estimation target and has the distance sensor perform a scan measurement, calculates the point at which the laser passes the height of the center of the coil of the position estimation target as a virtual passing point for each measurement point, extracts the virtual passing point closest to the position estimation target as a candidate end point that could be the coil end of the position estimation target in the cross-sectional direction, and derives the position of the position estimation target based on the coil end estimated from the candidate end point. The control calculation unit estimates the height of the top of the coil, which is the target of position estimation, based on measurement data obtained by the distance sensor measuring from multiple positions, before calculating the virtual waypoint, and derives the height of the center of the coil, which is the target of position estimation, based on the estimated height, in a position estimation system.

2. The position estimation system according to Claim 1, wherein the control calculation unit derives the height of the coil center of the position estimation target before calculating the virtual passing point, and calculates the height of the coil center by subtracting the coil radius, which has been acquired as information in advance, from the estimated height of the apex of the coil.

3. The position estimation system according to claim 1 or 2, wherein the control calculation unit estimates each of the coil ends using the measurement data and performs statistical processing on the estimated values ​​of the plurality of coil ends to identify the coil end to be used for estimating the position of the position to be estimated.

4. The position estimation system according to claim 1 or 2, wherein the control calculation unit divides the plurality of virtual waypoints into a plurality of groups based on their proximity to each other, identifies the group closest to the position estimation target among the plurality of groups based on the calculation results of the virtual waypoints calculated for each plurality of frames, and identifies the coil end to be used for estimating the position of the position estimation target based on the candidate end points extracted from the plurality of virtual waypoints included in the identified group.

5. The position estimation system according to claim 1 or 2, wherein the control calculation unit determines the amount of displacement between the virtual passing point at the height of the coil center and the coil end, which is caused by the tangent angle when the laser irradiated from the distance sensor comes into contact with the side surface of the position estimation target, and estimates the coil end by reflecting the amount of displacement in the candidate end point.

6. A position estimation method for estimating the position of a coil placed in a position where its central axis is parallel to the floor surface, A position estimation target identification step, which involves identifying at least one of the coils placed on the floor surface as a position estimation target, A measurement step in which, while moving the trolley on which the coil is suspended along the cross-sectional direction of the object to be positioned, a distance sensor installed on the trolley is used to scan and measure the distance to an object on the floor side using a laser, A vertex height estimation step is performed to estimate the height of the vertex of the coil, which is the object of position estimation, based on the measurement data obtained in the measurement step, A coil center height calculation step is performed to derive the height of the coil center of the position to be estimated based on the height of the vertex portion estimated in the vertex height estimation step, A virtual waypoint calculation step is performed to calculate, for each measurement point measured in the above measurement step, the point at which the laser passes the height of the coil center of the position estimation target as a virtual waypoint, An end candidate point extraction step in which the virtual passing point on the side closest to the position estimation target is extracted as an end candidate point that can become the coil end in the cross-sectional direction of the position estimation target, A position estimation method comprising: a coil position derivation step of deriving the position of the position to be estimated based on the coil end estimated from the candidate end point.

7. The position estimation method according to claim 6, wherein in the coil center height calculation step, the height of the coil center is calculated by subtracting the coil radius, which has been acquired as information in advance, from the height of the vertex portion estimated in the vertex height estimation step.

Images