Drainage determination method and hydrophobic material loading device

Abstract

To provide a drainage determination method which can reduce the burden of filling work with a hydrophobic material, efficiently form an underdrain, and eliminate the consumption of unnecessary hydrophobic material, thereby reducing wasteful costs.SOLUTION: A drainage determination method 60 can determines whether the drainage of the ground in a field is good or bad before a long culvert trench is formed in the field, and includes training data acquisition step S51, model generation step S52, N-value measurement step S61, first image for determination capturing step S62-1, input data for determination acquisition step S63-1, determination step S64, poor determination transmission step S65, hydrophobic material filling step S66, and closing step S67.SELECTED DRAWING: Figure 5

Description

[Technical Field]

[0001] The present invention relates to a drainage determination method for determining whether drainage is good or bad in an area where an underdrain is to be formed, and a hydrophobic material filling device used for the method. [Background technology]

[0002] Conventionally, drainage culverts have been constructed to improve drainage in agricultural fields. To construct the culverts, multiple long ditches are excavated in the land and hydrophobic materials such as rice husks and straw are poured into the ditches. However, because the hydrophobic material was often added to the trenches by hand, it was difficult to add it evenly to each trench. This resulted in waste of hydrophobic material. Furthermore, adding the hydrophobic material after the trenches had been excavated was inefficient. Therefore, in recent years, techniques have been developed to solve the above problems, and inventions relating to these have already been disclosed.

[0003] Patent Document 1 discloses an invention entitled "Bullet Culvert Forming Method" that relates to a method for forming a bullet culvert filled with a hydrophobic material. The invention disclosed in Patent Document 1 is characterized by comprising a work implement attachment section mounted and supported on a blade provided at the front of a bulldozer, a hydrophobic material hopper provided integrally with the work implement attachment section, a hydrophobic material feeding section connected to the bottom of the hydrophobic material hopper, a groove forming blade provided ahead of the hydrophobic material feeding section in the direction of excavation of the bullet culvert, a hydrophobic material feeding inlet provided below the hydrophobic material feeding section behind the direction of excavation of the bullet culvert, a bullet provided behind the hydrophobic material feeding inlet in the direction of excavation of the bullet culvert, a hydrophobic material supply hopper provided above the hydrophobic material hopper, and a conveying means for feeding the hydrophobic material from the hydrophobic material feed hopper into the hydrophobic material hopper, the operation of the conveying means being linked to the operation of forming the bullet culvert so that the hydrophobic material is fed into the hydrophobic material hopper only when the bullet culvert is formed at a predetermined position. According to the invention having such a configuration, the hydrophobic material can be efficiently and uniformly poured and filled into the bullet culvert at the same time as the formation of the bullet culvert, and wasteful consumption of the hydrophobic material can be suppressed. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent No. 4997324 Summary of the Invention [Problem to be solved by the invention]

[0005] In the invention disclosed in Patent Document 1, waste of hydrophobic material is reduced by limiting the timing at which the hydrophobic material is added to the hydrophobic material hopper, but the hydrophobic material is added to all of the multiple excavated trenches. However, since the quality of drainage generally depends on the composition of the ground, etc., it is thought that there may be cases where filling the excavated trench with hydrophobic material is not necessary depending on the composition of the ground where the culvert is to be formed. Therefore, filling the entire trench with hydrophobic material without considering the properties of the ground not only imposes a heavy workload on the filling work and is inefficient, but also results in waste of money due to the consumption of unnecessary hydrophobic material. Therefore, the invention disclosed in Patent Document 1 may not be able to fully solve the problems of the prior art.

[0006] The present invention has been made in response to such conventional circumstances, and aims to provide a drainage assessment method and a hydrophobic material filling device that can reduce the burden of filling hydrophobic material, form a culvert efficiently, and eliminate the consumption of unnecessary hydrophobic material, by determining in advance whether the drainage of the ground where a culvert is to be formed is good or bad. [Means for solving the problem]

[0007] In order to achieve the above object, a first invention is a drainage assessment method for assessing whether the drainage of the ground in a field is good or poor before a long trench for a culvert is formed in the field, and is characterized by comprising: a model generation step in which a model generation unit performs machine learning to generate a model from training data in which a first image taken by a photographing means of the area where the culvert is to be formed and an N value measured in the area by a measuring instrument are used as input data, and the good or poor drainage of the area is used as output data; and a assessment step in which a judgment unit judges whether the drainage is good or poor based on the model, corresponding to the judgment input data including the first judgment image taken of the judgment area by the photographing means and the judgment N value measured in the judgment area by the measuring instrument, and outputs the judgment result.

[0008] In the invention configured as described above, the area refers to each of the multiple sections of the field into which one or more culverts will be formed, and the judgment area refers to the area where one or more new culverts are to be formed, and is the area for which the quality of drainage is to be judged. The first image is specifically a digital image taken by a camera. In this digital image, the more poorly the ground drains and the more water covers the ground, the more likely it is that the range showing high-brightness pixel values ​​will increase. Therefore, the first image is thought to indicate whether the drainage is good or poor.

[0009] Furthermore, the N-value is an engineering property of the ground obtained by a standard penetration test, and generally the harder the ground is compacted, the higher the value. Ground that contains a high proportion of clayey soil tends to have a low N-value and poor drainage. In contrast, ground that contains a high proportion of sandy soil tends to have a high N-value and good drainage. Therefore, the N-value is considered to be a factor that indicates the quality of drainage.

[0010] Therefore, by using machine learning to learn the relationship between good and poor drainage for the input data, which is a combination of the first image and the N value, and generating a model, it becomes possible to determine whether the drainage in the judgment area is good or poor based on this model.

[0011] The second invention is characterized in that, in the first invention, when a judgment result indicating poor drainage is output, a hydrophobic material filling process is performed after the judgment process, and the hydrophobic material filling process is characterized in that a drive unit drives a bottom lid that closes the lower end opening of a hopper into which the hydrophobic material is filled, changes its position to open the lower end opening, and fills the hydrophobic material into the long groove.

[0012] In the invention having such a configuration, in addition to the function of the first invention, when the drainage is judged to be poor, the drive unit drives the bottom cover to open the closed lower end opening and fill the long groove with hydrophobic material, whereas when the drainage is judged to be good, the drive unit does not drive the bottom cover, in which case the lower end opening remains closed and the long groove is not filled with hydrophobic material.

[0013] The third invention is characterized in that, in the first or second invention, the input data includes, in addition to the first image and the N value, a second image of the area photographed at a second time point later than the first time point when the first image was photographed, and weather conditions for at least a predetermined period before the first time point; the judgment input data includes, in addition to the first judgment image and the judgment N value, a second judgment image of the judgment area photographed at a second judgment time point later than the first judgment time point when the first judgment image was photographed, and judgment weather conditions for at least a predetermined period before the first judgment time point; and the weather conditions and judgment weather conditions each include at least precipitation.

[0014] In such an invention, in addition to the effects of the first or second invention, a model is generated by learning training data in which the combination of the first image, the N value, the second image, and weather conditions is used as input data, and the output data is whether drainage is good or bad. As described above, the reason why the input data further includes the second image and weather conditions is that when drainage is good, the ground dries easily, and when drainage is poor, the ground is difficult to dry, so the quality of drainage affects the amount of water that accumulates on the ground over time, and this change over time is thought to be affected by weather conditions such as the amount of precipitation.

[0015] The input data may include weather conditions for a predetermined period before the first time point, as well as weather conditions for a period from the first time point to the second time point. Similarly, the input data for determination may include weather conditions for determination for a predetermined period before the first time point, as well as weather conditions for determination for a period from the first time point to the second time point. The length of the predetermined period before the first time point and the length of the predetermined period before the first determination time point may or may not be the same. The weather conditions and the weather conditions for determination may include the outside temperature and wind speed in addition to the amount of precipitation.

[0016] Therefore, based on the above model, for example, even if there is a lot of precipitation, ground with a large change in the amount of water pooling over time has good drainage, and conversely, ground with a small change in the amount of water pooling over time even if there is little precipitation has poor drainage. In this way, whether drainage is good or poor is determined based on weather conditions.

[0017] The fourth invention is an invention used to implement the second invention, and comprises a bottom lid that is provided to open or close the bottom opening of the hopper, and a drive unit that drives the bottom lid to change its position, and is characterized in that when a determination result indicating poor drainage is output, the drive unit drives the bottom lid that closes the bottom opening, changes its position to open the bottom opening, and fills the long groove with hydrophobic material.

[0018] In the invention having such a configuration, the bottom cover may include, for example, a plate that covers the lower end opening from below, and a rotation shaft formed on a part of the periphery of the plate, and the drive unit may include, for example, a motor that rotates the rotation shaft of the bottom cover. In the invention of the above configuration, when a judgment result indicating poor drainage is output, the drive unit enables the hydrophobic material stored in the hopper to be discharged, thereby realizing the hydrophobic material filling process, which is the second invention. [Effects of the Invention]

[0019] According to the first invention, since it is possible to determine whether the drainage in the determination area is good or bad, it is possible to omit the work of filling long grooves that have good drainage and are considered to not require filling with hydrophobic material. Therefore, according to the first invention, the burden of filling work can be reduced and the underdrain can be formed efficiently, and unnecessary consumption of hydrophobic material can be eliminated, thereby reducing wasteful costs.

[0020] According to the second invention, in addition to the effects of the first invention, the hydrophobic material is filled into the long groove only when the drainage is judged to be poor, so that good drainage performance above a certain level and cost reduction can be achieved at the same time in all of the multiple culverts.

[0021] According to the third invention, in addition to the effects of the first or second invention, the quality of drainage is determined based on weather conditions, making it possible to more accurately determine whether or not filling is necessary.

[0022] According to the fourth aspect of the invention, it is possible to achieve the same effects as those of the second aspect of the invention. [Brief explanation of the drawings]

[0023] [Figure 1] 1 is a diagram illustrating the configuration of a drainage determining device and a hydrophobic material filling device for carrying out a drainage determining method according to an embodiment. [Figure 2] FIG. 2 is a side view of the hydrophobic material filling device. [Figure 3] 10 is a flowchart showing the process up to when the drainage determination device generates a model. [Figure 4] 10 is a list of training data used by the drainage determination device. [Figure 5] 1 is a flowchart of a drainage determination method according to an embodiment. [Figure 6] 10 is a list of training data used in a drainage determination method according to a modified example of the embodiment. [Figure 7] 10 is a flowchart of a drainage determination method according to a modified example of the embodiment. DETAILED DESCRIPTION OF THE INVENTION [Example]

[0024] First, a drainage determination device and a hydrophobic material filling device for carrying out a drainage determination method according to an embodiment of the present invention will be described with reference to Fig. 1. Fig. 1 is a configuration diagram of the drainage determination device and the hydrophobic material filling device for carrying out a drainage determination method according to an embodiment. As shown in Fig. 1, the drainage determination device 1 is a device that determines whether the drainage of the ground in a farm field is good or bad before forming a long underdrain trench in the farm field, and includes an acquisition unit 2, an output unit 3, a control unit 4, and a memory unit 5. Specifically, the drainage determination device 1 is a computer.

[0025] The control unit 4 also includes a model generation unit 6 and a determination unit 7. The control unit 4 is a central processing unit that controls all operations of the acquisition unit 2 through the storage unit 5. Furthermore, the storage unit 5 includes a teacher data storage unit 8, a model storage unit 9, a judgment input data storage unit 10, and a judgment result storage unit 11. Each component will be described below in order.

[0026] The acquisition unit 2 acquires training data that the model generation unit 6 uses to generate a model, and judgment input data for an object for which the judgment unit 7 judges whether the drainage is good or bad. The training data is a first image of the area where the underdrain is to be formed, and the drainage quality corresponding to the N value measured in this area. The input data for judgment is a judgment image of the judgment area, and the judgment N value measured in the judgment area.

[0027] The first image and the image for determination are digital images captured by the imaging means 20. The N value and the N value for determination are the same type of physical quantity measured by the measuring instrument 21. Specifically, the imaging means 20 is a camera, and the measuring instrument 21 is a cone penetrometer. Therefore, the acquisition unit 2 acquires the above-mentioned first image etc. and the N value etc. from the photographing means 20 and the measuring instrument 21 via the Internet N. In this embodiment, the weather data server 22 is not required.

[0028] The output unit 3 outputs the determination result of whether drainage is good or bad output by the determination unit 7. Specifically, the output unit 3 is a transmission unit that transmits the determination result that drainage is bad (hereinafter referred to as a bad determination) to the hydrophobic material filling device 30 described below. In addition, the output unit 3 includes a display screen and a printer. That is, the output unit 3 does not transmit the determination result that the drainage is good (hereinafter referred to as a good determination) to the hydrophobic material filling device 30. In this way, the control unit 4 determines whether or not to transmit the determination result to the hydrophobic material filling device 30 depending on the content of the determination result.

[0029] The model generation unit 6 generates a model by performing machine learning on the training data acquired by the acquisition unit 2. The judgment unit 7 judges whether the drainage corresponding to the judgment input data is good or bad based on the model generated by the model generation unit 6, and outputs a judgment of good or bad drainage.

[0030] The teacher data storage unit 8 stores the teacher data acquired by the acquisition unit 2, and the model storage unit 9 stores the model generated by the model generation unit 6. Furthermore, the judgment input data storage unit 10 stores the judgment input data acquired by the acquisition unit 2, and the judgment result storage unit 11 stores the judgment result output by the judgment unit 7.

[0031] Next, the hydrophobic material filling device 30 includes a bottom cover 31 , a driving unit 32 , and a receiving unit 33 . Of these, the receiver 33 receives the defective judgment output by the output unit 3 of the drainage judgment device 1 via the wireless communication path 34. Thereafter, the receiver 33 transmits the received defective judgment to the driver 32.

[0032] When the defective determination is transmitted, the drive unit 32 drives the bottom lid 31 to change its position. The bottom lid 31 is a plate provided at the bottom opening 44a (see FIG. 2) of a hopper 44 that stores the hydrophobic material. The detailed configurations and operations of the bottom lid 31 and the drive unit 32 will be described with reference to FIG. 2.

[0033] Next, the configuration and operation of the hydrophobic material filling device will be described with reference to Fig. 2. Fig. 2 is a side view of the hydrophobic material filling device. As shown in Figure 2, the hydrophobic material filling device 30 includes a bottom lid 31 that is arranged to open or close the lower end opening 44a of an existing hopper 44 that stores hydrophobic material M, a drive unit 32 that drives the bottom lid 31 to change its position, a receiving unit 33 (see Figure 1) that receives a defect determination via a wireless communication path 34 (see Figure 1) and transmits it to the drive unit 32, and a holding unit 35 that holds the bottom lid 31 and the receiving unit 33 on one side 44b of the hopper 44. The hydrophobic material M is specifically rice husk.

[0034] Here, we will first explain the configuration of the hopper 44. The hopper 44 is attached to the tip of an arm 41 provided on a hydraulic excavator 40 that moves in a farm field F. In detail, the hopper 44 is held at the tip of the arm 41 via an attachment frame 42 together with an excavation blade 43 that excavates a long trench D of the culvert. The bottom opening 44a of the hopper 44 is inclined with respect to the vertical direction of the hopper 44, so that the hydrophobic material M can be discharged over a wide range in the depth direction of the long groove D. The bottom opening 44a has an inner space that is rectangular in plan view.

[0035] Furthermore, a bullet-shaped hole forming member 46 is attached to the lower end of the excavation blade 43 via a chain 45 of a predetermined length. Therefore, when the excavation blade 43 moves in the traveling direction X, the hole forming member 46 is pulled and a lower groove portion D1 is formed. Note that this lower groove portion D1 is not filled with the hydrophobic material M even when the hydrophobic material M is filled in the long groove D.

[0036] Next, a detailed description will be given of the configuration of the hydrophobic material filling device 30. The bottom cover 31 is made of metal and includes a rectangular plate 31a capable of closing the bottom opening 44a inside the hopper 44, and a cylindrical rotating shaft 31b formed on part of the periphery of the plate 31a. The rotating shaft 31b is disposed outside the hopper 44, and its axial direction is perpendicular to both side surfaces 44b, 44b of the hopper 44.

[0037] The drive unit 32 is an electric motor that includes a shaft 32a that is inserted through the inner space of the rotary shaft 31b and rotates the rotary shaft 31b around the axial center of the shaft 32a. The electric motor receives power from a battery that is newly installed or already installed in the hydraulic excavator 40 via a cable (not shown).

[0038] Furthermore, the holding unit 35 is made of metal and includes a cylindrical unit 35a that houses the drive unit 32, and a substantially trapezoidal fixing plate 35b that hangs and fixes the cylindrical unit 35a to one side surface 44b of the hopper 44. Note that the holding unit 35 may also include a cylindrical unit that rotatably holds the tip of the shaft 32a, and a fixing plate that hangs and fixes the cylindrical unit to the other side surface 44b.

[0039] In addition, the hydrophobic material filling device 30 is equipped with a level sensor 36 that detects the remaining amount of hydrophobic material M. This level sensor 36 is provided on the inner wall of the hopper 44. Specifically, the level sensor 36 is installed above the bottom cover 31 that closes the lower end opening 44a, and outputs a hydrophobic material shortage signal when the remaining amount of hydrophobic material M is almost gone.

[0040] Therefore, in the hydrophobic material filling device 30, when the judgment unit 7 outputs a defective judgment, the drive unit 32 drives the rotation shaft 31b of the bottom cover 31 that closes the lower end opening 44a in the Y1 direction to change the position of the plate body 31a and open the lower end opening 44a. As a result, the hydrophobic material M stored in the hopper 44 is filled into the long groove D.

[0041] Furthermore, the drive unit 32 is configured to rotate the shaft 32a to close the bottom opening 44a when the receiver 33 receives the hydrophobic material shortage signal output by the level sensor 36. Therefore, after filling, when the hydrophobic material M in the hopper 44 is almost gone, the rotation shaft 31b of the bottom cover 31 rotates in the Y2 direction, and the plate 31a closes the bottom opening 44a again. Furthermore, the drive unit 32 is configured so that the bottom cover 31 opens or closes the bottom opening 44a of the hopper 44 regardless of the determination result when the receiver 33 receives a switch operation by the operator of the hydraulic excavator 40. This makes it possible to open or close the bottom opening 44a as needed, for example, when the hydraulic excavator 40 is moving or in an emergency, or when the hydrophobic material M is being replenished into the hopper 44.

[0042] Next, a description will be given of the flow of processing up to the generation of a model by the drainage determination device 1. Fig. 3 is a flowchart up to the generation of a model by the drainage determination device. As shown in FIG. 3, a model generation method 50 executed by the drainage determination device 1 includes a training data acquisition step S51 and a model generation step S52. The teacher data acquisition step of S51 is a step in which the acquisition unit 2 acquires teacher data that is a combination of input data consisting of a first image and an N value, and output data that corresponds to the input data and indicates whether the drainage is good or bad. The acquired teacher data is stored in the teacher data storage unit 8.

[0043] The model generation step of S52 is a step in which the model generation unit 6 performs machine learning to generate a model from the input data and output data acquired by the acquisition unit 2. This machine learning is performed using a known learning method. The generated model is stored in the model storage unit 9.

[0044] Next, the contents of the training data will be explained using Fig. 4. Fig. 4 is a list of training data used by the drainage determination device. As shown in FIG. 4, the "first image number" in the first column is a serial number assigned to the first image, and is a number that identifies each of the multiple areas where a culvert is to be formed. The "N value" in the second column is the N value measured for each area identified by the first image number. Therefore, multiple culverts to be built in one area will have the same N value, and multiple culverts to be built in different areas will have the same or different N values.

[0045] The third column, "Good or bad drainage," is data indicating the correct answer regarding drainage, corresponding to each combination of a plurality of "first image numbers" and a plurality of "N values." Therefore, by machine learning the training data as described above, the model generation unit 6 can correctly determine whether the drainage is good or poor for a new determination area where it is necessary to determine whether the drainage is good or poor.

[0046] Next, the steps of the drainage determining method carried out using the drainage determining device 1 and the hydrophobic material filling device 30 will be described with reference to Fig. 5. Fig. 5 is a flowchart of the drainage determining method according to the embodiment. As shown in Figure 5, the drainage assessment method 60 includes a training data acquisition process S51, a model generation process S52, an N-value measurement process for assessment S61, a first image capture process for assessment S62-1, an input data acquisition process for assessment S63-1, a assessment process S64, a defective assessment transmission process S65, a hydrophobic material filling process S66, and a closing process S67.

[0047] The training data acquisition step S51 and the model generation step S52 of the model generation method 50 may be performed before the determination step S64 is started. Also, once the model generation method 50 is performed, it does not necessarily have to be performed each time the drainage determination method 60 is performed. Furthermore, the bottom cover 31 of the hydrophobic material filling device 30 closes the bottom opening 44a of the hopper 44 when the drainage determining method 60 is started. Each step will be described below in order.

[0048] The judgment N-value measurement step of S61 is a step in which an operator measures the judgment N-value in the judgment region using the measuring device 21. As this judgment N-value, for example, a representative value of multiple measurement values ​​measured at multiple locations in the judgment region can be used.

[0049] The first determination image capturing step of S62-1 is a step in which an operator captures an image of the determination area using the image capturing means 20 to obtain a first determination image. Alternatively, the first determination image may be obtained by remotely operating the image capturing means 20 mounted on an aircraft such as a drone, instead of the operator.

[0050] The determination input data acquisition step of S63-1 is a step in which the acquisition unit 2 acquires determination input data consisting of a determination N value and a first determination image. The acquired determination input data is stored in the determination input data storage unit 10.

[0051] The judgment step of S64 is a step in which the judgment unit 7 judges whether the drainage is good or bad, corresponding to the first judgment image and the judgment input data including the judgment N value, based on the model generated by the model generation unit 6, and outputs a judgment result of either good or bad. The judgment result is stored in the judgment result storage unit 11. If a defective judgment is output, the defective judgment transmission step of S65 is executed. In this defective judgment transmission step of S65, the output unit 3 transmits the defective judgment to the hydrophobic material filling device 30. On the other hand, if a good judgment is output, the output unit 3 does not transmit this good judgment to the hydrophobic material filling device 30. Therefore, the steps after the hydrophobic material filling step of S66 are not executed, and the drainage judgment method 60 ends.

[0052] The hydrophobic material filling process of S66 is a process in which the drive unit 32 drives the bottom cover 31 of the hydrophobic material filling device 30, which closes the lower end opening 44a of the hopper 44, changes its position to open the lower end opening 44a, and fills the hydrophobic material M into the long groove D.

[0053] The closing process of S67 is a process in which the drive unit 32 drives the bottom cover 31 with the lower end opening 44a open in response to a hydrophobic material shortage signal output by the level sensor 36 or a switch operation by the operator, thereby closing the lower end opening 44a again.

[0054] As explained above, according to the drainage assessment method 60, by using the drainage assessment device 1 and the hydrophobic material filling device 30, only when the assessment unit 7 outputs a defective assessment, the drive unit 32 of the hydrophobic material filling device 30 opens the lower end opening 44a of the hopper 44 to enable the filling of the hydrophobic material M. Therefore, it is possible to omit the work of filling the long grooves D that have good drainage and are considered unnecessary to fill with the hydrophobic material M. Therefore, according to the drainage determination method 60, the burden of filling work can be reduced and an underdrain can be formed efficiently, and unnecessary consumption of the hydrophobic material M can be eliminated, thereby reducing unnecessary costs.

[0055] Furthermore, the hydrophobic material M is filled into the long groove D only when the judgment unit 7 outputs a poor judgment, so that good drainage performance above a certain level and cost reduction can be achieved simultaneously in all of the multiple culverts. In addition, the judgment unit 7 judges whether the drainage in the judgment area is good or poor based on the first image that is thought to indicate good or poor drainage and a model generated using training data with the N value as input data, so that it is possible to obtain accurate judgment results that reflect the characteristics of the ground in the judgment area.

[0056] Next, a drainage determination method according to a modified example of the embodiment will be described with reference to Figures 6 and 7. Figure 6 is a list of training data used in the drainage determination method according to a modified example of the embodiment. In the drainage determination method according to the modified example of the embodiment, training data different from the training data used in the drainage determination method 60 shown in Fig. 4 is used. As a result, the model generated by the model generation unit 6 is also different from the model used in the drainage determination method 60.

[0057] As shown in Figure 6, in the drainage determination method according to the modified example, the input data of the training data includes a first image (first column) identified by a first image number, an N value (second column), a second image (fifth column) identified by a second image number, and weather conditions (third and fourth columns). Of these, the first image numbers "A1 to A5" identify areas of one field, and "A100 to A104" identify areas of fields different from the first field. Furthermore, the second image is taken at a second time point that is later than the first time point at which the first image was taken, and is of the same area as the first image, and the second image numbers "B1 to B5" and "B100 to B104" correspond to the first image numbers "A1 to A5" and "A100 to A104", respectively, in terms of the areas taken.

[0058] Furthermore, the weather conditions are the amount of precipitation and the outside temperature for a predetermined period before the first time point, and are acquired by the acquisition unit 2 from the weather data server 22 (see FIG. 1) via the Internet N. However, the weather conditions do not have to include the outside temperature, and may also include other weather conditions such as wind speed. Note that any period can be selected as the predetermined period. The rightmost column, "Drainage good / poor," which is the output data of the training data, is data that indicates the correct answer regarding drainage, corresponding to each combination of multiple "first image numbers," multiple "second image numbers," multiple "N values," and multiple "weather conditions." In the drainage determination method according to the modified example, the generation method up to the point where the drainage determination device 1 generates a model using the above-mentioned training data is the same as the model generation method 50.

[0059] Therefore, the input data for determination also includes the second image for determination and the meteorological conditions for determination during a predetermined period before the first time point for determination, in addition to the first image for determination and the N value for determination. Of these, the second judgment image is an image of the same judgment area as in the first judgment image, taken at a second judgment time point that is later than the first judgment time point at which the first judgment image was taken. The weather conditions for determination are meteorological information for a predetermined period before the first time point for determination, and the route, type, and predetermined period for acquisition are the same as those for the weather conditions included in the input data.

[0060] Next, the steps of the method for determining drainage according to the modified example of the embodiment will be described with reference to Fig. 7. Fig. 7 is a flowchart of the method for determining drainage according to the modified example of the embodiment. 7, a drainage assessment method 60A according to a modified example of the embodiment includes a second assessment image capturing step of S62-2 and an assessment input data acquisition step of S63-2 between the assessment input data acquisition step of S63-1 and the assessment step of S64 in the drainage assessment method 60. Note that the model generation method 50 is not shown in the figure. Furthermore, the drainage determining method 60A is carried out using the drainage determining device 1 and the hydrophobic material filling device 30, similar to the drainage determining device 1.

[0061] In drainage assessment method 60A, the step of acquiring input data for assessment in S63-1 is a step in which acquisition unit 2 acquires input data for assessment at a first time point for assessment. However, this input data for assessment further includes weather conditions for assessment for a predetermined period before the first time point for assessment. Other than this, this step is the same as the step of acquiring input data for assessment in S63-1 of drainage assessment method 60.

[0062] The second determination image capturing step of S62-2 is a step in which the image capturing means 20 captures an image of the same determination area as in the first determination image at the second determination time point. Other than this, this step is the same as the first determination image capturing step of S62-1 in the drainage determination method 60. Furthermore, the step of acquiring input data for determination in S63-2 is a step in which the acquisition unit 2 acquires second input data for determining the time point for determination. This input data for determination includes only the second image for determination and does not include weather conditions for determination. Other than this, it is the same as the step of acquiring input data for determination in S63-1 of the drainage determination method 60. The other configurations of the drainage determination method 60A are the same as those of the drainage determination method 60.

[0063] According to drainage assessment method 60A, the input data includes first and second images taken at a period of time and weather conditions, so that assessment results can be output based on a model that is thought to reflect the temporal changes in the amount of water pooling on the ground and the factors that affect these temporal changes. Therefore, since the quality of drainage is determined based on weather conditions, it is possible to more accurately determine whether filling is necessary. Other effects of drainage determination method 60A are the same as those of drainage determination method 60.

[0064] The drainage determination device 1, the hydrophobic material filling device 30, and the drainage determination method 60, 60A according to the present invention are not limited to those shown in the examples. For example, the drainage determination device 1 may have the teacher data storage unit 8 provided in a computer separate from the drainage determination device 1. Further, the hydrophobic material filling device 30 may use a slat around which the shaft body 32a of the drive unit 32 can be wound, instead of the bottom cover 31 in the form of a rotating plate. Furthermore, in the drainage determination method 60, the order of execution of the determination N-value measurement step of S61 and the first determination image capture step of S62-1 may be reversed. Additionally, in drainage determination method 60A, the input data may include weather conditions for a period from a first time point to a second time point. In this case, the input data for determination includes weather conditions for determination for a period from the first time point for determination to the second time point for determination, and acquisition unit 2 acquires the weather conditions for determination in the input data for determination acquisition step of S63-2. [Industrial Applicability]

[0065] INDUSTRIAL APPLICABILITY The present invention can be used as a drainage determination method for determining whether drainage is good or poor in an area where an underdrain is to be formed, and as a hydrophobic material filling device used for the method. [Explanation of symbols]

[0066] 1...Drainage determination device 2...Acquisition unit 3...Output unit 4...Control unit 5...Memory unit 6...Model generation unit 7...Determination unit 8...Teacher data memory unit 9...Model memory unit 10...Determination input data memory unit 11...Determination result memory unit 20...Photographing means 21...Measuring instrument 22...Weather data server 30...Hydrophobic material filling device 31...Bottom cover 31a...Plate body 31b...Pivoting shaft 32...Drive unit 32a...Shaft body 33...Receiver 34...Wireless communication path 35...Holding unit 35a...Cylindrical part 35b...Fixing plate 36...Level sensor 40...Hydraulic excavator 41...Arm 42...Mounting frame 43...Digging blade 44...Hopper 44a...Lower end opening 44b...Side surface 45...Chain 46...Hole forming member 50...Model generation method 60,60A...Drainage assessment method F...Field M...Hydrophobic material D...Long ditch D1...Lower ditch section

Claims

1. A drainage assessment method for assessing whether the drainage of ground in a field is good or bad before forming a long trench of an underdrain in the field, comprising: a model generation step in which a model generation unit performs machine learning to generate a model from training data that includes a first image taken by a photographing means of the area where the culvert is to be formed and an N value measured in the area by a measuring instrument as input data, and that includes the good or bad drainage in the area as output data; A drainage assessment method characterized by comprising a assessment step in which the assessment unit assesses the goodness or poorness of the drainage based on the model, corresponding to assessment input data including a first assessment image of the assessment area taken by the photographing means and a assessment N value measured in the assessment area by the measuring instrument, and outputs the assessment result.

2. When the determination result that the drainage is poor is output, a hydrophobic material filling step is executed after the determination step, The drainage determination method described in claim 1, characterized in that the hydrophobic material filling process involves a drive unit driving a bottom lid that closes the lower end opening of a hopper into which the hydrophobic material is filled, changing its position to open the lower end opening, and filling the hydrophobic material into the long groove.

3. the input data includes, in addition to the first image and the N value, a second image of the area captured at a second time point after the first time point at which the first image was captured, and weather conditions for at least a predetermined period before the first time point; the input data for determination includes, in addition to the first image for determination and the N value for determination, a second image for determination obtained by photographing the region for determination at a second time point for determination that is later than the first time point for determination at which the first image for determination was photographed, and meteorological conditions for determination for at least a predetermined period prior to the first time point for determination; 3. The drainage determining method according to claim 1, wherein the weather conditions and the weather conditions for determination each include at least an amount of precipitation.

4. A hydrophobic material filling device used to implement the drainage determination method according to claim 2, a bottom cover provided to open or close the bottom opening of the hopper; a drive unit that drives the bottom cover to change its position, The hydrophobic material filling device is characterized in that when the judgment result that the drainage is poor is output, the drive unit drives the bottom cover that closes the lower end opening, changes its position to open the lower end opening, and fills the hydrophobic material into the long groove.

Images