KR20230107793A - Job judgment system, job judgment method and recording medium - Google Patents

Abstract

A job judgment system, a job judgment method, and a recording medium capable of estimating the type of work performed on the work vehicle even when the operation information of the work vehicle cannot be acquired. According to one embodiment, the work judgment system includes external information preparation means and work estimation means. The external information preparation means prepares external information that can be prepared outside of the work vehicle that moves in the field and performs work, and is independent of the operation status of the work vehicle. The job estimating means estimates the type of job based on information recorded on the movement of the work vehicle and external information.

Description

Job judgment system, job judgment method and recording medium

The present invention relates to a job determination system for estimating the type of job of a work vehicle, a job determination method, and a recording medium.

In farming, it is desirable to record work performed as agricultural work in order to evaluate the management of the current year and establish an improvement plan for the next year. However, in the case of using a tractor, which is a general-purpose work vehicle, labor and time are required for such recording work. This is because it is possible to perform a variety of agricultural work by mounting a variety of work machines on one tractor, and there are many choices of objects to be input as work records.

Instead of manually recording the work, the type of work is estimated by a method based on machine learning using operation information indicating the operation status transmitted from a tractor having a predetermined communication function, and based on the estimation result Based on this, there is a system that automatically records work.

In connection with the above, in Patent Document 1 (Japanese Unexamined Patent Publication No. 2018-190292), based on positional information and operation information acquired when a working device such as a tractor moves while moving a field and performs agricultural work, , An agricultural work support system is disclosed that calculates agricultural work performed by a work device and manages the agricultural work.

In addition, in Patent Document 2 (Japanese Unexamined Patent Publication No. 2019-088227), a work machine having a work device attached to a vehicle body such as a tractor travels on a field and performs work, and the positional information detected when the work machine is formed, Disclosed is a management device that acquires work-related items based on information received from an RFID (Radio Frequency IDentification) tag in which work machine information indicating a model of the device is set.

Japanese Unexamined Patent Publication No. 2018-190292 Japanese Unexamined Patent Publication No. 2019-088227

In the above technique, operation information such as engine speed of a tractor or PTO (Power Take Off) speed is used as information used for estimating the type of work. However, in order to acquire such operation information, it is necessary that the tractor has a function of outputting operation information. In other words, since the types of tractors that output operation information are limited, it becomes difficult to estimate the type of work by the same method when the tractors owned by the user are different.

In the case of using a tractor of a model not compatible with the above method, the user may manually input data indicating the type of work before starting the work or the like, and record the data by manually inputting the data to the terminal of the tractor or the like. However, the operation of manually inputting work data each time a different work is performed by exchanging a working machine mounted on a tractor is cumbersome, and it is frequently possible to forget to input, making it difficult to record which work was performed. situation is likely to occur.

Accordingly, an object of the present invention is to provide a job judgment system, a job judgment method, and a recording medium that enable estimation of the type of work performed on the work vehicle even when the operation information of the work vehicle cannot be acquired. Other problems and novel features will become clear from the description of this specification and accompanying drawings.

According to one embodiment for solving the above problem, the work judgment system includes external information preparation means and work estimation means. The external information preparation means prepares external information that can be prepared outside of the work vehicle that moves in the field and performs work, and is independent of the operation status of the work vehicle. The job estimating means estimates the type of job based on the movement information recorded on the movement of the work vehicle and the external information.

According to an embodiment for solving the above problems, the work judgment method can be prepared outside of a work vehicle that moves in a field and performs work, and also prepares external information independent of the operation status of the work vehicle. and estimating the type of work based on movement information recorded on the movement of the work vehicle and external information.

According to one embodiment for solving the above problem, a recording medium stores a job determination program. The job determination program is a job determination program for realizing a function for determining the type of job performed by the work vehicle by being executed by the arithmetic device. This function can be prepared outside the work vehicle that moves in the field and performs work, and also prepares external information independent of the operation status of the work vehicle, movement information recording the movement of the work vehicle, It involves estimating the type of work based on external information.

According to the above aspect, even when operation information of the work vehicle cannot be acquired, the type of work performed in the work vehicle can be estimated.

1 is a diagram showing an example of a configuration of a work judgment system according to an embodiment.
2A is a diagram showing an example of a configuration of a work vehicle according to an embodiment.
2B is a block circuit diagram showing an example of a configuration of a data collection unit according to an embodiment.
3 is a block circuit diagram showing an example of a configuration of an on-vehicle terminal according to an embodiment.
4 is a block circuit diagram showing an example of a configuration of a server according to an embodiment.
5 is a block circuit diagram showing an example of a configuration of a terminal according to an embodiment.
6 is a flowchart showing an example of a configuration of a job determination method and a job determination program according to an embodiment.
7 is a diagram illustrating an example of display by a terminal according to an embodiment.
8 is a block circuit diagram showing an example of a configuration of a server according to an embodiment.
9 is a flowchart showing an example of a configuration of a job determination method and a job determination program according to an embodiment.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT With reference to the accompanying drawings, modes for implementing a job judgment system, a job judgment method, and a recording medium according to the present invention will be described below.

(1st Embodiment)

The work judgment system, work judgment method, and recording medium according to the present embodiment can estimate the type of work performed by the work vehicle based on information other than information indicating the internal state of the work vehicle during work. As a result, even if the work vehicle does not support a specific support system, it is possible to simplify and more reliably record the work performed in the past. Here, the type of work includes, for example, planting, competition, harrowing, planting, fertilization, control, harvesting, and the like.

As shown in FIG. 1 , the job determination system 1 includes an on-vehicle terminal 3 and a server 5 . The on-vehicle terminal 3 is mounted on the work vehicle 2 . The on-vehicle terminal 3 and the server 5 are connected to each other via the network 4 . Connection between the on-vehicle terminal 3 and the network 4 is realized by wireless communication. A mobile phone line may be used for this radio communication, and in that case, the mobile phone line may be included in the network 4. In addition, a wireless local area network (LAN) may be used for this wireless communication. The connection between the server 5 and the network 4 is realized by wired communication. Terminals 6 other than the on-vehicle terminal 3 may be additionally connected to the network 4 .

As shown in FIG. 2A , the work vehicle 2 includes a body 21, a work machine 22, a pair of front wheels 23, a pair of rear wheels 24, a steering unit 25, , and a data collection unit 26. An on-vehicle terminal 3 is mounted on the vehicle body 21 .

The work machine 22 is, for example, a device for performing agricultural work, and is supported by the vehicle body 21 so as to be able to move up and down freely. In the example of Fig. 2A, the work vehicle 2 is a tractor, and the work machine 22 is a tilling device that tills the field. The vehicle body 21 is supported by front wheels 23 and rear wheels 24, and moves forward or backward by driving the rear wheels 24 by a power source (not shown). The steering means 25 is a steering or a steering wheel, and an operator can adjust the driving direction of the vehicle body 21 by adjusting the direction of the front wheels 23 by manipulating the steering means 25.

As a modified example of the configuration shown in Fig. 2A, the work vehicle 2 may be of an integral type such as a combine. In other words, the work vehicle 2 of the present embodiment is not limited to a combination of a tractor and a work implement 22 detachably attached to the tractor.

As shown in Fig. 2B, the data collection unit 26 includes a bus 28, an interface device 27, and a plurality of sensors 29-1, 29-2, ..., 29-N. . When a plurality of sensors 29-1, 29-2, ..., 29-N are not distinguished, these are simply written as sensors 29. Each sensor 29 is connected to an interface device 27 via a bus 28 . The interface device 27 is configured to be connectable to the on-vehicle terminal 3 .

The plurality of sensors 29 generate sensor signals representing the states of a plurality of devices formed in the work vehicle 2 . In other words, the sensor 29 generates a sensor signal as operation information representing the operation condition of the work vehicle 2 . The operating condition of the work vehicle 2 is, at least, an internal state related to the work of the work vehicle 2, such as the driving state of the driving source of the tractor and the operating state of the transmission mechanism that transmits the driving force of the driving source to the work machine 22. includes The operation information indicating the operating state of the work vehicle 2 includes, for example, drive source information indicating the driving state of the drive source, transmission mechanism information indicating the operating state of the transmission mechanism, engine speed indicating the engine speed of the work vehicle 2 Rotation speed information, PTO rotation speed information indicating the number of revolutions of PTO (Power Take Off), hitch height information and lift arm angle information indicating the lifting state of the implement 22 relative to the work vehicle 2, etc. include

However, in this embodiment, the case where the kind of work is estimated without using operation information is demonstrated. In this case, the data collection unit 26 can be omitted. The case where the type of work is estimated using the data collection unit 26 and operation information will be described later as another embodiment.

As shown in FIG. 3 , the on-vehicle terminal 3 includes a bus 31, an operation information acquisition unit 32, an arithmetic unit 33, a storage unit 34, an external storage unit 35, and , a positional information acquisition means 36 and a communication device 37 are provided. Operation information acquisition means 32, arithmetic device 33, storage device 34, external storage device 35, positional information acquisition means 36, and communication device 37 communicate with each other via bus 31. Communication is possible.

The on-vehicle terminal 3 may be a so-called computer. In other words, the functions of the on-vehicle terminal 3 may be realized when the arithmetic unit 33, such as a CPU (Central Processing Unit), executes a predetermined program stored in the storage device 34. . This program may be read out from the recording medium 351 by the external storage device 35 and stored in the storage device 34, or may be received from the outside via the communication device 37 and stored in the storage device 34. It may be stored in .

The operation information acquisition means 32 is configured to be able to acquire sensor signals from the data collection unit 26 . Communication between the on-vehicle terminal 3 and the data collection unit 26 of the work vehicle 2 is realized by wired communication or wireless communication between the operation information acquisition means 32 and the interface device 27 .

The positional information acquisition means 36 measures the position of the on-vehicle terminal 3 using GNSS (Global Navigation Satellite System) or the like. The positional information acquisition means 36 measures the position of the on-vehicle terminal 3 by measuring the latitude and longitude of the on-vehicle terminal 3 . Since the on-vehicle terminal 3 is mounted on the work vehicle 2, the position of the work vehicle 2 can be measured by measuring the position of the on-vehicle terminal 3. The arithmetic unit 33 acquires positional information indicating the measured latitude and longitude and stores it in the storage unit 34 .

The arithmetic device 33 stores the positional information obtained at a plurality of times in the storage device 34 in association with the measurement time. These plurality of times may have periodicity. As an example, the arithmetic device 33 acquires positional information at a cycle of 1 second, and stores the positional information in the storage device 34 in correspondence with time information indicating the time at which the positional information was acquired.

The start and end of measurement of the positional information may be automatically controlled in conjunction with starting and stopping the engine of the work vehicle 2, respectively.

The communication device 37 transmits the positional information stored in the storage device 34 to the server 5 via the network 4 .

As shown in FIG. 4 , the server 5 includes a bus 51, an arithmetic unit 53, a storage unit 54, an external storage unit 55, and a communication unit 56. The arithmetic unit 53 , the storage unit 54 , the external storage unit 55 , and the communication unit 56 are connected via a bus 51 so that communication with each other is possible.

The arithmetic unit 53 realizes the functions of the server 5 by executing programs stored in the storage unit 54 . In other words, the functions of the server 5 are realized through the cooperation of programs stored in the arithmetic unit 53 and the storage unit 54. This program may be read out from the recording medium 551 by the external storage device 55 and stored in the storage device 54, or may be received from the outside via the communication device 56 and stored in the storage device 54. It may be stored in .

In FIG. 4, the functions of the server 5 realized in this way are shown as means provided to the arithmetic unit 53 for convenience. The arithmetic device 53 includes an external information preparation means 531 and a job estimation means 533 .

External information preparation means 531 is used for estimation of the type of work and prepares external information obtained from the outside of the work vehicle 2 to be readable and output by the arithmetic device 53 . The external information is information independent of the operation status of the work vehicle 2 and is information prepared outside the work vehicle 2 .

One or more pieces of information may be included in the external information. One or more pieces of information included in the external information are referred to as one or more pieces of individual external information. The one or more pieces of individual external information includes one or more pieces of performance information indicating the performance of the work vehicle 2 in terms of specifications, one or more pieces of pavement geographic information indicating the geographical characteristics of the pavement on which the work vehicle 2 works, and land conditions in the pavement. At least one piece of information may be included among one or more pieces of paved land attribute information indicating the characteristics of the pavement and one or more pieces of paved weather information indicating the weather in the pavement. The paved land attribute information includes data based on information obtained by observing the land surface from above with, for example, artificial satellites, airplanes, drones, and the like.

One or more pieces of performance information include maximum horse power information indicating the maximum horse power of the work vehicle 2, driving method information indicating the driving method in which the work vehicle 2 travels with tires and/or crawlers, and work vehicle 2 ) may include at least one of vehicle height information indicating the vehicle height and travel width information indicating the travel width of the work vehicle 2. The piece of pavement geographic information may include at least one of pavement elevation information indicating the elevation of the pavement, pavement soil information indicating the type of soil of the pavement, and pavement area information indicating the area where the pavement is located. One or more types of pavement property information include pavement surface information indicating the nature of the pavement surface (eg, information identifying whether it is concrete, soil, water, or vegetation), and pavement vegetation indicating the cover rate or growth state of the pavement plants. Information (e.g., type of crop, NDVI (Normalized Difference Vegetation Index: Normalized Vegetation Index), feeding rate, visible light, estimated nitrogen content, etc.) . One or more pieces of pavement meteorological information include: pavement precipitation information indicating the amount of precipitation during a predetermined period of the pavement, pavement integrated solar radiation information indicating the accumulated insolation amount of the pavement during a predetermined period of time, and pavement mean temperature indicating an average temperature of the pavement during a predetermined period. Among the information, pavement maximum temperature information indicating the highest temperature of the pavement during a predetermined period, pavement minimum temperature information indicating the lowest temperature of the pavement during a predetermined period, and pavement integrated temperature information indicating the integrated temperature of the pavement during a predetermined period. You may contain at least one.

The job estimating means 533 uses the decision model stored in the decision model storage means 541 of the storage device 54 to estimate the type of job based on positional information and external information. A decision model is a mathematical model used to estimate the type of work by AI (Artificial Intelligence). The decision model is built by machine learning and learns features included in external information. For example, in the case of supervised learning, methods such as support vector machines, random forests, neural networks, logistic regression, and topological data analysis are used, and external information is used as training data to estimate the type of work. In addition, the learning data here includes at least one piece of one or more pieces of individual external information included in the external information as a correct answer label indicating a correct answer to the estimation of the type of job.

The user can use the terminal 6 to confirm the result of estimating the type of job by the job determination system 1 according to the present embodiment. The terminal 6 may be, for example, a tablet type information terminal.

As shown in FIG. 5 , the terminal 6 includes a bus 61, an input/output device 62, an arithmetic device 63, a storage device 64, an external storage device 65, and a communication device. (67) is provided. The input/output device 62 , the arithmetic device 63 , the storage device 64 , the external storage device 65 , and the communication device 67 are connected via a bus 61 so that communication with each other is possible.

Like the on-vehicle terminal 3, the terminal 6 may be a so-called computer. In other words, the function of the terminal 6 may be realized when the arithmetic unit 63, such as a CPU, executes a predetermined program stored in the storage device 64. This program may be read out from the recording medium 651 by the external storage device 65 and stored in the storage device 64, or acquired from the outside via the network 4 and the communication device 67. It may be stored in the storage device 64.

The input/output device 62 includes a touch panel device that is integrated with the display device and allows the user to input. The input/output device 62 may further include input devices such as keys, buttons, and switches for operating the terminal 6 .

The communication device 67 is connected to the server 5 via the network 4 and is capable of obtaining data from the server 5, and conversely, data stored in the storage device 64 is stored in the server 5. ) may be sent.

Referring to Fig. 6, an example of the operation of the job judgment system 1 having the above configuration, that is, a configuration example of the job judgment method according to the present embodiment, will be described. In each step of the job determination method, the arithmetic unit 33 of the on-vehicle terminal 3, the arithmetic unit 53 of the server 5, or the arithmetic unit 63 of the terminal 6 determines the job according to the present embodiment. realized by running the program.

In the first step S01, when the work of the work vehicle 2 is finished and the position information of the work vehicle 2 during work is transmitted from the on-vehicle terminal 3 to the server 5, the server 5 performs the work. The determination method is disclosed. When the job judgment method is started, a judgment model is prepared in the judgment model storage means 541 .

In the second step S02, various types of data used for estimation of the type of work are prepared in the storage device 54. The data prepared here includes position information of the work vehicle 2 and external information prepared outside the work vehicle 2 .

The positional information of the work vehicle 2 is acquired by the positional information acquiring means 36 at a plurality of times while the work vehicle 2 is moving the pavement and performing work, and the on-vehicle terminal 3 stored in the storage device 34. This operation is completed before starting the job determination method in the first step S01.

As described above, the positional information is transmitted from the computing device 33 of the on-vehicle terminal 3 to the computing device 53 of the server 5 via the network 4 . The arithmetic unit 53 of the server 5 stores the received positional information in the storage device 54, and prepares the arithmetic unit 53 to enable reading and output.

As a modified example of positional information, vehicle speed information may be used. In short, the type of work may be estimated using vehicle speed information indicating the vehicle speed of the work vehicle 2 at each of a plurality of times. Here, vehicle speed information may be used instead of location information or may be used in addition to location information.

The calculation of the vehicle speed information may be performed by arithmetic unit 53 of server 5 as vehicle speed calculating means, or by arithmetic unit 33 of on-vehicle terminal 3 . Among a plurality of times at which the location information is measured, the vehicle speed information at a certain first time T1 is, for example, the first time T1 and the second time T2 following that time (or the 0th time T0 before that time). and the first positional information measured at time T1 and the second positional information (or zeroth positional information) measured at second time T2 (or zeroth time T0). Also, the average value of the vehicle speed at each time may be used as the vehicle speed information.

The position information and the vehicle speed information are collectively referred to as movement information recording the movement of the work vehicle 2 . In the case where the vehicle speed information can be used in conjunction with the location information or in the case where the vehicle speed information can be used instead of the location information, the location information or the vehicle speed information may be read as movement information.

As described above, the external information is prepared so that the external information preparation means 531 of the server 5 can be read and output by the arithmetic unit 53.

In third step S03, the job estimating means 533 of the server 5 estimates the type of job of the work vehicle 2. At this time, the job estimating means 533 inputs individual external information and positional information into the decision model read out from the decision model storage means 541 in order to estimate the type of job.

In the above example, the job estimation means 533 inputs maximum horsepower information as individual external information, time information derived from positional information, and vehicle speed information into the third decision model C. Here, the maximum horsepower information indicates that the maximum horsepower in the specification of the work vehicle 2 is "100 PS (horsepower)", the time information indicates that the season is "spring", and the vehicle speed information indicates that the work vehicle 2 A case in which the average vehicle speed during work indicates that "2 kilometers per hour" will be described.

Further, the maximum horsepower information is, for example, based on an identification signal transmitted from the on-vehicle terminal 3 to the server 5, from a database previously stored in the storage device 54 of the server 5, the server The arithmetic unit 53 of (5) may read out. Here, for example, the identification signal includes model number information indicating the model number of the work vehicle 2, and the model number information is read out by the on-vehicle terminal 3 from the data collection unit 26, and is stored in the database. Information indicating maximum horsepower information for each model number may be included. In addition, the time information is obtained from the time information included in the positional information transmitted from the on-vehicle terminal 3 to the server 5 and a database previously stored in the storage device 54 of the server 5. ) may be read out by the arithmetic unit 53. Here, the date on which the positional information was obtained may be included in the time information, and information indicating the season for each date may be included in the database.

Under the precondition that the value of pavement area information is "Kanto region" and the value of pavement soil information is "water and cabbage", the third judgment model C has the value of visual information is "spring", and , In the case where the value of the vehicle speed information is "2 kilometers per hour", the frequency of the maximum horsepower of the work vehicle 2 being "100 PS or more" is high when performing competition work, and when performing stationary work, this maximum horsepower is " 50 horsepower to 70 horsepower or more” is learned by machine learning. As a result, the 3rd judgment model C outputs if the certainty (certainty) that the type of job is "stop" is 55% and the certainty that the type of job is "competition" is 45%. Record each of the combinations in which the degree of certainty corresponds to the type of outputted work as an estimation result.

In fourth step S04, the job estimating means 533 of the server 5 records each estimation result in the storage device 54.

In fifth step S05, the job estimating means 533 of the server 5 determines the type of job. In detail, the job estimating means 533 determines that among the estimation results obtained in the third step S03 and recorded in the fourth step S04, the type of job associated with the highest degree of certainty is the type of the actually performed job. and this determination result is recorded in the storage device 54.

In sixth step S06, the terminal 6 displays and informs the user of the determination result. In detail, the job estimating means 533 of the server 5 transmits to the terminal 6 a signal for notifying the terminal 6 that the determination result is prepared. Upon receiving this signal, the terminal 6 receives the determination result and estimation result from the storage device 54 of the server 5 and stores them in the storage device 64. The terminal 6 notifies the user that the determination result and estimation result are ready to be displayed, by means of a display or sound. The user operates the terminal 6 to display the determination result on the touch panel of the input/output device 62 .

As shown in Fig. 7, the input/output device 62 of the terminal 6 displays map information 621 indicating the pavement on which the work has been performed, and a window 622 showing the result of determination of the type of work performed on this pavement. may indicate In this window 622, for all estimation results including judgment results, the name of the type of estimated job, the degree of certainty of this estimation, and the names and values of representative information input to the judgment model for making this estimation. may be displayed. In this window 622, the date and time zone on which the estimated job was performed may be further displayed.

In the example of Fig. 7, it is shown that the type of the estimated job is "stopped" and that the degree of certainty of this estimation is "55%". Further, as another result of estimation, it is shown that the type of the estimated job is "competition" and that the degree of certainty of this estimation is "45%". Here, in order to show that "stop" is the judgment result among these estimation results, the terminal 6 surrounds the portion of "stop 55%" with a frame and highlights it.

In the example of FIG. 7 , as representative information used for estimation, the window 622 indicates that the vehicle speed is “2”, the maximum horsepower is “100”, and the season is “Spring”.

Here, the user can correct the judgment result of the type of job if it is wrong while confirming the display on the terminal 6. In the example of FIG. 7 , the user performs an operation such as tapping the “Competition 45%” part on the touch panel to correct the job type determination result from “Stop” to “Competition”. According to the user's operation, the computing device 63 of the terminal 6 transmits to the server 5 a correction signal for correcting the judgment result of the job to "match". The server 5 corrects the determination result of the type of job stored in the storage device 54 from "stop" to "competition" based on this correction signal.

When 6th step S06 is completed, the flowchart of FIG. 6 ends in 7th step S07.

As described above, the job judgment system 1 according to the present embodiment realizes each step of the job judgment method by executing the job judgment program, so that even if the operation information of the work vehicle 2 cannot be acquired, the work judgment system 1 executes the job judgment program. Based on the location information of the vehicle 2 and external information, the type of work of the work vehicle 2 can be estimated. In addition, the recording of jobs performed in the past can be simplified and performed more reliably.

(Second Embodiment)

This embodiment is a modified example of the first embodiment to which the following changes are added. That is, in the first embodiment, the description has been given of a configuration in which only one decision model is used for job estimation, but in the present embodiment, a plurality of decision models are prepared in the decision model storage means 541, and these plurality of decision models are provided. One decision model is selected from among them and used for estimation of the task.

As shown in FIG. 8 , the server 5 according to the present embodiment is obtained by adding the following changes to the server 5 according to the first embodiment shown in FIG. 4 . That is, the arithmetic device 53 further has a judgment model selection means 532. Decision model selection means 532 selects a decision model to be used for estimating the type of job from among a plurality of decision models stored in decision model storage means 541 .

A plurality of decision models differ in external information used as training data. For example, in the first judgment model A of the judgment models, learning is performed using learning data in which pavement soil information is colored. Moreover, the 2nd judgment model B different from the 1st judgment model A is learning using data for learning whose field soil information is water rice and cabbage.

In the present embodiment, when the decision model selection means 532 selects a decision model from among a plurality of decision models, the external information preparation means 531 selects a decision model that has been learned using training data suitable for individual external information prepared. do. By using the judgment model selected in this way for job estimation, it can be expected that excellent accuracy is obtained in job estimation.

Referring to Fig. 9, an example of the operation of the job judgment system 1 having the above configuration, that is, a configuration example of the job judgment method according to the present embodiment, will be described. In each step of the job determination method, the arithmetic unit 33 of the on-vehicle terminal 3, the arithmetic unit 53 of the server 5, or the arithmetic unit 63 of the terminal 6 determines the job according to the present embodiment. realized by running the program.

1st step S11 and 2nd step S12 are the same as 1st step S01 and 2nd step S02 of 1st Embodiment shown in FIG. 6, respectively.

In the third step S13, the decision model selection means 532 of the server 5 selects a decision model to be used for estimating the type of job. Decision model selection means 532 selects, from a plurality of decision models stored in decision model storage means 541, a decision model suitable for individual external information included in the external information stored in storage device 54. .

When a decision model in which learning is performed by using individual external information and correct answer labels as training data is used, the type of task can be estimated with high accuracy even when individual external information not used for learning is input to the judgment model. can The process of estimating the type of work from such external information is expressed as the individual external information being suitable for the correct answer label.

As an example, a case in which three decision models each having different combinations of individual external information suitable for the correct answer label are stored in the decision model storage means 541 will be described. The first decision model A is suitable for the correct answer label of the combination of pavement soil information being water and pavement area information being Kyushu region. The second decision model B is suitable for the correct answer label of a combination in which the field soil information is water rice and cabbage, and the field area information is the Kyushu region. The third judgment model C is suitable for the correct answer label of a combination in which the field soil information is water rice and cabbage and the field area information is the Kanto region. Here, field soil information indicates crops that have been cultivated in the field in the past, and indicates that the field has soil characteristics suitable for growing the crop. Further, the pavement area information indicates the district where the pavement is located.

In the above example, among the individual external information prepared for the task for which the type is to be estimated, if the pavement area information is "Kanto region" and the pavement soil information is "water and cabbage", individual external information suitable for the correct answer label. The decision model selection means 532 selects the third decision model C with which ?

In fourth step S14, the job estimating means 533 of the server 5 estimates the type of job of the work vehicle 2. 4th step S14 of this embodiment adds the following change to 3rd step S03 of 1st embodiment. That is, the job estimating means 533 inputs individual external information and positional information into the decision model selected in the third step S13 in order to estimate the type of job.

5th step S15, 6th step S16, 7th step S17, and 8th step S18 are 4th step S04, 5th step S05, 6th step S06, and 7th step of 1st embodiment shown in FIG. 6, respectively. Same as S07.

As described above, the job determination system 1 according to the present embodiment can select a determination model suitable for a correct answer label from among a plurality of determination models as a determination model used to estimate the type of job. Therefore, compared to the case of the first embodiment, it can be expected that excellent precision is obtained in the estimation of the work.

(Third Embodiment)

In the present embodiment, as information used to estimate the type of work of the work vehicle 2, in addition to the location information of the work vehicle 2 and external information prepared outside the work vehicle 2, the work vehicle (2) Operation information indicating the operation status of is additionally used.

The configuration of the job judgment method and the job judgment program according to the present embodiment is obtained by adding the following changes to the flow chart of the first embodiment shown in FIG. 6 or the flow chart of the second embodiment shown in FIG. . First, while the work vehicle 2 is working, the on-vehicle terminal 3 acquires operation information in addition to positional information. This operation is performed in 1st step S01 of FIG. 6 or 1st step S11 of FIG.

The operation information includes, as an example, drive source information indicating the driving state of the drive source, transmission mechanism information indicating the operating state of the transmission mechanism, engine speed information indicating the engine speed of the work vehicle 2, and engine speed of the work vehicle 2. Engine load factor information indicating the load factor, PTO rotation speed information indicating the number of revolutions of the PTO that transmits power from the work vehicle 2 to the work machine 22, and a hitch indicating the lifting state of the work machine 22 relative to the work vehicle 2 Height and lift arm angle information are included.

Next, in 2nd step S02 of FIG. 6 or 2nd step S12 of FIG. 9, operation information is transmitted to arithmetic unit 53 of server 5 similarly to positional information, and is stored in storage unit 54. do. In 3rd step S13 of FIG. 9, decision model selection means 532 uses the operation information stored in the memory|storage device 54 in addition to the positional information and external information when selecting a decision model. In 3rd step S03 of FIG. 6 or 4th step S14 of FIG. 9, the job estimation means 533 inputs operation information into a judgment model in addition to positional information and external information, and estimates the kind of job. In sixth step S06 of FIG. 6 or seventh step S17 of FIG. 9 , arithmetic unit 63 of terminal 6 displays operation information as representative information used for estimation together with the result of determination in window 622 . may be displayed on

The work judgment system 1 according to the present embodiment realizes each step of the work judgment method by executing a work judgment program, thereby determining the work of the work vehicle 2 with higher precision when operation information cannot be obtained. type can be inferred. In addition, the recording of jobs performed in the past can be simplified and performed more reliably.

As mentioned above, although this invention was concretely demonstrated based on each embodiment, this invention is not limited to the above-described embodiment, and various changes are possible within the range which does not deviate from the summary. In addition, each feature described in the above-described embodiment can be freely combined within a range that does not contradict technically.

As a modified example of the work judgment system 1 according to each embodiment, the on-vehicle terminal 3 and the terminal 6 may be the same terminal. In other words, the terminal 6 detachably connected to the work vehicle 2 may be used as the on-vehicle terminal 3 . When the terminal 6 is used as the on-vehicle terminal 3, the terminal 6 additionally includes various components provided to realize the functions of the on-vehicle terminal 3. The input/output device 62 further includes an interface device for connecting to the data collection unit 26.

As a modification of the first embodiment in which operation information is not used, when the data collection unit 26 is omitted, the operation information acquisition means 32 of the on-vehicle terminal 3 can also be omitted.

Each of the recording medium 351, recording medium 551, and recording medium 651 may be a non-transitory tangible medium.

A plurality of methods for the external information preparation unit 531 to prepare external information in the storage device 54 will be described. The external information preparation means 531 may obtain external information from another server connected to the network 4 via the network 4 and the communication device 56 and store it in the storage device 54, or may store the information in a recording medium ( 551), external information may be acquired via the external storage device 55 and stored in the storage device 54, or may be read out from a database previously stored in the storage device 54. External information may be registered in the storage device 54 in advance. In either case, the external information preparation means 531 can prepare external information by reading out the external information from the storage device 54 .

As a modified example of the sixth step S06 of the first embodiment and the seventh step S07 of the second embodiment, when the job type determination result by the job estimation means 533 of the server 5 is modified by the user. Alternatively, the judgment model used for estimation of the type of work may be updated. Updating the decision model may also include subjecting the decision model to additional learning using other training data. Moreover, in 2nd Embodiment, update of a decision model may also include a change of the criterion and/or method for selecting the decision model to be used from among a plurality of decision models in the third step S13.

This application claims priority based on Japanese Patent Application No. 2020-191565 for which it applied on November 18, 2020, and takes in all of the indication here.

Claims (12)

External information preparation means for preparing external information that can be prepared outside of a work vehicle that moves in the field and performs work, and that is independent of the operation status of the work vehicle;
and a job estimation means for estimating the type of the job based on movement information recording movement of the work vehicle and the external information. According to claim 1,
The movement information,
including positional information indicating a position of the work vehicle at each of a plurality of times;
The job determination system further comprising a positional information acquiring means for acquiring the positional information. According to claim 1 or 2,
The movement information,
A work determination system including vehicle speed information indicating a vehicle speed of the work vehicle at each of a plurality of times. According to any one of claims 1 to 3,
The external information includes one or more individual external information,
The one or more pieces of individual external information may include one or more pieces of performance information indicating the performance of the work vehicle, one or more pieces of pavement geographic information indicating the geographical characteristics of the pavement, and information obtained by observing an indicator from above to determine the pavement. A work judgment system comprising at least one of one or more pieces of pavement land attribute information indicating characteristics of land conditions in the field and one or more pieces of pavement weather information indicating conditions in the field. According to claim 4,
The one or more pieces of performance information include maximum horsepower information indicating maximum horsepower in the specifications of the work vehicle, driving method information indicating a driving method of the work vehicle, garage information indicating a garage of the work vehicle, and Includes at least one piece of driving width information indicating a driving width;
The one or more pavement geographical information includes at least one of pavement elevation information indicating the elevation of the pavement, pavement soil information indicating the type of soil of the pavement, and pavement area information indicating a region where the pavement is located; ,
The one or more pavement meteorological information includes pavement precipitation information indicating the amount of precipitation in the pavement for a predetermined period, pavement integrated insolation information indicating the accumulated insolation amount of the pavement for a predetermined period, and average temperature of the pavement for a predetermined period of time. Pavement mean temperature information indicating, pavement maximum temperature information indicating the highest temperature of the pavement during a predetermined period, pavement minimum temperature information indicating the lowest temperature of the pavement during a predetermined period, and accumulated temperature of the pavement during a predetermined period A work judgment system including at least one of the pavement integration temperature information indicating According to claim 4 or 5,
further comprising decision model storage means for storing a decision model that has been trained to estimate the type of the task using learning data including at least one of the one or more pieces of individual external information as a correct answer label for supervised learning;
wherein the job estimation means estimates the type of the job using the determination model. According to claim 4 or 5,
Decision model storage means for storing a plurality of decision models that have been trained to estimate the type of the task using learning data including at least one of the one or more pieces of individual external information as a correct answer label for supervised learning;
further comprising decision model selection means for selecting one decision model from among the plurality of decision models based on the individual external information included in the external information;
wherein the job estimation means estimates the type of the job using the judgment model selected by the judgment model selection means. According to any one of claims 1 to 7,
wherein the job estimating means estimates the type of the job further based on operation information indicating the operation status of the work vehicle. According to claim 8,
The operation information includes engine speed information representing the engine speed of the working vehicle, engine load factor information representing the engine load factor of the working vehicle, and PTO (Power Take Off) for transmitting power from the working vehicle to an external work machine. : PTO rotation speed information indicating the number of revolutions of power take-off), hitch height information and lift arm angle information indicating the lifting state of the work machine relative to the work vehicle. According to any one of claims 1 to 9,
A display device for displaying the result of estimating the type of work together with the name and value of the information used for the estimation;
The job judgment system further comprising an input device that accepts an input for correcting a result of estimating the type of the job. Preparing external information that can be prepared outside of a work vehicle that moves in the field and performs work, and is independent of the operation status of the work vehicle;
and estimating the type of the job based on movement information recording the movement of the work vehicle and the external information. A recording medium storing a job determination program for realizing a function for determining a type of job by a work vehicle by being executed by an arithmetic device,
The function is
Preparing external information that can be prepared outside of the work vehicle that moves in the field and performs the work, and is independent of the operation status of the work vehicle;
and estimating the type of the work based on the movement information recording the movement of the work vehicle and the external information.

Images