Field management system, field management method, and field management program

Abstract

To provide a system that enables automatic detection of water leakage in a field.SOLUTION: A field management system including an acquisition unit 10 that acquires the water level of a paddy field P for every preset period shorter than one day, a calculation unit 11 that computes a reduction rate of water, representing the speed at which the water of the paddy field P decreases, from the acquired water level, and a determination unit 12 that determines, according to a relationship between the current reduction rate of water computed by the calculation unit 11 and a past reduction rate of water in the paddy field P, whether leakage occurs in the paddy field P.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a farm field management system, a farm field management method, and a farm field management program. [Background technology]

[0002] Technologies for automatically managing fields such as rice paddies have been proposed. For example, Patent Document 1 discloses a technology for automating water management in a field. The field management system disclosed in Patent Document 1 calculates the amount of water supply required based on the water supply status of the field, and supplies the required amount of water when the water demand depth of the field falls within a predetermined allowable range.

[0003] Water leakage from the field is a major issue in field water management. When water leakage occurs from the field, it becomes difficult to accurately manage the water level. Furthermore, a decrease in water level due to leakage leads to the proliferation of weeds, which impairs rice growth and leads to increased field management costs. Therefore, Patent Document 2 discloses a field management system that notifies the user when the water level does not reach a set upper limit due to water supply to the field, so that the user can check the field themselves and determine whether water leakage is occurring from the ridges or whether the water outlet is not properly sealed. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2021-193924 [Patent Document 2] Japanese Patent Application Publication No. 2022-174447 Summary of the Invention [Problem to be solved by the invention]

[0005] However, with conventional technology, only information about the water level in the field is presented, and the user who manages the field has to determine whether or not there is a water leak in the field based on their own knowledge and experience. Therefore, with conventional technology, it has been difficult to automatically detect the occurrence of a water leak in the field.

[0006] The present invention has been made to solve the above-mentioned problems, and has an object to automatically detect the occurrence of water leakage in a farm field. [Means for solving the problem]

[0007] In order to solve the above-mentioned problems, the field management system of the present invention is characterized by comprising an acquisition unit that acquires the water level of the field for each set period of less than one day, a calculation unit that calculates the water reduction rate, which is the rate at which water in the field is decreasing, based on the acquired water level, and a determination unit that determines whether or not there is a water leak in the field based on the relationship between the current water reduction rate calculated by the calculation unit and the past water reduction rate in the field.

[0008] In addition, in the farm field management system of the present invention, the judgment unit may determine that a leak has occurred in the farm field if the current water reduction rate calculated by the calculation unit is higher than the past water reduction rate.

[0009] In addition, in the field management system of the present invention, the judgment unit determines that a leak has occurred in the field if the current water reduction rate calculated by the calculation unit is not within a set range from the average value of the past water reduction rates, and the set range may be a range of water reduction rate values set taking into account the daily water reduction depth of the field.

[0010] In addition, in the farm field management system of the present invention, the judgment unit may include a classification unit that provides the current water reduction rate and the past water reduction rate calculated by the calculation unit as unknown inputs to a trained classifier, performs calculations on the trained classifier, and outputs a classification class indicating whether or not a leak has occurred in the farm field, and the judgment unit may output the classification class output by the classification unit as a judgment result.

[0011] In addition, the farm field management system of the present invention further includes a learning unit that uses training data in which the current water reduction rate and the past water reduction rate are labeled with information regarding the occurrence of water leakage in the field as a correct answer label to learn the relationship between the current water reduction rate and the past water reduction rate in the field and the occurrence of water leakage, and constructs the trained classifier, and a memory unit that stores the trained classifier constructed by the learning unit, and the classifier may read out the trained classifier from the memory unit and perform calculations on the trained classifier.

[0012] Furthermore, the farm land management system according to the present invention may further include a sensor installed in the farm land to measure the water level in the farm land, and the acquisition unit may acquire the water level from the sensor.

[0013] The farm land management system according to the present invention may further include a notification unit that generates notification information indicating that a water leak is occurring in the farm land, based on the determination result by the determination unit.

[0014] In order to solve the above-mentioned problems, the field management method of the present invention is characterized by comprising a first step of acquiring the water level of the field for each set period of less than one day, a second step of calculating the water reduction rate, which is the rate at which water in the field is decreasing, based on the acquired water level, and a third step of determining whether or not there is a water leak in the field based on the relationship between the current water reduction rate calculated in the second step and the past water reduction rate in the field.

[0015] In addition, in the field management method of the present invention, the third step may determine that a water leak is occurring in the field if the current water reduction rate calculated in the second step is higher than the past water reduction rate.

[0016] In addition, in the field management method of the present invention, the third step may determine that a leak has occurred in the field if the current water loss rate calculated in the second step is not within a set range from the average value of the past water loss rates, and the set range may be a range of water loss rate values set taking into account the daily water loss depth of the field.

[0017] In addition, in the field management method of the present invention, the third step may include a fourth step of providing the current water loss rate calculated in the second step and the past water loss rate as unknown inputs to a trained classifier, performing calculations on the trained classifier, and outputting a classification class indicating whether or not a leak has occurred in the field, and the third step may output the classification class output in the fourth step as a judgment result.

[0018] In addition, the field management method of the present invention further includes a fifth step of learning the relationship between the current water reduction rate and the past water reduction rate in the field and the presence or absence of water leakage using training data in which information regarding the presence or absence of water leakage in the field is attached as a correct answer label to the current water reduction rate and the past water reduction rate, and constructing the trained classifier, and a sixth step of storing the trained classifier constructed in the fifth step in a memory unit, and the fourth step may read out the trained classifier from the memory unit and perform calculations on the trained classifier.

[0019] Furthermore, the farm field management method according to the present invention may further include a sensor installed in the farm field to measure the water level in the farm field, and the first step may include acquiring the water level from the sensor.

[0020] Furthermore, the farm land management method according to the present invention may further include a seventh step of generating notification information indicating that a water leak has occurred in the farm land, based on the determination result in the third step.

[0021] In order to solve the above-mentioned problems, the field management program of the present invention causes a computer to execute a first step of acquiring the water level of the field for each set period of less than one day, a second step of calculating the water reduction rate, which is the rate at which water in the field is decreasing, based on the acquired water level, and a third step of determining whether or not there is a water leak in the field based on the relationship between the current water reduction rate calculated in the second step and the past water reduction rate in the field. [Effects of the Invention]

[0022] According to the present invention, the water reduction rate, which is the rate at which water in a field decreases, is calculated based on the water level for each set period of less than one day, and the presence or absence of water leakage in the field is determined based on the relationship between the calculated current water reduction rate and the past water reduction rate in the field, thereby making it possible to automatically detect water leakage in the field. [Brief explanation of the drawings]

[0023] [Figure 1] FIG. 1 is a block diagram showing the configuration of a farm land management system according to a first embodiment of the present invention. [Figure 2] FIG. 2 is a diagram illustrating an overview of the farm land management system according to the first embodiment. [Figure 3] FIG. 3 is a diagram illustrating an overview of the farm land management system according to the first embodiment. [Figure 4] FIG. 4 is a diagram for explaining the determining unit of the farm land management system according to the first embodiment. [Figure 5] FIG. 5 is a block diagram showing the hardware configuration of a paddy field sensor included in the farmland management system according to the first embodiment. [Figure 6]FIG. 6 is a block diagram showing the hardware configuration of a farm land management device included in the farm land management system according to the first embodiment. [Figure 7] FIG. 7 is a sequence diagram illustrating the operation of the farm land management system according to the first embodiment. [Figure 8] FIG. 8 is a block diagram showing the configuration of a farm land management device included in a farm land management system according to the second embodiment. [Figure 9] FIG. 9 is a diagram illustrating a determination unit included in the farm land management system according to the second embodiment. [Figure 10] FIG. 10 is a sequence diagram illustrating the operation of the farm land management system according to the second embodiment. [Figure 11] FIG. 11 is a block diagram showing the configuration of a farm land management device included in a farm land management system according to a modified example. [Figure 12] FIG. 12 is a flowchart illustrating the operation of a farm land management system according to a modified example. [Figure 13] FIG. 13 is a flowchart illustrating the operation of a farm land management system according to a modified example. DETAILED DESCRIPTION OF THE INVENTION

[0024] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will now be described in detail with reference to FIGS.

[0025] [First embodiment] Fig. 1 is a block diagram showing the configuration of a farmland management system according to a first embodiment of the present invention. The farmland management system according to this embodiment is a system that detects the occurrence of water leakage in a paddy field P where rice is cultivated. Water leakage in a paddy field P occurs due to holes made by small animals in the ridges f, deterioration of the ridges f, or cracks in the soil surface of the paddy field P. In this embodiment, when water leakage occurs in a paddy field P, this also includes a state in which water loss in the paddy field P increases and the set water level cannot be maintained or does not rise to the set water level.

[0026] [Field management system configuration] First, an overview of a farm land management system according to an embodiment of the present invention will be described. As shown in Fig. 1, the farm land management system includes a farm land management device 1 and a paddy field sensor (sensor) 2 installed in a paddy field P to be managed, which are connected via a network via a base station GW.

[0027] For example, in a farm field management system, a paddy field sensor 2 is connected to a farm field management device 1 on the network side via a base station gateway via a 3G, LTE, or 5G wireless line, Ethernet (registered trademark), or other line. The paddy field sensor 2 on the terminal side and the base station gateway can communicate wirelessly with low power consumption and long distance communication using LPWA (Low Power Wide Area) such as LoRaWAN (registered trademark).

[0028] The base station GW is equipped with a communication module that supports 3G, LTE, 5G communication and LPWA such as LoRaWAN (registered trademark), and transfers sensor data from the paddy field sensor 2 to the cloud system, which is the farmland management device 1. Note that in this embodiment, an example will be described in which one paddy field sensor 2 is installed in a paddy field P, but multiple paddy field sensors 2 can be placed depending on the area of the paddy field P to be managed, etc.

[0029] Water supply channels and drainage channels are provided in the paddy fields P, and water is supplied and drained for each paddy field P. A water supply device (not shown) that supplies irrigation water to the paddy fields P is provided in the water supply channels. A drainage device (not shown) that drains the water from the paddy fields P is provided in the drainage channels. For example, the water supply device and the drainage device may be configured to be connected to a water management device (not shown) that manages water supply and drainage via a base station GW. The functions of the water management device may also be provided in the farm field management device 1.

[0030] For example, the water management device can send water level control instructions to the water supply device and the drainage device via the base station GW in order to set the water level in the paddy field P to a predetermined level. The water supply device and the drainage device that receive the water level control instructions can be configured to supply water to the paddy field P, stop water flow, and drain water, respectively, based on the control instructions. For example, the water supply device opens and closes a gate, takes in an amount of irrigation water into the paddy field P according to the gate opening degree, and sets the water level in the paddy field P to a set level according to the area of the paddy field P and the state of rice growth. Note that in this embodiment, water management for the paddy field P is performed assuming that no water leakage occurs in the paddy field P.

[0031] In the farm land management system, a communication terminal UE used by a user who manages a farm land connects to the farm land management apparatus 1 via a network NW such as the Internet. The communication terminal UE is configured by a smartphone, tablet terminal, or the like, and outputs information related to farm land management from a display screen or speaker. Specifically, the farm land management apparatus 1 provides information to the communication terminal UE via an API (Application Programming Interface).

[0032] By pre-installing the application program of the farm field management system on the communication terminal UE, the user can remotely monitor changes in the water level in the managed paddy fields P, changes in the rate of water decline (described later), whether or not water leakage has occurred, and even changes in air temperature and water temperature, and can also operate the farm field management device 1.

[0033] [Rice paddy sensor configuration] The paddy field sensor 2 is installed in a paddy field P and measures the water level of the paddy field P. The paddy field sensor 2 has a structure in which, for example, a sensor unit fixed to the ground side of a support pole driven into the paddy field P and a communication unit fixed to the other end of the support pole are connected by wiring. As shown in FIG. 5, the paddy field sensor 2 includes, for example, a substrate thermometer 20, a water thermometer 21, a water level gauge 22, a battery 23, and a communication interface 24. The substrate thermometer 20, the water thermometer 21, and the water level gauge 22 make up the sensor unit of the paddy field sensor 2.

[0034] Furthermore, the communication interface 24 constitutes the communication unit of the paddy field sensor 2. The communication interface 24 is, for example, a wireless communication interface that complies with a communication standard such as LoRaWAN (registered trademark). The substrate thermometer 20 is provided on the other end side of the support, that is, on the outside air side.

[0035] In the paddy field sensor 2, power is supplied to the sensor unit and communication unit by a battery 23, such as a lithium-ion secondary battery or a manganese dioxide lithium primary battery. The paddy field sensor 2 transmits the water level measured by the substrate thermometer 20, water thermometer 21, and water level gauge 22 to the farmland management device 1 at a predetermined cycle, such as every 30 minutes. In addition to the water level in the paddy field P, the paddy field sensor 2 can also transmit the air temperature and water temperature in the paddy field P measured by the substrate thermometer 20 to the farmland management device 1 at a predetermined cycle.

[0036] The water level gauge 22 can measure the water level at the timing of the data transmission cycle. The sensor data also includes information such as the identification information of the paddy field sensor 2 and the date and time of measurement. The paddy field sensor 2 only needs to measure at least the water level, and may be configured to measure air temperature and / or water temperature depending on the design.

[0037] [Configuration of farm management device] Next, the functional blocks of the farm land management device 1 will be described with reference to Fig. 1. As shown in Fig. 1, the farm land management device 1 includes an acquisition unit 10, a calculation unit 11, a determination unit 12, a notification unit 13, and a storage unit 14.

[0038] The acquisition unit 10 acquires the water level of the paddy field P for each set period of less than one day. The acquisition unit 10 can use, for example, six hours as the set period of less than one day. Specifically, the acquisition unit 10 receives the water level of the paddy field P measured by the paddy field sensor 2 at intervals of 30 minutes, for example, via the base station GW, and calculates the average water level of the paddy field P for each six-hour period. The water level of the paddy field P acquired by the acquisition unit 10 is time-series data in which the average water level for each set period has been accumulated. Note that, as long as the set period is less than one day, a period longer or shorter than six hours can be used.

[0039] The calculation unit 11 calculates the water reduction rate [cm / day], which is the rate at which the water in the paddy field P decreases per day, based on the water level in the paddy field P acquired by the acquisition unit 10. Here, the water reduction rate calculated by the calculation unit 11 will be explained with reference to Fig. 2. Fig. 2(a) is a cross-sectional view of the paddy field P showing the water level d1 of the paddy field P at a certain date and time. Fig. 2(b) shows a cross-sectional view of the paddy field P in a state where the water level d1 of the paddy field P has decreased to a water level d2 after a certain amount of time has passed.

[0040] Specifically, Figure 2(a) shows the average water level d1 [cm] of paddy field P from 12 hours ago to 6 hours ago, with the current time as the base time. Furthermore, Figure 2(b) shows the average water level d2 [cm] of paddy field P from 6 hours ago to the current time. The water decline rate [cm / day] is calculated using the following formula (1) from the difference between the average water level d2 [cm] for the most recent set period, which is the 6-hour period from 6 hours ago to the current time, and the average water level d1 [cm] for the 6-hour period from 12 hours ago to 6 hours ago, which is the set period before that. Water reduction rate = {(d1 - d2) / set period [hours]} × 24 [hours] (1)

[0041] Assume that the set period is 6 hours, and the average water levels are d1 = 10 [cm] and d2 = 9.5 [cm]. Substituting these values into the above formula (1), the water decline rate is calculated as 2 [cm / day]. Note that if the water decline rate is a negative value, i.e., if the average water level is increasing, the water decline rate may be set to 0 [cm / day]. The water decline rate calculated by the calculation unit 11 is stored in the memory unit 14, which will be described later.

[0042] Returning to FIG. 1, the determination unit 12 determines whether or not there is a leak in the paddy field P based on the relationship between the current water reduction rate calculated by the calculation unit 11 and the past water reduction rate in the paddy field P. The past water reduction rate in the paddy field P is the past water reduction rate calculated by the calculation unit 11 using the above formula (1) based on the time series data of the average water level in the paddy field P every six hours acquired by the acquisition unit 10. In this embodiment, the water reduction rate is treated as a value specific to the paddy field P, and the current and past water reduction rates use the value of the same paddy field P.

[0043] More specifically, when the current water reduction rate calculated by the calculation unit 11 is higher than the past water reduction rate, the determination unit 12 determines that water leakage has occurred in the paddy field P. The determination unit 12 reads out the past water reduction rates stored in the memory unit 14 and performs a determination process.

[0044] Here, Figure 3 is a graph showing time series data of the water level and water reduction rate of paddy field P. The horizontal axis shows time [month / day hour:minute], the right vertical axis shows the water reduction rate [cm / day] in paddy field P, and the left vertical axis shows the water level [cm] in paddy field P. The dotted curve shows the change in the water level D of paddy field P.

[0045] In the example shown in Figure 3, water supply to paddy field P began on June 1st, reached the set water level on June 2nd, and then dropped to 0 cm by June 4th. After that, a second water supply to paddy field P began on June 5th, reached the set water level on June 6th, and then dropped again to 0 cm by June 7th. Also, in the example shown in Figure 3, water leakage occurred from paddy field P's ridges f and other areas between June 5th and June 6th.

[0046] The solid curve in Figure 3 shows the change in the water reduction rate S in the paddy field P. Compared to the water reduction rate s1 when the first water supply was made on June 1, the water reduction rate s2 when the second water supply was made on June 5 increased. The judgment unit 12 judges this increase in water reduction rate to be the occurrence of a leak. Note that in Figure 3, the time difference between the value of the water level D and the value of the water reduction rate S is due to the measurement cycle of the paddy field sensor 2.

[0047] Here, the process of determining whether or not there is a water leak performed by the determination unit 12 will be explained in more detail with reference to Figure 4. The horizontal axis of Figure 4 represents time, and the vertical axis represents the water reduction rate. Water reduction rates a, b, and c are plot values of the past water reduction rate and the current water reduction rate in the paddy field P, respectively. Water reduction rate a indicates that the past water reduction rate and the current water reduction rate are constant. Water reduction rate b indicates that the current water reduction rate is increasing compared to the past water reduction rate. Furthermore, water reduction rate c indicates that the current water reduction rate is decreasing compared to the past water reduction rate.

[0048] The determination unit 12 can determine that no water leak has occurred when the past value and the current value are constant, as indicated by water reduction rate a. The determination unit 12 also determines that a water leak has occurred when the current value is increasing compared to the past value, as indicated by water reduction rate b. Note that, when the current water reduction rate is decreasing compared to the past value, as indicated by water reduction rate c, no determination can be made or the determination can be invalid.

[0049] Returning to FIG. 1 , the notification unit 13 generates notification information indicating that a water leak has occurred in the paddy field P based on the determination result by the determination unit 12. When the determination unit 12 determines that a water leak has occurred, the notification unit 13 can send information indicating the occurrence of the water leak, information on the date and time when the leak occurred, and identification information of the paddy field P to the communication terminal UE linked to the user who is the manager of the paddy field P. The notification unit 13 can generate notification information in the form of text, image, sound, vibration, or a combination of these.

[0050] The memory unit 14 stores time-series data that accumulates the average water level for a set period acquired by the acquisition unit 10. The memory unit 14 also stores time-series data of the water reduction rate calculated by the calculation unit 11. Furthermore, the memory unit 14 stores identification information of the paddy field P to be managed, identification information of the paddy field sensor 2, etc. Furthermore, the memory unit 14 can store location data of the paddy field sensor 2, including the latitude and longitude.

[0051] [Hardware configuration of farm management device] Next, an example of a hardware configuration for realizing the farm land management device 1 having the above-described functions will be described with reference to FIG.

[0052] 6, the farm land management device 1 can be realized by, for example, a computer including a processor 102, a main memory device 103, a communication interface 104, an auxiliary memory device 105, and an input / output (I / O) 106, which are connected via a bus 101, and a program that controls these hardware resources. The farm land management device 1 can also include a display device 107 connected via the bus 101.

[0053] The main memory device 103 pre-stores programs for the processor 102 to perform various controls and calculations. The processor 102 and the main memory device 103 implement the functions of the farm land management device 1, such as the acquisition unit 10, calculation unit 11, determination unit 12, and notification unit 13 shown in FIG.

[0054] The communication interface 104 is an interface circuit for network connection between the farm land management device 1 and various external electronic devices such as a base station GW and a communication terminal UE. The communication interface 104 can perform wireless communication using, for example, LPWA (Low Power Wide Area), LPWAN (Low-Power Wide Area Network), WiFi (registered trademark), etc. The communication interface 104 can also perform wireless communication using a mobile phone communication network or a data communication network such as 3G, 4G / LTE, or 5G. The communication interface 104 realizes some of the functions of the acquisition unit 10 and the notification unit 13 shown in FIG. 1.

[0055] The auxiliary storage device 105 is composed of a readable / writable storage medium and a drive for reading and writing various information such as programs and data from and to the storage medium. The auxiliary storage device 105 can use a semiconductor memory such as a hard disk or flash memory as the storage medium.

[0056] The auxiliary storage device 105 has a program storage area for storing the farm land management program executed by the farm land management device 1. The auxiliary storage device 105 realizes the storage unit 14 described in Fig. 1. Furthermore, the auxiliary storage device 105 may have, for example, a backup area for backing up the data and programs described above.

[0057] The input / output I / O 106 is configured by an I / O terminal for inputting signals from an external device and outputting signals to an external device.

[0058] The display device 107 is configured by a liquid crystal display or the like.

[0059] [Operation sequence of the farm management system] Next, the operation of the farm land management system having the above-described configuration will be described with reference to the sequence diagram of FIG.

[0060] First, when an operation input for performing initial setup is received at the user's communication terminal UE, the initial setup for field management is performed in the field management system (step S1). The setup information includes position data and identification information of the paddy fields P to be managed and the paddy field sensors 2 installed in the paddy fields P.

[0061] Next, the paddy field sensor 2 installed in the paddy field P measures the water level of the paddy field P (step S2). Next, the paddy field sensor 2 transmits the measured water level to the farm field management device 1 (step S3). For example, the paddy field sensor 2 transmits the measured water level to the farm field management device 1 at intervals of 30 minutes.

[0062] Next, the acquisition unit 10 of the farm land management device 1 acquires the water level of the paddy field P for each set period of less than one day (step S4). More specifically, the acquisition unit 10 calculates the average water level for each six-hour period based on the water level of the paddy field P measured by the paddy field sensor 2 at 30-minute intervals.

[0063] Next, the calculation unit 11 of the farmland management device 1 calculates the water reduction rate, which is the rate at which water in the paddy field P decreases per day, based on the water level in the paddy field P for each set period acquired in step S4 (step S5). Specifically, the calculation unit 11 calculates the water reduction rate per day from the difference between the average water level for the most recent six hours and the average water level for the six hours before that, using the above formula (1). The calculated water reduction rate is stored in the memory unit 14.

[0064] Next, if the current water reduction rate in the paddy field P calculated in step S5 is higher than the past water reduction rate (step S6: YES), the determination unit 12 of the farm field management device 1 determines that water leakage has occurred in the paddy field P (step S7). The determination unit 12 reads out the past water reduction rate in the paddy field P stored in the memory unit 14 and performs a determination process.

[0065] If the current water reduction rate calculated in step S5 and the past water reduction rate are constant (step S6: NO), the determination unit 12 determines that no water leak has occurred and can proceed to step S8. Similarly, if the current water reduction rate calculated in step S5 is lower than the past water reduction rate (step S6: NO), the determination unit 12 makes no determination or the determination is invalid and proceeds to step S8.

[0066] Next, the notification unit 13 of the farm land management device 1 generates notification information indicating that a water leak has occurred in the paddy field P based on the determination result in step S7 (step S8). After that, the notification unit 13 transmits the notification information to the communication terminal UE (step S9). The communication terminal UE, which has received the notification information, displays the notification information on its display screen (step S10). For example, the communication terminal UE can display, on its display screen, identification information of the paddy field P, text and an image indicating that a water leak has occurred, and the date and time when the water leak occurred. In addition, the communication terminal UE can display the notification information together with time-series data of the water level of the paddy field P and time-series data of the rate of water reduction.

[0067] As described above, the farmland management system according to the first embodiment calculates the daily water loss rate based on the water level for a set period of less than one day. Furthermore, the calculated current water loss rate is compared with the past water loss rate, and if the current water loss rate is higher than the past water loss rate, it is determined that a leak has occurred. This allows for automatic detection of water leakage in the paddy field P.

[0068] [Second embodiment] Next, a second embodiment of the present invention will be described. In the following description, the same components as those in the first embodiment will be denoted by the same reference numerals, and the description thereof will be omitted.

[0069] In the first embodiment, it is determined that a water leak has occurred when the current water reduction rate of the paddy field P is higher than the past water reduction rate. In contrast, in the second embodiment, it is determined that a water leak has occurred in the paddy field P when the current water reduction rate of the paddy field P is not within a set range from the average value of the past water reduction rates.

[0070] [Function block of farm management device] As shown in Fig. 8, the farm land management apparatus 1A according to the second embodiment includes an acquisition unit 10, a calculation unit 11, a determination unit 12A, a notification unit 13, and a storage unit 14. The farm land management apparatus 1A according to the second embodiment differs from the farm land management apparatus 1 according to the first embodiment in the configuration of the determination unit 12A. The following description will focus on the configuration that differs from the first embodiment.

[0071] The determination unit 12A determines that water leakage is occurring in the paddy field P if the current water loss rate of the paddy field P calculated by the calculation unit 11 is not within a set range from the average value of past water loss rates. The set range is a range of values for the water loss rate set in consideration of the daily water loss depth of the paddy field P. Here, the water loss depth is a value that represents the sum of the evapotranspiration rate and infiltration rate of the field expressed in units of water depth, and is a value that represents the degree of water loss in a single field. The daily water loss depth also indicates the quality of the permeability of the field, and in Japan, it is known to be generally around 1 to 2 cm / day.

[0072] The determination unit 12A can determine that water leakage has occurred when the current water loss rate of the paddy field P exceeds at least the value obtained by adding the daily water loss depth to the average past water loss rate of the paddy field P. Furthermore, when the current water loss rate is approximately twice the average water loss rate of the field over the past two weeks or so, there is a high possibility that water leakage has occurred in the field. For this reason, a range can be adopted that takes into account the average past water loss rate and the typical daily water loss depth of 1 to 2 cm / day, for example, a range from 2 to 2.5 times the average water loss rate over the past two weeks. Note that the set range can be set to a value between 1.5 and 4 times the average past water loss rate, taking into account the individual circumstances of each paddy field P.

[0073] Fig. 9 is a diagram for explaining the determination process by the determination unit 12A. The horizontal axis of Fig. 9 represents time, and the vertical axis represents the water reduction rate. Fig. 9 shows the average value g of the past water reduction rate of the paddy field P, which is, for example, the average value of the water reduction rate over the past two weeks.

[0074] Furthermore, the water reduction rate e shown in FIG. 9 indicates the water reduction rate at each date and time calculated by the calculation unit 11, and the most recent water reduction rate e is not within the range th set from the average value g of the water reduction rate over the past two weeks. In the example shown in FIG. 9, the set range is set to a range up to twice the average value g of the water reduction rate of the paddy field P over the past two weeks. In this way, the determination unit 12A can determine that a leak has occurred in the paddy field P by performing threshold processing on the most recent water reduction rate e.

[0075] Here, the determination process by the determination unit 12A will be further explained using the example of Fig. 3 mentioned above. As an example, the average value of the water reduction rate for the past two weeks is given as the water reduction rate s1. In this case, since the water reduction rate s2 from June 6th to June 7th is more than twice the value of the past water reduction rate s1, the determination unit 12A determines that leakage has occurred in the paddy field P.

[0076] [Operation sequence of the farm management system] Next, the operation of the farm land management system having the above-described configuration will be described using the sequence diagram of Fig. 10. Note that, among the operations of the farm land management system according to this embodiment, the processes other than step S6A are the same as the steps in the operations of the farm land management system according to the first embodiment. It is also assumed that the set range values used by the determination unit 12A in the determination process are stored in advance in the memory unit 14. The following description will focus on the operation of the farm land management device 1A.

[0077] 10, the acquisition unit 10 of the farm land management device 1A acquires the water level of the paddy field P (step S4), and then the calculation unit 11 calculates the water reduction rate of the paddy field P (step S5). The calculation unit 11 calculates the average value of the water reduction rate in the paddy field P over a set past period, for example, the past two weeks, and stores the average value in the memory unit 14.

[0078] Next, if the current water reduction rate calculated by the calculation unit 11 in step S5 is not within a set range from the average value of past water reduction rates (step S6A: NO), the judgment unit 12A of the field management device 1A determines that water leakage has occurred in the paddy field P (step S7).

[0079] More specifically, the determination unit 12A reads out the set range from the memory unit 14 and determines that a leak has occurred if the current water reduction rate of the paddy field P calculated in step S5 is not within a range of, for example, twice the average value of the water reduction rate in the paddy field P over the past two weeks. Note that if the current water reduction rate of the paddy field P calculated in step S5 is within the set range of the average value of the past water reduction rates (step S6A: YES), it is determined that a leak has not occurred in the paddy field P, and the process can proceed to step S8.

[0080] Next, the notification unit 13 of the farm land management apparatus 1A generates notification information indicating that a water leak has occurred in the paddy field P based on the determination result in step S7 (step S8). Thereafter, the notification unit 13 transmits the notification information to the communication terminal UE (step S9). The communication terminal UE, which has received the notification information, displays the notification information on its display screen (step S10). For example, the communication terminal UE can display the notification information indicating that a water leak has occurred together with time-series data of the water level in the paddy field P, time-series data of the rate of water reduction, and the threshold value used by the determination unit 12A for the determination process.

[0081] As described above, the farmland management system according to the second embodiment determines that a water leak has occurred if the current water loss rate in the paddy field P is not within a range set based on the average past water loss rate and taking into account the daily water loss depth of the paddy field P. Therefore, water leakage in the paddy field P can be automatically detected.

[0082] [Variations] Next, a description will be given of modified examples of the first and second embodiments of the present invention. In the first embodiment, the determination unit 12 determines that a leak has occurred when the current water reduction rate of the paddy field P is higher than the past water reduction rate. In the second embodiment, the determination unit 12A determines that a leak has occurred when the current water reduction rate of the paddy field P is not within a set range from the average value of the past water reduction rates. In contrast, in this modified example, the determination units 12 and 12A are configured by a classification unit 16 that uses a machine learning model.

[0083] [Function block of farm management device] 11 is a block diagram showing the configuration of a farm land management apparatus 1B according to this modification. Farm land management apparatus 1B differs from the farm land management apparatuses 1, 1A according to the first and second embodiments in that it includes a learning unit 15 and a classification unit 16. The following description will focus on the configuration that differs from the first and second embodiments.

[0084] The learning unit 15 uses training data in which information regarding the presence or absence of water leakage in the paddy field P is attached as a correct answer label to the water reduction rate in the paddy field P to learn the relationship between the water reduction rate in the paddy field P and the presence or absence of water leakage, and constructs a trained classifier. Specifically, the learning unit 15 provides the current water reduction rate in the paddy field P calculated by the calculation unit 11 and the past water reduction rate as input to the classifier, and learns the presence or absence of water leakage in the paddy field P using the training data. In other words, the learning unit 15 uses the classifier to learn the relationship between changes in the water reduction rate in the paddy field P and the presence or absence of water leakage.

[0085] The learning unit 15 can use a neural network model consisting of an input layer, a hidden layer, and an output layer as a classifier. The current and past water reduction rates of the paddy field P calculated by the calculation unit 11 are provided to the input nodes of the input layer. For the current and past water reduction rates of the paddy field P provided to the input layer, an activation function is applied to the weighted sum of the inputs, and the output determined by threshold processing is passed to the output layer.

[0086] For example, two classification classes, "water leakage" and "water leakage not present," can be used as each output node of the output layer. The output layer can also output the probability that the input data belongs to each classification class.

[0087] The learning unit 15 adjusts the weights of the connections between nodes so that the water loss rate of the paddy field P matches the classification class value of "leakage" or "no leakage" indicated in the correct label of the training data. The learning unit 15 uses error backpropagation or the like to compare the obtained output value with the given input value, examine the error of each weight, and propagate it backward, ultimately determining parameters such as weights. Through this learning process, the learning unit 15 constructs a trained classifier.

[0088] The classification unit 16 provides the water reduction rate of the paddy field P calculated by the calculation unit 11 as an unknown input to the trained classifier, performs calculations on the trained classifier, and outputs a classification class indicating whether or not water leakage has occurred in the paddy field P.

[0089] The storage unit 14 stores the trained classifier constructed by the training unit 15.

[0090] [Farm management system operation] Next, the operation of the farm land management system having the above-described configuration will be described with reference to the flowcharts in Fig. 12 and Fig. 13. Fig. 12 is a flowchart showing the learning process in the farm land management system. Fig. 13 is a flowchart showing the classification process in the farm land management system. Of the steps shown in Fig. 13, the steps other than step S25 and step S26 are the same as the processes shown in the first and second embodiments described above.

[0091] First, as shown in FIG. 12, the learning unit 15 of the farmland management device 1B prepares training data (step S20). Specifically, the learning unit 15 prepares training data in advance in which correct labels indicating the presence or absence of water leakage are attached to the water reduction rate of the paddy field P. Specifically, training data in which the current and past water reduction rates of the paddy field P are labeled with the presence or absence of water leakage can be used. Note that the learning unit 15 can create the training data using actual data on the water reduction rate of the paddy field P obtained by the acquisition unit 10 and the calculation unit 11.

[0092] Next, the learning unit 15 uses the training data prepared in step S20 to learn the relationship between the water reduction rate in the paddy field P and the occurrence of leakage, and constructs a trained classifier (step S21). Thereafter, the memory unit 14 stores the trained classifier (step S22).

[0093] Next, the classification process by the farm land management system will be described with reference to Fig. 13. First, the acquisition unit 10 of the farm land management device 1B acquires the water level of the paddy field P (step S4), and then the calculation unit 11 calculates the water reduction rate of the paddy field P (step S5). The water reduction rate of the paddy field P calculated in step S5 is stored in the memory unit 14.

[0094] Next, the classification unit 16 loads the trained classifier from the storage unit 14 (step S25). Thereafter, the classification unit 16 provides the water reduction rate of the paddy field P calculated by the calculation unit 11 in step S5 as an unknown input to the trained classifier, performs calculations on the trained classifier, and outputs a classification class indicating whether or not leakage has occurred in the paddy field P (step S26). Specifically, the current water reduction rate and past water reduction rates of the paddy field P are used as unknown inputs.

[0095] Next, the notification unit 13 of the farm land management device 1B generates notification information indicating that a water leak has occurred in the paddy field P based on the classification result in step S26 (step S8). Thereafter, the notification unit 13 transmits the notification information to the communication terminal UE (step S9).

[0096] As described above, according to the farmland management system of this modified example, the water reduction rate of the paddy field P is provided as an unknown input to the trained classifier, and the trained classifier performs calculations to output a classification class indicating whether or not water leakage has occurred in the paddy field P. Therefore, water leakage in the paddy field P can be automatically detected.

[0097] In the above-described embodiment, a neural network is used as the classifier. However, in addition to the above-described neural network model, logistic regression or the like can be used as the classifier. Other classifiers that can be used include SVM, Naive Bayes, Random Forest, Decision Tree, and even deep learning, which is a multi-layered neural network. Furthermore, a recurrent network that handles time-series data can be used as the neural network.

[0098] In addition, in the modified example described above, as an example, a classifier is constructed for each paddy field P, but a classifier can be constructed that corresponds to all paddy fields P that are managed by the field management system.

[0099] The above describes embodiments of the farm field management system and farm field management method of the present invention, but the present invention is not limited to the described embodiments, and various modifications that a person skilled in the art can conceive are possible within the scope of the invention described in the claims. [Explanation of symbols]

[0100] 1...field management device, 2...paddy field sensor, 10...acquisition unit, 11...calculation unit, 12...judgment unit, 13...alarm unit, 14...memory unit, 20...substrate thermometer, 21...water thermometer, 22...water level gauge, 23...battery, 101...bus, 102...processor, 103...main memory device, 24, 104...communication interface, 105...auxiliary memory device, 106...input / output I / O, 107...display device, NW...network, GW...base station, P...paddy field, UE...communication terminal, f...ridge.

Claims

1. an acquisition unit that acquires the water level of the field for each set period of less than one day; a calculation unit that calculates a water reduction rate, which is a rate at which water in the field is reduced, based on the acquired water level; and a determination unit that determines whether or not there is a water leak in the field based on a relationship between the current water reduction rate calculated by the calculation unit and a past water reduction rate in the field; A field management system comprising:

2. The farmland management system according to claim 1, The determination unit determines that a water leak is occurring in the field when the current water reduction rate calculated by the calculation unit is higher than the past water reduction rate. A farm field management system characterized by the above.

3. The farmland management system according to claim 1, the determination unit determines that a water leak has occurred in the field when the current water reduction rate calculated by the calculation unit is not within a set range from an average value of the past water reduction rates; and The set range is a range of values of the water reduction rate set in consideration of the daily water reduction depth of the field. A farm field management system characterized by the above.

4. The farmland management system according to claim 1, the determination unit includes a classification unit that provides the current water reduction rate and the past water reduction rate calculated by the calculation unit as unknown inputs to a trained classifier, performs calculations on the trained classifier, and outputs a classification class that indicates whether or not a water leak has occurred in the field; The determination unit outputs the classification class output by the classification unit as a determination result. A farm field management system characterized by the above.

5. The farmland management system according to claim 4, a learning unit that uses training data in which information on the occurrence of water leakage in the field is attached as a correct answer label to the current water reduction rate and the past water reduction rate, and learns the relationship between the current water reduction rate and the past water reduction rate in the field and the occurrence of water leakage, and constructs the trained classifier; a storage unit that stores the trained classifier constructed by the learning unit; Equipped with The classification unit reads out the trained classifier from the storage unit and performs calculations on the trained classifier. A farm field management system characterized by the above.

6. The farm land management system according to any one of claims 1 to 5, Further, a sensor is installed in the field and measures the water level in the field, The acquisition unit acquires the water level from the sensor. A farm field management system characterized by the above.

7. The farmland management system according to claim 6, Further, a notification unit is provided that generates notification information indicating that a water leak has occurred in the field based on the determination result by the determination unit. A farm field management system characterized by the above.

8. A first step of acquiring the water level of a field for each set period of less than one day; a second step of calculating a water reduction rate, which is a rate at which water in the field is reduced, based on the acquired water level; a third step of determining whether or not there is a water leak in the field based on a relationship between the current water reduction rate calculated in the second step and a past water reduction rate in the field; A field management method comprising:

9. The farmland management method according to claim 8, The third step determines that a water leak is occurring in the field when the current water reduction rate calculated in the second step is higher than the past water reduction rate. A farm field management method characterized by the above.

10. The farmland management method according to claim 8, the third step determining that a water leak has occurred in the field when the current water reduction rate calculated in the second step is not within a set range from an average value of the past water reduction rates; The set range is a range of values of the water reduction rate set in consideration of the daily water reduction depth of the field. A farm field management method characterized by the above.

11. The farmland management method according to claim 8, the third step includes a fourth step of providing the current water reduction rate calculated in the second step and the past water reduction rate as unknown inputs to a trained classifier, performing calculations on the trained classifier, and outputting a classification class indicating whether or not a water leak has occurred in the field; The third step outputs the classification class output in the fourth step as a determination result. A farm field management method characterized by the above.

12. The farmland management method according to claim 11, a fifth step of learning a relationship between the current and past water reduction rates in the field and the presence or absence of water leakage using training data in which information regarding the presence or absence of water leakage in the field is attached as a correct answer label to the current and past water reduction rates, and constructing the trained classifier; a sixth step of storing the trained classifier constructed in the fifth step in a storage unit; Equipped with The fourth step reads out the trained classifier from the storage unit and performs a calculation on the trained classifier. A farm field management method characterized by the above.

13. The farmland management method according to any one of claims 8 to 12, Further, a sensor is installed in the field and measures the water level in the field, The first step is to obtain the water level from the sensor. A farm field management method characterized by the above.

14. The farmland management method according to claim 13, Further, a seventh step of generating notification information indicating that a water leak has occurred in the field based on the determination result in the third step is provided. A farm field management method characterized by the above.

15. On the computer, A first step of acquiring the water level of a field for each set period of less than one day; a second step of calculating a water reduction rate, which is a rate at which water in the field is reduced, based on the acquired water level; a third step of determining whether or not there is a water leak in the field based on a relationship between the current water reduction rate calculated in the second step and a past water reduction rate in the field; A field management program to implement the above.

Images