Plant protection system

Abstract

To provide a plant protection system for predicting a damage risk by pests and weeds corresponding to a crop kind and a growth stage of the crop.SOLUTION: A plant protection system S1 includes: a terminal device T for receiving farming data from a user and displaying a damage risk; and a server C1 connected to the terminal device through a network. The server C1 includes: a crop growth prediction part 1 for predicting a growth stage of an object crop on the basis of data for growth prediction including the farming data inputted to the terminal device T and weather data at an object point; a pest occurrence prediction part 2 for predicting at least one kind and its occurrence amount of pests and weeds of the object crop at the object point; and a damage risk prediction part 3 for predicting a damage risk on the basis of a growth prediction result by the crop growth prediction part 2 and an occurrence prediction result by the damage risk prediction part 3. The damage risk prediction part 3 predicts a damage risk for each of at least one kind of the pests and the weeds in accordance with the growth stage of the object crop.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present disclosure relates to a plant protection system.

Background Art

[0002] Generally, for crops and the like, information on where, when, what pests and weeds exist, and how much they exist is basic information for pest control and understanding their dynamics. If such information is obtained, it is possible to understand the ecology of pests and the transmission status of pathogens and viruses that cause diseases, determine the appropriate control timing, and also utilize the information for decision-making on issuing warning information regarding pests.

[0003] For example, crop damage caused by the massive occurrence of Spodoptera litura has become a global problem. In Japan as well, the large occurrence of the brown planthopper has caused great damage to rice cultivation. In particular, due to recent climate change and the like, the occurrence of new migratory pests such as the beet armyworm is also regarded as a problem. Therefore, accurately monitoring the occurrence of these pests and leading to appropriate control is an important technical issue in agricultural technology.

[0004] In addition, research is being conducted in various countries around the world to predict the future occurrence of pests by using machine learning and statistical models with environmental data including monitored current or past pest occurrence data and meteorological data as explanatory variables. The applicant of the present application has also proposed various technologies not only for collecting labor-saving pest occurrence information related to pest monitoring devices but also for predicting the occurrence and movement of pests (for example, Non-Patent Documents 1 and 2).

Prior Art Documents

Non-Patent Documents

[0005]

Non-Patent Document 1

[0006] Incidentally, while crops are susceptible to various diseases and pests, the degree (magnitude) of damage suffered by crops when infected with diseases or pests differs depending on the type of crop and its growth stage. For example, in the case of wheat, the risk of damage from Fusarium head blight, a mycotoxin, is highest when it rains from the flowering stage to the milky stage (maturity stage), and the risk of such damage is expected to be little or no before that time. Since the spores of the Fusarium head blight pathogen are dispersed by wind and rain, it is assumed that there is a strong correlation between the risk of such damage and meteorological factors. As another example, in the case of soybeans, the period from the pod elongation stage to the seed enlargement stage is a time when they are susceptible to damage from stink bugs that suck the seeds. If a large number of stink bugs are predicted to occur during this period, the risk of soybean damage from stink bugs is expected to be high. On the other hand, even if a large number of stink bugs are predicted to occur at other times, soybeans are not expected to suffer much damage, so the risk of such damage is expected to be low. Thus, since crops are sometimes sensitive to pest and disease outbreaks and sometimes not, it is desirable to predict the risk of damage from pests and diseases according to the crop's growth stage in order to effectively implement appropriate control measures.

[0007] This disclosure is made in view of the above, and its purpose is to provide a plant protection system that predicts the risk of damage from pests, diseases, and weeds according to the type of crop and the growth stage of the crop. [Means for solving the problem]

[0008] To achieve the above objectives, this disclosed technology incorporates crop growth prediction data into disease and pest occurrence prediction data to predict the risk of damage from diseases and pests according to the crop's growth stage.

[0009] Specifically, this disclosure relates to a plant protection system that predicts the risk of damage to a target crop caused by at least one type of pest, disease, or weed at a target location. The system comprises a terminal device into which farming data is input by a user and in which the damage risk is displayed, and a server connected to the terminal device via a network. The server comprises a crop growth prediction unit that predicts the growth stage of the target crop based on growth prediction data including the farming data and weather data at the target location input to the terminal device, a pest and disease outbreak prediction unit that predicts at least one type of pest and disease and weed and their occurrence at the target location, and a damage risk prediction unit that predicts the damage risk based on the growth prediction results from the crop growth prediction unit and the occurrence prediction results from the pest and disease outbreak prediction unit. The damage risk prediction unit is characterized in that the damage risk is predicted for at least one type of pest and disease and weed according to the growth stage of the target crop.

[0010] In the plant protection system disclosed herein, when a user inputs farming data such as the type of target crop and sowing date into a terminal device, the crop growth prediction unit predicts the growth stage of the target crop, and the pest and disease occurrence prediction unit predicts at least one type of pest and disease and weed that affects the target crop, along with their occurrence amounts. Based on this, the damage risk prediction unit predicts the damage risk from at least one type of pest and disease and weed affecting the target crop, and the prediction result is displayed on the terminal device. Since the damage risk displayed as a prediction result is information that takes into account (adjusts) the growth stage of the target crop, the user can obtain damage risk information on when, what kind of pest and disease and weed damage risk will occur and to what extent during the growth of the target crop, and based on this, can prepare pest and disease control plans and necessary materials in advance.

[0011] The pest and disease occurrence prediction unit may be configured to predict at least one type of pest and disease and weed and their occurrence amount for the target crop at the target location, based on the occurrence amounts of at least one type of pest and disease and weed at multiple locations, including the target location and at least one other location different from the target location. A configuration that uses pest and disease occurrence data (preferably environmental data) from multiple locations including other locations is expected to be able to predict pest and disease occurrence with higher accuracy compared to a configuration that uses pest and disease occurrence data from a single location (the target location only).

[0012] The crop growth prediction unit may be configured to identify control importance data indicating the importance of controlling at least one type of pest, disease, and weed according to the growth stage of the predicted target crop. In this configuration, one of the growth prediction results identifies control importance data for pests and diseases according to the growth stage of the target crop.

[0013] The pest and disease outbreak prediction unit may be configured to identify outbreak urgency data indicating the urgency of outbreaks for at least one type of pest, disease, or weed corresponding to the growth stage, based on the predicted occurrence and amount of at least one type of pest, disease, or weed of the target crop and the growth stage of the target crop predicted by the crop growth prediction unit. In this configuration, outbreak urgency data for pests and diseases corresponding to the growth stage of the target crop is identified as one of the outbreak prediction results.

[0014] The aforementioned control importance data and the occurrence urgency data are each identified by numerical parameter data, and the damage risk prediction unit may be configured to predict the damage risk by combining parameters indicating the control importance data and parameters indicating the occurrence urgency data, which correspond to the growth stage of the target crop and at least one type of pest, disease, and weed. In this configuration, a method for predicting damage risk is concretized.

[0015] The server may further include a specification unit that identifies pest and disease control information, including at least one type and amount of pests, diseases, and weeds of the target crop predicted by the pest and disease occurrence prediction unit, and / or at least one pest and disease control plan and necessary materials linked to the damage risk predicted by the damage risk prediction unit. In this configuration, pest and disease control information such as pest and disease control plans and necessary materials based on damage risk is displayed on the terminal device. This allows the user to obtain information such as effective types of pesticides and how to use them.

[0016] The terminal device receives data on materials used by the user, and the server further includes a pest control effect prediction unit that predicts the effectiveness of pest control using the materials based on the data on materials used, at least one type of pest and weed of the target crop and its occurrence, and / or growth information of the target crop. The specific unit may be configured to sequentially update the pest and disease control information based on the pest control effect prediction results from the pest control effect prediction unit. In this configuration, the pest and disease control information is sequentially updated based on the prediction results from the pest control effect prediction unit, so the user can obtain optimal pest and disease control information. [Effects of the Invention]

[0017] As explained above, this disclosure provides a plant protection system that predicts the risk of damage from pests, diseases, and weeds according to the type of crop and the growth stage of the crop. [Brief explanation of the drawing]

[0018] [Figure 1] Figure 1 is a diagram showing a schematic configuration of a plant protection system according to the first embodiment of this disclosure. [Figure 2] Figure 2 shows an example of a method by which the pest and disease outbreak prediction unit, which constitutes the plant protection system according to the first embodiment, predicts the occurrence of pests and diseases using pest and disease outbreak data from multiple locations. [Figure 3] Figure 3 shows an example of a method by which the damage risk prediction unit constituting the plant protection system according to the first embodiment predicts the damage risk of a target crop caused by pests and diseases using control importance data and occurrence urgency data. [Figure 4] Figure 4 shows an example of the damage risk prediction results output by the plant protection system according to the first embodiment. [Figure 5] Figure 5 shows an example of pest and disease data output by the plant protection system according to the first embodiment. [Figure 6] Figure 6 is a diagram showing the schematic configuration of a plant protection system according to the second embodiment of this disclosure, and corresponds to Figure 1. [Figure 7]FIG. 7 is a diagram showing an example of a method for a specific part constituting the plant protection system according to the second embodiment to specify pest control information. [Figure 8] FIG. 8 is a diagram showing a schematic configuration of the plant protection system according to the third embodiment of the present disclosure, and is a diagram corresponding to FIG. 1.

MODE FOR CARRYING OUT THE INVENTION

[0019] Hereinafter, the present embodiment will be described in detail based on the drawings. The following description of the preferred embodiment is merely illustrative in nature and is not intended to limit the present invention, its applications, or its uses in any way.

[0020] (First Embodiment) FIGS. 1 to 5 show the plant protection system S1 according to the present embodiment. As shown in FIG. 1, the plant protection system S1 is a system for predicting the risk of damage caused by at least one of pests and weeds of target crops, and includes a terminal device T and a server C1. The terminal device T and the server C1 are connected via a network.

[0021] <Terminal Device> The terminal device T is, for example, a farmer (a user of the system S1) who uses the plant protection system S1 to input necessary information (data) and display the damage risk information output from the server C1. Therefore, the terminal device T has a function of transmitting the input data input by the user to the server C1 and outputting and displaying prediction results such as damage risk information received from the server C1. The terminal device T is not particularly limited as long as it can communicate with the server C1 and display damage risk information and the like. For example, generally commercially available electronic devices such as mobile phones (smartphones), tablet terminals, and personal computers can be used.

[0022] The input data is information necessary to predict the risk of damage to crops (target crops) from pests and diseases, and includes, for example, farming data and pest and weed occurrence data at the target site. Farming data includes, for example, the type of crop, its sowing date data (farming work data), and location data (location information of the farmland (target site), crop growth information). Location data for the target site can be determined, for example, by identifying the region (target site) to which the access point belongs from the IP address of the terminal device T that is using (accessing) the plant protection system S1, or it can be input directly from the terminal device T. Pest and disease occurrence data includes, for example, the types of pests and diseases currently occurring at the target site and their occurrence amounts. Pest and disease occurrence data and crop growth information can also be, for example, images taken by cameras or drones, or satellite images.

[0023] <server> Server C1 is a computer that manages the entire plant protection system S1, and as shown in Figure 1, it comprises a crop growth prediction unit 1, a pest and disease occurrence prediction unit 2, and a damage risk prediction unit 3. Server C1 also comprises a database 4 that manages various data in a read-write manner as needed, and a communication unit 5 that has the function of sending and receiving various data with terminal devices T (such as measuring devices for measuring environmental data and pest monitoring devices M, which will be described later as needed).

[0024] [Crop Growth Prediction Department] Crop growth is primarily determined by meteorological factors such as temperature and day length, environmental conditions such as soil fertility, and cultivation management factors such as the type and amount of fertilizer applied. Therefore, the crop growth prediction unit 1 predicts the growth stage of the crop based on meteorological data, environmental conditions, and cultivation management conditions from the sowing date onward, according to the type of crop and sowing date entered by the user into the terminal device T. Specifically, the crop growth prediction unit 1 predicts the growth stage of the target crop based on the growth prediction data entered into the crop growth prediction unit 1. In addition, the crop growth prediction unit 1 identifies data indicating the importance of pest and disease control according to the predicted growth stage (hereinafter also referred to as "pest and disease control importance data," see Figure 3).

[0025] The growth prediction data includes, for example, farming data, weather data, growth stage data for the target crop, and pest control importance data. As shown in Figure 1, the growth prediction data is managed in the growth prediction database 41 as needed. The growth prediction database 41 may include, for example, a farming database, an environmental database, a growth stage database, a pest control importance database, and so on.

[0026] Weather data may be calculated based on sowing date data and location data entered by the user into terminal device T, or it may be obtained using an external system, or it may be obtained from the Japan Meteorological Agency or various websites. Examples of weather data include temperature, precipitation, and wind (wind volume, wind speed, wind direction) at the target location of the cultivated crop and / or its surrounding environment (e.g., the same or neighboring country, prefecture, city, town, or village). Weather data may be stored in an environmental database as needed and may be configured to be modifiable as appropriate.

[0027] The growth stage data for the target crop represents a growth algorithm specific to the crop and serves as a benchmark for predicting the crop's growth stage. Examples of growth stage data include mathematical formulas and numerical parameter data (tables, matrices) managed for each crop. The growth stage data may be stored in a growth stage database as needed, and may be predetermined based on academic information, or it may be configured to be modifiable as appropriate.

[0028] Pest and disease control importance data refers to data indicating the importance of controlling each pest and disease according to the growth stage of the target crop. As shown in Figure 3, examples of pest and disease control importance data include parameter data (tables, matrices) or mathematical formulas that quantify the occurrence of each type of pest and disease that occurs at each growth stage of the crop using a 6-level scale from 0 to 5. The score for the pest and disease control importance data (a parameter indicating the degree of importance of pest and disease control) may be stored in the pest and disease control importance database as needed, and may be predetermined based on academic information, or it may be configured to be changeable as appropriate.

[0029] Methods for predicting crop growth stages include, for example, using input data such as crop type, sowing date, and weather data to calculate daily growth based on a growth algorithm (growth stage data for the target crop), and assuming that a predetermined growth stage is reached when a certain value is obtained, then calculating and predicting the day on which each growth stage will be reached.

[0030] Furthermore, as described above, the crop growth prediction unit 1 identifies one of the pest control importance data managed in the pest control importance database, according to the predicted growth stage of the target crop.

[0031] As shown in Figure 1, the predicted growth stage data for the target crop and the identified control importance data are managed, as needed, in the growth prediction results database 42, which includes, for example, a growth stage prediction database and a specific control importance database.

[0032] After completing the above calculations, the crop growth prediction unit 1 outputs (transmits) the growth prediction results, including predicted growth stage data for the target crop and identified control importance data, to the damage risk prediction unit 3 and, if necessary, to the pest and disease occurrence prediction unit 2, as shown in Figure 1.

[0033] [Pest and Disease Occurrence Prediction Section] The pest and disease occurrence prediction unit 2 predicts the future occurrence of at least one type of pest, disease, and weed at the target site and target crop. The types of pests and diseases are determined appropriately according to the target crop and are not particularly limited, but migratory pests are also included in the prediction in the plant protection system S1. Examples of migratory pests include lepidoptera (moths, etc.), hemiptera (stink bugs, aphids, etc.), beetles (longhorn beetles, scarab beetles), thrips, diptera (flies, etc.), and orthoptera (grasshoppers, etc.). Weeds are also included in the prediction in the plant protection system S1. Weeds, like pests and diseases, have the property of spreading to other locations and need to be controlled in the same way as pests and diseases, so it is preferable to predict their types and occurrence amounts. The types of weeds are determined appropriately according to the target crop, its farmland (target site), growing area, and the grasses growing around it, and are not particularly limited. The pest and disease outbreak prediction unit 2 predicts the future occurrence of at least one pest and disease and weed based on the pest and disease outbreak prediction data input to the pest and disease outbreak prediction unit 2.

[0034] The data for predicting pest and disease outbreaks includes, for example, pest and disease outbreak data (including weed data as needed); environmental data; and pest and disease data. As shown in Figure 1, the data for predicting pest and disease outbreaks is managed in the outbreak prediction database 43 as needed. The outbreak prediction database 43 may include, for example, a pest and disease outbreak database, an environmental database, and a pest and disease database.

[0035] Pest and disease occurrence data includes, for example, pest and disease occurrence data entered by the user into terminal device T, and weed data as needed; current pest and disease occurrence data such as pest and disease occurrence data collected independently; and past pest and disease occurrence data observed at the target location and target crop. Pest and disease occurrence data is stored in the pest and disease occurrence database as needed. In addition, it is also possible to obtain and use pest and disease occurrence forecast data collected individually by public research stations in the surrounding area and environment of the cultivated crop (for example, in the same or neighboring country, prefecture, city, town, or village).

[0036] In the case of insect pests, many pests are migratory, and understanding the relationship between migratory pests and their source of migration (information about the source of migration) allows us to determine certain flight patterns. Similarly, in the case of disease outbreaks, the occurrence situation in similar environments can serve as a reference for the risk of outbreak. Therefore, as shown in Figure 2, it is preferable to use disease and pest outbreak data from multiple locations, including the target location x and at least one other location (other locations) a, b, c… that is different from the target location x. In other words, from the viewpoint of improving the accuracy of disease and pest outbreak prediction, it is preferable for the disease and pest outbreak prediction unit 2 to predict the future amount of disease and pest outbreaks at the target location and target crop based on disease and pest outbreak data from multiple locations. Examples of other locations a, b, c… include locations adjacent to the target location x, locations within the growth range of disease and pests, and locations within the migration distance range of migratory pests. Other locations a, b, c… may be appropriately determined based on meteorological data, environmental data, etc.

[0037] The independently collected pest and disease outbreak data refers to pest and disease monitoring data using a pest monitoring device M (see Figure 1) that attracts and kills pests using pheromone traps that utilize pheromones (sex pheromones) that specifically attract male pests, light traps that utilize attracting lights, and color traps that utilize colors that attract pests generally (non-specifically) or specifically, and measures the number of pests caught. The pest monitoring device M may be installed at multiple locations, preferably including the target location x and at least one other location a, b, c…. As shown in Figure 1, it is preferable that the pest monitoring device M is equipped with wireless communication means capable of transmitting monitoring results (various data such as pest images and the number of captured pests) to the server C1 via a network.

[0038] Environmental data includes meteorological data (parameters related to the movement and growth of pests, especially wind and temperature) at the target site, preferably at multiple locations including the surrounding environment; data on the types of pests and diseases, cultivated crops, weeds, topography, altitude, and geographical requirements such as urban and mountainous areas; and data on the types of work performed according to the cultivated crop. Hereinafter, this data (including meteorological data) will be simply referred to as "environmental data." Environmental data may be obtained, for example, by installing commercially available measuring devices at the target site, preferably at multiple locations, or by obtaining it from the Japan Meteorological Agency or various websites. Environmental data is stored in an environmental database as needed.

[0039] Pest and disease data refers to data related to the movement (arrival) of pests and diseases according to their species. Specifically, it refers to parameter data that quantifies the type of migratory pest, body size, and wing size. By using the biological classification of migratory pests and wing size as one of the data for predicting pest and disease outbreaks, the possible flight (migration) distance per unit day can be calculated and predicted. Pest and disease data is stored in the pest and disease database as needed.

[0040] One method for predicting future pest and disease outbreaks at a target location and target crop is to integrate (combine) pest amounts, environmental data, and, if necessary, past pest outbreak data for each of the multiple locations (target location x, other locations a, b, c…), for each predetermined period (e.g., every day), or for each predetermined number of days (e.g., one week), and then use a predictive model generated using, for example, a known statistical model or AI technology such as machine learning, to analyze the future pest outbreak at target location x. Alternatively, as shown in Figure 7, the future pest outbreak at target location x can be predicted by combining the pest outbreak amount calculated using various models (calculation algorithms) with environmental conditions such as cultivated crops A, B… and work α, β… corresponding to cultivated crops A, B… at each of the multiple locations or at each of the multiple locations.

[0041] Furthermore, the pest and disease outbreak prediction unit 2 predicts the urgency of outbreaks for each pest and disease according to the growth stage of the target crop, based on the predicted pest and disease outbreak prediction data and the growth stage data of the target crop predicted by the crop growth prediction unit 2 (by combining these data). Specifically, for each type of pest and disease predicted to occur, the pest and disease outbreak prediction unit 2 identifies the predicted amount of outbreaks for each growth stage as data indicating the urgency of the pest and disease outbreak (hereinafter also referred to as "outbreak urgency data," see Figure 3).

[0042] As for the urgency of outbreak data, as shown in Figure 3, examples include parameter data (tables, matrices) or formulas that quantify the predicted occurrence amount of pests and diseases in 51 stages from 0 to 50, depending on the type of pest or disease predicted to occur for each predicted growth stage of the target crop. As for methods of calculating the urgency of outbreak data based on the predicted occurrence amount of pests and diseases, examples include methods that use past weather data and pest and disease occurrence data (hereinafter collectively referred to as "past data"). For example, as shown in Figure 3, when predicting the occurrence amount of pests and diseases on "June 30th," which is predicted to be the "early vegetative growth stage" of the target crop, past data from the previous day (June 29th) to 10 days prior (June 29th to June 20th) is used. Similarly, when predicting the occurrence amount of pests and diseases on the following day, "July 1st," past data from the previous day (June 30th) to 10 days prior (June 30th to June 21st) is used. Since the occurrence of pests and diseases is expected to autocorrelate with past occurrences, future occurrences of pests and diseases can be predicted by sequentially using past data (or predicted data) in this manner. This predicted future occurrence of pests and diseases (numerical value) may be output directly as occurrence urgency data, or a value obtained by combining (multiplying or adding) the occurrence amount (numerical value) with a predetermined coefficient may be used as occurrence urgency data. The higher these values, the greater the predicted number of flying pests and diseases and the number of pests that will appear. The score for occurrence urgency data (a parameter indicating the degree of urgency of pest and disease outbreaks) may be stored in the occurrence urgency database as needed, and may be predetermined based on academic information, etc., or may be configured to be changeable as appropriate.

[0043] As shown in Figure 1, the predicted pest and disease outbreak prediction data and the identified outbreak urgency data are managed, as needed, in the outbreak prediction results database 44, which includes, for example, a pest and disease outbreak prediction database and an outbreak urgency database.

[0044] After completing the above calculations, the pest and disease outbreak prediction unit 2 outputs (transmits) the outbreak prediction results, including future pest and disease outbreak prediction data for the predicted target locations and target crops, and identified outbreak urgency data, to the damage risk prediction unit 3, as shown in Figure 1.

[0045] [Damage Risk Prediction Department] The damage risk prediction unit 3 takes into account the crop growth prediction results of the target crop calculated and predicted by the crop growth prediction unit 1 and the future occurrence amounts of pests and diseases calculated and predicted by the pest and disease occurrence prediction unit 2 to predict when, what kind of pests and diseases will cause damage and to what extent during crop growth. Therefore, in the damage risk prediction unit 3, damage risk is predicted for each pest and disease corresponding to the growth stage of the target crop. Specifically, as shown in Figure 1, the damage risk prediction unit 3 predicts the degree of damage risk from pests and diseases of the target crop based on the growth stage prediction data (especially the importance of control data) input from the crop growth prediction unit 1 and the pest and disease occurrence prediction data (especially the urgency of occurrence data) input from the pest and disease occurrence prediction unit 2.

[0046] One method for predicting damage risk is to combine the parameters (matrix components) of the control importance data identified by the crop growth prediction unit 1, which correspond to the growth stage (row) and the type of pest (column) of the target crop, with the parameters (matrix components) of the occurrence urgency data identified by the pest occurrence prediction unit 2. Methods for combining these parameters include multiplication and addition of multiple parameters. For example, in the example shown in Figure 3, if the growth stage is "flowering / pod count determination stage" and the pest is "beet armyworm" (hereinafter also simply referred to as "armyworm"), the product of the control importance parameter "3" and the occurrence urgency parameter "42" is identified as the damage risk parameter (damage risk amount) as "126". The damage risk amount indicates the degree of damage caused by the pest according to the growth stage. As shown in Figure 3, the damage risk amount X may be displayed (output) as a numerical value, but it may also be displayed as letters such as the alphabet based on a predetermined standard. In the example shown in Figure 3, for example, the damage risk amounts are indicated as follows: X < 10: "E", 10 ≤ X < 30: "D", 30 ≤ X < 50: "C", 50 ≤ X < 100: "B", and 100 ≤ X: "A". In this case, the damage risk is predicted to be highest in the order of E to A. As shown in Figure 1, the predicted and identified damage risk data is stored and managed in the damage risk database 45 as needed.

[0047] As shown in Figure 1, after the damage risk prediction unit 3 completes the above calculation process, the communication unit 5 outputs (transmits) output data to the terminal device T, which includes the predicted damage risk data from pests and diseases, and, if necessary, the various data mentioned above. The terminal device T displays the output data according to the user's input.

[0048] <Summary> In the plant protection system S1 configured as described above, the user inputs farming data such as the type of target crop, sowing date, and target location into a terminal device T, for example. The crop growth prediction unit 1 of system S1 then predicts the future growth stage of the target crop, and based on the prediction results, the importance of pest and disease control is identified. The pest and disease occurrence prediction unit 2 of system S1 also predicts the occurrence of pests and diseases in the target crop and target location, and based on the prediction results, the urgency of the pest and disease occurrence is identified. Finally, the damage risk prediction unit 3 of system S1 predicts the damage risk from pests and diseases according to the growth stage of the target crop, based on the data predicted and identified by the crop growth prediction unit 1 and the pest and disease occurrence prediction unit 2. The predicted damage risk is displayed on the terminal device T.

[0049] As an example of the damage risk prediction results output by the plant protection system S1 (as shown in Figure 4 on the terminal device T), the predicted growth stage of "Sachiyutaka" (the target crop) and letters A to E indicating the degree of damage risk from pests and diseases corresponding to that growth stage are displayed (E to A indicates the highest damage risk). Users who view the growth prediction results and pest and disease risks for "Sachiyutaka" can obtain necessary information for control measures against, for example, "armyworms" and "stink bugs" around July 25th, which is predicted to be the mid-vegetative growth stage; "armyworms," ​​"stink bugs," and "purple spot disease" (especially "armyworms") around August 14th, which is predicted to be the flowering and pod count determination stage; "armyworms," ​​"stink bugs," and "scarab beetles" around September 13th, which is predicted to be the seed enlargement stage; and "stink bugs" around October 23rd, which is predicted to be the final maturation stage. This allows users to prepare for pest and disease control in advance, such as by creating a pest control plan and purchasing necessary materials.

[0050] In the plant protection system S1, as shown in Figure 2, the pest and disease outbreak prediction unit 2 uses pest quantity and environmental data from multiple locations, including the target location x and other locations a, b, c, etc., to predict future pest outbreaks at the target location x. This allows for the prediction of outbreaks of migratory pests, which are difficult to predict using pest quantity data at a single location (target location x only) because they are carried by the wind, and also enables outbreak predictions that take climate change into account. In particular, since migratory pests have a certain movement pattern, using pest quantity and environmental data from the source locations can be expected to improve the accuracy of outbreak predictions. Thus, the plant protection system S1 can perform outbreak predictions that take into account the network structure of pest movement between locations (relationships regarding pest outbreaks between locations), and as a result, the accuracy of the damage risk prediction unit 3's prediction of pest and disease damage risk according to the growth stage of the target crop can be improved. Furthermore, since pest outbreak predictions can be made at multiple locations, wide-area control can also be expected.

[0051] Furthermore, the plant protection system S1 may be configured to indicate, based on the predicted damage risk, when and which crops should be cultivated (sown) to minimize the risk of damage from pests and diseases, making cultivation possible. In this configuration, users can obtain information on the optimal cultivation time for their desired crops before cultivation (sowing). In this case, for example, the user would only need to input the type of crop, location, etc., as farming data into the terminal device T before sowing.

[0052] Furthermore, as shown in Figure 5, the plant protection system S1 may be configured to output and display pest and disease data stored in the pest and disease database to the terminal device T in response to user input. In this configuration, users can immediately check pest and disease information for their crops, thus increasing the value of using system S1.

[0053] (Second embodiment) Figures 6 and 7 show a plant protection system S2 according to the second embodiment. This plant protection system S2 differs from the plant protection system S1 according to the first embodiment described above in that, as shown in Figure 6, it further includes a configuration for outputting and displaying pest and disease control information to a terminal device T. Other aspects are the same as those of the first embodiment described above, so a detailed explanation is omitted here. Also, components the same as in the first embodiment are denoted by the same reference numerals and their descriptions are omitted. Note that all the configurations of the first embodiment are also applicable to the plant protection system S2.

[0054] Pest and disease control information includes, for example, pest and disease control plans and materials necessary for pest and disease control (essential materials). Pest and disease control plans include, for example, information on measures to prepare for pest and disease damage. Essential materials include, for example, information on the type of pesticide, the amount to use, and the timing of use. Pest and disease control information only needs to include at least one of these pieces of information, and may also include other types of information.

[0055] <server> Server C2 is a computer that manages the entire plant protection system S2, and as shown in Figure 6, it includes a crop growth prediction unit 1, a pest and disease occurrence prediction unit 2, a damage risk prediction unit 3, a database 4, a communication unit 5, and a specific unit 6.

[0056] [Database] As shown in Figure 6, database 4 further comprises a planning database 46 for storing pest and disease control plan data and a materials database 47 for storing necessary materials data, along with the various databases 41 to 45 described above. The pest and disease control plan data and necessary materials data are linked to the pest and disease outbreak prediction data stored in the outbreak prediction result database 44 and / or the damage risk data stored in the damage risk database 45. Here, "linked" means, for example, the control plans and necessary materials corresponding to the pest and disease outbreak prediction and / or damage risk.

[0057] [Specific section] The identification unit 6 identifies preferred pest control plans and necessary materials from the planning database 46 and the materials database 47. One method for identifying these pest control plans and necessary materials is to identify them by analyzing the types and amounts of pests and diseases of the target crop predicted by the pest and disease occurrence prediction unit 2. In the example shown in Figure 7, the pest and disease occurrence prediction unit 2 predicts the future amount of pests at target site x by combining the amount of pests calculated using various models (calculation algorithms) with environmental conditions such as cultivated crops A, B, etc. and work α, β, etc. corresponding to cultivated crops A, B, etc. at multiple sites x or at multiple sites x. Subsequently, the identification unit 6 identifies pest and disease control information such as the necessity (amount used) and type of corresponding pesticide materials P, Q, etc. for each predicted future amount of pests.

[0058] Alternatively, control plans and necessary materials may be identified by analyzing the damage risk predicted by the damage risk prediction unit 3. For example, multiple control plans and necessary materials corresponding to parameters such as numerical values ​​or letters indicating the degree of damage risk may be stored in each database 46, 47, or control plans and necessary materials may be analyzed based on these parameters. Furthermore, control plans and necessary materials may be identified by combining disease and pest occurrence prediction data and damage risk data.

[0059] As shown in Figure 6, after the identification unit 6 completes the above calculation process, the communication unit 5 outputs (transmits) pest control information, including the identified control plan and necessary materials, to the terminal device T. In other words, the output data from the terminal device T includes pest control information in addition to the predicted pest damage risk data mentioned above. The terminal device T displays the output data according to the user's input.

[0060] <Summary> In the plant protection system S2 configured as described above, the specific unit 6 constituting system S2 identifies pest and disease control information, such as a control plan and necessary materials, based on at least one data point of pest and disease outbreak prediction and damage risk, and presents (displays) this information on the terminal device T. Users can implement appropriate pest control measures through the advertisement and presentation of information such as pesticides to be used and how to use them. Furthermore, system S2 can encourage users to purchase and use necessary materials.

[0061] (Third embodiment) Figure 8 shows a plant protection system S3 according to the third embodiment. This plant protection system S3 differs from the plant protection system S2 according to the second embodiment described above in that, as shown in Figure 8, the server C3 further includes a pest control effect prediction unit 7. Other aspects are the same as those of the second embodiment described above, so a detailed explanation is omitted here. Also, components the same as those of the second embodiment are denoted by the same reference numerals and their descriptions are omitted. Note that the configurations of the first and second embodiments are all applicable to the plant protection system S3.

[0062] In pest and disease control methods where producers (users) spray pesticides based on the appearance of pests, diseases, and weeds, it is actually difficult to objectively determine which pesticides are effective and to what extent at the individual regional level. As a result, pest control (such as the selection of pesticides) is currently carried out based on the producer's own experience and intuition. In contrast, by using the plant protection system S3, it is possible to monitor the amount of pests, weeds, and pathogens caused by them. Based on the pesticide spraying information entered by the producer into the terminal device T and the amount of pests and diseases (or the amount of diseases and weeds) and / or crop growth information obtained by the pest monitoring device M, the effectiveness of pest control can be predicted.

[0063] <Terminal device> In the plant protection system S3, the user inputs material usage data into the terminal device T as one of the input data. Material usage data refers to information about the materials actually used by the user, and includes information such as the type of pesticide used, the amount used, and the timing of use. The input material usage data is managed in the material usage database 48, which will be described later, as needed.

[0064] <server> Server C3 is a computer that manages the entire plant protection system S3, and as shown in Figure 8, it includes a crop growth prediction unit 1, a pest and disease occurrence prediction unit 2, a damage risk prediction unit 3, a database 4, a communication unit 5, a specific unit 6, and also a control effectiveness prediction unit 7.

[0065] [Database] As shown in Figure 8, database 4 further includes, along with the various databases 41 to 47 described above, a materials usage database 48 for storing materials usage data and a pest control effect prediction result database 49 for storing the prediction results from the pest control effect prediction unit 7.

[0066] [Pest Control Effect Prediction Unit] The pest control effectiveness prediction unit 7 predicts the effectiveness of pest control materials by combining information on the pesticide materials used by the user of the plant protection system S3 with information on the amount of pests and diseases that have occurred around the crops cultivated by the user, and / or information on the growth of the crops. Specifically, as shown in Figure 8, the pest control effectiveness prediction unit 7 predicts the degree of effectiveness of pest control by materials based on the material usage data entered into the terminal measure T, at least one type of pest, disease, and weed of the target crop and their occurrence amount managed in the occurrence prediction database 43 or occurrence prediction result database 44 (which may be received from the pest and disease occurrence prediction unit 2 as needed), and / or the growth prediction information of the target crop managed in the growth prediction database 41 or growth prediction result database 42 (which may be received from the crop growth prediction unit 1 as needed). In other words, the input data for predicting the effectiveness of pest control materials in the pest control effect prediction unit 7 may be a combination of material usage data and at least one type of pest, disease, and weed of the target crop and their occurrence (for example, predicting the effectiveness of pest control materials from the increase or decrease in the occurrence of pests, disease, and weeds before and after the use of materials); material usage data and target crop growth prediction information may be a combination (for example, predicting the effectiveness of pest control materials from the growth status of the crop before and after the use of materials); or material usage data, at least one type of pest, disease, and weed of the target crop and their occurrence, and target crop growth prediction information may be a combination (using both data on the increase or decrease in the occurrence of pests, disease, and weeds before and after the use of materials and the growth status of the crop to comprehensively predict the effectiveness of pest control materials). Methods for predicting the effectiveness of pest control using materials include, for example, calculating the ratio of "types of pests and weeds and their occurrence amounts" (actual values) to "types of pests and weeds and their occurrence amounts" (predicted values) predicted by the pest and weed occurrence prediction unit 2 on the reference date (actual occurrences) in relation to the growth stage of the target crop predicted by the crop growth prediction unit 1 (reference date, crop growth information); calculating the ratio of "actual crop growth information" (actual values) to "predicted results of the growth stage of the target crop predicted by the crop growth prediction unit 1" (predicted values); or methods that combine both of these methods.When crop growth prediction information is used as input data, the impact of materials used on crop growth can also be predicted. "Actual crop growth information" is entered by the user as one of the input data into, for example, terminal device T. The predicted control effect (control effect prediction result) may be displayed (output) as a numerical value such as a percentage, or it may be displayed as letters such as the alphabet based on predetermined criteria. The control effect prediction result is stored and managed in the control effect prediction result database 49 as needed.

[0067] After completing the above calculations, the pest control effectiveness prediction unit 7 outputs (transmits) the predicted pest control effectiveness result to the identification unit 6, as shown in Figure 8. The pest control effectiveness prediction unit 7 may also output (transmit) the pest control effectiveness prediction result to the terminal device T as needed. In this case, the communication unit 5 outputs (transmits) the pest control effectiveness prediction result to the terminal device T. That is, the output data of the terminal device T includes the pest control effectiveness prediction result in addition to the predicted pest and disease damage risk data and pest and disease control information. The terminal device T displays the output data according to the user's input operation.

[0068] [Specific section] In the plant protection system S3, the identification unit 6 sequentially updates pest and disease control information based on the pest and disease control effectiveness prediction results from the pest and disease control effectiveness prediction unit 7. Specifically, the identification unit 6 has the function of automatically and sequentially updating at least one of the planning database 46 and the material database 47 in response to the pest and disease control effectiveness prediction results. This optimizes the pest and disease control plan and necessary materials identified by the identification unit 6. For example, if the pest and disease control effectiveness prediction result is poor (low or no control effect), the necessary materials such as the type of pesticide, the amount used, and the timing of use are optimized, and the pest and disease control plan is also revised accordingly. On the other hand, if the pest and disease control effectiveness prediction result is good (effective control), the pest and disease control information may be maintained as is.

[0069] As shown in Figure 8, after the specific unit 6 completes the above calculation process, the communication unit 5 outputs (transmits) the updated pest control information, including the pest control plan and necessary materials, to the terminal device T.

[0070] <Summary> In the plant protection system S3 configured as described above, the user inputs data on materials used, such as the type, amount, and timing of application of pesticides actually used, as well as crop growth information as needed, into a terminal device T. The pest control effect prediction unit 7 of system S3 then predicts the degree of effectiveness of the pesticides used at a given growth stage of the target crop. Furthermore, the identification unit 6 of system S3 continuously updates the pest and disease control information displayed on the terminal device T according to the pest control effect prediction results from the pest control effect prediction unit 7. As a result, the user can always implement optimal pest control based on the continuously updated pest and disease control information. The user can also learn the degree of effectiveness of the materials used. Moreover, the user can learn the extent of the impact of the materials used on crop growth.

[0071] <Applications of plant protection systems> The following uses are envisioned for the plant protection systems S1, S2, and S3 disclosed herein. For research purposes, for example, they can be used to elucidate the detailed migration routes of migratory pests and the mechanisms of their mass outbreaks; to understand the ecology of migratory pests and predict their outbreaks; and to predict damage risks. For social applications, for example, they can be used to monitor and predict the outbreaks of pests on field crops, vegetables, fruit trees, flowers, and trees; to control and monitor pests that transmit infectious diseases; to use them as a foundational technology in plant quarantine; to support pesticide application decision-making based on pest outbreak data (smart agriculture technology); and to predict pest outbreaks and the resulting damage risks using high-resolution outbreak data. [Examples]

[0072] The present disclosure will be described below based on examples. However, the present disclosure is not limited to the following examples, and the following examples can be modified or changed in accordance with the spirit of the present disclosure, and such modifications do not exclude them from the scope of the present disclosure.

[0073] (Example 1) In Example 1, two pest outbreak prediction methods were developed for the plant protection system comprising the plant protection system described herein. These methods predict the amount of Spodoptera litura (a type of armyworm) at a given location (target location) based on past outbreak history. These methods include "multiple-location prediction," which incorporates outbreak information from other locations along with the target location, and "single-location prediction," which considers only the target location. The prediction accuracy of both methods was then compared. For the analysis, pest quantity data collected from June to September (a four-month period) was used from capture data of male adult Spodoptera litura collected between 2008 and 2018 at three fields in Hiroshima Prefecture (K town in the northwest, M city in the south, and F city in the southeast. The straight-line distance from the field in K town to the field in F city, the furthest away, is approximately 70-80 km). This pest quantity data was collected approximately every five days, a total of six times per month, by setting up traps in the fields that attracted male adult Spodoptera litura using pheromones. In other words, the pest abundance data used in the analysis includes data points recording the occurrence amount at 24 time points per year (6 records / per month × 4 months / per year). In the analysis, pest abundance data from 10, 20, 30, and 40 days prior (4 time points) was used, and in "multi-site prediction," pest abundance data from other locations at the same 4 time points was also added as an explanatory variable. In other words, in "multi-site prediction," pest abundance data from 3 locations was used, and when predicting the occurrence of the beet armyworm at one location (target location), pest abundance data from the other 2 locations (past occurrence data from 4 time points × the other 2 locations) was also used. Furthermore, the beet armyworm pest abundance data used in the analysis was logarithmically transformed, and all explanatory variables were standardized.

[0074] In Example 1, four types of machine learning models were used as examples of models for estimating the occurrence of the beet armyworm: support vector regression (SVR), general regression model (OLS), Lasso regression, and Ridge regression. Pest mass data collected from 2008 to 2016 was used for model calibration. On the other hand, pest mass data collected from 2017 to 2018 was used as test data to evaluate the extrapolability of the model.

[0075] To determine whether the model's prediction accuracy improves with or without pest quantity data from other locations, we compared the prediction accuracy on the test data using "multi-location prediction" (with pest quantity data from other locations included as an explanatory variable in the model) and "single-location prediction" (without the data). The Mean Absolute Error (MAE) was used as the evaluation method. In other words, a smaller MAE value indicates higher prediction accuracy.

[0076] Table 1 shows the results of the analysis of the prediction accuracy (MAE) of four different beet armyworm population prediction models using the methods described above. Table 2 shows the average prediction accuracy (MAE) at three locations.

[0077] [Table 1]

[0078] [Table 2]

[0079] Table 1 shows that in all four models, the "multi-site prediction" method, which uses pest quantity data from other locations as an explanatory variable, was able to predict the occurrence of the beet armyworm with higher accuracy compared to the "single-site prediction" method, which does not use pest quantity data from other locations as an explanatory variable. Furthermore, Table 2 shows that, looking at the average prediction accuracy at the three locations, the MAE for "single-site prediction" was around 0.74 to 0.76 across the models, while the MAE for "multi-site prediction" was around 0.70 to 0.71, indicating an improvement in prediction accuracy of approximately 0.03 to 0.06 in terms of MAE value. From these results, it is highly likely that a pest occurrence prediction method that considers pest quantities at multiple locations can contribute to improving the accuracy of pest occurrence prediction. Moreover, by applying a pest occurrence prediction method that considers pest quantities at multiple locations to the pest and disease occurrence prediction unit, it is expected that the prediction accuracy of damage risk in the damage risk prediction unit will improve. [Industrial applicability]

[0080] This disclosure can be applied to systems that predict the risk of crop damage caused by pests and diseases. [Explanation of symbols]

[0081] S1, S2, S3 Plant Protection System C1, C2, C3 Servers T terminal device M Pest Monitoring Device 1. Crop Growth Prediction Section 2. Pest and Disease Occurrence Prediction Section 3. Damage Risk Prediction Department 4 Databases 41. Database for predicting plant growth 42 Growth Prediction Results Database 43 Database for predicting outbreaks 44 Occurrence Prediction Result Database 45 Damage Risk Database 46. ​​Planning Database 47. Materials Database 48. Materials Used Database 49. Database of predicted control effectiveness results (as needed) 5 Communications Department 6 Specific part 7. Pest control effectiveness prediction unit

Claims

1. A plant protection system that predicts the risk of damage to target crops from at least one type of pest, disease, or weed at a target location, A terminal device into which farming data is entered by the user and in which the damage risk is displayed, The terminal device comprises a server connected to it via a network, The aforementioned server, A crop growth prediction unit predicts the growth stage of the target crop based on growth prediction data including the farming data and weather data at the target location input to the terminal device, A pest and disease occurrence prediction unit predicts at least one type of pest and disease and weed of the target crop at the aforementioned target location, and the amount of occurrence thereof. The system includes a damage risk prediction unit that predicts the damage risk based on the growth prediction results from the crop growth prediction unit and the occurrence prediction results from the pest and disease occurrence prediction unit, In the damage risk prediction unit, the damage risk is predicted for at least one type of pest, disease, and weed according to the growth stage of the target crop. A plant protection system characterized in that, in the crop growth prediction unit, control importance data indicating the importance of controlling at least one type of pest, disease, and weed corresponding to the growth stage of the predicted target crop is identified.

2. The plant protection system according to claim 1, characterized in that the pest and disease occurrence prediction unit predicts at least one type of pest and disease and weed and their occurrence amount for the target crop at the target location based on the occurrence amount of at least one type of pest and disease and weed at a plurality of locations including the target location and at least one other location different from the target location.

3. The plant protection system according to claim 1, characterized in that the pest and disease occurrence prediction unit identifies occurrence urgency data indicating the urgency of occurrence for at least one type of pest and disease and weed corresponding to the growth stage, based on the predicted occurrence amount of at least one type of pest and disease and weed of the target crop and the growth stage of the target crop predicted by the crop growth prediction unit.

4. The aforementioned data on the importance of pest control and the aforementioned data on the urgency of the outbreak are each identified by numerical parameter data. The plant protection system according to claim 3, characterized in that the damage risk prediction unit predicts the damage risk by calculating a combination of parameters indicating the importance of control data and parameters indicating the urgency of occurrence data, which correspond to the growth stage of the target crop and at least one type of pest, disease, and weed.

5. The plant protection system according to claim 1, wherein the server further comprises a specification unit that identifies pest and disease control information including at least one type and amount of pests and weeds of the target crop predicted by the pest and disease occurrence prediction unit, and / or at least one pest and disease control plan and necessary materials linked to the damage risk predicted by the damage risk prediction unit.

6. The terminal device receives data on materials used by the user. The server further includes a pest control effect prediction unit that predicts the effectiveness of pest control using the materials based on the material usage data, at least one type of pest and weed of the target crop and its occurrence amount, and / or growth information of the target crop. The plant protection system according to claim 5, characterized in that the pest and disease control information is updated sequentially in the specified unit based on the pest control effect prediction results in the pest control effect prediction unit.

Images