JP2024027358A - Pallet position detection method and pallet position detection system - Google Patents

Abstract

An object of the present invention is to detect the position of a pallet with high accuracy. [Solution] A position data acquisition step of emitting light from an optical sensor 14 toward a plurality of column parts 23 and acquiring position data of a plurality of measurement points P before a guided vehicle 10 enters an entry space 24; , an inner point extraction step of extracting an inner point, which is a point on the surface 23a of the column 23 facing the entry space 24, from a plurality of measurement points P; and an inner point extracted in the inner point extraction step. a pallet position calculation step of calculating the position of the pallet 20 based on the position data of the pallet 20. In the inner point extraction step, for each of the plurality of pillar parts 23, a candidate point group for the inner point is selected from a plurality of measurement points P, and it is determined whether the measurement points included in the candidate point group are appropriate as the inner point. Judgment is made based on predetermined conditions. [Selection diagram] Figure 4

Description

The present invention relates to a technique for detecting the position of a pallet transported by a transport vehicle.

For example, in the premises of a steelworks, it has been conventional practice to place heavy items such as large coils and steel materials on pallets, and to transport these pallets using transport vehicles. A typical pallet has a structure in which a platform on which a heavy object is placed is supported by left and right legs. Then, by entering the carrier into the entry space between the left and right legs of the pallet and then raising the loading platform of the carrier, it is possible to transport the pallet with the pallet supported from below by the loading platform. It has become.

In recent years, transportation vehicles have been automated, and for this purpose, it is necessary to accurately detect the position of the pallet so that the transportation vehicle can automatically and appropriately enter the pallet entry space. Usually, the legs of a pallet are made of a truss structure or a rigid frame structure made of section steel, and have a plurality of columns. In Patent Document 1, light is irradiated toward these pillars from an optical sensor (approach sensor) provided on a transport vehicle, and position data of measurement points irradiated with light is acquired based on the reflected light. are doing. Then, an inner point (a point on the surface of the column facing the entry space) is extracted from among the plurality of measurement points, and the position of the pallet is calculated based on the position data of the inner point.

Specifically, in Patent Document 1, the interior normal points are extracted by the following steps 1 to 4.
1. Calculate distances for all combinations of measurement points on the left pillar and measurement points on the right pillar 2. 3. Let P(i) be the number of the measurement point on the left column that is the shortest distance from the measurement point A(i) on the right column. 4. Let Q(j) be the number of the measurement point on the right column that is the shortest distance from the measurement point B(j) on the left column. For measurement point number i on the right column, if the formula i=Q(P(i)) holds, then measurement point A(i) and measurement point B(P(i)) are treated as a pair of internal points. Extraction Hereinafter, this process of extracting internal points will be referred to as pairing.

Patent No. 4427360

When extracting inner points by pairing as in Patent Document 1, there are the following problems. For example, if the position of the optical sensor is shifted from the center of the pallet in the left-right direction, the number of measurement points may be significantly different between the left column and the right column. The upper limit of the number of pairs of internal points that can be extracted by pairing is the number of measurement points in the column section with fewer measurement points. For this reason, for the column with a larger number of measurement points, it is not possible to effectively utilize the position data of all measurement points, and the number of internal point data that can be used to calculate the pallet position is limited. It becomes less.

Furthermore, in a column part far away from the optical sensor, the number of measurement points tends to decrease due to the angle at which light is irradiated. In pairing, if a pair of left and right measurement points with the minimum distance from each other is found, it is extracted as an inner point. Therefore, if pairing is performed with a small number of measurement points, a pair of measurement points with the minimum distance from each other will be extracted as inner points, even if they are not actually inner points.

Due to the above-mentioned problems, when detecting the position of the pallet based on pairing, the error may become large when detecting the position of the pallet due to the small amount of data or incorrect extraction of the internal point. .

The present invention has been made in view of the above-mentioned problems, and an object of the present invention is to enable the position of a pallet to be detected with high accuracy.

The pallet position detection method according to the present invention allows the pallet, which is configured to support a platform with left and right legs having a plurality of pillars, to enter an entry space between the left and right legs. A method for detecting the position of a pallet by using an optical sensor provided on a transport vehicle, the method comprising: detecting the position of a pallet by using an optical sensor provided on a transport vehicle, wherein before the transport vehicle enters the entry space, light is emitted from the optical sensor toward the plurality of pillars. a positional data acquisition step of irradiating the beam and acquiring positional data of a plurality of measurement points; and extracting an internal point, which is a point on a surface of the column facing the entry space, from the plurality of measurement points. a normal point extraction step; and a pallet position calculation step of calculating the position of the pallet based on the position data of the internal normal point extracted in the internal normal point extraction step, the internal normal point extraction step Then, for each of the plurality of pillar parts, a candidate point group for the inner normal point is selected from the plurality of measurement points, and a predetermined determination is made as to whether the measurement point included in the candidate point group is appropriate as the inner normal point. It is characterized by making judgments based on conditions.

A pallet position detection system according to the present invention is a pallet position detection system that detects the position of a pallet configured to support a base with left and right legs having a plurality of pillars, an optical sensor provided on a carrier that can enter an entry space between the legs; and a control device that uses the optical sensor to calculate the position of the pallet, and the control device a position data acquisition process of acquiring position data of a plurality of measurement points based on light irradiated from the optical sensor toward the plurality of column parts before entering the entry space; An internal point extraction process that extracts an internal point that is a point on a surface of the column facing the approach space from a point, and the position data of the internal point extracted by the internal point extraction process. pallet position calculation processing for calculating the position of the pallet based on the above, and in the inner normal point extraction processing, for each of the plurality of column parts, the inner normal point is calculated from the plurality of measurement points. The method is characterized in that a group of candidate points is selected, and it is determined based on a predetermined condition whether the measurement point included in the group of candidate points is appropriate as the inner point.

According to the present invention, when extracting an inner point from a plurality of measurement points, first, a group of candidate points for the inner point is selected for each column, and then the measurement points included in each candidate point group are selected from the inner point. It is determined whether the point is valid or not based on predetermined conditions. In other words, when extracting internal points, pairing is not performed between the left and right pillars, but whether or not the candidate point group of each pillar is appropriate as an internal point is determined individually. Therefore, even if the number of measurement points differs between the left and right columns, the amount of usable data does not decrease, and the acquired data can be utilized to the fullest. Furthermore, when pairing is performed with a small number of measurement points, it is possible to avoid the problem of erroneous measurement points being extracted as inner points. Therefore, the position of the pallet can be detected with high accuracy.

In the present invention, the predetermined condition may be such that the distance between the two most distant measurement points among the measurement points included in the candidate point group is equal to or greater than a predetermined value.

When calculating the position of the pallet based on the position data of the inner points, for example, the inner normal line (straight line connecting the inner points) may be approximately determined from the measurement points extracted as the inner points. . When determining the inner normal line approximately in this way, it is preferable that the measurement points be appropriately scattered, and if the measurement points are closely spaced, errors tend to increase during approximation. Therefore, by setting the above-mentioned conditions, it is possible to ensure that the measurement points are scattered to some extent, and the inner normal line can be determined with high accuracy.

In the present invention, the predetermined condition may be that the number of measurement points included in the candidate point group is equal to or greater than a predetermined number.

As described above, when the inner normal line is approximately determined from the measurement points extracted as the inner normal points, if the number of measurement points is too small, errors tend to increase during the approximation. Therefore, by providing the above-mentioned conditions, a sufficient number of measurement points can be secured and the inner normal line can be determined with high accuracy.

In the present invention, it is preferable that a first region including a pillar is set for each of the plurality of pillars based on the position information of the pallet, and the candidate point group is selected from the first region.

In this way, by limiting the range from which the candidate point group is selected to the first region in advance, measurement points unrelated to the column can be excluded from the calculation, and the calculation load can be reduced.

In the present invention, a second region is set to include the measurement point closest to the approach space among the measurement points included in the first region, and the measurement points included in the second region are set to include the measurement points included in the candidate point group. It is best to select it as

In this way, it is possible to easily select a measurement point that is likely to be an inner point from among the measurement points included in the first region.

In the present invention, the optical sensor may be disposed at a position offset from the center of the carrier in the width direction of the carrier.

Since the present invention does not require pairing, it is only necessary to obtain a sufficient number of measurement points from either the left or right column parts. As described above, by deliberately shifting the position of the optical sensor, there is an advantage that many measurement points can be obtained for at least one of the left and right pillar sections.

According to the present invention, the position of the pallet can be detected with high accuracy by extracting the inner point without performing pairing.

It is a perspective view showing the appearance of a carrier and a pallet. 3 is a flowchart showing a pallet position detection process. 3 is a flowchart showing a pallet position detection process. FIG. 3 is a schematic plan view for explaining pallet position detection processing. FIG. 3 is a schematic plan view for explaining pallet position detection processing. FIG. 3 is a schematic plan view for explaining pallet position detection processing. FIG. 3 is a schematic plan view for explaining pallet position detection processing. FIG. 2 is a schematic plan view showing inner points extracted in a conventional technique using pairing.

Embodiments of the present invention will be described below with reference to the drawings.

(transport vehicle)
FIG. 1 is a perspective view showing the appearance of the transport vehicle 10 and the pallet 20. Transport vehicle 10 is a vehicle for transporting pallets 20. The transport vehicle 10 is configured to be able to travel automatically based on the position information of the vehicle based on the Global Navigation Satellite System (GNSS). The transport vehicle 10 includes a driver's cab 11, a plurality of wheels 12, a loading platform 13, an optical sensor 14, and a control device 15. The optical sensor 14 and the control device 15 constitute a "pallet position detection system" according to the present invention. In order to enable the guided vehicle 10 to travel automatically, instead of or in addition to the method of using GNSS, a method of using an inertial measurement unit (IMU), a method of image recognition of a predetermined tag, a method of magnetic A method using a marker (RFID: Radio Frequency Identifier), a method using the optical sensor 14 to detect the position of the vehicle, etc. may be adopted.

The driver's cabin 11 is a room in which the driver gets in and operates the transport vehicle 10 when the driver needs to operate the transport vehicle 10 . For example, the driver's cab 11 is used when there is a problem with the automatic travel function of the guided vehicle 10. In addition, in this embodiment, the case where the guided vehicle 10 is automatically traveling unmanned is explained, and it is assumed that the driver's cab 11 is not particularly used. Further, it is not essential to provide the driver's cab 11, and it is also possible to omit the driver's cab 11.

The loading platform 13 is configured to be movable up and down, and can support the pallet 20 in a lifted state from below.

The optical sensor 14 can measure the distance to an object by emitting light from a projector (not shown) toward an object and receiving reflected light from the object with a receiver (not shown). It is a sensor. Optical sensors 14 are provided on both the front and rear sides of the transport vehicle 10, and one of the optical sensors 14 is used depending on the situation. The optical sensor 14 is arranged at the center of the transport vehicle 10 in the width direction, and can scan a wide range horizontally in an arc. Note that although the optical sensor 14 of this embodiment uses laser light, it is also possible to use a sensor that uses light other than laser light. Further, the optical sensor 14 may be placed on the loading platform 13.

The control device 15 performs various controls regarding the transport vehicle 10. The control device 15 includes an automatic travel control section 16, a pallet position calculation section 17, and a storage section 18. The automatic travel control unit 16 executes automatic travel of the guided vehicle 10 by appropriately controlling a steering device, a drive device, etc. (all not shown). The pallet position calculation unit 17 calculates the position of the pallet 20 based on data acquired by scanning with the optical sensor 14, as will be described in detail later. The storage unit 18 stores various data (for example, map information, size information of the pallet 20, etc.) required for calculation processing by the automatic travel control unit 16 and the pallet position calculation unit 17.

(palette)
The pallet 20 has a configuration in which a platform 21 on which a heavy object is placed is supported by left and right legs 22. The leg portion 22 has a truss structure using H-shaped steel, and has a plurality of pillar portions 23 arranged in front and behind. However, it is not essential that the leg portions 22 have a truss structure of H-shaped steel; other steel sections may be used, or a rigid frame structure may be used instead of a truss structure. An entry space 24 into which the transport vehicle 10 can enter is formed between the left and right legs 22 . In other words, the approach space 24 is a space surrounded by the platform 21, the left and right legs 22, and the ground. The transport vehicle 10 enters the entry space 24 with the loading platform 13 lowered, and by raising the loading platform 13 after entering, the platform 21 of the pallet 20 can be lifted from below by the loading platform 13. As a result, the pallet 20 and the heavy objects placed on the pallet 20 can be transported by the transport vehicle 10.

In order for the transport vehicle 10 to appropriately enter the entry space 24 automatically, it is necessary to accurately detect the position of the pallet 20. The transport vehicle 10 detects the position of the pallet 20 by using the optical sensor 14 to detect the left and right legs 22 of the pallet 20, more specifically, the left and right columns 23. Below, the position detection process of the pallet 20 will be explained in detail.

(Pallet position detection processing)
2 and 3 are flowcharts showing a process for detecting the position of the pallet 20. 4 to 7 are schematic plan views for explaining the position detection process of the pallet 20. FIG. Note that in FIGS. 4 to 7, only four pillars 23 are shown for each of the left and right legs 22 due to space limitations. Furthermore, to simplify the explanation, the number of measurement points P illustrated in FIGS. 4 to 7 is smaller than the actual number.

The automatic travel control unit 16 directs the transport vehicle 10 toward the pallet 20 to be transported (hereinafter simply referred to as the target) based on the position information of the vehicle based on GNSS, map information stored in the storage unit 18, etc. Make it run automatically. Then, when the transport vehicle 10 reaches a predetermined position near the target pallet 20, the automatic travel control unit 16 stops the transport vehicle 10. This predetermined position is set as a position where laser light can be irradiated onto each of the left and right columns 23 of the target pallet 20 by the optical sensor 14 provided on the transport vehicle 10. The following pallet 20 position detection processing (processing in steps S1 to S13) is executed by the pallet position calculation unit 17.

When the transport vehicle 10 stops at a predetermined position, scanning by the optical sensor 14 is started. The pallet position calculation unit 17 acquires position data (position coordinates) of each measurement point P (see FIG. 4) irradiated with laser light (step S1, corresponding to the position data acquisition step and position data acquisition process of the present invention). ). As shown in FIG. 4, scanning by the optical sensor 14 is performed by horizontally rotating the optical sensor 14 within a range of 180 degrees forward and emitting laser light at regular intervals (equal angular intervals). The positional coordinates of each measurement point P can be calculated from the distance from the optical sensor 14 to the measurement point P and the irradiation angle of the laser beam. Note that the calculation of the position coordinates of each measurement point P may be performed by a controller (not shown) of the optical sensor 14 instead of the pallet position calculation unit 17.

Here, each column portion 23 made of H-shaped steel has a flange surface 23a on the side of the entrance space 24, a flange surface 23b on the opposite side of the entrance space 24, and a web surface 23c connecting the flange surfaces 23a and 23b. In the position detection process of the pallet 20 of this embodiment, the position of the pallet 20 is calculated based on the position data of the inner point, which is a point on the flange surface 23a facing the entry space 24. For this reason, it is necessary to extract the inner point from a plurality of measurement points P. In steps S2 to S11, each process for extracting an inner mean point from a plurality of measurement points P is performed, and corresponds to the inner mean point extraction step and the inner mean point extraction process of the present invention.

Of the measurement points P, the measurement point P0 (see FIG. 5) is a measurement point obtained by the laser beam of the optical sensor 14 not being irradiated onto each column part 23 but being irradiated onto an object other than the target pallet 20. It is. Such position data of the measurement point P0 is not particularly necessary for detecting the position of the pallet 20. Therefore, as shown in FIG. 5, the pallet position calculation unit 17 sets, for each column 23, a first area A1 including the column 23 (step S2). Note that step S2 may be executed before step S1. Each first area A1 is set to be larger than the pillar part 23 so that it can reliably include the pillar part 23 based on the position information of the pallet 20 by GNSS and the dimension information of the pallet 20 stored in the storage unit 18. It is set wide.

Eliminate measurement points P0 (indicated by white circles in FIG. 5) that are not included in each first area A1, and target only measurement points P1 (indicated by black circles in FIG. 5) included in each first area A1. Therefore, the calculation load can be reduced. Note that, since the storage location of the pallet 20 is basically fixed in many cases, previously set data may be used for the first area A1. In preparation for such a case, setting data regarding the first area A1 may be stored in the storage unit 18, and this setting data may be used from next time onwards. Alternatively, once the position of the pallet 20 has been successfully detected, the second and subsequent settings of the first area A1 may be performed based on the result of the previous position detection of the pallet 20.

Extraction of the inner normal point is performed by sequentially searching each first area A1. If there are still unsearched first areas A1 (NO in step S3), one first area A1 is selected from among the unsearched first areas A1 (step S4).

As shown in FIG. 6, the pallet position calculation unit 17 sets the second area A2 to include the measurement point P1 closest to the entrance space 24 within the selected first area A1 (step S5), and A point group consisting of the measurement points P1 in the area A2 is set as a candidate point group G that is a candidate for the inner normal point (step S6). In FIG. 6, among the measurement points P1 in the first area A1, the measurement points included in the candidate point group G are denoted by P2, and are indicated by black circles.

The measurement points P1 in the first area A1 include not only the inner point on the flange surface 23a but also the points on the flange surface 23b and the points on the web surface 23c. Therefore, among the measurement points P1 in the first area A1, the candidate point group G is set by setting the second area A2 based on the measurement point P1 closest to the approach space 24, which is most likely to be an internal point. It becomes easier to select efficiently. In the present embodiment, the second area A2 is set as a rectangular area with predetermined lengths and widths centered on the measurement point P1 closest to the entry space 24. For example, the vertical (front and rear) length is about 2.5 times Lmax (the front and back length of the flange surface 23a), and the horizontal (left and right) length is about 0. It may be about 5 times as large. However, how the second area A2 is set can be changed as appropriate.

Next, the pallet position calculation unit 17 determines whether the number N of measurement points P2 included in the candidate point group G is greater than or equal to a predetermined number N1 (N1=3 in this embodiment) (step S7). If N≧N1 (YES in step S7), then the maximum distance L between the measurement points P2 included in the candidate point group G, that is, the most distant measurement point P2 included in the candidate point group G. It is determined whether the distance L between the two measurement points P2 is equal to or greater than a predetermined value L1 (L1=Lmax/2 in this embodiment) (step S8). If L≧L1 (YES in step S8), the measurement point P2 included in the candidate point group G is adopted as the inner point (step S9), and will be used later in the position calculation of the pallet 20. . In FIG. 7, among the measurement points P2 included in the candidate point group G, the measurement points adopted as the inner normal points are designated by the symbol P3, and are indicated by black circles.

If N<N1 (NO in step S7) or L<L1 (NO in step S8), the measurement point P2 included in the candidate point group G is not adopted as the inner point. (Step S10) is not used in the position calculation of the pallet 20. The conditions N≧N1 and L≧L1 correspond to predetermined conditions of the present invention, respectively. Note that the above N1 and L1 are only examples, and can be changed as appropriate.

When the processing in steps S4 to S10 is completed, the pallet position calculation unit 17 determines that the first area A1 has been searched (step S11), and returns to step S3. If an unsearched first area A1 remains (NO in step S3), steps S4 to S11 are repeated for the unsearched first area A1. On the other hand, if all the first areas A1 have been searched (YES in step S3), the inner normal line (straight line connecting the inner normal points) is calculated as an approximate straight line from the measurement point P3 adopted as the inner normal point. (Step S12), and the position of the pallet 20 is calculated based on the normal line therein (Step S13). Steps S12 and S13 correspond to a pallet position calculation step and a pallet position calculation process of the present invention. The automatic travel control unit 16 performs automatic travel control so that the guided vehicle 10 can appropriately enter the entry space 24 while referring to the position of the pallet 20 calculated by the pallet position calculation unit 17.

(effect)
In this embodiment, when extracting the inner normal point from a plurality of measurement points P, first, after selecting the candidate point group G of the inner normal point for each column 23, the measurement points included in each candidate point group G are selected. It is determined based on predetermined conditions whether P2 is appropriate as an internal point. In other words, when extracting the internal points, pairing is not performed between the left and right pillars 23, but whether the candidate point group G of each pillar 23 is appropriate as an internal point is determined individually. Therefore, even if the number of measurement points P differs between the left and right pillar sections 23, the amount of usable data does not decrease, and the acquired data can be utilized to the fullest. Furthermore, when pairing is performed with a small number of measurement points P, it is possible to avoid the problem of erroneous measurement points being extracted as inner points. Therefore, the position of the pallet 20 can be detected with high accuracy.

This will be specifically explained with reference to FIGS. 7 and 8. In FIG. 7, the measurement point P3 extracted as the inner standard point in this embodiment is illustrated by a black circle. On the other hand, in FIG. 8, when the measurement point P is obtained as in FIG. 4, the measurement point P4 extracted as the inner point by conventional pairing is illustrated by a black circle.

As is clear from FIG. 8, by performing the pairing, only the same number of measurement points P4 are extracted as inner normal points on the left and right pillar sections 23. For example, in the first to third pillars 23 from the bottom on the right side of FIG. 8, there are fewer measurement points on the left pillar 23 that are paired, so the measurement point P4 is adopted as the inner point. It's limited. On the other hand, in this embodiment, as shown in FIG. 7, since the inner point is extracted individually for each column 23, many measurement points P3 can be employed as the inner point. Furthermore, in the past, as long as pairing was established, the point was adopted as the inner point, so for example, in the column 23 at the top right in FIG. It has been adopted. On the other hand, in the present embodiment, by imposing a predetermined condition, it is possible to prevent inappropriate measurement points from being adopted as inner measurement points.

In this embodiment, the predetermined condition is that the distance L between the two most distant measurement points P2 among the measurement points P2 included in the candidate point group G is equal to or greater than a predetermined value L1. When determining the inner normal line approximately from the measurement points P3 extracted as inner normal points, it is preferable that the measurement points P3 are appropriately scattered, and if the measurement points P3 are crowded together, errors may occur during approximation. tends to become large. Therefore, by setting the above-mentioned conditions, it is possible to ensure that the measurement points P3 are scattered to some extent, and the inner normal line can be determined with high accuracy.

In this embodiment, the predetermined condition is that the number N of measurement points P2 included in the candidate point group G is greater than or equal to the predetermined number N1. When the inner normal line is approximately determined from the measurement points P3 extracted as inner normal points, if the number of measurement points P3 is too small, errors tend to increase during the approximation. Therefore, by providing the above-mentioned conditions, a sufficient number of measurement points P3 can be ensured, and the inner normal line can be determined with high accuracy.

In this embodiment, a first area A1 including the pillar part 23 is set for each of the plurality of pillar parts 23 based on the position information of the pallet 20, and a candidate point group G is selected from the first area A1. In this way, by limiting the range from which the candidate point group G is selected to the first area A1 in advance, the measurement point P0 that is unrelated to the pillar portion 23 can be excluded from the calculation, and the calculation load can be reduced. Can be done.

In this embodiment, the second area A2 is set to include the measurement point P1 closest to the approach space 24 among the measurement points P1 included in the first area A1, and the measurement point P2 included in the second area A2 is selected as a candidate. Select it as a point group G. In this way, it is possible to easily select the measurement point P2 that is likely to be an inner point from among the measurement points P1 included in the first area A1.

(Other embodiments)
A modification example in which various changes are made to the above embodiment will be described.

In the embodiment described above, the first area A1 and the second area A2 are set, and the candidate point group G is selected from the plurality of measurement points P. However, in order to select the candidate point group G, it is not essential to set the first area A1 and the second area A2, and the candidate point group G may be selected by other methods.

In the above embodiment, two conditions, L≧L1 and N≧N1, are imposed as predetermined conditions for determining whether it is appropriate to adopt the measurement point P2 included in the candidate point group G as an inner standard point. did. However, as the predetermined condition, only one of the two conditions L≧L1 and N≧N1 may be imposed, or other conditions may be adopted.

In the embodiment described above, it is assumed that the control device 15 provided in the transport vehicle 10 performs control regarding automatic travel and detects the position of the pallet 20. However, a host system for controlling the operation of the transport vehicle 10 may be provided separately, and this host system may control automatic travel and/or detect the position of the pallet 20. In this case, the control device 15 may be configured to receive map information, size information of the pallet 20, etc. from the host system on time.

In the embodiment described above, after all the first areas A1 have been searched, the inner normal line is calculated, and the guided vehicle 10 is caused to enter the pallet 20 based on the inner normal line. However, for example, only several first areas A1 from the front are searched, the inner normal line is calculated from the result, the transport vehicle 10 is moved forward a little, and then the first area A1 on the back side is searched. You can do it like this.

In the embodiment described above, the process of detecting the position of the pallet 20 is performed in a state where the carrier 10 is offset from the center of the pallet 20 in the width direction, that is, when the optical sensor 14 is offset from the center of the pallet 20 in the width direction. An example was explained. In this case, as is clear from FIG. 4, due to the angle of the laser beam irradiated from the optical sensor 14, the number of measurement points P obtained on one (left) column 23 is reduced, but on the other (right) In the column part 23, the number of measurement points P obtained increases. As a result, for the right column portion 23, many inner normal points could be extracted and the inner normal line could be calculated more accurately. In order to achieve the same effect as this, the optical sensor 14 may be arranged not at the center of the transport vehicle 10 in the width direction but at a position offset from the center. By intentionally shifting the position of the optical sensor 14 in the width direction, it becomes easier to extract many inner points as described above for at least one of the left and right pillar portions 23. If at least one of the left and right inner normals can be calculated with high accuracy, the position detection accuracy of the pallet 20 can be improved.

10: Transport vehicle 15: Control device 20: Pallet 21: Platform 22: Legs 23: Column 24: Entry space P, P1, P2, P3: Measurement point A1: First area A2: Second area

Claims (7)

The position of a pallet configured to support a platform with left and right legs having a plurality of pillars is determined using an optical sensor installed on a carrier that can enter the space between the left and right legs. A method for detecting the position of a pallet, the method comprising:
Before the guided vehicle enters the entry space, a position data acquisition step of irradiating light from the optical sensor toward the plurality of pillars and acquiring position data of the plurality of measurement points;
an inner point extraction step of extracting an inner point, which is a point on a surface of the column facing the entry space, from the plurality of measurement points;
a pallet position calculation step of calculating the position of the pallet based on the position data of the inner point extracted in the inner point extraction step;
Equipped with
In the inner normal point extraction step, for each of the plurality of column parts, a candidate point group for the inner normal point is selected from the plurality of measurement points, and the measurement point included in the candidate point group is selected as the inner normal point. A pallet position detection method characterized by determining whether or not it is appropriate based on predetermined conditions. 2. The pallet according to claim 1, wherein the predetermined condition is that a distance between two farthest measurement points among the measurement points included in the candidate point group is a predetermined value or more. Location detection method. 2. The pallet position detection method according to claim 1, wherein the predetermined condition is that the number of the measurement points included in the candidate point group is greater than or equal to a predetermined number. Claims 1 to 3, characterized in that a first region including a pillar is set for each of the plurality of pillars based on positional information of the pallet, and the candidate point group is selected from the first region. The method for detecting the position of a pallet according to any one of the above. A second area is set to include a measurement point closest to the approach space among the measurement points included in the first area, and the measurement points included in the second area are selected as the candidate point group. The pallet position detection method according to claim 4. A pallet position detection system that detects the position of a pallet configured to support a platform with left and right legs having a plurality of pillars,
an optical sensor provided on a carrier that can enter the entry space between the left and right legs;
a control device that calculates the position of the pallet using the optical sensor;
Equipped with
The control device includes:
Before the guided vehicle enters the entry space, a position data acquisition process that acquires position data of a plurality of measurement points based on light irradiated from the optical sensor toward the plurality of pillars;
An inner point extraction process that extracts an inner point, which is a point on a surface of the column facing the entry space, from the plurality of measurement points;
pallet position calculation processing that calculates the position of the pallet based on the position data of the inner normal point extracted in the inner normal point extraction process;
is executable,
In the inner normal point extraction process, for each of the plurality of column parts, a candidate point group for the inner normal point is selected from the plurality of measurement points, and the measurement point included in the candidate point group is selected as the inner normal point. A pallet position detection system that determines whether or not it is appropriate based on predetermined conditions. 7. The pallet position detection system according to claim 6, wherein the optical sensor is arranged at a position offset from the center of the carrier in the width direction of the carrier.

Images