WO2010106573A1

WO2010106573A1 - Target ic tag discrimination system, target ic tag discrimination apparatus, target ic tag discrimination method and computer program

Info

Publication number: WO2010106573A1
Application number: PCT/JP2009/001190
Authority: WO
Inventors: 英人木原; 博靖菅野; 一樹松井
Original assignee: 富士通株式会社
Priority date: 2009-03-17
Filing date: 2009-03-17
Publication date: 2010-09-23
Also published as: JPWO2010106573A1; JP5196007B2

Abstract

Whether an IC tag is a target to be read or not is determined more surely than conventionally done. An actual read result is obtained as a result of reading an IC tag on a path a plurality of times (#12), an unnecessary tag read result which is a result of reading an IC tag assumed to be outside the path a plurality of times is estimated on the basis of the actual read result (#13), a typical read result which is a typical result obtained when an IC tag on the path is read a plurality of times is prepared (#14), and then first similarity which is the degree of similarity between the unnecessary tag read result and the actual read result is calculated and second similarity having a value higher than that of the first similarity is calculated. An object read result which is a result of reading an IC tag to be discriminated a plurality of times is obtained (#16), third similarity which is the degree of similarity between the object read result and the typical read result is calculated (#17), and then if the third similarity is higher than the second similarity, the IC tag to be discriminated is determined as a target (if Yes in #18, #19) and if the third similarity is lower than the second similarity, the IC tag to be discriminated is not determined as a target (if No in #18, #21).

Description

Objective IC tag discrimination system, objective IC tag discrimination apparatus, objective IC tag discrimination method, and computer program

The present invention relates to a system and method for determining whether or not an IC tag is an object.

In recent years, RFID (Radio Frequency Identification) has been used in various fields. RFID is a technique for reading data including unique identification information stored in an IC (Integrated Circuit) tag or writing data in an IC tag by wireless communication.

IC tags include an active IC tag with a built-in battery and self-supplying power, and a passive IC tag that operates by obtaining power from high-frequency radio waves transmitted from a reader / writer device. Since the passive IC tag does not have a built-in battery, it is offered at a lower price than the active IC tag. Therefore, it is expected to be used in various fields including the logistics field.

Several frequency bands are used for RFID depending on the application, but when UHF (Ultrahigh Frequency) band, that is, 860 to 960 MHz band is used as the frequency band, the reading range is different even if it is passive type. It is wider than the frequency band and can read multiple tags at once. For this reason, for example, in the physical distribution field, a plurality of tags attached to a large number of articles can be collectively read and inspected.

However, if the reading range is widened, IC tags that are not intended by the administrator may also enter the range, and unnecessary data may be read.

For example, when receiving and shipping in a warehouse, etc., an electromagnetic wave reflected by chance on a forklift that has passed near the article arrives at the IC tag of the article that is usually placed away from the gate and cannot be read. It is possible that data is read from the IC tag. In addition, when a plurality of gates are juxtaposed, data may be read from an IC tag of an article that has entered the adjacent gate.

In order to solve such a problem, it is conceivable to eliminate data read from an IC tag having an irrelevant ID by performing filtering based on the ID read from the IC tag.

For example, if the ID is hierarchized according to the type of item (whether it is a pallet or an individual item), if you know the hierarchical structure in advance, depending on what item you want to read, Filtering can be performed.

A method is also well known in which multiple readings are executed intensively and data that cannot be read continuously for a predetermined number of times or more are excluded as accidental readings (see Patent Document 1).
JP 2005-275960 A

However, if the IC tag that is to be read and the IC tag that is not to be read are attached to the same type of article, the necessity determination based on the above hierarchy cannot be performed. Moreover, the environment which can apply the method of filtering by whether it was able to detect continuously is limited.

In view of such problems, the present invention has an object to make it possible to more reliably determine whether or not an IC tag is a target object than before.

A target IC tag discrimination system according to the present invention is a target IC tag discrimination system for discriminating whether or not a first IC tag is a target object, and is a first location that is a predetermined location for reading. An actual reading result acquisition means for transmitting a signal to the second IC tag at a plurality of times and acquiring an actual reading result as a result of success or failure of receiving a response from the second IC tag for the signal of each time And when a signal is transmitted a plurality of times to a third IC tag assumed to be in a second location other than the first location, the signal from the third IC tag for each signal is transmitted from the third IC tag. Obtained when unnecessary tag read result estimation means for estimating an unnecessary tag read result as a result of success or failure of response reception based on the actual read result, and when a signal is transmitted to the second IC tag a plurality of times. Typical for each signal of interest A typical reading result storage means for storing a typical reading result that is a result of success or failure of reception of a response from the second IC tag, and a degree of similarity between the unnecessary tag reading result and the actual reading result. A first similarity calculation unit that calculates one similarity, a second similarity calculation unit that calculates a second similarity that is higher than the first similarity, and transmits a signal multiple times And a target reading result acquisition means for acquiring a target reading result that is a success or failure of receiving a response from the first IC tag for each signal, and a degree of similarity between the target reading result and the typical reading result And a third similarity calculating means for calculating a third similarity, and if the third similarity is higher than the second similarity, the first IC tag is determined to be the object. And the third similarity is lower than the second similarity. Said first IC tag is determined that the non-target product has an attribute discriminating means.

Preferably, the actual reading result acquisition unit alternately transmits a first signal that is a first intensity signal and a second signal that is a second intensity signal lower than the first intensity. The actual reading result is acquired, and the unnecessary tag reading result estimation unit is configured to output the third signal when the first signal and the second signal are alternately transmitted as the unnecessary tag reading result. When the response from the IC tag is received, the typical reading result storage means transmits the first signal and the second signal alternately as the typical reading result. The result of success or failure of reception of the response from the second IC tag is stored.

According to the present invention, it is possible to more reliably determine whether or not an IC tag is an object than before.

It is a figure which shows the example of the whole structure of a receipt / shipment management assistance system. It is a figure which shows the example of the hardware constitutions of IC tag reader / writer. It is a figure which shows the example of the hardware constitutions of a management apparatus. It is a figure which shows the example of a functional structure of a management apparatus. It is a figure which shows the example of an inventory database. It is a figure which shows the example of reading result data. It is a figure which shows the example of the positional relationship of an unnecessary tag, IC tag reader / writer, and a trolley | bogie, and the range which a search command reaches. It is a figure which shows the example of an assumption pattern. It is a figure which shows the example of a reading pattern. It is a figure which shows the example of offset. It is a figure which shows the example of the matching of the elements of both patterns. It is a flowchart explaining the example of the flow of a whole similarity calculation process. It is a flowchart explaining the example of the flow of a matching process. It is a flowchart explaining the example of the flow of a temporary similarity calculation process. It is a flowchart explaining the example of the flow of a temporary similarity calculation process. It is a figure which shows the example of a structure of a parameter adjustment part. It is a flowchart explaining the 1st example of the flow of an initial threshold value calculation process. It is a flowchart explaining the 2nd example of the flow of an initial threshold value calculation process. It is a figure which shows the example of an unnecessary pattern. It is a flowchart explaining the example of the flow of a 1st unnecessary pattern production | generation process. It is a flowchart explaining the example of the flow of a 2nd unnecessary pattern production | generation process. It is a flowchart explaining the example of the flow of a 3rd unnecessary pattern production | generation process. It is a flowchart explaining the example of the flow of a 4th unnecessary pattern production | generation process. It is a flowchart explaining the example of the flow of a threshold value adjustment process. It is a flowchart explaining the example of the flow of a reset process. It is a figure which shows the example of reading pattern Q. It is a flowchart explaining the example of the flow of adjustment processes, such as a parameter. It is a flowchart explaining the example of the flow of the whole process in a management apparatus.

1 is a diagram illustrating an example of the overall configuration of a receipt / shipment management support system SYS, FIG. 2 is a diagram illustrating an example of a hardware configuration of an IC tag reader / writer 2, and FIG. 3 is an example of a hardware configuration of the management device 1. FIG. 4 is a diagram illustrating an example of a functional configuration of the management apparatus 1, and FIG. 5 is a diagram illustrating an example of an inventory database 1DB.

The receipt / shipment management support system SYS is a system for supporting the management of the presence / absence (ie, inventory) of luggage in a warehouse or the like, and as shown in FIG. 1, a management device 1, an IC (Integrated Circuit) tag reader / writer 2 And a passage sensor 3. Hereinafter, the case where the receipt / shipment management support system SYS is used in the warehouse X at the relay point where the packages shipped from various places are relayed to the destination will be described as an example.

Luggage is stored in one of the containers (pallets, containers) 4 depending on the recipient or destination. One or a plurality of containers 4 are placed on a cart (cart) 6 and are loaded or unloaded.

One IC tag 5 is attached to one container 4. A unique ID (Identification) is recorded in the IC tag 5. In the present embodiment, a passive IC tag is used as the IC tag 5.

The IC tag reader / writer 2 is a device that reads data from the IC tag 5 or writes data to the IC tag 5, and one IC tag reader / writer 2 is installed near the entrance gate and the exit gate of the warehouse X.

As shown in FIG. 2, the IC tag reader / writer 2 includes a control device 20a, a RAM (Random Access Memory) 20b, a ROM (Read Only Memory) 20c, a communication device 20d, a wireless transmission / reception device 20e, an antenna 20f, and the like. Each unit shown in FIG. 2 is connected to each other through a bus or the like.

The ROM 20c stores a program for controlling each part of the IC tag reader / writer 2 so as to read data from the IC tag 5.

The RAM 20b is an SRAM (Static RAM) or a flash memory. A program stored in the ROM 20c is appropriately read into the RAM 20b. In addition, the RAM 20b temporarily stores data necessary for executing the program and data (such as read data and various parameters) generated by executing the program.

The control device 20a is a CPU (Central Processing Unit) or MPU (Micro Processing Unit), and executes the program read to the RAM 20b. If the control device 20a is an MPU, the RAM 20b and the ROM 20c may be integrated into the control device 20a.

The communication device 20d is a device for communicating with the management device 1 via a network. In particular, as will be described later, it is used for receiving a command from the management apparatus 1 and transmitting data read from the IC tag 5 to the management apparatus 1. As the communication device 20d, a wired or wireless LAN (Local Area Network) apparatus or the like is used. Alternatively, an interface device such as USB (Universal Serial Bus) or IEEE (Institute Electrical Electronic Electronics Engineer) 1394 may be used.

The wireless transmission / reception device 20e is a device for transmitting a command to the IC tag 5 wirelessly and receiving a response thereto. In the present embodiment, it is particularly used for transmission / reception of a search command for reading an ID from the IC tag 5 and a response to the search command. Reading of the ID from the IC tag 5 is performed as follows.

The wireless transmission / reception device 20e searches (inventory) for the IC tag 5 existing in a range where communication with the antenna 20f is possible. At this time, a search command is transmitted. When the IC tag 5 receives the search command radio wave, a current is generated, a voltage is supplied to each part of the IC tag 5, and its own ID is transmitted as a response to the search command. The wireless transmission / reception device 20e receives this ID.

When there are a plurality of IC tags 5 in a range where communication by the antenna 20f is possible, these IC tags 5 transmit responses to search commands (that is, IDs) almost simultaneously. In this case, a situation (collision) in which the wireless transmission / reception device 20e cannot receive a response from the IC tag 5 may occur due to interference between the responses. In order to avoid this, various types of collision avoidance functions have been studied and implemented in the wireless transmission / reception device 20e and the IC tag 5.

As the IC tag reader / writer 2 and the IC tag 5, for example, an IC tag reader / writer and an IC tag having a communication frequency band of 860 to 960 MHz are used.

Referring back to FIG. 1, the management device 1 is a device for centrally managing data on the presence / absence of luggage in the warehouse X (that is, inventory), and is installed in, for example, a management department.

3, the management device 1 includes a control device 10a, a RAM 10b, a ROM 10c, a hard disk 10d, a display 10e, a keyboard 10f, a pointing device 10g, a reader / writer communication device 10h, a sensor communication device 10i, and the like. 3 are connected to each other via a bus or the like.

As shown in FIG. 4, software such as an operating system SW0, middleware SW1, and inventory management application SW2 is stored in the ROM 10c or the hard disk 10d.

The operating system SW0 performs system management of the entire management apparatus 1. It also provides a basic user interface and driver.

The middleware SW1 is software for selecting necessary and unnecessary IDs read from the IC tag 5 by the IC tag reader / writer 2, and includes a passage signal detection unit 101, a reading control unit 102, and a reading result. Functions such as the buffer 103, the end determination unit 104, the similarity calculation unit 105, the tag attribute determination unit 106, the tag attribute response unit 107, the output level determination unit 121, the parameter storage unit 141, and the group data storage unit 142 are realized. Further, a parameter adjustment tool is assigned to the middleware SW1. The parameter adjustment tool is a module for adjusting various parameters, and implements the parameter adjustment unit 122 and the like. Details of each part will be described later.

The inventory management application SW2 manages an inventory database 1DB indicating the status of the container 4 received in the warehouse X as shown in FIG. Details of the inventory database 1DB will be described later.

Each module included in these software is read into the RAM 10b as necessary and executed by the control device 10a.

The control device 10a is composed of a CPU, MPU, or the like, similar to the control device 20a of the IC tag reader / writer 2. Similar to the RAM 20b of the IC tag reader / writer 2, the RAM 10b includes an SRAM or a flash memory.

The display 10e includes a screen showing the status of the container 4, a screen showing the status of the operation of the management apparatus 1, a screen for inputting commands or data, a screen showing information to be notified to the operator, and the like. Is displayed.

The keyboard 10f and the pointing device 10g are input devices for the user to input commands and data into the management device 1.

The reader / writer communication device 10h is a device for communicating with the IC tag reader / writer 2 via a network. A wired or wireless LAN device or the like is used as the reader / writer communication device 10h. An interface device such as USB or IEEE1394 may be used.

The anti-sensor communication device 10 i is a device for communicating with the passage sensor 3. An interface device such as USB or IEEE1394 is used as the sensor communication device 10i.

When the communication interface between the IC tag reader / writer 2 and the passage sensor 3 is the same, one device may be shared as the reader / writer communication device 10h and the sensor communication device 10i.

As the management device 1, a personal computer, a workstation, a host computer, or the like is used.

Referring back to FIG. 1, one passage sensor 3 is installed near the entrance gate and the exit gate of the warehouse X. The passage sensor 3 provided at the warehousing gate detects that the carriage 6 has passed through the warehousing gate, and transmits a warehousing signal SI to the management device 1. On the other hand, the passage sensor 3 provided at the exit gate detects that the cart 6 has passed through the exit gate and transmits the exit signal SO to the management device 1. As the passage sensor 3, an optical sensor such as an infrared sensor or an ultrasonic sensor is used.

Hereinafter, the IC tag reader / writer 2 provided near the entrance gate and the IC tag reader / writer 2 provided near the exit gate are described separately from “IC tag reader / writer 2A” and “IC tag reader / writer 2B”, respectively. There is. Similarly, the passage sensor 3 provided near the warehousing gate and the passage sensor 3 provided near the warehousing gate may be described separately from “passing sensor 3A” and “passing sensor 3B”, respectively.

The IC tag reader / writer 2A and the passage sensor 3A are arranged so that the passage sensor 3A detects the carriage 6 before the IC tag reader / writer 2A detects the container 4 when the carriage 6 with the container 4 is loaded. Has been. Similarly, the IC tag reader / writer 2B and the passage sensor 3B are arranged so that the passage sensor 3B detects the carriage 6 before the IC tag reader / writer 2B detects the container 4 when the carriage 6 carrying the container 4 is delivered. Has been placed.

FIG. 6 is a diagram showing an example of read pattern data 7Q, FIG. 7 is a diagram showing an example of the positional relationship between unnecessary tags, IC tag reader / writer 2A, and cart 6 and the reach of search command CI, and FIG. FIG. 9 is a diagram illustrating an example, and FIG. 9 is a diagram illustrating an example of a reading pattern.

Next, taking as an example the case where the container 4 is stored in the warehouse X, each part of the management apparatus 1 shown in FIG. 4, each part of the IC tag reader / writer 2 shown in FIG. .

When the cart 6 carrying the container 4 passes through the warehousing gate, the passing sensor 3A first detects the cart 6. Then, the passage sensor 3 </ b> A transmits the warehousing signal SI to the management device 1.

Then, the reading control unit 102 transmits a start command CS for instructing start of ID reading to the IC tag reader / writer 2A. At this time, the reading control unit 102 notifies the IC tag reader / writer 2A of two levels of the strength of the transmitted radio wave. The two levels are different. The level is determined by the output level determination unit 121. The level determination method will be described later.

In addition, when the passage signal detection unit 101 detects the delivery signal SO transmitted from the passage sensor 3B, the reading control unit 102 transmits a start command CS to the IC tag reader / writer 2B. Further, the reading result buffer 103 is reset before starting to receive the reading result from the IC tag reader / writer 2A.

In FIG. 2, when the communication device 20d of the IC tag reader / writer 2A receives the start command CS, the control device 20a controls the wireless transmission / reception device 20e so that reading of the ID is started.

Then, the radio transmission / reception device 20e is notified of a radio wave of a search command (hereinafter referred to as a search command CI) from the management device 1 for a predetermined time (hereinafter referred to as “reading time Tr”). Alternately at predetermined levels and at predetermined time intervals. Or, it repeats a predetermined number of times.

By the way, the carriage 6 passes through a predetermined route. At this time, after passing through the warehousing gate and being detected by the passage sensor 3A, it gradually approaches the IC tag reader / writer 2A, passes in front of the IC tag reader / writer 2A, and moves away from the IC tag reader / writer 2A. Therefore, the IC tag 5 placed on the carriage 6 is gradually changed from a state in which it is difficult to receive the search command CI until it passes in front of the IC tag reader / writer 2A, and after that, the search is gradually performed. It changes to a state where it is difficult to receive the command CI. Therefore, the wireless transmission / reception device 20e changes from a state where it is difficult to read the ID to a state where it is easy to read until the carriage 6 passes in front of the IC tag reader / writer 2A, and gradually changes to a state where it is difficult to read after passing.

Further, the search command CI reaches the IC tag 5 of the container 4 placed on the other cart 6 or the container 4 already received (that is, the IC tag 5 of the container 4 that is not the target), or the IC tag 5 of the container 4 that is not the target. The ID may reach the IC tag reader / writer 2A.

Also, the search command CI with a higher level is easier to be received by the IC tag 5 than the search command CI with a lower level. Hereinafter, the search command CI having a higher level and the search command CI having a lower level may be described separately from the “high search command CIH” and the “low search command CIL”, respectively. The pattern of the read result for one IC tag 5 is referred to as “read pattern Q”.

The wireless transmission / reception device 20e repeatedly transmits the high search command CIH and the low search command CIL, thereby obtaining a reading pattern Q as shown in FIG. For the sake of space, FIG. 6 shows each read pattern Q in two parts. The reading pattern Q has the following tendency.
Overall, the success rate of reading near the center of the reading time Tr is higher than that at the beginning and end of the reading time Tr.
Overall, the read success rate with the high search command CIH is higher than that with the low search command CIL.
More reading patterns Q of the IC tags 5 than the number of containers 4 placed on the cart 6 may be obtained.

The read pattern data 7Q (7Q1, 7Q2,...) Indicating the read pattern Q of each IC tag 5 is stored in the read result buffer 103 and the read log storage unit 143.

The reading pattern Q shown in the reading pattern data 7Q stored in the reading result buffer 103 is for determining whether or not each IC tag 5 is the original reading target (that is, the IC tag 5 of the container 4 to be stored). Used for. A part of the storage medium (for example, the RAM 10 b) having a high access speed is used for the read result buffer 103 so that the determination is made promptly. The read result buffer 103 exchanges data using a FIFO (First-In-First-Out) method.

On the other hand, the reading pattern Q shown in the reading pattern data 7Q stored in the reading log storage unit 143 is used for correcting a threshold value β described later. The read log storage unit 143 may be realized by using a part of the hard disk 10d.

In FIG. 6, “reading order number” indicates the number of times of reading (sending a search command CI) after a certain start command CS is received. In the example of FIG. 6, the search command CI is sent 32 times during the reading time Tr. Therefore, the control device 20a issues reading order numbers “1” to “32”.

Further, the control device 20a generates one read pattern data 7Q for each ID read by the wireless transmission / reception device 20e even once in the reading time Tr. In the read pattern data 7Q, a sequence code attached for convenience is shown in addition to the read ID. Further, the result of reading the ID for each reading order number is shown. “H” means that it was able to be read by the high search command CIH (successful reading). “L” means that it could be read by the low search command CIL. “-” Means that reading could not be performed (reading failed).

The read pattern data 7Q of each ID obtained by the wireless transmission / reception device 20e is transmitted to the management device 1 by the communication device 20d under the control of the control device 20a.

The end determination unit 104 determines the timing for ending the reading of the IC tag 5 of the container 4 placed on one cart 6. For example, with respect to the cart 6 to be stored, if the reading time Tr elapses after the cart 6 is detected by the passage sensor 3A, it is determined that the timing for completing the reading has come. Alternatively, when the reading order number reaches a predetermined value, it may be determined that the timing for ending reading has come. Then, the reading control unit 102 causes the IC tag reader / writer 2A to stop reading.

By the way, a typical pattern can be seen in the result of the success or failure of reading the ID from a certain IC tag 5 at the reading time Tr. Before starting the operation of the receipt / shipment management support system SYS, the cart 6 loaded with the container 4 under the standard (non-regular, normal) conditions is received from the receipt gate in advance, and the IC as described above. A typical pattern is acquired by causing the tag reader / writer 2A to perform ID reading processing. Here, a method for obtaining a typical pattern will be described in detail.

First, the output level (intensity) of radio waves when transmitting the high search command CIH and the low search command CIL is determined as follows. Assume that the IC tags 5 of the container 4 are placed in four places around the IC tag reader / writer 2A as shown in FIG. Hereinafter, these IC tags 5 may be referred to as “unnecessary tags”.

The position of the first unnecessary tag and the position of the second unnecessary tag are almost directly in front of the IC tag reader / writer 2A. Further, both positions are beyond the path of the carriage 6 when viewed from the IC tag reader / writer 2A. However, the distance between the first unnecessary tag and the IC tag reader / writer 2A is longer than the distance between the second unnecessary tag and the IC tag reader / writer 2A.

On the other hand, the position of the third unnecessary tag and the position of the fourth unnecessary tag are shifted from the front of the IC tag reader / writer 2. The distance between the third unnecessary tag and the IC tag reader / writer 2A is longer than the distance between the first unnecessary tag and the IC tag reader / writer 2A, but the distance between the second unnecessary tag and the IC tag reader / writer 2A. Shorter than. The distance between the fourth unnecessary tag and the IC tag reader / writer 2A is longer than the distance between the second unnecessary tag and the IC tag reader / writer 2A.

The output level of the radio wave when the low search command CIL is transmitted is almost certainly reached by the first unnecessary tag, but the second unnecessary tag and the fourth unnecessary tag. Is not reached at all, and the level is determined so as to reach or not reach the third unnecessary tag. The level of the radio wave output when transmitting the high search command CIH is set to such a level that the high search command CIH is barely delivered to the second unnecessary tag and the high search command CIH is occasionally received to the fourth unnecessary tag. decide. For example, the radio wave output level when transmitting the high search command CIH is determined to be 57 dBm, and the radio wave output level when transmitting the low search command CIL is determined to be 51 dBm. However, the above-mentioned method of arrival may not be exact. The output level is appropriately corrected as will be described later.

Hereinafter, the output level of the radio wave when transmitting the high search command CIH is simply referred to as “the output level of the high search command CIH”. The same applies to the low search command CIL.

When the output level of each of the high search command CIH and the low search command CIL is determined, a cart 6 with five containers 4 as shown in FIG. Then, as described above, the passage signal detector 101 and the passage sensor 3A detect that the carriage 6 has passed, and the IC tag 5 of each container 4 placed on the carriage 6 is detected by the reading controller 102 and the IC tag reader / writer 2A. Read. However, no other IC tag 5 is placed in the warehouse X. Alternatively, if another IC tag 5 is placed in the warehouse X, the IC tag 5 is moved to a position sufficiently away from an area where radio waves transmitted from the IC tag reader / writer 2A reach.

Note that the high search command CIH and the low search command CIL are alternately transmitted, but the output level determination unit 121 prepares for the start of operation of the receipt / shipment management support system SYS when the high search command CIH and the low search command CIL are prepared. The output level of each command CIL is determined to be an output level determined by assuming the above four unnecessary tags. Therefore, the IC tag reader / writer 2A alternately transmits a high search command CIH of 57 dBm and a low search command CIL of 51 dBm. After the start of operation, the output level is updated as necessary, which will be described later.

If such a warehousing operation is performed several times (for example, about 10 to 30 times), about 100 patterns can be obtained. As shown in FIG. 9A, “−” indicating that reading has failed continues in some parts from the beginning and each part from the end of each pattern. As shown in FIG. 9B, both portions are excluded from the read pattern. However, if the beginning of the pattern is “L” as a result of the exclusion, “-” is added to the beginning. When the end of the pattern is “H”, one “−” is added to the end.

• Each pattern that has been processed near the beginning and near the end can be grouped together with similar patterns. A typical pattern is obtained by selecting a representative pattern for each group. For example, three typical patterns are obtained as shown in FIG. These typical patterns can be assumed to be similar or consistent with the results of reading for the IC tag 5 being read. Therefore, hereinafter, a typical pattern is referred to as an “assumed pattern P”.

Then, assumed pattern data indicating each assumed pattern P (P 1, P 2, P 3) is generated and stored in the group data storage unit 142.

Now, after the start of the operation of the receipt / shipment management support system SYS, every time the container 4 is placed on the carriage 6 and received, it is detected that the carriage 6 has passed by the passage signal detection unit 101 and the passage sensor 3A and read. The IC tag 5 of each container 4 placed on the carriage 6 is read by the control unit 102 and the IC tag reader / writer 2A.

After the start of operation, there may be other containers 4 that have already been received. Therefore, as described with reference to FIG. 6, not only the IC tag 5 of the container 4 placed on the carriage 6 but also other IC tags 5 may be read.

The similarity calculation unit 105 calculates the degree of similarity (similarity) between the read pattern Q obtained by the IC tag reader / writer 2A and stored in the read result buffer 103 and the assumed pattern P registered in the group data storage unit 142. ,calculate.

As described above, three assumed patterns P are prepared in advance in the present embodiment. The similarity calculation unit 105 calculates the similarity RB between a certain reading pattern Q and each of the three assumed patterns P, and calculates the average value of the three similarities RB as the one reading pattern Q and the assumed pattern P. The similarity RA is determined.

10 is a diagram illustrating an example of offset, FIG. 11 is a diagram illustrating an example of matching between elements of both patterns, FIG. 12 is a flowchart illustrating an example of the flow of overall similarity calculation processing, and FIG. 13 is a flow of matching processing FIG. 14 and FIG. 15 are flowcharts illustrating an example of the flow of temporary similarity calculation processing.

Here, a method for calculating the similarity between one read pattern Qm and one assumed pattern Pn will be described with reference to FIG.

As shown in FIG. 10, the similarity calculation unit 105 offsets (shifts) the correspondence between the elements (members) of the read pattern Qm and the elements of the assumed pattern Pn, and the similarity RC (RC1, RC2,. RC3,... Are calculated one after another. Then, the largest similarity RC among the similarities RC1, RC2, RC3,... Is determined as the similarity RB between the read pattern Qm and the assumed pattern Pn. The similarity RC is calculated as follows.

As shown in FIG. 11, the similarity calculation unit 105 performs the first, second, third,..., Qth elements of the read pattern Qm and the first, second, third,. Compare elements of the th element.

When both of the s-th (1 ≦ s ≦ q) elements indicate “H”, that is, when the high search command CIH indicates successful reading, the similarity between both elements (hereinafter, “partial similarity”). Is described as “weight α1” (where weight α1> 0. For example, weight α1 = 1.0). Alternatively, when both indicate “L”, that is, when reading by the low search command CIL is successful, the partial similarity between the two elements is “weight α2” (however, weight α2> 0. For example, weight α2 = 2.0). Alternatively, when only one of them indicates “H”, the partial similarity between both elements is determined as “weight α3” (where weight α3 <0, for example, weight α3 = −1.0). Alternatively, when only one of them indicates “L”, the partial similarity between both elements is determined as “weight α4” (where weight α4 <0, for example, weight α4 = −1.5). In other cases (such as when both indicate “−” and when the reading result is not displayed on one side), the partial similarity between both elements is determined to be “0”.

Then, the average value is calculated by summing up the partial similarities of each number of pairs and dividing by the number of pairs (number of comparisons). This average value is the similarity RC.

In addition, as will be described later, association is performed so that an element indicating the reading result by the high search command CIH and an element indicating the reading result indicating the reading result by the low search command CIL are not compared. Therefore, it is impossible that one element indicates “H” and the other element indicates “L”. In principle, | weight α1 | <| weight α2 | and | weight α3 | <| weight α4 |.

The weights α1 to α4 are stored in the parameter storage unit 141. In addition, the parameter storage unit 141 stores output levels of the high search command CIH and the low search command CIL, a threshold value β and a weighting coefficient γ, which will be described later, the size of the reading result buffer 103, and various other coefficients. The initial value of the weighting factor γ is “1”.

In the example of FIG. 11, by comparison, 5 combinations in which both elements indicate “H” are found, 2 combinations in which both elements indicate “L” are found, and 0 combinations in which only one element indicates “H” are found. , Only three of the combinations showing “L” are found. The number of comparisons between elements is “10”. Therefore, the similarity RC is
{1.0 × 5 + 2.0 × 2 + (− 1.0) × 0 + (− 1.5) × 3} /10=0.45
It is.

The similarity calculation unit 105 calculates the similarity RB (RB1, RB2, RB3) between the read pattern Q and each of the three assumed patterns P by the method described with reference to FIGS. Then, an average value of the three similarities RB is calculated, and this average value is determined as the similarity RA.

More specifically, the similarity calculation unit 105 calculates the similarity RA between the read pattern Qm and the assumed pattern Pn according to the procedure shown in the flowcharts of FIGS.

12, the similarity calculation unit 105 defines a variable SUM having an initial value “0” and a variable “x” having an initial value “1” (# 151, # 152).

The similarity calculation unit 105 calculates the similarity RB (RB1) between the read pattern Qm and the assumed pattern Px (initially the assumed pattern P1) by the method described above with reference to FIGS. 11 and 10 (# 154). . More specifically, the calculation process of the procedure shown in FIG. 13 is performed.

That is, as in the first matching example of FIG. 10, the similarity calculation unit 105 reads the read result by the first high search command CIH of the read pattern Qm and the read result by the last high search command CIH of the assumed pattern Px. (# 161 in FIG. 13) and a similarity RB that is a variable is defined (# 162). The initial value of the similarity RB is “−∞”. Alternatively, the minimum value that can be stored in the similarity RB is set as the initial value. Further, an offset variable FS that is a variable for offset is defined (# 163). The initial value of the offset variable FS is a value obtained by multiplying the length of the assumed pattern P by “−1”.

Until the offset variable FS exceeds the length of the read pattern Qm, as shown in FIG. 10, the similarity RC (RC1, RC2,...) Between the two is successively calculated while offsetting the assumed pattern Px by two characters. (# 165, # 169). At this time, if the largest similarity RC is obtained (Yes in # 167), the similarity RC is substituted into the similarity RB (# 168). A specific method for calculating the similarity RC is shown in FIGS. 14 and 15.

The similarity calculation unit 105 performs an initialization process such as defining a cumulative similarity Ro that is a variable having an initial value of “0” (# 171 to # 177 in FIG. 14).

The similarity calculation unit 105 sequentially compares the reading results indicated by the elements of the assumed pattern Px with the reading results of the elements of the corresponding reading pattern Qm (# 179). When both are “H” (Yes in # 181), the weight α1 is added to the cumulative similarity Ro (# 182). When both are “L” (Yes in # 183), the weight α2 is added to the cumulative similarity Ro (# 184). If only one is “H” (Yes in # 185), the weight α3 is added to the cumulative similarity Ro (# 186). If only one is “L” (Yes in # 187), the weight α4 is added to the cumulative similarity Ro (# 188).

Then, the similarity calculation unit 105 sets the value obtained by dividing the cumulative similarity Ro by the number of comparisons as the similarity RC (# 190).

Note that the similarity calculation unit 105 determines the similarity between elements of a predetermined number (for example, 10% at the beginning and 10% at the end of the assumed pattern P) of the assumed pattern P and elements of the read pattern Q ( (Partial similarity) is weighted with a weighting factor γ. Specifically, in steps # 182, # 184, # 186, and # 188 in FIG. 15, the partial similarity of these elements is set to a value obtained by multiplying the weights α1 to α4 by the weight coefficient γ. That is, a value obtained by multiplying the weights α1 to α4 by the weight coefficient γ is added to Ro.

Referring back to FIG. 12, the similarity calculation unit 105 similarly calculates the similarity RB (RB2, RB3) between the read pattern Qm and each of the assumed pattern P2 and the assumed pattern P3 (# 154). Then, the average value of the similarity RB between the read pattern Qm and each of the assumed patterns P1 to P3 is calculated and set as the similarity RA (# 157).

16 is a diagram illustrating an example of the configuration of the parameter adjustment unit 122, FIG. 17 is a flowchart illustrating a first example of the flow of the initial threshold value calculation process, and FIG. 18 illustrates a second example of the flow of the initial threshold value calculation process. 19 is a diagram illustrating an example of the unnecessary pattern F, FIG. 20 is a flowchart illustrating an example of the flow of the first unnecessary pattern generation process, and FIG. 21 is an example of the flow of the second unnecessary pattern generation process. FIG. 22 is a flowchart for explaining an example of the flow of the third unnecessary pattern generation process, and FIG. 23 is a flowchart for explaining an example of the flow of the fourth unnecessary pattern generation process.

Returning to FIG. 4, the tag attribute determination unit 106 determines the attribute of each IC tag 5. Specifically, if the similarity RA between the read pattern Q of the IC tag 5 and the assumed pattern P calculated by the similarity calculation unit 105 is equal to or greater than the threshold value β, the IC tag 5 is the original reading target ( That is, it is determined that the container 4 is to be received. Hereinafter, the IC tag 5 that is the original reading target is referred to as a “necessary tag”. On the other hand, if it is less than the threshold value β, it is determined that the tag is erroneously read although it is not a necessary tag. Hereinafter, such an IC tag 5 is referred to as an “unnecessary tag”. The threshold β is set by a threshold setting unit 130 described later.

Note that the attribute determination result of the IC tag 5 by the tag attribute determination unit 106 is displayed on the display 10 e of the management apparatus 1 together with the ID of the IC tag 5.

The tag attribute response unit 107 responds the determination result by the tag attribute determination unit 106 to the inventory management application SW2. The determination result is used by the inventory management application SW2. In the above example, the tag attribute determination unit 106 determines whether the IC tag 5 detected near the warehousing gate is a necessary tag or an unnecessary tag. The ID of the IC tag 5 determined to be a necessary tag is related to the container 4 newly received in the warehouse X. Therefore, the inventory management application SW2 newly adds a record with the ID to the inventory database 1DB of FIG.

Note that the “issue flag” of this record indicates whether or not the container 4 related to this record has been issued. The default is “off”.

Whether the IC tag 5 read by the IC tag reader / writer 2 (2B) at the exit gate is also a necessary tag or an unnecessary tag is determined by the above-described processing. Then, the inventory management application SW2 updates the delivery flag of the record of the ID determined to be a necessary tag among the IDs of the IC tag 5 read by the IC tag reader / writer 2B to “ON”.

As shown in FIG. 16, the parameter adjustment unit 122 includes a threshold setting unit 130, a read parameter adjustment unit 137, an assumed pattern reacquisition control unit 138, and the like.

The threshold setting unit 130 includes an initial threshold calculation unit 131, a threshold adjustment unit 132, a threshold recalculation unit 133, an unnecessary pattern generation method database 133, an unnecessary pattern generation unit 134, and the like, and is a threshold used by the tag attribute determination unit 106. Calculate or adjust β.

The initial threshold value calculation unit 131 calculates the initial value of the threshold value β according to the procedure shown in the flowchart of FIG.

幾つか Try placing several (for example, five) containers 4 on the carriage 6 and entering the warehouse X. At this time, the initial threshold value calculation unit 131 requests the reading control unit 102 to read the IC tag 5 of each container 4. Then, as in the case of acquiring the assumed pattern P, the reading control unit 102 gives the output levels of the high search command CIH and the low search command CIL determined by the output level determination unit 121 to the IC tag reader / writer 2A. Notification is made to read the IC tag 5 of each container 4 (# 201 in FIG. 17). As a result, five read patterns Q as illustrated in FIG. 9A are obtained.

The initial threshold value calculation unit 131 calls the three assumed patterns P from the group data storage unit 142 (# 202). Then, the similarity with the assumed pattern P is calculated for each of the five reading patterns Q (# 203), and the lowest similarity is determined as the threshold value β (# 204). The method of calculating the similarity is as described above with reference to FIGS. The threshold value β is stored in the parameter storage unit 141.

Alternatively, the initial threshold value calculation unit 131 calculates the initial value of the threshold value β in accordance with the procedure shown in the flowchart of FIG. 18 in cooperation with the unnecessary pattern generation method database 133 and the unnecessary pattern generation unit 134.

In FIG. 18, the initial threshold value calculation unit 131 performs the same processing as steps # 201 to # 203 in FIG. 17, and includes five read patterns Q and three assumed patterns P as illustrated in FIG. Similarities are calculated (# 211 to # 213). Then, the lowest similarity is selected (# 214). Unlike the case of FIG. 17, the selected similarity does not become the threshold value β.

Incidentally, the unnecessary pattern generation method database 133 stores unnecessary pattern generation method data 8F indicating the procedure for generating the unnecessary pattern F (see FIGS. 20 to 23). The “unnecessary pattern F” is a pattern of the reading result as shown in FIG. 19 when the ID is read from the unnecessary tag shown in FIG. However, in the present embodiment, the unnecessary pattern F is estimated by performing the process of the procedure shown in the unnecessary pattern generation method data 8F on the reading pattern Q of the IC tag 5 of the container 4 placed on the cart 6. Generate. The unnecessary pattern F may be acquired by actually placing and reading each unnecessary tag shown in FIG. 7, but according to the unnecessary pattern generation unit 134, the administrator actually places the IC tag 5. The unnecessary pattern F can be obtained easily.

The initial threshold value calculation unit 131 requests the unnecessary pattern generation unit 134 to generate a plurality of unnecessary patterns F. Then, the unnecessary pattern generation unit 134 estimates and generates the unnecessary pattern F according to the procedure shown in FIGS. 20 to 23 based on the unnecessary pattern generation method data 8F.

The unnecessary pattern generation unit 134 generates four unnecessary patterns F by duplicating four read patterns Q obtained in step # 211 of FIG. 18 and performing the processing shown in FIGS. 20 to 23 on each of them. To do.

In FIG. 20, the unnecessary pattern generation unit 134 sequentially checks the reading results of each element included in the first copy of the reading pattern Q. When the reading is successful, that is, for an element indicating “H” or “L” (Yes in # 343), the reading result is replaced with “−” (# 344).

On the other hand, when reading failure is indicated, that is, an element indicating “−” (No in # 343), when it is an element indicating a reading result by the high search command CIH, that is, when it is an odd-numbered element (# If the element is an even-numbered element (No in # 345), it is replaced with “L” (# 347).

In FIG. 21, the unnecessary pattern generation unit 134 sequentially checks the reading result of each element included in the second copy of the reading pattern Q, and replaces the reading result of the element indicating “L” with “−”. (Yes in # 303, # 304).

Further, in FIG. 22, the unnecessary pattern generation unit 134 sequentially checks the reading result of each element included in the third copy of the reading pattern Q. When the reading is successful, that is, for the element indicating “H” or “L” (Yes in # 313), the following processing is performed for each element.

The unnecessary pattern generation unit 134 generates a random number from 0 to less than 1 (# 314) and defines a constant t (# 315). The constant t is a value of 0 or more and less than 1, and for example, “0.3” is used. If the read result of the element is “H” (Yes in # 316), if the generated random number exceeds the constant t (Yes in # 318), the read result of the element is “−”. (# 319). If the read result of the element is “L” (No in # 316), if the generated random number exceeds half the constant t (Yes in # 317 and # 318), the element The read result is replaced with “−” (# 319).

That is, the unnecessary pattern generation unit 134 replaces about 70% of “H” with “−” and about 85% of “L” with “−”.

Further, in FIG. 23, the unnecessary pattern generation unit 134 replaces the reading results of each element included in the fourth copy of the reading pattern Q with “−” (# 331) and reads 1 or more. An odd random number that is less than the length of the pattern Q is generated (# 332). Then, the reading result of the element corresponding to the generated random number is replaced with “H” (# 333). For example, when the random number is “15”, the reading result of the fifteenth element from the front is replaced with “−”.

The duplication in which such replacement is performed becomes an unnecessary pattern F. Note that the unnecessary pattern F generated by the algorithm shown in FIG. 20 represents the pattern of the reading result of the first unnecessary tag shown in FIG. An unnecessary pattern F generated by the algorithm shown in FIG. 21 represents the pattern of the reading result of the second unnecessary tag shown in FIG. The unnecessary pattern F generated by the algorithm shown in FIG. 22 represents the pattern of the reading result of the third unnecessary tag shown in FIG. The unnecessary pattern F generated by the algorithm shown in FIG. 23 represents the pattern of the reading result of the fourth unnecessary tag shown in FIG.

Returning to FIG. 18, the unnecessary pattern generation unit 134 similarly generates four unnecessary patterns F for the remaining read patterns Q and performs the processing shown in FIGS. 20 to 23 for each of them. To do. Therefore, in step # 215, four unnecessary patterns F are generated based on each of the five read patterns Q, and therefore 20 unnecessary patterns F are generated.

The initial threshold value calculation unit 131 calculates the similarity with the assumed pattern P for each generated unnecessary pattern F (# 216), and selects the highest similarity (# 217). The calculation method is the same as the method for calculating the degree of similarity between the read pattern Q and the assumed pattern P, as described above with reference to FIGS. 11, 10, and 12.

Then, the initial threshold value calculation unit 131 determines the value between the similarities selected in steps # 214 and # 217 as the threshold value β (# 218). For example, a value obtained by adding these similarities and dividing by 2 is determined as the threshold value β. Alternatively, a value obtained by adding a slight value (for example, a value of about 0.01) to the similarity selected in step # 217 is determined as the threshold value β. In this case, the processes of steps # 213 and # 214 are not necessary.

24 is a flowchart for explaining an example of the flow of the threshold adjustment process, FIG. 25 is a flowchart for explaining an example of the flow of the resetting process, FIG. 26 is a diagram showing an example of the reading pattern Q, and FIG. It is a flowchart explaining the example of a flow.

4 and 16, after the operation of the receipt / shipment management support system SYS is started, the tag attribute determination unit 106 uses the threshold value β calculated by the initial threshold value calculation unit 131 to set the attribute of the IC tag 5. Determine. However, at the initial threshold value β, the attribute of the IC tag 5 may be erroneously determined. Therefore, the threshold adjustment unit 132 performs the process of adjusting the threshold β according to the procedure shown in FIG.

The manager of the warehouse X appropriately checks whether there is an error in the discrimination result by comparing the discrimination result displayed on the display 10e with the container 4 actually received. Then, the manager of the warehouse X inputs the ID of the IC tag 5 whose determination result is incorrect into the management device 1 by operating the keyboard 10f.

The threshold adjustment unit 132 receives the input ID (# 221 in FIG. 24). If all of the IDs belong to any of the IC tags 5 that are determined to be unnecessary tags even though they are necessary tags (Yes in # 222), the IC tags 5 having the respective IDs. The threshold value β is lowered to the lowest similarity RA among the similarities RA between the read pattern Q and the assumed pattern P (# 223). Alternatively, the threshold value β may be lowered to a value slightly lower than the lowest similarity RA (for example, a value obtained by subtracting a predetermined value from the lowest similarity RA).

Alternatively, if all of the IDs belong to any of the IC tags 5 that are determined to be necessary tags even though they are unnecessary tags (Yes in # 224), the IC tag 5 having each of those IDs The threshold value β is raised to the highest similarity RA among the similarities RA between the read pattern Q and the assumed pattern P (# 223). Alternatively, the threshold value β may be raised to a value slightly higher than the highest similarity RA (for example, a value obtained by adding a predetermined value to the lowest similarity RA).

However, it is for the IC tag 5 that is determined to be an unnecessary tag even though some of the input IDs are necessary tags, and is necessary even though the other part is an unnecessary tag. In the case of the IC tag 5 determined to be a tag, the threshold β cannot be adjusted by the processing by the threshold adjustment unit 132 shown in FIG.

When the threshold value β cannot be adjusted by the threshold value adjusting unit 132, it is necessary to improve the accuracy of determining the attribute of the IC tag 5 by adjusting a setting value other than the threshold value β. The read parameter adjustment unit 137 cooperates with the unnecessary pattern generation method database 133, the unnecessary pattern generation unit 134, the output level determination unit 138, the assumed pattern reacquisition control unit 138, and the like to improve the accuracy of attribute determination. For adjustment. Here, the contents of the adjustment processing will be described with reference to the flowchart of FIG.

For example, during operation of the receipt / shipment management support system SYS, when four containers 4 are placed on the same cart 6 and received, five reading patterns Q (Q1, Q2,..., Q5 as shown in FIG. ) Is obtained by the IC tag reader / writer 2A. The reading patterns Q1 to Q3 are those of the IC tag 5 of the container 4 placed on the cart 6, and these IC tags 5 are correctly determined as necessary tags by the tag attribute determining unit 106. However, although the read pattern Q4 is not that of the IC tag 5 of the container 4 placed on the cart 6, the IC tag 5 is erroneously determined to be a necessary tag by the tag attribute determination unit 106. Further, although the read pattern Q5 is for the IC tag 5 of the container 4 placed on the carriage 6, the IC tag 5 is erroneously determined to be an unnecessary tag by the tag attribute determination unit 106. .

When the administrator finds these erroneous determination results, the administrator inputs the IDs of the IC tags 5 of the read patterns Q4 and Q5.

The reading parameter adjustment unit 137 causes the unnecessary pattern generation unit 134 to generate an unnecessary pattern F for each of the required tag reading patterns Q (that is, the reading patterns Q1 to Q3 and Q5) (# 231 in FIG. 25). The method for generating the unnecessary pattern F is as described above with reference to FIGS. Since four unnecessary patterns F are generated for each reading pattern Q, a total of 16 unnecessary patterns F are generated.

The reading parameter adjusting unit 137 calculates the similarity between the reading pattern Q corresponding to each ID input by the administrator (that is, the reading patterns Q4 and Q5) and each unnecessary pattern F (# 232), and the reading pattern Q4. For each Q5, the unnecessary pattern F having the highest similarity is selected (# 233). The method for calculating the degree of similarity between two patterns is the same as the method for calculating the degree of similarity between the read pattern Q and the assumed pattern P described above with reference to FIGS.

The reading parameter adjustment unit 137 uses an algorithm (that is, one of the algorithms shown in FIGS. 20 to 23) used when the selected unnecessary pattern F is generated, or an outline thereof, and adjustment according to the algorithm. The method or the outline thereof is displayed on the display 10e and presented to the administrator (# 234).

Specifically, the read parameter adjustment unit 137 obtains the selected unnecessary pattern F if it is generated from the algorithm shown in FIG. 21 (that is, read from the second unnecessary tag in FIG. 7). The first adjustment method is presented.

The first adjustment method is a method for increasing the output level of the low search command CIL to improve the accuracy of discrimination. The reason for presenting the first adjustment method is that it is considered that the reason why the attribute is erroneously determined is that reading by the low search command CIL is not suitably performed.

Alternatively, when the selected unnecessary pattern F is generated by the algorithm shown in FIG. 21 and generated from the read pattern Q4, it is used together with the first adjustment method or instead of the first adjustment method. In addition, a second adjustment method may be presented. The second adjustment method is a method in which the weight of the read result by the low search command CIL, that is, the absolute values of the weights α2 and α4 shown in FIG. Alternatively, when the selected unnecessary pattern F is generated by the algorithm shown in FIG. 21 and generated from the read pattern Q5, it is used together with the first adjustment method or instead of the first adjustment method. In addition, a third adjustment method may be presented. The third adjustment method is a method for improving the accuracy of discrimination by increasing the weight of the read result by the low search command CIL, that is, the absolute values of the weights α2 and α4 shown in FIG.

The reading parameter adjusting unit 137 is assumed to be obtained if the selected unnecessary pattern F is read by the algorithm shown in FIG. 22 (that is, read from the third unnecessary tag in FIG. 7). Pattern) and generated from the read pattern Q4, a fourth adjustment method is presented.

The fourth adjustment method is a method in which the weights at both ends of the assumed pattern P and in the vicinity thereof are made higher than the weights of the other portions, that is, the weight coefficient γ is increased to improve the determination accuracy. . The reason for presenting the fourth adjustment method is that the cause of misjudgment is a read pattern Q4 in which the result of successful reading is shown only in a short part (for example, 10% of the whole) and a long assumed pattern P ( For example, it is considered that the similarity with the assumed pattern P1) in FIG. 8 is higher than expected.

The reading parameter adjusting unit 137 is assumed to be obtained if the selected unnecessary pattern F is read by the algorithm shown in FIG. 22 (that is, read from the third unnecessary tag in FIG. 7). Pattern) and generated from the read pattern Q5, the fifth adjustment method is presented.

The fifth adjustment method is a method in which the weights at both ends of the assumed pattern P and in the vicinity thereof are made lower than the weights of the other portions, that is, the weighting factor γ is reduced to improve the discrimination accuracy. . The reason for presenting the fifth adjustment method is that the cause of misclassification is that the portion of the read pattern Q5 where the result of successful reading appears is shorter than ideal, so the similarity to the assumed pattern P is lower than expected. It is because it is considered that it is in the point that has become.

The reading parameter adjusting unit 137 is assumed to be obtained if the selected unnecessary pattern F is generated by the algorithm shown in FIG. 23 (that is, read from the fourth unnecessary tag in FIG. 7). Pattern), the sixth adjustment method is presented.

The sixth adjustment method is a method of increasing the output level of the high search command CIH to improve the accuracy of discrimination. The reason for presenting the sixth adjustment method is that the reason why the attribute is erroneously determined is considered to be that reading by the high search command CIH is not suitably performed.

It is assumed that the read parameter adjustment unit 137 is obtained if the selected unnecessary pattern F is read by the algorithm shown in FIG. 20 (that is, read from the first unnecessary tag in FIG. 7). Pattern), the seventh adjustment method is presented.

The seventh adjustment method is a method for increasing the size of the read result buffer 103 and improving the accuracy of discrimination. The reason for presenting the seventh adjustment method is that the reason why the attribute is erroneously determined is that the data of the read pattern Q necessary for comparison with the assumed pattern P cannot be stored in the read result buffer 103. This is because it can be considered.

Alternatively, the eighth adjustment method may be presented together with the seventh adjustment method or instead of the seventh adjustment method. The eighth adjustment method is a method of increasing the accuracy of discrimination by increasing the distance between the antenna of the IC tag reader / writer 2A and the passage sensor 3A.

The administrator selects one of the presented adjustment methods. Then, the reading parameter adjustment unit 137 executes a process as shown in FIG. 27 (# 235).

When the first adjustment method is selected (Yes in # 351 in FIG. 27), the read parameter adjustment unit 137 slightly sets the output level of the low search command CIL stored in the parameter storage unit 141 (for example, 1 dBm) (# 352).

When the second adjustment method is selected (Yes in # 353), the read parameter adjustment unit 137 brings the weight α2 and the weight α4 stored in the parameter storage unit 141 close to zero. That is, the weight of the read result of the low search command CIL is reduced. For example, the weight α2 and the weight α4 are updated by a coefficient larger than 0 and smaller than 1 (# 354).

When the third adjustment method is selected (Yes in # 355), the reading parameter adjustment unit 137 separates the weight α2 and the weight α4 stored in the parameter storage unit 141 from zero. That is, the weight of the read result of the low search command CIL is increased. For example, the weight α2 and the weight α4 are updated by multiplying by a coefficient larger than 1 (# 356).

If the fourth adjustment method is selected (Yes in # 357), the reading parameter adjustment unit 137 increases the weighting coefficient γ stored in the parameter storage unit 141 (# 358). For example, the weighting coefficient γ is updated by a coefficient larger than 1.

If the fifth adjustment method is selected (Yes in # 359), the read parameter adjustment unit 137 lowers the weighting coefficient γ stored in the parameter storage unit 141 (# 360). For example, the weighting coefficient γ is updated by multiplying by a coefficient larger than 0 and smaller than 1.

If the sixth adjustment method is selected (Yes in # 361), the read parameter adjustment unit 137 slightly increases the output level of the high search command CIH (for example, 1 dBm) (# 362).

When the seventh adjustment method is selected (Yes in # 363), the read parameter adjustment unit 137 increases the size of the read result buffer 103. Specifically, for example, in cooperation with the operating system, the area for the read result buffer 103 provided in the RAM 10b is expanded (# 364).

When the eighth adjustment method is selected (Yes in # 365), the read parameter adjustment unit 137 presents to the administrator so as to widen the distance between the antenna of the IC tag reader / writer 2A and the passage sensor 3A (# 366). The manager adjusts the arrangement of the antenna of the IC tag reader / writer 2A and the passage sensor 3A based on the presented contents.

Referring back to FIG. 25, the assumed pattern reacquisition control unit 138 re-acquires the assumed pattern P based on various setting values such as the output level adjusted by the read parameter etc. adjustment unit 137 and the unnecessary pattern F. Is re-calculated, and the threshold β is recalculated based on the assumed pattern P and the unnecessary pattern F (# 237). In addition, what is necessary is just to perform these processes as needed.

FIG. 28 is a flowchart illustrating an example of the overall processing flow in the management apparatus 1.

Next, an overall processing procedure in the management apparatus 1 for receiving the container 4 will be described with reference to the flowchart of FIG.

In FIG. 28, the management apparatus 1 performs a preparation process for starting the operation of the receipt / shipment management support system SYS (# 11 to # 14).

That is, first, the management device 1 stores initial values of various settings such as the output level of each of the high search command CIH and the low search command CIL, the weighting coefficient γ, and the size of the read result buffer 103 (# 11).

* With no unnecessary tags, place the container 4 on the cart 6 and pass it through a predetermined route. At this time, the management apparatus 1 causes the IC tag reader / writer 2A to read the IC tag 5 of the container 4 (# 12). Thereby, a plurality of patterns showing the result of reading as shown in FIG.

Based on the pattern obtained by the IC tag reader / writer 2A, the management apparatus 1 generates and stores a typical pattern obtained from the IC tag 5 of the container 4 at the time of warehousing, that is, an assumed pattern P (see FIG. 8) (#). 13).

And the management apparatus 1 produces | generates the unnecessary pattern F, the similarity of the reading pattern Q and the assumption pattern P which are obtained from the IC tag 5 of the container 4 mounted on the cart 6, those reading pattern Q and the assumption pattern P The threshold value β is calculated based on the similarity to and stored (# 14). The procedure for calculating the threshold value β is as described above with reference to FIG.

Note that the pattern obtained in step # 12 may be used for a part or all of the read pattern Q.

When the preparation process is completed, the management apparatus 1 waits for the container 4 to be received. When the passing sensor 3A detects that the carriage 6 has passed (Yes in # 15), the management device 1 instructs the IC tag reader / writer 2A to perform reading (# 16). Then, a plurality of read patterns Q are obtained by the IC tag reader / writer 2A. Some of these are obtained from the IC tag 5 (that is, necessary tag) of the container 4 to be stored, and some are obtained from the IC tag 5 (that is, unnecessary tag) that is not.

The management apparatus 1 calculates the similarity between each read pattern Q and the assumed pattern P (# 17), and the IC tag 5 of the read pattern Q whose similarity to the assumed pattern P is equal to or greater than the threshold value β is a necessary tag. It is discriminated that there is (Yes in # 18, # 19), and the IDs of those IC tags 5 are written in the inventory database 1DB (# 20). On the other hand, the management apparatus 1 determines that the IC tag 5 of the read pattern Q whose similarity with the assumed pattern P is less than the threshold value β is an unnecessary tag (No in # 18, # 21).

The management apparatus 1 displays the identification results of the IDs of the IC tags 5 read by the IC tag reader / writer 2A and the attributes of the IC tags 5 (# 22). Here, when the manager finds that there is an error in the attribute determination, the administrator inputs the ID of the IC tag 5 that has been erroneously determined.

When the ID of the IC tag 5 discriminated incorrectly is input (Yes in # 23), the management device 1 performs a process of adjusting the threshold value β (# 23). The procedure of this processing is as described above with reference to FIG.

However, when the ID of the IC tag 5 determined to be a necessary tag despite being an unnecessary tag and the ID of the IC tag 5 determined to be an unnecessary tag despite being a necessary tag are input, FIG. The threshold value β cannot be adjusted by the process shown in FIG. In this case, the management apparatus 1 searches for an unnecessary pattern F having the highest similarity with the read pattern Q of the IC tag 5 that has been erroneously determined, and performs processing according to the searched unnecessary pattern F, for example. Various set values are adjusted so as to increase the accuracy of (# 26). The procedure of this process is as described above with reference to FIG.

Then, the management device 1 performs processing such as reading of the IC tag 5 of the container 4 to be stored, based on a set value such as the threshold β adjusted as appropriate.

In addition, the management apparatus 1 performs the same process as the process for the IC tag 5 of the container 4 to be received for the IC tag 5 of the container 4 to be released.

According to this embodiment, it is possible to more reliably determine whether or not an IC tag is an original reading target than before.

Furthermore, it is possible to improve the accuracy of discrimination by adjusting various setting values without the administrator conducting a detailed investigation of the reading status when a misjudgment occurs. Moreover, it is possible to easily cope with the case where the situation on the site changes after the assumed pattern is created.

In the present embodiment, the first to eighth adjustment methods are presented to the administrator and selected, but the management device 1 may select based on a predetermined rule.

In the present embodiment, the management apparatus 1 performs processing such as the determination of the attribute of the IC tag 5 and the adjustment of the setting value necessary for it, but the IC tag reader / writer 2 may perform these processing. In this case, the functions of the middleware SW1 shown in FIG.

In addition, the overall configuration of each receipt / shipment management support system SYS, management device 1, and IC tag reader / writer 2, each processing unit, processing contents, processing order, data configuration, and the like can be appropriately changed in accordance with the spirit of the present invention.

Claims

A target IC tag determination system for determining whether or not a first IC tag is a target,
Result of success / failure of receiving a response from the second IC tag to the second IC tag in the first place which is a predetermined place for reading a plurality of times and transmitting the signal to the second IC tag each time An actual reading result acquisition means for acquiring an actual reading result,
When a signal is transmitted a plurality of times to a third IC tag assumed to be in a second location other than the first location, a response from the third IC tag to the signal of each time An unnecessary tag reading result estimation means for estimating an unnecessary tag reading result as a result of success or failure of reception based on the actual reading result;
A typical reading result that is a result of success or failure of reception of a response from the second IC tag for each time of the signal, which is typically obtained when a signal is transmitted to the second IC tag a plurality of times, is stored. Typical reading result storage means;
First similarity calculation means for calculating a first similarity that is a degree of similarity between the unnecessary tag reading result and the actual reading result;
Second similarity calculation means for calculating a second similarity that is higher than the first similarity;
A target reading result acquisition means for transmitting a signal a plurality of times and acquiring a target reading result that is a success or failure of receiving a response from the first IC tag for the signal of each time;
Third similarity calculation means for calculating a third similarity that is a degree of similarity between the target reading result and the typical reading result;
If the third similarity is higher than the second similarity, the first IC tag is determined to be the object, and if the third similarity is lower than the second similarity Attribute discrimination means for discriminating that the first IC tag is not the object;
A target IC tag discrimination system having
The actual reading result acquisition means alternately transmits a first signal that is a first intensity signal and a second signal that is a second intensity signal lower than the first intensity, Get the actual reading result,
The unnecessary tag reading result estimation means determines whether or not the response from the third IC tag has been successfully received when the first signal and the second signal are alternately transmitted as the unnecessary tag reading result. Guess the result of
The typical reading result storage unit stores a result of success or failure of reception of a response from the second IC tag when the first signal and the second signal are alternately transmitted as the typical reading result. To
The target IC tag discrimination system according to claim 1.
The first similarity calculation means determines whether the response to the second signal is successful among the results of successful responses to the respective signals included in the unnecessary tag reading result and the actual reading result. The first similarity is calculated by placing more importance on the result of the response than the result of success or failure in receiving the response to the first signal.
The target IC tag discrimination system according to claim 2.
The fourth similarity degree, which is the degree of similarity between the typical reading result and the actual reading result, is a result of success or failure of receiving a response to each signal included in each of the typical reading result and the actual reading result. A fourth similarity calculating means for calculating a result of success or failure of reception of the response to the second signal, more importantly than a result of success or failure of reception of the response to the first signal,
The second similarity calculation means calculates a value that is higher than the first similarity and equal to or lower than the fourth similarity as the second similarity.
The objective IC tag discrimination system according to claim 2 or claim 3.
Storage means for storing the second similarity,
The attribute determining means determines whether or not the first IC tag is the object based on the second similarity stored in the storage means,
When the attribute determination unit erroneously determines that the first IC tag is the target object, the attribute determination unit updates the second IC card so as to reduce the second similarity, and the attribute determination unit changes the first IC tag. If it is erroneously determined that it is not the object, it has a similarity update means for updating to increase the second similarity,
The target IC tag discrimination system according to any one of claims 1 to 4.
The first intensity, the second intensity, the second similarity, a first weight indicating a degree of emphasis on the first signal, and a degree of emphasis on the second signal Storage means for storing a second weight and a third weight indicating a degree of emphasis on the other portions of the head portion, the tail portion, and the vicinity thereof of the typical reading result;
When the attribute discriminating means erroneously discriminates whether or not the first IC tag is the object, the first intensity, the second intensity, and the first stored in the storage means Updating means for updating any one of the second similarity, the first weight, and the second weight,
The first signal and the second signal are transmitted by an IC tag reader;
The first similarity calculation means calculates the first similarity according to the first weight, the second weight, and the third weight stored in the storage means,
The actual reading result acquisition unit transmits the first signal and the second signal based on the first intensity and the second intensity stored in the storage unit to obtain the actual reading result. Acquired,
The target reading result acquisition unit transmits the first signal and the second signal based on the first intensity and the second intensity stored in the storage unit to obtain the target reading result. Acquired,
The attribute determining means determines whether or not the first IC tag is the object based on the second similarity stored in the storage means,
The updating unit is configured to determine the first intensity, the second intensity, and the second intensity according to the unnecessary tag reading result having the highest third similarity with the target reading result of the first IC tag. Updating one of the second similarity, the first weight, the second weight, and the third weight;
The target IC tag discrimination system according to any one of claims 1 to 4.
The unnecessary tag read result inference means is the third tag in the case where the second location is the front of the IC tag reader and the first signal almost certainly reaches but the second signal is difficult to reach. A second unnecessary tag reading result indicating the success or failure of receiving the response from the IC tag is estimated as the unnecessary tag reading result,
The updating means stores the unnecessary tag reading result having the highest third similarity with the target reading result of the first IC tag in the storage means when the unnecessary tag reading result is the second unnecessary tag reading result. Update to increase the stored second intensity,
The target IC tag discrimination system according to claim 6.
The unnecessary tag read result inference means is the third tag in the case where the second location is the front of the IC tag reader and the first signal almost certainly reaches but the second signal is difficult to reach. A second unnecessary tag reading result indicating the success or failure of receiving the response from the IC tag is estimated as the unnecessary tag reading result,
The updating means includes the second unnecessary tag reading result, the unnecessary tag reading result having the highest third similarity with the target reading result of the first IC tag, and the first IC tag. Is mistakenly determined to be the object, the second weight stored in the storage means is updated so as to be lowered.
The target IC tag discrimination system according to claim 6.
The unnecessary tag read result inference means is the third tag in the case where the second location is the front of the IC tag reader and the first signal almost certainly reaches but the second signal is difficult to reach. A second unnecessary tag reading result indicating the success or failure of receiving the response from the IC tag is estimated as the unnecessary tag reading result,
The updating means includes the second unnecessary tag reading result, the unnecessary tag reading result having the highest third similarity with the target reading result of the first IC tag, and the first IC tag. Is mistakenly determined not to be the object, the second weight stored in the storage means is updated to increase.
The target IC tag discrimination system according to claim 6.
The unnecessary tag reading result estimation means is a place where the second location is a location deviated from the front of the IC tag reader and the first signal can be delivered almost certainly but the second signal is difficult to reach. A third unnecessary tag reading result indicating a result of success or failure of reception of a response from the third IC tag is estimated as the unnecessary tag reading result,
The updating means includes the unnecessary tag reading result having the highest third similarity with the target reading result of the first IC tag is the third unnecessary tag reading result, and the first IC tag. Is mistakenly determined to be the object, it is updated to increase the third weight stored in the storage means.
The target IC tag discrimination system according to claim 6.
The unnecessary tag reading result estimation means is a place where the second location is a location deviated from the front of the IC tag reader and the first signal can be delivered almost certainly but the second signal is difficult to reach. A third unnecessary tag reading result indicating a result of success or failure of reception of a response from the third IC tag is estimated as the unnecessary tag reading result,
The updating means includes the unnecessary tag reading result having the highest third similarity with the target reading result of the first IC tag is the third unnecessary tag reading result, and the first IC tag. Is mistakenly determined not to be the object, the third weight stored in the storage means is updated so as to decrease.
The target IC tag discrimination system according to claim 6.
The unnecessary tag reading result estimation means is a response from the third IC tag when the second location is a location shifted from the front of the IC tag reader and the first signal hardly reaches. A fourth unnecessary tag reading result indicating the success or failure result of the reception is estimated as the unnecessary tag reading result,
The updating unit stores the storage unit when the unnecessary tag reading result having the highest third similarity with the target reading result of the first IC tag is the fourth unnecessary tag reading result. Update to increase the stored first intensity,
The target IC tag discrimination system according to claim 6.
A target reading result buffer for storing the target reading result;
The third similarity calculating means calculates the third similarity based on the target reading result stored in the target reading result buffer;
The unnecessary tag read result inference means is configured such that the second location is a front surface of the IC tag reader and the location where both the first signal and the second signal reach almost certainly. A fifth unnecessary tag reading result indicating a result of success or failure of reception of the response from the IC tag is estimated as the unnecessary tag reading result;
When the unnecessary tag reading result having the highest third similarity with the target reading result of the first IC tag is the fifth unnecessary tag reading result, the size of the target reading result buffer is increased. To
The target IC tag discrimination system according to claim 6.
A target IC tag discrimination device for discriminating whether or not a first IC tag is a target,
This is a result of success or failure of receiving a response from the second IC tag to the second IC tag in the first place, which is a predetermined place for reading, and receiving a response from the second IC tag for each signal. An actual reading result acquisition control means for causing an IC tag reader to execute processing for acquiring an actual reading result;
When the IC tag reader transmits a signal a plurality of times to a third IC tag assumed to be in a second location other than the first location, the third IC tag corresponding to the signal of each time Unnecessary tag reading result estimation means for estimating an unnecessary tag reading result which is a result of success or failure of reception of a response from the actual reading result;
Typical reading result that is a result of success or failure of receiving a response from the second IC tag in response to the signal of each time, obtained when the IC tag reader transmits a signal to the second IC tag a plurality of times. Storing typical reading result storage means;
First similarity calculation means for calculating a first similarity that is a degree of similarity between the unnecessary tag reading result and the actual reading result;
Second similarity calculation means for calculating a second similarity that is higher than the first similarity;
A target reading result acquisition control means for causing the IC tag reader to execute a process of acquiring a target reading result that is a success or failure of receiving a response from the first IC tag to the signal for each time multiple times;
Third similarity calculation means for calculating a third similarity that is a degree of similarity between the target reading result and the typical reading result;
If the third similarity is higher than the second similarity, the first IC tag is determined to be the object, and if the third similarity is lower than the second similarity Attribute discrimination means for discriminating that the first IC tag is not the object;
A target IC tag discriminating apparatus.
An object IC tag determination method for determining whether or not a first IC tag is an object,
Result of success / failure of receiving a response from the second IC tag to the second IC tag in the first place which is a predetermined place for reading a plurality of times and transmitting the signal to the second IC tag each time Obtaining an actual reading result of
When a signal is transmitted a plurality of times to a third IC tag assumed to be in a second location other than the first location, a response from the third IC tag to the signal of each time Estimating an unnecessary tag reading result that is a result of success or failure of reception based on the actual reading result;
A step of preparing a typical reading result, which is a typical result obtained when a signal is transmitted to the second IC tag a plurality of times, and is a result of success or failure of receiving a response from the second IC tag for each signal. When,
Calculating a first similarity that is a degree of similarity between the unnecessary tag reading result and the actual reading result;
Calculating a second similarity that is higher than the first similarity;
Transmitting a signal a plurality of times, obtaining a target reading result that is a success or failure of receiving a response from the first IC tag for the signal of each time;
Calculating a third similarity that is a degree of similarity between the target reading result and the typical reading result;
If the third similarity is higher than the second similarity, the first IC tag is determined to be the object, and if the third similarity is lower than the second similarity Determining that the first IC tag is not the object;
A target IC tag discrimination method having
A computer program used in a computer for determining whether or not a first IC tag is an object,
In the computer,
This is a result of success or failure of receiving a response from the second IC tag to the second IC tag in the first place, which is a predetermined place for reading, and receiving a response from the second IC tag for each signal. The process of controlling the IC tag reader is executed so that the process of acquiring the actual reading result is performed,
When the IC tag reader transmits a signal a plurality of times to a third IC tag assumed to be in a second location other than the first location, the third IC tag corresponding to the signal of each time An unnecessary tag reading result, which is a result of success or failure of reception of a response from, is executed based on the actual reading result,
Typical reading result that is a result of success or failure of receiving a response from the second IC tag in response to the signal of each time, obtained when the IC tag reader transmits a signal to the second IC tag a plurality of times. To store in the storage means,
Executing a process of calculating a first similarity that is a degree of similarity between the unnecessary tag reading result and the actual reading result;
Executing a process of calculating a second similarity that is higher than the first similarity,
A process of controlling the IC tag reader so as to perform a process of acquiring a target reading result that is a success or failure of receiving a response from the first IC tag to the signal for each time and transmitting the signal multiple times;
Executing a process of calculating a third similarity which is a degree of similarity between the target reading result and the typical reading result;
If the third similarity is higher than the second similarity, the first IC tag is determined to be the object, and if the third similarity is lower than the second similarity A process of determining that the first IC tag is not the object;
Computer program.