METHOD FOR PREPARING A PRODUCT IN AN ADEQUATE QUANTITY FOR LOCALIZED TREATMENT WITHIN A PLOT

Abstract

METHOD FOR PREPARING A PRODUCT IN THE APPROPRIATE QUANTITY FOR LOCALIZED TREATMENT IN A PLOT. The invention relates to the field of agricultural spraying. It concerns a method for preparing a treatment product in the quantity necessary for treating a plot using a localized spraying system.According to the invention, the method (100) comprises: a step (120) of generating a vegetation prediction map, said map being generated from a prior vegetation map and a plant development model representing the development of the cultivated plants in the plot; a step (130) of generating a spray prediction map, said map being generated from the vegetation prediction map, with a quantity of treatment product to be sprayed being determined for each zone of the spray prediction map; and a step (140) of determining a total quantity of treatment product, said quantity being calculated based on the quantities of treatment product to be sprayed in the different zones of the spray prediction map. Figure to be published: Figure 1.

Description

Description Title of the invention: METHOD FOR PREPARING A PRODUCED IN ADEQUATE QUANTITY FOR A LOCALIZED TREATMENT WITHIN A PLOT technical field

[0001] = The invention lies in the field of agricultural spraying and, more specifically, localized spraying based on real-time data. concerns a process for preparing a treatment product in the necessary quantity for the treatment of a plot of land using a localized spraying system carried by a agricultural machinery. The invention also relates to a computer program and a recording medium including instructions leading to implementation of this process. Finally, it concerns a tank filling system. localized spraying including a configured data processing device to implement the process of preparing the treatment product. PREVIOUS STATE OF THE ART

[0002] — Agricultural spraying aims to apply to different crops treatment products generally aimed at increasing their growth, their yield and / or their quality. Treatment products may include, in particular, used for weed control, disease control, and insect or parasite infestations, and provide the nutrients necessary for the proper development of crops.

[0003] A spraying system conventionally comprises a tank arranged for contain a treatment product, possibly diluted, a spray boom comprising a plurality of spray sections, each equipped with a nozzle of spraying, and a hydraulic circuit connecting the tank to the various spraying sections risation. The spray boom generally extends along a transverse axis by relative to a longitudinal direction according to which the agricultural machine moves on the plot. The hydraulic circuit may include, in particular, a pump arranged for draw the treatment product from the tank and direct it to the spray boom adjustment, and a pressure regulator arranged to maintain pressure in the circuit hydraulically at a predetermined threshold pressure. Each spray nozzle is arranged to spray the treatment product over a predetermined width of the plot defined along the transverse axis.

[0004] With the aim of reducing the use of treatment products, the systems of Spraying methods have been adapted to allow for localized treatment of plots. By localized treatment, meaning spraying the product only on the affected areas the plot actually requiring treatment. To this end, each section of pul- The system is also equipped with a distributor arranged to assume an open position, allowing product to flow from the tank to the corresponding spray nozzle, and a closed position, blocking this flow. The individual distributors can be controlled individually. The spraying system also includes an image acquisition system and a control unit. The image acquisition system is mounted on the agricultural machinery and includes at least one camera arranged to capture images of the field a few seconds before the spraying system passes over it. Typically, it comprises multiple cameras arranged along a second axis transverse to the longitudinal direction, so as to cover the entire width that can be treated by the spraying system.The control unit is configured to determine effective areas to be treated using real-time image processing on images acquired by the image acquisition system, and to individually control each distributor according to the effective areas to be treated. Image processing algorithms used to identify the necessary application rates and the accuracy of spraying systems have continuously improved. Localized spraying thus effectively addresses the issue of using treatment products judiciously based on actual needs at the time of application. However, localized spraying introduces a new challenge in managing treatment products. Since the requirements for treatment products are determined only at the time of application, or at most a few seconds beforehand, the quantity of treatment product needed for a given plot is not known in advance. It is therefore not possible to fill the spraying system's tank with the precise amount needed to treat the plot.This problem is compounded by the fact that many treatment products have a limited duration of effectiveness after preparation for spraying. In particular, treatment products may be stored in concentrated form and diluted shortly before use. After dilution, depending on the quality and temperature of the diluent, typically water, the duration of use of the treatment product may be limited to a few days or even a few hours. An agricultural operator is thus faced with the impossibility of knowing the quantity of treatment product to prepare for treating a plot, and the risk of overuse of treatment products remains significant. One solution for precisely determining the amount of treatment product to apply for localized treatment in a plot of land is to first scan the plot with the image acquisition system and identify the areas to be treated. The required amount of treatment product is then precisely determined. This However, this solution is not feasible for relatively large plots and / or plots relatively far from the storage location of the treatment products for obvious reasons of efficiency and cost. Another solution would be to use a satellite image acquired shortly before the planned treatment of the plot. However, this solution is not always applicable due to the variability in the availability of satellite images, particularly due to cloud cover, and insufficient resolution for certain types of processing. In light of the above, the invention aims to provide a solution for preparing a quantity of treatment product for a given plot that is appropriate to the actual needs of that plot. This solution must be able to rely on data acquired prior to the passage of the agricultural machinery over the plot and relatively easily accessible. Description of the invention The invention is based on the use of a map representing the state of a plot, zone by zone, at a given date prior to the planned application of the treatment, and a model capable of determining the new state of the plot at the planned treatment date. The planned treatment date may be the current date or a future date. Within a plot, the state of each zone can be determined by the state of the cultivated plants present in that zone. In this case, the map representing the plot is a vegetation map, and the model used to estimate the new state of the plot at a later date is a model of cultivated plant development. The state of the different zones of a plot may also relate to the presence of a pest or disease.In this case, the map representing the plot is a map of the presence of the pest, and the model used to estimate the new state of the plot at a later date is a model of the pest's evolution. More specifically, a first object of the invention is a method for preparing a treatment product for treating a plot of land using a localized spraying system mounted on an agricultural machine, the method comprising: a vegetation forecast map generation step, wherein a vegetation forecast map is generated from a previous vegetation map and a plant development model modeling the development of cultivated plants in the plot, the vegetation forecast map and the previous vegetation map being a graphical representation of the plot at a planned treatment date and at a date prior to the planned treatment date, respectively, each map spatially dividing the plot into a set of zones of vegetation, each vegetation zone being associated with a vegetation indicator representing the state of the cultivated plants present in said vegetation zone, a step of generating a spray forecast map, in which a spray forecast map is generated from the vegetation forecast map, the spray forecast map being a graphical representation of the plot spatially dividing the plot into a set of spray zones, each spray zone spatially corresponding to a vegetation zone and being associated with a quantity of treatment product to be sprayed according to the vegetation indicator of the corresponding vegetation zone, and a step of determining a total quantity of treatment product, necessary for the treatment of the plot, in which the total quantity of treatment product is calculated based on the quantities of treatment product to be sprayed from the different spraying areas. The condition of cultivated plants can be identified in particular by a stage of growth, by plant height, leaf area or spectral distribution of reflected radiation. A second object of the invention is a method for preparing a treatment product for treating a plot of land using a localized spraying system mounted on agricultural machinery, the method comprising: a step of generating a pest presence prediction map, in which a pest presence prediction map is generated from a previous pest presence map and a pest evolution model modeling the evolution of said pest, the pest presence prediction map and the previous pest presence map being a graphical representation of the plot at a planned treatment date and at a date prior to the planned treatment date, respectively, each map spatially dividing the plot into a set of pest zones, each pest zone being associated with a pest indicator representative of a rate of presence and / or a rate of development of the pest in said pest zone, a step of generating a spray prediction map, in which a spray prediction map is generated from the pest presence prediction map, the spray prediction map being a graphical representation of the plot spatially dividing the plot into a set of spray zones, each spray zone spatially corresponding to a pest zone and being associated with a quantity of treatment product to be sprayed according to the pest indicator of the corresponding pest zone, and a step of determining a total quantity of treatment product, necessary for the treatment of the plot, in which the total quantity of treatment product is calculated based on the quantities of treatment product to be sprayed from the different spraying areas. A pest is defined as a living organism that harms the development of cultivated plants in the plot. A pest can be a plant pathogen (for example, a fungus, bacterium, or virus), an animal pest (for example, a predator or parasite), or a weed. Generally, the previous vegetation map and the previous pest presence map can be established at a date between 1 and 60 days prior to the planned treatment date. Preferably, this period is between 5 and 25 days. The quantity of treatment product can be expressed in mass or volume. In one particular embodiment, during the step of generating a pest presence prediction map, the pest presence prediction map is also generated using information relating to the pest's presence rate and / or development rate in one or more surrounding plots. Taking this information into account makes it possible to anticipate the arrival and evolution of the pest's presence when it is present in low or no numbers in the plot under consideration. The plant development model and the pest evolution model can take into account, at a minimum, a period of time elapsed between the earlier date on which the previous map was established and the planned treatment date. In one particular embodiment, the model used to update the vegetation map or the pest presence map takes into account agronomic data specific to the plot. Specifically, the plant development model is designed to determine a vegetation indicator in each vegetation zone at a second date, based on a vegetation indicator in that zone at a first date prior to the second date, and on agronomic data relating to that vegetation zone. During the vegetation forecast map generation step, the plant development model then considers the date associated with the previous vegetation map as the first date, and the planned treatment date as the second date.Similarly, the pest evolution model allows for the determination of a pest indicator in each pest zone at a second date, based on a pest indicator in that zone at a first date, prior to the second date, and agronomic data relating to said pest zone. During the step of generating the pest presence forecast map, the pest evolution model then considers... The first date is the date associated with the previous map of the presence of the bioaggressor, and the second date is the planned treatment date. Agronomic data include, for example, meteorological data covering the period between the previous date and the planned treatment date, a previous tillage date, soil physicochemical parameters, a planting date for cultivated plants, data relating to a previous application of a treatment product, and / or data relating to a crop previously grown on the plot. Meteorological data may include, in particular, rainfall, sunshine duration, average temperature, the number of days a temperature threshold was exceeded, and / or air and / or soil humidity levels. The plant development model and the pest evolution model may take into account one or more types of agronomic data. During the vegetation forecast map generation step, the forecast map can be generated from multiple previous vegetation maps and the plant development model. The previous vegetation maps are then a graphical representation of the plot at different distinct dates prior to the planned treatment date. Similarly, during the pest presence forecast map generation step, the pest presence forecast map can be generated from multiple previous pest presence maps and the pest development model. The previous pest presence maps are then a graphical representation of the plot at different distinct dates prior to the planned treatment date. Each previous vegetation map or each previous pest presence map can be generated from at least one satellite image and / or images acquired during the passage of an image acquisition system through the plot on the previous date in question. Each image can be generated to determine radiation intensity in one or more wavelength bands of the visible, ultraviolet, and / or infrared spectrum. According to a particular embodiment, during the spray forecast map generation step, the spray forecast map is generated from the vegetation forecast map and one or more predetermined state thresholds. Each spray zone is then associated with a quantity of product to be sprayed based on the vegetation indicator of the corresponding vegetation zone and the predetermined state threshold(s). In particular, each spray zone can be associated with an initial quantity of product to be sprayed when the vegetation indicator of the corresponding vegetation zone is less than a certain threshold. a predetermined state threshold, and a second quantity of product to be sprayed when the vegetation indicator of the corresponding vegetation zone is greater than or equal to the predetermined state threshold. The first quantity of product to be sprayed or the second quantity of product to be sprayed may be zero. In a particular embodiment, during the spray prediction map generation step, the spray prediction map is generated from the pest presence prediction map and one or more predetermined presence thresholds. Each spray zone is then associated with a quantity of product to be sprayed based on the pest indicator of the corresponding pest zone and the predetermined presence threshold(s). Specifically, each spray zone can be associated with a first quantity of product to be sprayed when the pest indicator of the corresponding pest zone is below a predetermined presence threshold, and with a second quantity of product to be sprayed when the pest indicator of the corresponding pest zone is above or equal to the predetermined presence threshold.The first quantity of product to be sprayed or the second quantity of product to be sprayed can be zero. According to yet another particular embodiment, the step of determining the total quantity of treatment product includes: a substep for determining an estimation error margin, in which an estimation error margin is determined based on a reliability index associated with the plant development model or the pest evolution model, and a sub-step of calculating the total quantity of treatment product, in which the total quantity of treatment product is calculated based on the quantities of treatment product to be sprayed from the different spraying areas and the margin of error of estimation. The reliability index associated with the plant development model or the pest evolution model can depend, in particular, on the agronomic data considered by the model, and / or the time elapsed between the date on which the previous vegetation map or pest presence map was established and the planned treatment date. During the substep of calculating the total quantity of treatment product, a quantity of treatment product corresponding to the margin of error in the estimation can be calculated and added to the quantities of treatment product to be sprayed in the different spraying zones. According to another particular embodiment, the step of determining the total quantity of treatment product includes: a sub-step of determining a functional safety margin, in which A functional safety margin is determined based on parameters of the localized spraying system and / or meteorological data on the scheduled treatment date, and a sub-step of calculating the total quantity of treatment product, in which the total quantity of treatment product is calculated based on the quantities of treatment product to be sprayed from the different spraying areas and the functional safety margin. The meteorological data should preferably be specific to the plot in question. In particular, the functional safety margin may depend on wind conditions. During the substep of calculating the total quantity of treatment product, a quantity of treatment product corresponding to the functional safety margin can be calculated and added to the quantities of treatment product to be sprayed in the different spraying zones. The parameters of a localized spraying system include, for example, the distance between adjacent spray nozzles, a spray width corresponding to the ground width covered by each spray nozzle, the agricultural vehicle's travel speed, a travel speed reliability index, and / or a latency period in establishing a nominal flow rate at each spray nozzle. These parameters can, in particular, lead to the establishment of a spraying time to be applied before reaching each localized treatment area and a spraying time to be applied after leaving each area. These additional spraying times result in additional product consumption, which is therefore taken into account when determining the quantity of product required. The process of preparing a treatment product may also include a step of filling a tank of the localized spraying system with the total quantity of treatment product. The treatment product can be a biostimulant product, for example a fertilizer or a product to stimulate the natural defenses of the cultivated plant, or a biocontrol product, for example a weed killer, an insecticide or a fungicide. The invention also relates to a computer program comprising instructions which, when executed by a computer, lead the computer to implement the process described above. The invention also relates to a computer-readable recording medium comprising instructions which, when executed by a computer, lead the computer to implement the process described above. Another object of the invention is a filling system for filling a tank of a localized spraying system mounted on an agricultural vehicle. The filling system includes: a hydraulic circuit arranged to connect, in a removable manner, a tank containing a treatment product to the tank of the localized spraying system, a measuring device arranged to measure a quantity of treatment product injected into the tank, and a data processing device configured to implement the process described above. The filling system can be installed on a farm or outside of any farm, between different plots of land. According to a particular embodiment, the measuring means comprises a flow meter installed in the hydraulic circuit and a calculation unit receiving flow information from the flow meter and calculating a quantity of treatment product by time integration. In a particular embodiment, the filling system comprises a controlled valve installed in the hydraulic circuit and a control device configured to operate the controlled valve. The controlled valve is configured to assume an open or closed position based on a control signal issued by the control device. The control device is configured to receive, on the one hand, information regarding the total quantity of treatment product determined for the treatment of a plot and, on the other hand, information regarding the quantity of treatment product injected into the tank. It is further configured to issue a control signal to close the controlled valve when the quantity of treatment product injected into the tank reaches the total quantity of treatment product. Brief description of the drawings Other features, details, and advantages of the invention will become apparent from the following description, given solely by way of example and with reference to the accompanying drawings in which: [fig.1] represents an example of a process for preparing a treatment product according to the invention; [fig.2] represents an example of a step in determining the total quantity of treatment product implemented in the process of [fig.1]. DETAILED DESCRIPTION The invention aims to estimate the quantity of treatment product required for a given treatment of a plot using a localized spraying system. The treatment product may be directly beneficial to the cultivated plants or act on their environment. The product used in the localized spraying system can be a biostimulant, such as a fertilizer or a product that stimulates the natural defenses of the cultivated plant, or a biocontrol product, such as a herbicide, insecticide, or fungicide. The system includes a tank designed to hold at least one treatment product, a spray boom with multiple spray sections, and a hydraulic circuit connecting the tank to the spray sections. The hydraulic circuit may include a pump designed to draw the treatment product from the tank and deliver it to the spray boom. It may also include a pressure regulator designed to maintain the pressure in the hydraulic circuit at a predetermined threshold pressure. Localized spraying requires a preliminary diagnostic step to determine, for each area of ​​the plot, whether or not a treatment product application is necessary. This determination is at least qualitative and can potentially be quantitative. It is generally performed by analyzing images acquired by an image acquisition system comprising multiple cameras, most often mounted on the front of an agricultural machine. The cameras may be mounted on a boom extending transversely to the longitudinal direction in which the agricultural machine moves across the plot, thus covering a width of several meters. Image analysis can rely on various image processing algorithms. In particular, it can analyze plant shape and their electromagnetic spectrum.The differentiated application of the treatment to the various elementary zones is ensured by multiple spray sections, each comprising a spray nozzle and a distributor. The spray boom also extends transversely to the longitudinal direction of travel of the agricultural machinery. It is positioned behind the camera-carrying boom to allow image processing before the spray boom passes over it. This image processing is referred to as real-time processing. In practice, it can take up to a few seconds. Each spray nozzle is arranged to spray the treatment product over a predetermined width of the plot defined along the transverse axis.Each dispenser is arranged to assume an open position, in which product can flow from the tank to the corresponding spray nozzle, and a closed position, in which this flow is blocked. The dispensers are individually controlled by a control unit. They are activated in the open position when the corresponding spray nozzle passes over a treatment area, and in the closed position otherwise. Figure 1 represents an example of a process for preparing a product of treatment according to the invention. The process 100 includes a step 110 of acquiring agronomic data relating to the plot to be treated, a step 120 of generating a vegetation prediction map and / or a prediction map of the presence of a bioaggressor, a step 130 of generating a spraying prediction map, a step 140 of determining the total quantity of treatment product and a step 150 of filling the tank. The method according to the invention relies on the use of a vegetation map or a pest presence map. A vegetation map spatially divides the plot into a set of areas, called "vegetation zones," each zone being associated with a vegetation indicator representative of the state of the cultivated plants present in said vegetation zone. The state of the cultivated plants can be identified, in particular, by a growth stage, plant height, leaf area, or spectral distribution of reflected radiation. The vegetation map thus includes data relating to the spatial distribution of a plant state on the plot. This data is typically obtained from satellite imagery or from images acquired during a previous pass of the image acquisition system over the plot. This pass may have been carried out a few days or a few weeks before the planned treatment date.The vegetation map is then called the "previous vegetation map" and the date on which the images were acquired is called the "previous date". Similarly, a pest presence map spatially divides the plot into a set of zones, called "pest zones," each zone being associated with a pest indicator representing the presence and / or development rate of the pest in that zone. This data is generally obtained from images acquired during a previous pass of the image acquisition system over the plot. This pass may have been carried out a few days or weeks before the planned treatment date. The pest presence map is then called the "previous pest presence map," and the date on which the images were acquired is called the "previous date." Step 110 of the agronomic data acquisition process for the plot to be treated involves collecting information on one or more parameters likely to influence the growth of cultivated plants or the development of a pest or disease. The agronomic data ideally covers the period between the previous date and the date on which the treatment is scheduled to be applied, referred to as the "scheduled treatment date." This information can be general, covering the entire plot, or localized, meaning it varies according to specific areas within the plot. General agronomic data might include, for example, a working date. The data collected includes previous soil conditions in the plot, the sowing date of cultivated plants, data relating to a previous application of a treatment product, data relating to a crop previously grown on the plot, and / or meteorological data. Meteorological data may include, for example, rainfall, sunshine duration, average temperature, the number of days a temperature threshold was exceeded, and / or air and / or soil humidity levels. Localized agronomic data may include, for example, the physicochemical parameters of the soil, i.e., its composition. Step 120 of generating a vegetation forecast map or a pest presence forecast map involves updating the previous vegetation map or pest presence map based on, respectively, a plant development model or a pest evolution model, and the agronomic data acquired in step 110. The map is updated to accurately represent the state of the plot on the planned treatment date. It is called a "vegetation forecast map" or a "pest presence forecast map." During the vegetation forecast map generation step, for each vegetation zone, the vegetation indicator from the previous date and the agronomic data are fed into the plant development model, which then generates a vegetation indicator for the planned treatment date.Similarly, during the generation step of the pest presence forecast map, for each pest zone, the pest indicator at the previous date and the agronomic data are fed into the pest evolution model, which then generates a pest indicator at the planned treatment date. The plant development model and the pest evolution model can take into account a single type of agronomic data or several types of agronomic data. Step 130 of generating a spray forecast map involves creating a spray forecast map from the vegetation forecast map or the pest presence forecast map. The spray forecast map is a graphical representation of the plot, spatially dividing it into different zones called "spray zones." Each spray zone corresponds spatially to a vegetation zone or a pest zone and is associated with a specific quantity of treatment product to be sprayed. This quantity is determined for each zone based on the vegetation indicator of the corresponding vegetation zone or the pest indicator of the corresponding pest zone. According to a particular embodiment, the spraying forecast map is generated from the vegetation forecast map and a predetermined state threshold. When, in a vegetation zone, the vegetation indicator is below the predetermined threshold, the corresponding spray area can be associated with an initial quantity of product to be sprayed. Conversely, when, in a vegetation zone, the vegetation indicator is equal to or greater than the predetermined threshold, the corresponding spray area can be associated with a second quantity of product to be sprayed. For example, when the vegetation indicator represents a relatively early stage of growth, a quantity of product in accordance with the treatment product manufacturer's recommendations can be associated with the corresponding spray area; and, when the vegetation indicator represents a relatively advanced stage of growth, a zero quantity of product can be associated with the corresponding spray area. In another particular embodiment, the spraying prediction map is generated from the pest presence prediction map and a predetermined presence threshold. When, in a pest zone, the pest indicator has a value below the predetermined presence threshold, the corresponding spraying zone can be associated with an initial quantity of product to be sprayed. Conversely, when, in a pest zone, the pest indicator has a value greater than or equal to the predetermined presence threshold, the corresponding spraying zone can be associated with a second quantity of product to be sprayed.For example, when the bioaggression indicator is representative of a relatively low presence of insect pests, a zero quantity of product can be associated with the corresponding spray area; and, when the bioaggression indicator is representative of a relatively high presence of insect pests, a quantity of product in accordance with the recommendations of the treatment product manufacturer can be associated with the corresponding spray area. Step 140, determining the total quantity of treatment product, consists of determining the total quantity of treatment product required for the localized treatment of the plot, based on the quantities of treatment product in the different spraying zones of the spraying forecast map. In one particular embodiment, the total quantity of treatment product is determined as the sum of the quantities of treatment product in all the spraying zones of the spraying forecast map. Figure 2 represents another specific embodiment of step 140, which involves determining the total quantity of treatment product. Step 140 comprises a substep 141 for acquiring parameters of the localized spraying system, a substep 142 for determining a functional safety margin, and a substep 143 for determining an estimation margin of error and a sub-step 144 for calculating the total quantity of treatment product. Substep 141 of spray system parameter acquisition consists of acquiring parameters relating to the arrangement and / or properties of the spraying system. These parameters include, for example, the distance between the spray nozzles of the different adjacent spray sections along the axis of the spray arm, a spray width, a travel speed of the agricultural machinery and therefore of the spraying system, a reliability index of the travel speed, and / or a latency in establishing a nominal flow rate in each spray nozzle. Substep 142 of determining a functional safety margin involves estimating the additional quantity of treatment product required due to uncertainties related to the physical parameters of the spraying system and / or the weather conditions during the application of the treatment product. Specifically, in the presence of wind, it may be decided to apply more treatment product upstream and downstream of each elementary zone to be treated. The additional quantity of treatment product, called the "functional safety margin," is determined, for example, by calculating the spray duration required before reaching each elementary zone to be treated and the spray duration required after leaving each zone. Based on the estimated number of zones to be treated and the flow rate of each spray nozzle, it is then possible to calculate the functional safety margin. Substep 143, determining the margin of error in estimation, aims to quantify the maximum potential discrepancy between the total quantity of treatment product estimated by the process and the quantity of treatment product that will actually be used to treat the plot. It involves determining an additional quantity of treatment product, called the "margin of error in estimation," based on a reliability index associated with the plant development model or the pest evolution model used. This index may vary depending on the agronomic data used by the model and / or the time elapsed between the previous vegetation map and the planned treatment date. Substep 144 of the calculation of the total quantity of treatment product consists of calculating the total quantity of treatment product based on the quantities of treatment product in the different spraying zones of the spraying prediction map, the functional safety margin, and the margin of error in the estimation. In practice, the total quantity of treatment product can be calculated by adding the quantities of treatment product in all the spraying zones of the spray prediction map, functional safety margin and estimation error margin. Calculation step 140 may omit substep 141 for acquiring parameters of the localized spraying system, substep 142 for determining a functional safety margin, or substep 143 for determining an estimation error margin. Substep 144 for calculating the total quantity of treatment product then only considers the functional safety margin or the estimation error margin. Furthermore, it should be noted that substep 143 for determining an estimation error margin may be performed before, during, or after substep 141 for acquiring parameters and substep 142 for determining a functional safety margin. When the treatment product is to be applied in diluted form, step 140, which determines the total quantity of treatment product, may also include a substep for determining the quantity of diluent to be used with the total quantity of treatment product. The quantity of diluent is preferably determined according to the recommendations of the treatment product supplier. The diluent is, for example, water. Reference is again made to [Fig. 1]. Step 150, tank filling, consists of filling the tank of the localized spraying system with the total quantity of treatment product and, where applicable, the quantity of diluent, determined in step 140. In the latter case, step 150 may include a substep of mixing the treatment product with the diluent. This substep may be carried out directly in the tank of the localized spraying system, or upstream. Step 150, the filling step, can be carried out using a filling system for an agricultural installation. The filling system may include one or more storage tanks, a hydraulic circuit, and a measuring device. Each storage tank may contain a treatment product and be connected to a manual or controlled valve to regulate the flow of the treatment product through the hydraulic circuit. The hydraulic circuit is designed to connect each tank to the spraying system tank in a removable manner. The measuring device is designed to measure the quantity of treatment product injected into the tank. It may include, in particular, a flow meter installed in the hydraulic circuit and a processing unit that receives flow information from the flow meter and calculates the quantity of treatment product dispensed by time integration. The filling system may also include a control device configured to control the valve based on information relating to the quantity of treatment product measured by the measuring means and information relating to the total quantity of treatment product determined in the step 140. In this case, the control device can be configured to close the controlled valve when the measured quantity of treatment product reaches the total quantity of treatment product. Furthermore, the filling system may include a data processing device arranged to implement the other steps 110, 120, 130, and 140 of process 100 according to the invention. The device may, in particular, include a user interface for entering agronomic data and the parameters of the localized spraying system, in accordance with steps 110 and 141. It may also include a processor arranged to implement step 120, which generates a vegetation prediction map and / or a pest presence prediction map; step 130, which generates a spray prediction map; and step 140, which determines the total quantity of treatment product. In a particular embodiment, the calculation unit of the measuring means and / or the control device associated with the controlled valve are integrated into the data processing device.

Claims

Demands

1. Method for preparing a treatment product for treatment of a plot by a localized spraying system carried by a agricultural machinery, the process (100) comprising: a step (120) in generating a vegetation forecast map, in which a vegetation forecast map is generated from from a previous vegetation map and a development model of plant modeling the development of cultivated plants in the plot, the vegetation forecast map and the previous vegetation map Getation being a graphical representation of the plot at a date planned treatment and at a date prior to the planned date of processing, respectively, each map spatially dividing the plot into a set of vegetation zones, each vegetation zone- Getation being associated with a vegetation indicator representative of a state of cultivated plants present in said vegetation zone, a step (130) of generating a spray prediction map, in which a spraying forecast map is generated from from the vegetation forecast map, the spray forecast map risation being a graphical representation of the parcel dividing spa- Termally the plot into a set of spraying zones, each spray zone spatially corresponding to a zone of ve- gestation and being associated with a quantity of treatment product at spray according to the vegetation indicator of the area to be treated. corresponding gestation, and a step (140) of determining a total quantity of product of treatment, necessary for the treatment of the plot, in which the The total quantity of treatment product is calculated based on the quantities of treatment product to be sprayed in the different areas of spray.

2. Method for preparing a treatment product for treatment of a plot by a localized spraying system carried by a agricultural machinery, the process (100) comprising: a step (100) of generating a presence prediction map of a bioaggressor, in which a map predicting the presence of the bio- aggressor is generated from a previous bio- presence map aggressor and a model of the evolution of the bioaggressor modeling the evolution of said bioaggressor, the map predicting the presence of the bio- aggressor and the previous map of the bioaggressor's presence being a re- graphical representation of the plot at a planned treatment date and at a date prior to the scheduled treatment date, respectively each map spatially divides the plot into a set of zones of bioaggression, each bioaggression zone being associated with an in- indicator of bioaggression representative of a rate of presence and / or a development rate of the bioaggressor in said bio- zone assault, a step (130) of generating a spray prediction map, in which a spraying forecast map is generated from from the pest presence prediction map, the map of spray forecast being a graphical representation of the plot spatially dividing the plot into a set of zones spraying, each spraying zone corresponding spatially to a bioaggression zone and being associated with a quantity of product treatment to be sprayed according to the bioaggression indicator of the corresponding bioaggression zone, and a step (140) of determining a total quantity of product of treatment, necessary for the treatment of the plot, in which the The total quantity of treatment product is calculated based on the quantities of treatment product to be sprayed in the different areas of spray.

3. A method according to claim 2, wherein, during step (120) generating a map predicting the presence of a pest, the In addition, a map predicting the presence of the bioaggressor is generated at based on information relating to an attendance rate and / or a rate of dis- development of the pest in one or more surrounding plots ronnantes.

4. A method according to any one of the preceding claims, wherein the plant development model is arranged to determine an in- vegetation indicator in each vegetation zone at a second date from a vegetation indicator in this area to a first date, prior to the second date, and agro- data standards relating to said vegetation zone, or in which the model The evolution of the bioaggressor is arranged to determine an indicator of bioaggression in each bioaggression zone at a second date to starting from a bioaggression indicator in this area to a first date, prior to the second date, and agronomic data relating to the said bioaggression zone.

5. A method according to claim 4, wherein the agronomic data include meteorological data covering a period between said earlier date and said scheduled treatment date, a date of previous soil preparation, physico-chemical parameters of the soil, a sowing date of cultivated plants, data relating to an ap- previous application of a treatment product, and / or data relating to a crop previously grown on the plot.

6. A method according to any one of the preceding claims, wherein, at during step (120) of generating a forecast map of ve- The vegetation forecast map is generated from a plurality of previous vegetation maps and development model- plant mapping, previous vegetation maps being a representation graphic representation of the plot at different distinct dates, an- prior to the scheduled date of treatment, or in which, during the step of generating a map predicting the presence of a bio- aggressor, the bioaggressor presence prediction map is generated based on a plurality of previous maps of the presence of the bioaggressor and the evolution model of the bioaggressor, the previous maps of The presence of the bioaggressor is a graphical representation of the plot on different distinct dates, prior to the scheduled date of treatment.

7. A method according to any one of the preceding claims, wherein each previous vegetation map or each previous map of presence of the bioaggressor is generated from at least one satellite image and / or of images acquired during a pass through an acquisition system of images in the plot at the earlier date considered.

8. A method according to any one of the preceding claims, wherein the step (140) determination of the total quantity of treatment product understand : a substep (143) of determining a margin of error estimation, in which a margin of error in the estimation is de- completed in [function of a reliability index associated with the development model- plant development or the evolutionary model of the pest, and a substep (144) of calculating the total quantity of product treatment, in which the total quantity of treatment product is calculated based on the quantities of treatment product to be sprayed of the different spraying zones and the margin of error estimation.

9. A method according to any one of the preceding claims, wherein the step (140) determination of the total quantity of treatment product understand : a substep (142) of determining a functional safety margin functional, in which a functional safety margin is defined completed according to spraying system parameters localized and / or meteorological data as of the expected date of treatment, and a substep (144) of calculating the total quantity of product treatment, in which the total quantity of treatment product is calculated based on the quantities of treatment product to be sprayed of the different spraying zones and the functional safety margin tional.

10. | Method according to claim 9, wherein the system parameters localized spraying includes a distance between nozzles of adjacent spray patterns, a spray width corresponding to a ground width covered by each spray nozzle, a speed of agricultural machinery movement, an indicator of speed reliability displacement, and / or latency in establishing a nominal flow rate in each spray nozzle.

11. A method according to any one of the preceding claims comprising, in in addition, a step (150) of filling a tank of the spraying system- localized adjustment with the total quantity of treatment product.

12. A method according to any one of the preceding claims, wherein the treatment product is a biostimulation product or a product of biocontrol.

13. | A computer program comprising instructions which, when they are executed by a computer, leading the computer to implement the method according to one of the preceding claims.

14. Computer-readable recording medium comprising ins- instructions which, when executed by a computer, lead this one to implement the process according to one of claims 1 to 12.

15. Filling system for filling a tank of a spraying system localized filling system carried by agricultural machinery including: a hydraulic circuit arranged to connect in a removable manner a reservoir containing a treatment product to the tank of the localized spraying system, a measuring device arranged to measure a quantity of treatment product injected into the tank, and a data processing device configured to implement the method according to any one of claims 1 to 12.