Pump operating condition estimation device, pump installation location estimation device, estimation method, estimation program, pump unit, and field groundwater level control system

Abstract

To provide a technology that improves deviation of a ground water level of a field.SOLUTION: A pump operation condition estimation device (100) is installed in a well in a field and estimates an operation condition of a pump (50) that drains ground water. The operation condition estimation device (100) comprises an operation condition estimation unit (12) that using an operation condition estimation model (33), inputs a saturated hydraulic conductivity near the well in a target field and water level information representing a current water level of the well in the field and a target water level of the well, and estimates the operation condition of the pump (50).SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a pump operation condition estimation device, a pump installation location estimation device, an estimation method, an estimation program, a pump unit, and a field groundwater level control system.

Background Art

[0002] There are various causes for poor drainage in a field, and corresponding measures are required according to the causes. Among the fields with poor drainage, particularly in the fields where there are water channels that are the flow paths of groundwater, there are significant biases in the groundwater level. In such fields, within one field, for example, there may be a problem that there are simultaneously places where crops are damaged by drought due to sunlight and places where there is so much water that the feet are submerged.

[0003] As a method for eliminating the bias in the groundwater level that affects such poor drainage, methods such as the construction of this underground canal or agricultural drainage are adopted. Also, a method of controlling the groundwater level of arable land using a pump has been reported (Non-Patent Documents 1 and 2).

[0004] Also, outside the field, a method for controlling the groundwater level has been reported. For example, in a rainwater drainage channel or the like, a method of monitoring and controlling a plurality of drainage pumps provided in the drainage channel has been reported (Patent Document 1).

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Non-Patent Documents

[0006]

Non-Patent Document 1

Non-Patent Document 2

[0007] Fields experiencing poor drainage due to uneven groundwater levels cannot be rectified by the pump-based solutions or agricultural drainage methods described in Non-Patent Documents 1 and 2. Therefore, such fields require fundamental civil engineering work, such as the construction of underground drainage systems. However, the construction of these systems is costly and not easily undertaken by farmers. Furthermore, the technology described in Patent Document 1 does not address the unevenness of groundwater levels in fields. Therefore, there is still a need for a technology that can prevent poor drainage in fields by improving the unevenness of groundwater levels.

[0008] One aspect of the present invention aims to provide a technique for improving the uneven distribution of groundwater levels in a field. [Means for solving the problem]

[0009] To solve the above problems, a pump operating condition estimation device according to one aspect of the present invention is a pump operating condition estimation device that estimates the operating conditions of a pump installed in a well in a field for draining groundwater, and includes an operating condition estimation unit that estimates the operating conditions of the pump by using an operating condition estimation model generated by machine learning using training data that associates the saturated hydraulic conductivity near the well, water level information including the water level of the well before, during, and after the pump is operated, and the operating conditions of the pump, with the saturated hydraulic conductivity near the well in the target field and water level information representing the current water level of the well in the field as input.

[0010] A pump unit according to one aspect of the present invention comprises at least one pump installed in a well within a field for draining groundwater from the field, and a control device that controls the operation of the pump based on operating conditions estimated by a pump operating condition estimation device according to one aspect of the present invention.

[0011] A pump installation location estimation device according to one aspect of the present invention is a pump installation location estimation device for estimating the installation location of a pump of a pump unit according to one aspect of the present invention, and includes an installation location estimation unit that estimates the pump installation location information by taking the subsurface environment geographic information of a target field and information representing the distribution of water level in the target field as input, using a pump installation location estimation model generated by machine learning using training data that associates subsurface environment geographic information related to the flow path of groundwater in a field, information representing the distribution of water level in the field after the installation of the pump, information representing the distribution of saturated hydraulic conductivity in the field, and pump installation location information in the field.

[0012] A field groundwater level control system according to one aspect of the present invention comprises a pump operating condition estimation device according to one aspect of the present invention, a pump unit according to one aspect of the present invention, and a pump installation location estimation device according to one aspect of the present invention.

[0013] A pump operating condition estimation method according to one aspect of the present invention is a pump operating condition estimation method for estimating the operating conditions of a pump installed in a well in a field for draining groundwater, and includes an operating condition estimation step in which the operating conditions estimation model is generated by performing machine learning using training data that associates the saturated hydraulic conductivity near the well, water level information including the water level of the well before, during, and after the pump is operated, and the operating conditions of the pump, and takes the saturated hydraulic conductivity near the well in the target field, water level information representing the current water level of the well in the field and the target water level of the well as input.

[0014] A pump installation location estimation method according to one aspect of the present invention is a pump installation location estimation method for estimating the installation location of a pump whose operation is controlled based on the operation conditions estimated by a pump operation condition estimation method according to one aspect of the present invention, and includes an installation location estimation step of estimating the pump installation location information using a pump installation location estimation model generated by machine learning using training data that associates subsurface environment geographic information related to the flow path of groundwater in the field, information representing the distribution of water level in the field after the installation of the pump, information representing the distribution of saturated hydraulic conductivity in the field, and pump installation location information in the field, with the subsurface environment geographic information of the target field and information representing the distribution of water level in the target field as input.

[0015] A pump operating condition estimation program for causing a computer to function as an operating condition estimation unit according to one aspect of the present invention is a pump operating condition estimation program for causing a computer to function as a pump operating condition estimation device according to one aspect of the present invention.

[0016] A pump installation location estimation program for causing a computer to function as an installation location estimation unit according to one aspect of the present invention is a pump installation location estimation program for causing a computer to function as a pump installation location estimation device according to one aspect of the present invention. [Effects of the Invention]

[0017] According to one aspect of the present invention, a technique can be provided to improve the uneven distribution of groundwater levels in a field. [Brief explanation of the drawing]

[0018] [Figure 1] This figure illustrates an example of the main components of a field groundwater level control system according to one aspect of the present invention. [Figure 2] This is a block diagram showing an example of the main components of a pump unit according to one aspect of the present invention. [Figure 3] This flowchart shows an example of the estimation process flow in an operating condition estimation device according to one aspect of the present invention. [Figure 4] It is a flowchart showing an example of the flow of estimation processing in an installation location estimation device according to one aspect of the present invention.

Embodiments for Carrying Out the Invention

[0019] 〔1. Field groundwater level control system〕 A field groundwater level control system according to one aspect of the present invention is a system that is installed in a well in a field and controls the groundwater level of the field by a pump that drains the groundwater of the field. An aspect of the groundwater level control system will be described with reference to FIG. 1. FIG. 1 is a block diagram showing an example of the main configuration of a field groundwater level control system according to one aspect of the present invention.

[0020] As shown in FIG. 1, the groundwater level control system 1000 includes an operation condition estimation device (pump operation condition estimation device) 100, an installation location estimation device (pump installation location estimation device) 200, and a pump unit 500. Further, the groundwater level control system 1000 includes an operation condition estimation model generation device 300, an installation location estimation model generation device 400, a storage device 60, an output device 70, and an input device 80.

[0021] The storage device 60 stores programs and data used in the groundwater level control system 1000. The storage device 60 stores, as an example, various data input via the input device 80. Further, the storage device 60 stores, as an example, teacher data and generated learning models used for generating learning models in the operation condition estimation model generation device 300 and the installation location estimation model generation device 400. Furthermore, the storage device 60 stores, as an example, learning models, input information, and output information used for prediction in the operation condition estimation device 100 and the installation location estimation device 200. The storage device 60 may have a database for storing various data on a cloud or a server.

[0022] The output device 70 outputs the results estimated by the operating condition estimation device 100 and the installation location estimation device 200. The mode of output by the output device 70 is not particularly limited. The output device 70 may be, for example, a display device that displays the information representing the estimation results as an image, or an audio output device that outputs the information as audio. Alternatively, the output device 70 may be the display of a mobile device such as a smartphone that displays the estimation results.

[0023] The input device 80 receives input operations from the user for the groundwater level control system 1000. For example, the input device 80 receives input of training data used to generate a learning model in the operating condition estimation model generation device 300 and the installation location estimation model generation device 400. The input device 80 also receives input of data used for estimation processing in the operating condition estimation device 100 and the installation location estimation device 200.

[0024] (Operating condition estimation device 100) The operating condition estimation device 100 is an estimation device installed in a well within a field and used to estimate the operating conditions of a pump that drains groundwater. The operating condition estimation device 100 uses the operating condition estimation model 33 generated by the operating condition estimation model generation device 300 (described later) to estimate the operating conditions of the pump 50 provided in the pump unit 500 (described later).

[0025] The operating condition estimation device 100 includes a control unit 10. The control unit 10 controls all parts of the operating condition estimation device 100. The control unit 10 is implemented, for example, by a processor and memory. In this example, the processor accesses storage (not shown), loads a program (not shown) stored in storage into memory, and executes a series of instructions contained in the program. This constitutes the various parts included in the control unit 10. These parts include an input data acquisition unit 11 and an operating condition estimation unit 12.

[0026] The input data acquisition unit 11 acquires the saturated hydraulic conductivity near a well installed in the target field for which the operating conditions of the pump 50 are to be estimated, as well as water level information representing the current water level of the well in the field and the target water level of the well. For example, the input data acquisition unit 11 receives input of saturated hydraulic conductivity and water level information from the input device 80. The input data acquisition unit 11 may also acquire the saturated hydraulic conductivity near a well installed in the target field and the current water level of the well in the field, which have been measured by the pump unit 500, from the pump unit 500. Furthermore, the input data acquisition unit 11 may read the saturated hydraulic conductivity and water level information stored in the storage device 60. The input data acquisition unit 11 may acquire saturated hydraulic conductivity and water level information based on an input signal representing an instruction to start acquiring input data from the input device 80. The input data acquisition unit 11 outputs the acquired saturated hydraulic conductivity and water level information to the operating condition estimation unit 12.

[0027] Here, the water level in wells within a field correlates with the groundwater level within the field. If the water levels in multiple wells within a field differ, it can be determined that there is a bias in the groundwater level within that field. On the other hand, if the water levels in multiple wells within a field are roughly the same, it can be determined that there is no bias in the groundwater level within that field and that the groundwater level is uniform. In this way, the distribution of the groundwater level within a field can be obtained based on the water levels in wells within the field.

[0028] Furthermore, the saturated hydraulic conductivity near a well installed in the field is intended to indicate the water permeability, which shows how easily groundwater infiltrates into the well, and can represent information that shows the delay time between the change in groundwater level inside the well and around the well. The saturated hydraulic conductivity may also be an apparent saturated hydraulic conductivity expressed by the amount of change in water level within a predetermined time period of water introduced into a well in the field. The saturated hydraulic conductivity is affected by the material and density of the soil and rock. Therefore, the saturated hydraulic conductivity may differ from point to point within the field. It is preferable to calculate the saturated hydraulic conductivity for each well equipped with a pump 50.

[0029] The wells installed in the field are not particularly limited and may consist of relatively small diameter pipes. Alternatively, existing wells in the field may be used, or new wells may be installed. From the viewpoint of easily changing the location of the pump 50, it is preferable that the wells be of a size that can be easily moved.

[0030] The operating condition estimation unit 12 estimates the operating conditions of the pump 50 based on the saturated hydraulic conductivity and water level information acquired by the input data acquisition unit 11, using the operating condition estimation model 33. When the operating condition estimation unit 12 acquires the saturated hydraulic conductivity and water level information, it acquires the operating condition estimation model 33 from the operating condition estimation model generation device 300 or the storage device 60.

[0031] As described later, the operating condition estimation model 33 is a trained model generated by machine learning using training data that associates the saturated hydraulic conductivity near the well with water level information including the water level of the well in the field before, during, and after operation of the pump 50, and the operating conditions of the pump 50. In other words, the operating condition estimation model 33 is a trained model for estimating the operating conditions of the pump 50 in order to bring the current water level of the well to the target water level in the target field. Furthermore, the operating condition estimation model 33 can also be a trained model for estimating the operating conditions of the pump 50 in order to equalize the groundwater level in the target field. Details of the operating condition estimation model 33 will be described later.

[0032] The operating condition estimation unit 12 inputs the saturated hydraulic conductivity near the well, the current water level of the well in the field, and water level information representing the target water level of the well as input information to the operating condition estimation model 33, thereby obtaining the operating conditions of the pump 50 as an estimated result. This makes it possible to estimate the operating conditions of the pump 50 to bring the current water level of the well to the target water level, taking into account the saturated hydraulic conductivity near the well.

[0033] The operating conditions for a pump used to change the groundwater level may vary depending on the water permeability of the groundwater. Furthermore, the water permeability of the groundwater may vary from field to field or from point to point within a field. The operating condition estimation device 100 estimates the operating conditions of the pump 50 by taking into account the saturated hydraulic conductivity, which represents the water permeability of the groundwater, and therefore it is possible to estimate the optimal operating conditions for the pump to change the groundwater level as desired.

[0034] The operating condition estimation unit 12 outputs information regarding the estimated operating conditions of the pump 50 to the pump unit 500. The operating condition estimation unit 12 may also store the information regarding the estimated operating conditions of the pump 50 in the storage device 60 or output it via the output device 70. The information regarding the operating conditions of the pump 50 may, for example, be the operating time of the pump 50 or the amount of wastewater discharged per unit time by the pump 50.

[0035] In the operating condition estimation unit 12, the water level information input to the operating condition estimation model 33 may include information on environmental factors of the field that affect the water level of the wells. In this case, the operating condition estimation model 33 is a trained model generated by machine learning using training data that associates the saturated hydraulic conductivity near the wells, the water level of the wells in the field before, during, and after the operation of the pump 50, as well as information on environmental factors of the field, with the operating conditions of the pump 50.

[0036] This allows the operating conditions of pump 50 to be estimated by further considering environmental factors of the field, in addition to the saturated hydraulic conductivity near the well and the water level in the well. Examples of information regarding environmental factors include the crops grown in the field and the groundwater level suitable for their growth, weather information in the field, the types of work performed in the field, the light-dark cycle in the field, and topographic information such as reservoirs and irrigation / drainage channels surrounding the field.

[0037] For example, if the water level information includes the crops grown in the field and the groundwater level suitable for their growth, the operating condition estimation unit 12 can estimate the pump operating conditions necessary to achieve the groundwater level suitable for each crop. Also, if the water level information includes meteorological information such as rainfall in the field, the unit can estimate the pump operating conditions necessary to achieve the desired groundwater level, taking the rainfall into consideration.

[0038] Furthermore, in the operating condition estimation unit 12, the water level information input to the operating condition estimation model 33 may also include information representing the distance between adjacent pumps 50 in a field where multiple pumps 50 are installed. In this case, the operating condition estimation model 33 is a trained model generated by machine learning using training data that associates the saturated hydraulic conductivity near the well, the water level of the wells in the field before, during, and after the operation of the pumps 50, water level information that further includes information representing the distance between adjacent pumps 50 in the field, and the operating conditions of the pumps 50.

[0039] The operating condition estimation unit 12 can more appropriately estimate the operating conditions of pump 50 by further considering the distance between adjacent pumps 50 in the field, in addition to the saturated hydraulic conductivity near the well and the water level in the well. The groundwater level at the location where one pump is installed may change due to the operation of other adjacent pumps. In a well, if the water in the well is emptied by the operation of pump 50, the difference in water level with the surrounding groundwater level becomes large, causing groundwater to flow towards the emptied well and groundwater to flow into the well. The hydraulic gradient, which is the driving force of groundwater flow caused by such a difference in groundwater level in the field, is affected by the presence of other adjacent pumps 50. That is, if the distance to other pumps is short, the influence of the operation of those pumps on the hydraulic gradient will be large, and if the distance is far, the influence of the operation of those pumps on the hydraulic gradient may be small. Since the operating condition estimation unit 12 estimates the operating conditions of pump 50 by considering the distance between adjacent pumps 50, it can estimate the optimal operating conditions of the pump to change the groundwater level as desired, taking into account the change in the hydraulic gradient.

[0040] Furthermore, the information input to the operating condition estimation model 33 in the operating condition estimation unit 12 may include information representing the operating method of the pump 50. In this case, the operating condition estimation model 33 is a trained model generated by machine learning using training data that associates the saturated hydraulic conductivity near the well, water level information including the water level of the well in the field before, during, and after the operation of the pump 50, information representing the operating method of the pump 50, and the operating conditions of the pump 50.

[0041] This allows the operating conditions of the pump 50 to be estimated by further considering the operating method of the pump 50, in addition to the saturated hydraulic conductivity near the well and the water level in the well. Examples of operating methods for the pump 50 include a method of continuous, slow operation, and an intermittent operation method that repeatedly starts and stops the pump. Since the operating condition estimation unit 12 estimates the operating conditions of the pump 50 by considering the operating method of the pump 50, it is possible to estimate the optimal operating conditions for the pump that match the operating method.

[0042] (Installation location estimation device 200) The installation location estimation device 200 is an estimation device that estimates the installation location of the pump 50 of the pump unit 500. The installation location estimation device 200 estimates the installation location of the pump 50, which is controlled to operate under the operating conditions estimated by the operating condition estimation device 100.

[0043] The installation location estimation device 200 includes a control unit 20. The control unit 20 controls all parts of the installation location estimation device 200. The control unit 20 is implemented, for example, by a processor and memory. In this example, the processor accesses storage (not shown), loads a program (not shown) stored in storage into memory, and executes a series of instructions contained in the program. This constitutes the various parts included in the control unit 20. These parts include an input data acquisition unit 21 and an installation location estimation unit 22.

[0044] The input data acquisition unit 21 acquires subsurface environment geographic information related to the groundwater flow path of the field to be used to estimate the installation location of the pump 50, distribution information representing the distribution of water levels within the target field, and distribution information representing the distribution of saturated hydraulic conductivity within the field. For example, the input data acquisition unit 21 receives input of subsurface environment geographic information and distribution information from the input device 80. Alternatively, the input data acquisition unit 21 may read subsurface environment geographic information and distribution information stored in the storage device 60. The input data acquisition unit 21 may acquire subsurface environment geographic information and distribution information based on an input signal representing an instruction to start acquiring input data from the input device 80. The input data acquisition unit 21 outputs the acquired subsurface environment geographic information and distribution information to the installation location estimation unit 22.

[0045] Here, subsurface environmental geographic information related to the groundwater channels in a field may include, for example, information representing a map showing the water channels in the field, information representing a map showing the field's construction history, aerial photographs of the field, and information regarding areas that farmers have judged to be unsuitable for crop growth or prone to water accumulation. In fields with poor drainage, the groundwater level is affected by the groundwater channels within the field, resulting in a distribution of varying groundwater levels, with some areas being high and others low, and the groundwater level is not uniform. Subsurface environmental geographic information for a field may correspond to this distribution of groundwater levels within the field.

[0046] The installation location estimation unit 22 uses the installation location estimation model (pump installation location estimation model) 43 to estimate the installation location for the pump 50 based on the underground environmental geographic information and distribution information acquired by the input data acquisition unit 21. Once the installation location estimation unit 22 has acquired the underground environmental geographic information and distribution information, it obtains the installation location estimation model 43 from the installation location estimation model generation device 400 or the storage device 60.

[0047] As described later, the installation location estimation model 43 is a trained model generated by machine learning using training data that associates subsurface environmental geographic information related to the flow path of groundwater in the field, the distribution of water level in the field after the installation of the pump 50, and the distribution of saturated hydraulic conductivity in the field with information on the installation location of the pump 50 in the field. In other words, the installation location estimation model 43 is a trained model for estimating the installation location of the pump 50 in order to achieve a desired distribution of groundwater level in the target field. Furthermore, the installation location estimation model 43 can be a trained model for estimating the installation location of the pump 50 in order to make the groundwater level in the target field uniform. Details of the installation location estimation model 43 will be described later.

[0048] The installation location estimation unit 22 inputs the subsurface environment geographic information of the target field and information representing the distribution of water levels within the field as input information to the installation location estimation model 43, thereby obtaining the installation location information of the pump 50 as an estimation result. This makes it possible to estimate the installation location of the pump 50 in order to make the distribution of the groundwater level in the field the desired distribution, taking into account the subsurface environment geographic information. The installation location of the pump to change the distribution of the groundwater level in the field may differ depending on the subsurface environment geographic information. Since the installation location estimation unit 22 estimates the installation location of the pump 50 taking into account the subsurface environment geographic information, it is possible to estimate the optimal installation location of the pump 50 in order to change the distribution of the groundwater level as desired.

[0049] The installation location estimation unit 22 may store the estimated installation location information of the pump 50 in the storage device 60, or it may output it via the output device 70. The information regarding the installation location of the pump 50 may, for example, be information representing position coordinates. Furthermore, the information regarding the installation location of the pump 50 may include information such as the number of pumps 50 to be installed in the field and the distance between adjacent pumps 50.

[0050] In the installation location estimation unit 22, the information input to the installation location estimation model 43 may include the number of pumps 50 to be installed. In this case, the installation location estimation model 43 is a trained model generated by machine learning using training data that associates subsurface environmental geographic information related to the flow path of groundwater in the field, the distribution of water level in the field after the installation of the pumps 50, and the distribution of saturated hydraulic conductivity in the field with the number of pumps 50 to be installed and information regarding the installation locations of the pumps 50 in the field.

[0051] This allows for the estimation of the installation location information for the pumps 50, taking into account the number of pumps 50 to be installed.

[0052] The installation location estimation device 200 can estimate the optimal installation location for the pump 50 to control the groundwater level in the field. For example, installing the pump in a location with a high groundwater level, where water easily accumulates, or where the groundwater has good permeability is more efficient for controlling the groundwater level in a field than installing the pump in a location with an appropriate groundwater level, where water does not easily accumulate, or where the groundwater has poor permeability. However, it is not easy to understand the distribution of groundwater levels and groundwater permeability and install the pump in the appropriate location. The installation location estimation device 200 estimates the installation location of the pump 50 by taking into account underground environmental geographic information, so it can estimate the optimal installation location for the pump to change the distribution of groundwater level as desired.

[0053] (Pump unit 500) The pump unit 500 is installed in a well within the field and includes a pump 50 for draining groundwater from the field, and a control device 51 that controls the operation of the pump 50 based on the operating conditions estimated by the operating condition estimation device 100. Details of the pump unit 500 will be described with reference to Figure 2. Figure 2 is a diagram illustrating an example of the main components of a pump unit 500 according to one embodiment of the present invention. As shown in Figure 2, the pump unit 500 may further include a water level gauge 52.

[0054] Pump 50 drains groundwater from the field. Pump 50 may consist of a single pump or, as shown in Figure 2, a group of multiple pumps. Pump 50 is installed at the location estimated by the installation location estimation device 200. The number of installed pumps 50 is also the number estimated by the installation location estimation device 200. The number of pumps 50 may also be predetermined considering the size of the field, cost, etc.

[0055] The type of pump 50 can be selected as appropriate, and examples include submersible drainage pumps, peristaltic pumps, and vacuum pumps. From the perspective of outdoor use, it is preferable that the pump 50 has a robust structure that is resistant to muddy water, has excellent weather resistance, and is energy-efficient.

[0056] The control device 51 controls the operation of the pump 50 based on the operating conditions estimated by the operating condition estimation device 100. The control device 51 also acquires the well water level obtained by the water level gauge 52 and outputs it to the operating condition estimation device 100 and the operating condition estimation model generation device 300. Furthermore, the control device 51 may calculate the saturated hydraulic conductivity of the soil in which the well is located based on the well water level obtained by the water level gauge 52. One control device 51 may be provided for each group of pumps 50, or one may be provided for each of the multiple pumps 50.

[0057] When multiple pumps 50 are installed in a field, as shown in Figure 2, the operation of one pump 50 affects the hydraulic gradient of the other pump 50 between adjacent pumps 50. In Figure 2, each symbol indicates the following: V: Volume of the well drilled in the field (L) 3 ) k: Hydraulic conductivity (L / T) h: Well water level (L) L: Distance between wells (L) Q: Specified drainage volume (L 3 ) = P × T Δq: Variable drainage volume (L 3 ) Furthermore, P and T used to calculate Q represent the following: P: Discharge volume per unit time by the pump (L 3 / T) T: Time required for drainage by pump (T).

[0058] Here, the specified drainage rate Q is a value calculated from the drainage rate P by the pump and the drainage time T by the pump, and depends on the permeability coefficient k, representing the amount of water that flows into the well under normal conditions. As shown in Figure 2, the water level h2 in a well equipped with pump 50b is affected by the operation of pump 50a. As a result, the specified drainage rate Q by pump 50b changes. Furthermore, the degree of this effect depends on the hydraulic gradient i, which is a value calculated by dividing the water level difference between adjacent pumps 50 by the distance between the pumps 50. The same applies to pumps 50c and 50d.

[0059] Furthermore, the variable drainage rate Δq is the amount of drainage that fluctuates due to weather conditions such as rainfall or drought. As shown in Figure 2, the amount of drainage from each pump 50 is determined by the specified drainage rate Q and the variable drainage rate Δq. In other words, the amount of drainage from each pump 50 is also affected by environmental factors of the field, such as weather conditions.

[0060] Thus, the amount of water drained by pump 50 is affected by the operation of adjacent pumps 50, and the amount of this influence depends on the hydraulic gradient between the pumps 50, as well as the environmental factors of the field. Therefore, by controlling pump 50 with the control device 51 based on the operating conditions estimated by the operating condition estimation device 100, taking into account the distance between adjacent pumps 50 and the environmental factors of the field, pump 50 can be appropriately controlled.

[0061] The water level gauge 52 measures the water level in the well. The water level gauge 52 may measure the water level continuously or periodically at predetermined intervals. The water level gauge 52 may output the measured water level to the control device 51 wirelessly or via a wired connection. The water level measured by the water level gauge 52 may be stored in the storage device 60 via the control device 51, thereby allowing the change in water level over time to be obtained. One water level gauge 52 may be provided for each group of pumps 50, or one may be provided for each of the multiple pumps 50.

[0062] The pump unit 500 may also include an evaluation unit (not shown) that evaluates the operating conditions of the pump 50 based on fluctuations in the groundwater level measured by a water level gauge 52. The control device 51 may also include an evaluation unit mode that evaluates the operating conditions of the pump 50 based on fluctuations in the groundwater level measured by a water level gauge 52. The evaluation unit or evaluation unit mode evaluates the operating conditions of the pump 50, as well as the drainage capacity and water level reduction effect of the pump unit 500 at the current installation site, based on the water level of the well before the pump 50 is operated and the water level of the well after the pump 50 is operated. The control device 51 may generate training data for generating an operating condition estimation model (pump operating condition estimation model) 33 by an operating condition estimation model generation device 300 based on the evaluation results from the evaluation unit or evaluation unit mode. The control device 51 may also generate training data for generating an installation site estimation model 43 by an installation site estimation model generation device 400 based on the evaluation results from the evaluation unit or evaluation unit mode. Such evaluation equipment may be removed when the evaluation of the operating conditions of the pump 50 is complete, or when the generation of the training data as described above is complete.

[0063] The pump unit 500 may also be equipped with a sprayer (not shown) for spraying the water drained by the pump 50. The sprayer can spray the water pumped up by the pump 50 to areas where the well water level is low, or spray during seasons when the water level is low. If a sprayer is provided, a water storage tank (not shown) for storing the water pumped up by the pump 50 may also be provided. Alternatively, the water may be directly injected into other wells.

[0064] The pump unit 500 may further include a location information acquisition unit (not shown) that acquires location information of the pump 50. This allows the location information of the pump 50 acquired by the location information acquisition unit to be used to acquire weather information for the field and to generate information on environmental factors.

[0065] The pump unit 500 controls the operation of the pump 50 based on the operating conditions estimated by the operating condition estimation device 100, so that the pump 50 can be operated under operating conditions that take into account the saturated hydraulic conductivity and water level near the well. As a result, the groundwater level in the field can be appropriately controlled.

[0066] Furthermore, the pump unit 500 controls the operation of each pump 50 based on operating conditions estimated by considering the mutual influence of multiple pumps 50, thereby enabling efficient operation of the pumps 50. In addition, the pump unit 500 controls the operation of the pumps 50 based on operating conditions estimated by considering the environmental factors of the field, thereby enabling operation of the pumps 50 under appropriate operating conditions.

[0067] Furthermore, in the pump unit 500, the pump 50 is installed at the location estimated by the installation location estimation device 200, so it is installed at a location that takes into account the groundwater flow path in the field. As a result, the groundwater level in the field can be appropriately controlled.

[0068] In other words, the pump unit 500 operates the pump 50 according to the operating conditions estimated by the operating condition estimation device 100, and the pump 50 is installed at the installation location estimated by the installation location estimation device 200, so that the groundwater level of the field can be controlled efficiently and appropriately. As a result, the groundwater level of the field can be improved without requiring large-scale construction work.

[0069] (Operating condition estimation model generation device 300) The operating condition estimation model generation device 300 generates an operating condition estimation model 33 for estimating the operating conditions of a pump 50 installed in a well within a field to drain groundwater. The operating condition estimation model generation device 300 generates the operating condition estimation model 33 by performing machine learning using training data that associates the saturated hydraulic conductivity near the well, water level information including the water level of the well in the field before, during, and after operation of the pump 50, and the operating conditions of the pump 50.

[0070] The operating condition estimation model generation device 300 includes a control unit 30. The control unit 30 controls all parts of the operating condition estimation model generation device 300. The control unit 30 is implemented, for example, by a processor and memory. In this example, the processor accesses storage (not shown), loads a program (not shown) stored in storage into memory, and executes a series of instructions contained in the program. This constitutes the various parts included in the control unit 30. These parts include a training data acquisition unit 31 and a learning model generation unit 32.

[0071] The training data acquisition unit 31 acquires training data that associates the saturated hydraulic conductivity near the well with water level information including the water level of the well in the field before, during, and after operation of the pump 50, and the operating conditions of the pump 50. For example, the training data acquisition unit 31 accepts training data input from the input device 80. The training data acquisition unit 31 may also acquire training data generated based on the evaluation results of the pump unit 500 by an evaluation machine from the pump unit 500. Furthermore, the training data acquisition unit 31 may read training data stored in the storage device 60. The training data acquisition unit 31 may also acquire training data based on an input signal from the input device 80 indicating an instruction to start training data acquisition. The training data acquisition unit 31 outputs the acquired training data to the learning model generation unit 32.

[0072] The training data acquired by the training data acquisition unit 31 may also include, as water level information, information on environmental factors of the field that affect the water level of the well. Furthermore, the training data acquired by the training data acquisition unit 31 may also include information representing the distance between adjacent pumps 50 in a field where multiple pumps 50 are installed. Furthermore, the training data acquired by the training data acquisition unit 31 may also include the volume V of the well in which the pump 50 is installed, the specified drainage rate Q, the fluctuating drainage rate Δq, etc.

[0073] The water level information may also include information on environmental factors of the field that affect the water level of the well. Examples of information on environmental factors include the crops grown in the field and the groundwater level suitable for their growth, weather information in the field, the types of work performed in the field, the light-dark cycle in the field, and topographic information such as reservoirs and irrigation / drainage channels surrounding the field.

[0074] The learning model generation unit 32 generates an operating condition estimation model 33 by performing machine learning using the training data acquired by the training data acquisition unit 31. The learning model generation unit 32 generates the operating condition estimation model 33 using known machine learning methods such as neural networks, decision trees, random forests, and support vector machines. The operating condition estimation model 33 takes the saturated hydraulic conductivity near the wells in the target field, the current water level of the wells in the field, and water level information representing the target water level of the well as input, and outputs information representing the operating conditions of the pump 50. The learning model generation unit 32 stores the generated operating condition estimation model 33 in the storage device 60.

[0075] The generation of the operating condition estimation model 33 by the operating condition estimation model generation device 300 may be performed based on an input signal from the input device 80 indicating a learning start instruction, or it may be performed periodically at predetermined intervals. The operating condition estimation model generation device 300 may also be configured to regenerate the operating condition estimation model 33 when the water level information included in the training data is updated, or it may regenerate the operating condition estimation model 33 daily. Furthermore, the operating condition estimation model generation device 300 may regenerate the operating condition estimation model 33 each time the season changes.

[0076] (Installation location estimation model generation device 400) The installation location estimation model generation device generates an installation location estimation model 43 that estimates the installation location of the pump 50 of the pump unit 500. The installation location estimation model generation device 400 generates the installation location estimation model 43 by performing machine learning using training data that associates subsurface environmental geographic information related to the flow path of groundwater in the field, the distribution of water level in the field after the installation of the pump 50, and the distribution of saturated hydraulic conductivity in the field with the installation location information of the pump 50 in the field.

[0077] The installation location estimation model generation device 400 includes a control unit 40. The control unit 40 comprehensively controls each part of the installation location estimation model generation device 400. The control unit 40 is implemented, for example, by a processor and memory. In this example, the processor accesses storage (not shown), loads a program (not shown) stored in storage into memory, and executes a series of instructions contained in the program. This constitutes the various parts included in the control unit 40. These parts include a training data acquisition unit 41 and a learning model generation unit 42.

[0078] The training data acquisition unit 41 acquires training data that associates subsurface environmental geographic information related to the flow path of groundwater in the field, the distribution of water level in the field after the installation of the pump 50, and the distribution of saturated hydraulic conductivity in the field with information on the installation location of the pump 50 in the field. For example, the training data acquisition unit 41 accepts training data input from the input device 80. Alternatively, the training data acquisition unit 31 may acquire training data generated based on the evaluation results of the pump unit 500 by an evaluation machine from the pump unit 500. Furthermore, the training data acquisition unit 41 may read training data stored in the storage device 60. The training data acquisition unit 41 may acquire training data based on an input signal from the input device 80 indicating an instruction to start training data acquisition. The training data acquisition unit 41 outputs the acquired training data to the learning model generation unit 42.

[0079] Subsurface environmental geographic information related to the flow of groundwater in a field may include, for example, information representing a map showing the water channels in the field, information representing a map showing the field's development history, aerial photographs of the field, and information regarding areas that farmers have judged to be difficult for crops to grow or prone to water accumulation.

[0080] The learning model generation unit 42 generates an installation location estimation model 43 by performing machine learning using the training data acquired by the training data acquisition unit 41. The learning model generation unit 42 generates the installation location estimation model 43 using known machine learning methods such as neural networks, decision trees, random forests, and support vector machines. The installation location estimation model 43 generated by the installation location estimation model generation device 400 takes environmental geographical information of the target field and information representing the distribution of water levels within the target field as input and outputs information representing the installation location of the pump 50. The learning model generation unit 42 stores the generated installation location estimation model 43 in the storage device 60.

[0081] The location estimation model 43 generated by the location estimation model generation device 400 may be performed based on an input signal from the input device 80 indicating a learning start instruction, or it may be performed periodically at predetermined intervals. In addition, the location estimation model generation device 400 may regenerate the location estimation model 43 when the environmental geographic information included in the training data is updated, or it may regenerate the location estimation model 43 daily.

[0082] (Flow of the process for estimating operating conditions) The flow of the operating condition estimation process (pump operating condition estimation method) by the operating condition estimation device 100 will be explained with reference to Figure 3. Figure 3 is a flowchart showing an example of the flow of the operating condition estimation process in the operating condition estimation device 100 according to one aspect of the present invention.

[0083] First, the input data acquisition unit 11 acquires the saturated hydraulic conductivity near the wells of the target field, the current water level of the wells in the field, and water level information representing the target water level of the well, based on the input signal from the input device 80 that indicates an instruction to start estimation processing (step S1).

[0084] Next, the operating condition estimation unit 12 inputs the acquired saturated hydraulic conductivity and water level information into the operating condition estimation model 33 and obtains the pump's operating conditions (step S2, operating condition estimation process). Then, the operating condition estimation unit 12 outputs the acquired pump operating conditions as the estimation result to the pump unit 500 (step S13), and the operating condition estimation process ends.

[0085] (Flowchart for estimating installation location) The flow of the installation location estimation process (pump installation location estimation method) by the installation location estimation device 200 will be explained with reference to Figure 4. Figure 4 is a flowchart showing an example of the flow of the installation location estimation process in the installation location estimation device 200 according to one aspect of the present invention.

[0086] First, the input data acquisition unit 21 acquires, based on an input signal from the input device 80 indicating an instruction to start estimation processing, subsurface environmental geographic information related to the groundwater flow path of the target field stored in the storage device 60, distribution information representing the distribution of water levels within the target field, and distribution information representing the distribution of saturated hydraulic conductivity within the field (step S11).

[0087] Next, the installation location estimation unit 22 inputs the acquired underground environmental geographic information and distribution information into the installation location estimation model 43 and obtains the output installation location information (step S12). Then, the installation location estimation unit 22 outputs the acquired installation location information as an estimation result to the output device 70 (step S13), and the installation location estimation process ends.

[0088] According to the groundwater level control system 1000, the pump 50 can be operated under operating conditions estimated by considering the saturated hydraulic conductivity and water level near the well of the pump 50, thus enabling efficient control of the groundwater level in the field. Furthermore, according to the groundwater level control system 1000, the pump 50 is installed in a location that takes into account the flow path of groundwater in the field, thus enabling appropriate control of the groundwater level in the field. Therefore, the groundwater level control system 1000 can improve uneven distribution of groundwater levels in the field.

[0089] According to the Groundwater Level Control System 1000, crop productivity can be improved and yields increased by correcting uneven groundwater levels in fields, thus contributing to Goal 2 of the United Nations-led Sustainable Development Goals (SDGs), "Zero Hunger."

[0090] [Examples of implementation using software] The control blocks (particularly control units 10, 20, 30, and 40) of the operating condition estimation device 100, the installation location estimation device 200, the operating condition estimation model generation device 300, and the installation location estimation model generation device 400 may be implemented by logic circuits (hardware) formed on an integrated circuit (IC chip) or the like, or by software.

[0091] In the latter case, the operating condition estimation device 100, the installation location estimation device 200, the operating condition estimation model generation device 300, and the installation location estimation model generation device 400 are equipped with a computer that executes instructions for a program, which is software that realizes each function. This computer is equipped with, for example, one or more processors and a computer-readable recording medium that stores the above program. The object of the present invention is achieved when the processor in the computer reads the above program from the recording medium and executes it. As the processor, for example, a CPU (Central Processing Unit) can be used. As the recording medium, a "tangible medium that is not temporary," such as ROM (Read Only Memory), can be used, as well as tape, disk, card, semiconductor memory, programmable logic circuit, etc. It may also be further equipped with RAM (Random Access Memory) for expanding the above program. Furthermore, the above program may be supplied to the computer via any transmission medium (such as a communication network or broadcast wave) capable of transmitting the program. In one aspect of the present invention, the above program can also be realized in the form of a data signal embedded in a carrier wave, which is embodied by electronic transmission. The pump operating condition estimation program, the pump installation location estimation program, the operating condition estimation model generation program, and the installation location estimation model generation program are also included in the scope of the present invention.

[0092] The present invention is not limited to the embodiments described above, and various modifications are possible within the scope of the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present invention. [Explanation of Symbols]

[0093] 12. Operating Condition Estimation Unit 22 Installation location estimation section 33. Driving Condition Estimation Model 43 Installation location estimation model 50 pumps 51 Control device 100 Operating Condition Estimation Device (Pump Operating Condition Estimation Device) 200 Installation location estimation device (pump installation location estimation device) 500 Pump Unit 1000 Groundwater Level Control System

Claims

1. A pump operating condition estimation device installed in a well within a field to estimate the operating conditions of a pump used to drain groundwater, Pump operating condition estimation device, comprising an operating condition estimation unit that uses an operating condition estimation model generated by machine learning using training data that associates the saturated hydraulic conductivity near the well, which is represented by information representing the delay time of groundwater level change between the inside of the well and the surrounding area, or the amount of water level change within a predetermined time of the water introduced into the well, with water level information including the water level of the well before, during, and after the operation of the pump, and the operating conditions of the pump, to estimate the operating conditions of the pump in order to equalize the groundwater level in the field, taking as input the saturated hydraulic conductivity near the well in the target field, the current water level of the wells in the field and water level information representing the target water level of the well.

2. The pump operating condition estimation device according to claim 1, wherein the water level information further includes information on environmental factors of the field that affect the water level of the well.

3. The pump operating condition estimation device according to claim 1 or 2, wherein the water level information further includes information representing the distance between adjacent pumps in a plurality of pumps installed in a field.

4. A well within the field is provided, and at least one pump is installed in the well to drain the groundwater of the field. A control device that controls the operation of the pump based on the operating conditions estimated by the pump operating condition estimation device according to any one of claims 1 to 3, A pump unit equipped with a pump unit.

5. A pump installation location estimation device for estimating the installation location of the pump of the pump unit described in claim 4, A pump installation location estimation device comprising an installation location estimation unit that estimates the pump installation location information by using a pump installation location estimation model generated by machine learning using training data that associates subsurface environmental geographic information related to the flow path of groundwater in the field, information representing the distribution of water level in the field after the installation of the pump, information representing the distribution of saturated hydraulic conductivity in the field, and information on the installation location of the pump in the field, taking the subsurface environmental geographic information of the target field and information representing the distribution of water level in the target field as input.

6. A pump operating condition estimation device according to any one of claims 1 to 3, The pump unit according to claim 4, The pump installation location estimation device according to claim 5 and A field groundwater level control system equipped with [a specific feature / feature].

7. A method for estimating the operating conditions of a pump installed in a well within a field to drain groundwater, A pump operating condition estimation method, comprising an operating condition estimation step, which involves using an operating condition estimation model generated by machine learning using training data that associates the saturated hydraulic conductivity near the well, which is represented by information indicating the delay time of groundwater level change between the inside of the well and the surrounding area, or the amount of water level change within a predetermined time of water introduced into the well, with water level information including the water level of the well before, during, and after the operation of the pump, and the operating conditions of the pump, to estimate the operating conditions of the pump in order to equalize the groundwater level in the field, taking the saturated hydraulic conductivity near the well in the target field, the current water level of the wells in the field and water level information indicating the target water level of the well as input.

8. A pump installation location estimation method for estimating the installation location of a pump whose operation is controlled based on the operating conditions estimated by the pump operating condition estimation method described in claim 7, A pump installation location estimation method comprising an installation location estimation step, in which a pump installation location estimation model is generated by machine learning using training data that associates subsurface environmental geographic information related to the flow path of groundwater in a field, information representing the distribution of water levels in the field after the installation of the pump, information representing the distribution of saturated hydraulic conductivity in the field, and information on the installation location of the pump in the field, and the model uses the subsurface environmental geographic information of a target field and information representing the distribution of water levels in the target field as input to estimate the installation location of the pump.

9. A pump operating condition estimation program for causing a computer to function as a pump operating condition estimation device according to any one of claims 1 to 3, wherein the pump operating condition estimation program causes a computer to function as the operating condition estimation unit.

10. A pump installation location estimation program for causing a computer to function as a pump installation location estimation device according to claim 5, wherein the pump installation location estimation program causes a computer to function as the installation location estimation unit.

Images