METHOD FOR GENERATING DATA FOR CONTROLLING AN AUTONOMOUS AGRICULTURAL MACHINE

Description

[0001] The present application relates to a method for generating data for controlling an autonomous agricultural machine, as well as a corresponding computer program product and its use.

[0002] In modern agriculture, there is a growing trend towards autonomous operation to increase productivity and efficiency. This is achieved through the use of so-called autonomous agricultural machinery, characterized by the absence of an operator in a cab. Consequently, with autonomous agricultural machinery, there is no operator present who understands the agricultural workflow with its sequence of steps and can perform or authorize individual control functions while the machine is operating on a farmland. Such autonomous agricultural machinery is therefore also referred to as unmanned agricultural machinery.

[0003] Systems for guiding vehicles are already known from the state of the art, which allow manned work machines to automatically follow a route on agricultural land.

[0004] For example, EP 1 527 667 A1 describes a system for guiding a vehicle, comprising a data storage device for storing elevation data and associated location data for an area to be worked, a position-determining receiver for determining the respective position of the vehicle within the area to be worked, a data processor comprising a roll angle estimator for estimating roll data corresponding to the respective position and a pitch angle estimator for estimating pitch data corresponding to the respective position, and a steering control that is operable to guide the vehicle using the estimated roll and pitch data so that the vehicle follows a desired path.

[0005] Although manned vehicles can be moved along a route by such systems, input from an operator located in the cab of the work machine is necessary for the execution of at least some control functions to carry out work steps while the route is being driven, so that, for example, work units adapted to the work machine can be raised or lowered, as when driving at the headland.

[0006] Autonomous agricultural machinery, however, does not have an operator with knowledge of the entire agricultural workflow, including all its steps. Rather, autonomous agricultural machinery is inherently incapable of performing any work step until it has acquired the data that enables it to carry out that step, or several steps, of an agricultural workflow.

[0007] Further methods for generating data for controlling an autonomous agricultural machine are known from publications DE 10 2016 121523 A1 and US 2019 / 289770 A1.

[0008] The publication DE 10 2016 121523 A1 describes a method for generating data for controlling an autonomous agricultural machine, which comprises the following process steps: generating digital data of an agricultural area to be cultivated by a processing unit and storing the digital data of the agricultural area to be cultivated in a database connected to the processing unit for data transmission; generating digital data of the autonomous agricultural machine by the processing unit and storing the digital data of the autonomous agricultural machine in the database;Simulating at least one of the digital data of the agricultural area to be processed and a work step of an agricultural workflow comprising a sequence of work steps by means of a mathematical model of the at least one work step by the processing device, wherein data are derived from the digital data of the autonomous agricultural machine, which are supplied as input variables to the mathematical model of the at least one work step, and wherein an instruction for the autonomous agricultural machine to carry out the work step is generated as an output variable of the simulation; transmitting the instruction generated as an output variable of the simulation to a control unit of the autonomous agricultural machine and executing the instruction by the control unit of the autonomous agricultural machine;wherein real-time data representing operating parameters and / or environmental parameters of the autonomous agricultural machine are generated by means of a sensor device during the execution of the work step by the autonomous agricultural machine.

[0009] Based on the aforementioned prior art, the object of the present invention is therefore to provide a method that enables an autonomous agricultural work machine to carry out a work step of an agricultural work process comprising a sequence of work steps on an agricultural area.

[0010] This problem is solved according to the invention by the features of independent claim 1, wherein advantageous further developments of the method according to the invention are the subject of the corresponding dependent claims 2 to 13.

[0011] Accordingly, the present invention relates to a method for generating data for controlling an autonomous agricultural machine. The method comprises generating a digital image of an agricultural area to be cultivated by a processing unit and storing the digital image of the agricultural area to be cultivated in a database connected to the processing unit for data transmission. Furthermore, the method comprises generating a digital image of the autonomous agricultural machine by the processing unit and storing the digital image of the autonomous agricultural machine in the database.The process is characterized by the simulation of at least one step of an agricultural workflow comprising a sequence of steps, using a mathematical model of that step. Data is derived from the digital representation of the agricultural area to be cultivated and the digital representation of the autonomous agricultural machine. This data serves as input to the mathematical model of the at least one step. The output of the simulation is an instruction for the autonomous agricultural machine to execute the work step.Furthermore, the procedure is characterized by transmitting the instruction generated as the output variable of the simulation to a control unit of the autonomous agricultural machine and executing the instruction by the control unit of the autonomous agricultural machine.

[0012] According to the invention, it is therefore possible to virtually run a step of a workflow comprising a sequence of steps before the actual real-world execution takes place on the agricultural land. First, digital twins of both the agricultural land to be cultivated and the autonomous agricultural machine are created, from which input variables for the simulation step are derived. The advantage of using digital twins lies in the fact that a nearly identical digital representation of the real agricultural land and autonomous agricultural machine is available, in which all parameters necessary for carrying out a work step are stored with reference to the agricultural land and the autonomous agricultural machine.In other words, the digital representations of the agricultural land and the autonomous agricultural machinery represent an essentially exact copy of the real agricultural land and the real autonomous agricultural machinery.

[0013] Based on the simulation of the work step, which is based on the mathematical model of at least one work step and uses data from the digital image of the agricultural area to be cultivated and the digital image of the autonomous agricultural machine as input variables, a work step to be carried out on the agricultural area to be cultivated can be simulated in advance with the autonomous agricultural machine, whereby instructions are generated as the output variable of the simulation, which enable the autonomous agricultural machine to subsequently carry out the previously simulated work step accordingly in reality.In other words, through simulation, which incorporates a model of the work step as well as digital representations of the agricultural area to be cultivated and the autonomous agricultural machine, the otherwise "unknowing" autonomous agricultural machine is equipped with the ability to perform the work step independently, i.e., autonomously, without an operator, on or near the machine. Based on the instructions, the control unit can then activate the individual working units of the autonomous agricultural machine and / or working units adapted to the autonomous agricultural machine, so that the previously simulated work step is carried out by the autonomous agricultural machine on the agricultural area to be cultivated.

[0014] The mathematical model of the work step defines relationships between the input variables or defines interacting links between the input variables with regard to the execution of a specific work step. When the specific data from the digital models are fed in, the instruction for carrying out the work step is generated as the output of the simulation. This allows the previously simulated work step to be carried out in reality by the autonomous agricultural machine, essentially as described, except for unforeseen events on site. The mathematical model can be based on simple and / or complex systems of mathematical equations that define the relationships, and on algorithms for solving these systems. In particular, the algorithms used to solve the systems of equations can be based on artificial intelligence, preferably an artificial neural network.

[0015] According to an advantageous embodiment of the invention, the processing device for generating the digital image of the agricultural area to be processed accesses reference data, wherein the reference data are agronomic data, in particular soil data, inventory data, yield data and / or area data, geodata and / or weather data, wherein the reference data are stored in the database and / or in a reference database independent of the database and the processing device is connected to the reference database for data transfer.

[0016] The use of reference data to generate the digital twin of the agricultural land to be cultivated ensures that the real agricultural area to be cultivated is accurately represented by the digital twin. Since the simulation of the work step uses data from the digital twin of the agricultural land to be cultivated, the use of reference data to generate the digital twin ensures that, apart from unforeseen events on-site during the execution of the work step, no deviations occur between the simulated and the actual execution of the work step.

[0017] According to an advantageous embodiment of the invention, the processing device for generating the digital image of the autonomous agricultural machine accesses data of the autonomous agricultural machine, in particular geometry data, operating status data and / or equipment data of the autonomous agricultural machine.

[0018] In accordance with the use of reference data to generate the digital image of the agricultural area to be cultivated, the access to data of the autonomous agricultural machine by the processing unit to generate the digital image of the autonomous agricultural machine ensures that the real autonomous agricultural machine to be used for cultivating the agricultural area is accurately represented by the digital twin.Since the simulation of the work step takes into account data derived from the digital twin of the autonomous agricultural machine as input variables, the use of the autonomous agricultural machine's data to generate the digital image ensures that, apart from unforeseen events on site due to the circumstances of the work step, there are no deviations between the simulated and the actual execution of the work step.

[0019] According to an advantageous embodiment of the invention, the method comprises generating a digital image of a working unit adaptable to the autonomous agricultural work machine by the processing device and storing the digital image of the working unit in the database.

[0020] In particular, data are derived from the digital image of the working unit adaptable to the autonomous agricultural machine, which are fed into the mathematical model of at least one work step as input variables for the simulation.

[0021] The creation of a digital image of a work unit adaptable to the autonomous agricultural machine, as well as the derivation of data from this digital image, which are fed into the mathematical model of the at least one work step as further input variables for the simulation, ensure that, even when carrying out work steps for which work units adapted to the autonomous agricultural machine are required, no deviations occur between the simulated and the real execution of the work step, except for situation-related unforeseeable events on site during the execution of the work step.The instruction transmitted to the control unit of the autonomous agricultural machine also ensures that the control unit can control the work units adapted to the autonomous agricultural machine in accordance with the simulated execution of the work step.

[0022] Preferably, the processing device for generating the digital image of the working unit adaptable to the autonomous agricultural work machine accesses data of the working unit, in particular geometry data, operating status data and / or equipment data of the working unit.

[0023] According to the use of reference data to generate the digital image of the agricultural area to be cultivated, as well as data from the autonomous agricultural machine to generate the digital image of the autonomous agricultural machine, the processing unit's access to data of the working unit adaptable to the autonomous agricultural machine to generate the digital image of the working unit ensures that the real working unit to be used by the autonomous agricultural machine to cultivate the agricultural area is accurately represented by the digital twin.Since the simulation of the work step takes into account data derived from the digital twin of the adaptable work unit as further input variables for the simulation, an extremely accurate correspondence between the simulated and real execution of the work step is ensured.

[0024] According to an advantageous further development of the invention, it is provided that at least one optimization goal, in particular an optimization goal relating to soil conservation, throughput or processing time, is specified in a user-specific manner, wherein the at least one optimization goal is supplied to the mathematical model of the at least one work step as a boundary condition of the simulation.

[0025] The use of optimization goals as boundary conditions in the simulation ensures that the execution of the work step can be simulated with regard to a specific objective defined by the user, for example, a farmer or agricultural contractor who owns the agricultural land to be cultivated, the autonomous agricultural machine, and / or an adaptable or adapted working unit. If, for example, the user selects an optimization goal concerning soil conservation, the instruction transmitted to the control unit will differ from one that would apply if an optimization goal concerning throughput and / or processing time were selected.It is also possible for the user to specify several optimization goals with varying weights, so that different objectives are considered proportionally in the simulation of the work step and an instruction taking these differently weighted objectives into account is generated as the output variable of the simulation, which is then used by the control unit of the autonomous agricultural work machine to control it for the actual execution of the work step.

[0026] According to an advantageous embodiment of the invention, the instruction for carrying out the work step includes a regulation that defines a route to be followed by the autonomous agricultural machine, a regulation that defines the control of a working unit in the autonomous agricultural machine, and / or a regulation that defines the control of at least one working unit adapted to the autonomous agricultural machine.

[0027] Depending on the agricultural area to be cultivated and the autonomous agricultural machine to be used, as well as any user-defined optimization goals, the processing unit generates a specific instruction for carrying out the work operation as the input variable of the simulation. This instruction contains one or more different regulations, which are then executed by the control unit of the autonomous agricultural machine in order to control the corresponding working units of the autonomous agricultural machine and / or the working units adapted to the autonomous agricultural machine according to the simulated execution of the work step.

[0028] According to an advantageous embodiment of the invention, it is provided that, as a further output variable of the simulation, an instruction for preparing the autonomous agricultural machine with regard to carrying out the work step is generated, wherein the instruction is transmitted to a person who executes the instruction for preparing the autonomous agricultural machine.

[0029] In particular, the instruction for preparing the autonomous agricultural machine includes a regulation defining a ballast date for the autonomous agricultural machine, a regulation defining a working unit to be adapted to the autonomous agricultural machine, and / or a regulation defining resources to supply the autonomous agricultural machine and / or the working unit to be adapted to the autonomous agricultural machine.

[0030] Depending on the agricultural area to be cultivated, the autonomous agricultural machine to be used, and any user-defined optimization goals, the processing unit generates a further input variable for the simulation: a specific instruction for preparing the autonomous agricultural machine for the execution of the work step. This instruction contains one or more different instructions, which are then carried out by a person, for example, the farmer, as a preparatory real-world measure for the actual execution of the work step by the autonomous agricultural machine. This preparation involves preparing the autonomous agricultural machine itself, its corresponding working units, and / or working units adapted to the autonomous agricultural machine.that the simulated execution of the work step can also be implemented in reality by the autonomous agricultural work machine.

[0031] According to the invention, it is provided that real-time data representing operating parameters and / or environmental parameters of the autonomous agricultural machine are generated by means of a sensor device during the execution of the work step by the autonomous agricultural machine.

[0032] According to the invention, the real-time data is transmitted to the processing unit for the adaptation of the digital image of the agricultural area to be cultivated and / or the digital image of the autonomous agricultural machine. Preferably, in addition, the control unit of the autonomous agricultural machine processes the real-time data in such a way that the instruction transmitted by the processing unit for the execution of the work step by the autonomous agricultural machine is automatically adapted by the control unit of the autonomous agricultural machine depending on the determined operating parameters and / or environmental parameters.

[0033] Transmitting real-time data to the processing unit allows the digital representation of the agricultural area to be cultivated and / or the autonomous agricultural machine to be continuously adapted to the actual, current conditions. For example, if the real-time data detects an obstacle in the route of the autonomous agricultural machine, the digital representation of the agricultural area to be cultivated, which does not yet contain this obstacle, can be modified so that it is recorded there and can be taken into account for future simulations of work steps.Furthermore, the autonomous agricultural machine can, based on the processing of real-time data by the control unit, react immediately and automatically to unforeseen events on-site during the execution of a work step. This is achieved by adapting the instruction transmitted by the processing unit to the control unit of the autonomous agricultural machine in real time, based on the available real-time data. For example, if the real-time data detects an obstacle in the autonomous agricultural machine's path, the control unit can automatically modify the route contained in the instruction to avoid the obstacle and then resume the original route.

[0034] According to an advantageous embodiment of the invention, the method further comprises simulating the agricultural workflow with a sequence of work steps by the processing device, wherein, as the output of the simulation of a work step of the sequence of work steps, an adapted digital image of the agricultural area to be cultivated is generated, which is stored in the database, wherein digital images of a plurality of autonomous agricultural machinery are stored in the database, wherein data are derived from the adapted digital images of the agricultural area to be cultivated and the digital images of the plurality of autonomous agricultural machinery, which are supplied as input variables to a respective mathematical model of the work step used for the simulation of a work step.and wherein, as the input variable of each simulation of a work step of the sequence of work steps, an instruction for the execution of the work step by at least one of the multitude of autonomous agricultural work machines (6) is generated.

[0035] Thus, an entire agricultural workflow, encompassing numerous sequential work steps, can be simulated using the processing device, taking into account an entire fleet of autonomous vehicles. The digital representation of the agricultural machinery being processed is updated after each simulation of a work step, reflecting the state of the field after the simulated execution. This ensures that subsequent simulations of the next work step always reflect the correct state of the agricultural area being processed. The various digital representations of the field can be stored in the database, allowing access for further simulations or analyses.After each simulation of a work step, an instruction is generated as the simulation's output variable for at least one of the numerous autonomous agricultural machines included in the simulation. This instruction is then transmitted to the control unit of the corresponding autonomous agricultural machine, enabling the control unit to execute it and for the autonomous agricultural machine to implement the previously simulated work step in reality on the agricultural area to be cultivated.

[0036] According to an advantageous embodiment of the invention, the simulation of the work step is graphically represented by means of a display device connected to the processing device.

[0037] This allows the simulation to be visualized for the user simultaneously, enabling them to follow the process during the simulation in order to visually identify any undesirable irregularities before the autonomous agricultural machine actually performs the work step and to make appropriate adjustments, for example to the mathematical model of a work step used for the simulation.The digital images of the agricultural area to be cultivated, the autonomous agricultural machine and any working units adapted to the autonomous agricultural machine include, in addition to the numerous data, an exact three-dimensional image of reality, so that these three-dimensional images are displayed together and interacting with each other during the simulation of the work step for the user according to the respective time during the execution of the work step.

[0038] The problem according to the invention is further solved by a computer program product according to independent claim 14 and by the use of a computer program product according to independent claim 15.

[0039] Accordingly, the invention also relates to a computer program product that includes commands which, when the program is executed by a processing device, cause it to execute the previously described inventive method for generating data for controlling an autonomous agricultural machine.

[0040] According to claim 15, it is provided that the execution of the program by the processing device is enabled if authorization is present.

[0041] The program that enables the processing device to execute the inventive method can therefore be activated for the user by paying a fee to a service provider who offers this service.

[0042] The present invention is explained in more detail below with reference to an embodiment shown in the figure.

[0043] FIG. 1Figure 1 shows a database-based management system 1 comprising a processing unit 2 and a database 3 connected to it for data transmission. The database 3 can be configured as a central database 3 or as a decentralized, i.e., distributed, database 3. If the database 3 is configured as a decentralized database 3, it is preferably configured as a blockchain database 3. It is also possible for the database 3 to be implemented as a cloud-based database 3. A display unit 4 can also be connected to the processing unit 2; this display unit can be either part of the management system 1 or independent of it.The display unit 4 is configured to show various graphical representations generated or transmitted by the processing unit 2 to a user 5 of the management system 1, for example, a farmer or a contractor who owns one or more autonomous agricultural machines 6, one or more adaptable or adapted work units 7, and / or one or more agricultural areas on which an agricultural workflow with a sequence of work steps is to be carried out. The processing unit 2 can also be connected to a reference database 8 for data transfer, in which reference data 9 can be stored, which the processing unit 2 can access.The communication between the various facilities, units, machines and aggregates 1, 2, 3, 4, 6, 7, 8, 9 for the transmission of data can optionally be wired and / or wireless.

[0044] Referring to the devices described above, the inventive method for generating data for controlling an autonomous agricultural machine 6 is described in detail below.

[0045] As already indicated at the outset, the control of autonomous agricultural machinery 6 is not intended to be carried out by operators who possess knowledge of a specific agricultural work process with a sequence of work steps. Rather, the autonomous agricultural machinery 6 is inherently "unknowing," meaning it is not capable of independently, i.e., autonomously, executing one or more specific work steps within the sequence of work steps. In other words, as long as the autonomous agricultural machinery 6 does not receive instructions enabling it to perform the work step or steps of the work process, it is unable to do so.In order for the autonomous agricultural machinery 6 to be able to perform a work step autonomously, instructions must therefore be generated that are transmitted to the autonomous agricultural machinery 6 so that it can carry out a work step on an agricultural area to be cultivated automatically, i.e. autonomously.

[0046] According to the invention, a digital image 10 of an agricultural area to be cultivated by the autonomous agricultural machine 6 is generated or created by means of the processing device 2 and stored in the database 3, which is connected to or communicates with the processing device 2 for data transmission. Furthermore, a digital image 11 of an autonomous agricultural machine 6 is generated or created by means of the processing device 2 and is also stored in the database 3, which is connected to or communicates with the processing device 2 for data transmission.Both the digital image 10 of the agricultural area to be cultivated and the digital image 11 of the autonomous agricultural machine 6 each represent a so-called digital twin of the real structures, i.e., the real agricultural area and the real autonomous agricultural machine 6, which identically reproduces the specific characteristics of the real structures. In addition to the digital image 10 of the agricultural area to be cultivated and the digital image 11 of the autonomous agricultural machine 6, a further digital image 12 of a working unit 7 adaptable to the autonomous agricultural machine 6, for example, a cultivator or plow, can be generated or created by means of the processing unit 2, which is then also connected to the processing unit 2 for data transmission.The data is stored in the communicating database 3. The digital images 10, 11, 12 can be designed solely as digital data models, but can optionally also include a three-dimensional virtual representation of the real structures, based on which the various characteristics of these structures become graphically visible to the user 5.

[0047] To generate the digital model 10 of the agricultural area to be processed, the processing unit 2 accesses the previously mentioned reference data 9, which includes agronomic data such as soil data (including soil compaction and soil preparation data), crop data (including crop-specific characteristics such as weed infestation, disease, and pest infestation data), yield data and / or area data (including agricultural indicators such as fertilization, irrigation, herbicide, and fungicide data), geodata, and / or weather data (including long-term climate data). As previously mentioned, the reference data 9 is stored in the reference database 8 or in the database 3 of the management system 1 and is accessible to the processing unit 2.

[0048] To generate the digital image 11 of the autonomous agricultural machine 6, the processing unit 2 accesses data 13 from the autonomous agricultural machine 6. This data 13 is primarily composed of geometric data, operating status data, and / or equipment data from the autonomous agricultural machine 6. For example, the data 13 may include data on the working units present in the autonomous agricultural machine 6, data on the dimensions of the autonomous agricultural machine 6, data on the fill levels of operating resources of the autonomous agricultural machine 6, data on the wear conditions of the working units present in the autonomous agricultural machine 6, or the like.

[0049] To generate the digital image 12 of a work unit 7 adaptable to the autonomous agricultural machine 6, the processing unit 2 accesses data 14 of the work unit 7 adaptable to the autonomous agricultural machine 6. This data 14 is also formed, in particular, by geometric data, operating condition data, and / or equipment data of the work unit 7. For example, the data 14 can also include data on the components present in the work unit 7, data on the dimensions of the work unit 7, data on the fill levels of operating fluids of the work unit 7, data on the wear conditions of the components present in the unit 7, or the like.

[0050] To generate an instruction that enables the autonomous agricultural machine 6 to perform a work step on an agricultural area in reality, such a work step of a sequence of work steps in an agricultural workflow is first digitally simulated by the processing unit 2 before it is actually carried out by the autonomous agricultural machine 6. For this purpose, the processing unit 2 accesses a mathematical model of such a work step as well as the generated and stored images 10, 11 of the agricultural area to be cultivated and the autonomous agricultural machine 6, whereby the corresponding work step is simulated based on the mathematical model of the work step. The mathematical model of the work step thus forms a necessary initial model for the simulation.To simulate the work step, data is derived from the digital model 10 of the agricultural area to be cultivated and the digital model 11 of the autonomous agricultural machine 6. This data is then fed into the mathematical model of the work step as input for the simulation. The output of the simulation is an instruction for the autonomous agricultural machine 6 to carry out the work step, enabling the autonomous agricultural machine 6 to perform the work step in reality on the agricultural area.

[0051] The mathematical model of the work step defines the relationships between the various input variables, or defines the interacting links between the various input variables with regard to the execution of a specific work step. Based on the mathematical model of the work step, the data derived as input variables from the digital image 10 of the agricultural area to be cultivated and the digital image 11 of the autonomous agricultural machine 6 are processed during the simulation in such a specific way with regard to a particular work step that a specific instruction for the execution of the work step by the autonomous agricultural machine 6 in reality is generated as an output variable.Depending on what kind of work step is to be simulated by the processing unit 2, a mathematical model specific to this work step is used by the processing unit 2.For example, if the work step to be simulated is the plowing of the agricultural area by the autonomous agricultural machine 6, the processing unit 2 selects a mathematical model specific for the simulation of the work step "plowing" from an available portfolio of mathematical models of work steps, by means of which the data derived from the digital image 10 of the agricultural area to be cultivated and the digital image 11 of the autonomous agricultural machine 6 are processed as input variables of the simulation in such a way that the instruction specific for carrying out the work step "plowing" is generated as an output variable for the autonomous agricultural machine 6.

[0052] The mathematical model can be based on simple and / or complex systems of mathematical equations that define the corresponding process-specific relationships of input variables, and on algorithms for solving these systems of equations. In particular, the algorithms used to solve the systems of equations can be based on artificial intelligence, preferably an artificial neural network. If, for the execution of the work step on the agricultural area to be cultivated, a working unit 7 has to be adapted to the autonomous agricultural machine 6, data are also derived from the digital image 12 of the adaptable working unit 7, which are fed into the mathematical model of the work step as further input variables for the simulation.

[0053] The instruction for the autonomous agricultural machine 6 to carry out the work step, obtained as an output variable by the processing unit 2 after completion of the simulation, is sent to a - in FIG. 1 The control unit (not shown) of the real autonomous agricultural machine 6 transmits the commands, preferably wirelessly, and executes them. This enables the autonomous agricultural machine 6 to carry out the previously simulated work step of the agricultural process automatically, i.e., autonomously, on the actual agricultural field.

[0054] The instruction obtained from the simulation for the autonomous agricultural machine 6 to carry out the work step comprises one or more rules that the control unit of the autonomous agricultural machine 6 processes to control the autonomous agricultural machine 6 after the instruction has been transmitted. The instruction may include a rule that defines a route to be followed by the autonomous agricultural machine 6 during the execution of the work step, a rule that defines the control of working units in the autonomous agricultural machine 6, and / or a rule that defines the control of at least one working unit 7 adapted to the autonomous agricultural machine 6.Thus, depending on the work step simulated in advance by the processing unit 2, a specific instruction can be transmitted to the autonomous agricultural machine 6 and executed by the control unit of the autonomous agricultural machine 6, ensuring that the execution of the work step in reality corresponds to the simulated execution.

[0055] In addition to the instruction for the autonomous agricultural machine 6 to carry out the work step, a further input for the simulation can be an instruction for preparing the autonomous agricultural machine 6 with regard to carrying out the work step by the processing unit 2. After its generation, this instruction is transmitted to a person, in particular the user 5 of the management system 1, who executes the transmitted instruction to prepare the autonomous agricultural machine 6 with regard to carrying out the work step.

[0056] The further instructions obtained from the simulation for preparing the autonomous agricultural machine 6 with regard to carrying out the work step also include one or more instructions to be carried out by the operator. These further instructions may include instructions defining the ballasting of the autonomous agricultural machine 6, instructions defining a working unit 7 to be adapted to the autonomous agricultural machine 6, and / or instructions defining the resources to be used to supply the autonomous agricultural machine 6 and / or the working unit 7 to be adapted to or adapted to the autonomous agricultural machine.Within the scope of the present invention, operating resources are considered to be all substances required for the operation of the autonomous agricultural machine 6, for example fuels, lubricants, hydraulic fluids or the like, as well as all substances required for the operation of the working unit 7 to be adapted or adapted to the autonomous agricultural machine 6, for example fuels, lubricants, hydraulic fluids, fertilizers, seeds, pesticides or the like.

[0057] In addition to the aforementioned digital representations and the data derived therefrom, at least one optimization goal 15 can be used as the basis for the simulation of the work step. The optimization goal 15 is a user-defined optimization goal 15 that, if specified, is fed into the mathematical model of the work step as a boundary condition of the simulation. The optimization goal 15 is preferably one relating to soil conservation, throughput, or processing time. It is possible for the user 5 to specify several such optimization goals 15 as boundary conditions in the simulation. In this case, it is possible for the user 5 to weight the optimization goals 15 specified by the user 5 against each other.If such an optimization goal 15 is specified by a user 5 as a boundary condition to be considered in the simulation, the simulation of the work step is carried out with the objective of achieving the specified optimization goal 15 or goals 15 according to their weighting. The instruction for carrying out the work step by the autonomous agricultural machine 6 and the instruction for preparing the autonomous agricultural machine 6 with regard to carrying out the work step, obtained as output variables, then also reflect compliance with the optimization goal(s) 15.

[0058] If the autonomous agricultural work machine 6, possibly with a work unit 7 adapted to it, is actually carrying out the previously simulated work step, it can be determined by means of a - in FIG. 1not shown - sensor device comprising one or more sensors which in turn are mounted on the autonomous agricultural work machine 6, on the adapted work unit 7 or on a - in FIG. 1 real-time data is generated by devices not shown, located in the immediate vicinity of the autonomous agricultural machine 6, in particular a drone, which represent operating parameters and / or environmental parameters of the autonomous agricultural machine 6.

[0059] This real-time data can be transmitted directly from the sensor device to the processing device 2 for the adaptation of the digital image 10 of the agricultural area to be processed, the digital image 11 of the autonomous agricultural machine 6 and / or the digital image 12 of the working unit 7 adaptable to the autonomous agricultural machine 6.

[0060] In addition or alternatively, the real-time data from the sensor device can be transmitted to the control unit of the autonomous agricultural machine 6 and processed by it in such a way that the instruction transmitted by the processing device 2 for the execution of the work step by the autonomous agricultural machine 6 is automatically adapted by the control unit of the autonomous agricultural machine 6 during the execution of the work step, depending on the determined operating parameters and / or environmental parameters.

[0061] The method according to the invention is not limited to the one-time simulation of a single work step by the processing device 2. Rather, it can be provided that an entire agricultural workflow with a multitude of successive work steps is to be simulated by the processing device 2. In addition to the instructions mentioned above, a customized digital image 10 of the agricultural area to be processed is generated as the input variable for the simulation and stored in the database 3. The customized digital image 10 of the agricultural area to be processed reflects the state of the agricultural area after the simulation of the work step.A simulation of a subsequent work step following the simulated work step is then carried out by the processing unit 2 using a specific mathematical model of the work step. Data is derived from the respective adapted digital image 10 of the agricultural area to be processed and fed into this mathematical model of the work step as input variables for this simulation. Furthermore, the database 3 contains digital images 11 of a large number of autonomous agricultural machines 6. Data is derived from these digital images 11 of the large number of autonomous agricultural machines 6 and also fed into the corresponding mathematical model of the respective work step as input variables for the respective simulation.Thus, for each work step to be simulated, data from a corresponding agricultural area to be cultivated during the execution of the work step and a corresponding autonomous agricultural machine 6 to be used for its execution can be derived and used as input variables for the simulation of the respective work step. The output variable of each simulation of a work step is an instruction to execute the respective work step by at least one of the numerous autonomous agricultural machines 6. Here, too, data can optionally be supplied as input variables to the mathematical model of the work step, if necessary. This data can be derived from digital images 12 of a large number of digital images 12 of work units 7 adaptable to the autonomous agricultural machines 6, which are stored in the database 3.An instruction for preparing the autonomous agricultural machine 6 with regard to carrying out the work step can also be generated as an output variable of the simulation.

[0062] Furthermore, the user 5 can follow the simulation of the work step visually. For this purpose, the simulation of the work step is graphically displayed by means of the display device 4 connected to the processing unit 2. To make this possible, the digital images 10, 11, 12, as already mentioned, comprise a three-dimensional digital image of the corresponding real structure, so that these three-dimensional digital images are displayed together and interacting with each other during the simulation of the work step, visually for the user 5 according to the respective point in time during the execution of the work step.

[0063] The display unit 4 can, for example, be part of a tablet, portable computer and / or smartphone that communicates with the processing unit 2.

[0064] Furthermore, a - in FIG. 1 A computer program product (not shown) is provided that includes commands which, when the program is executed by the processing unit 2, cause it to execute the previously described method according to the invention. Execution of the program by the processing unit 2 can be enabled if authorization is present. For example, such authorization can be obtained by paying a fee to a service provider that offers this program.

[0065] Finally, it should be noted that the embodiments described above serve only to describe the claimed teaching, but are by no means to be regarded as limiting or exhaustive. Reference symbol list

[0066] 1. Management system 2. Processing unit 3. Database 4. Display unit 5. User 6. Autonomous agricultural machine 7. Adaptable / adapted work unit 8. Reference database 9. Reference data 10. Digital representation of an agricultural area to be cultivated 11. Digital representation of the autonomous agricultural machine 12. Digital representation of an adaptable work unit 13. Data of the autonomous agricultural machine 14. Data of the adaptable / adapted work unit 15. Optimization goal

Claims

1. A method for generating data for controlling an autonomous agricultural working machine (6), which comprises the following method steps: - generating, by means of a processing device (2), a digital image (10) of an agricultural area to be processed in the form of a digital twin and storing the digital image (10) of the agricultural area to be processed in a database (3) which is connected to the processing device (2) for the transfer of data; - generating, by means of the processing device (2), a digital image (11) of the autonomous agricultural working machine (6) in the form of a digital twin and storing the digital image (11) of the autonomous agricultural working machine (6) in the database (3); - simulating, by means of the processing device (2), at least one working step of an agricultural working procedure comprising a sequence of working steps with the aid of a mathematical model of the at least one working step, wherein data are derived from the digital image (10) of the agricultural area to be processed and the digital image (11) of the autonomous agricultural working machine (6), the data being fed to the mathematical model of the at least one working step as input variables of the simulation, and wherein an instruction for the implementation of the working step by the autonomous agricultural working machine (6) is generated as an output variable of the simulation; and - transmitting the instruction generated as an output variable of the simulation to a control device of the autonomous agricultural working machine (6) and execution of the instruction by the control device of the autonomous agricultural working machine (6); wherein real time data are generated by means of a sensor device during the implementation of the working step by the autonomous agricultural working machine (6), the real time data representing operating parameters and / or environmental parameters of the autonomous agricultural working machine (6); wherein the real time data are transmitted to the processing device (2) in order to adjust the digital image (10) of the agricultural area to be processed and / or in order to adjust the digital image (11) of the autonomous agricultural working machine (6).

2. The method according to claim 1, characterized in that in order to generate the digital image (10) of the agricultural area to be processed, the processing device (2) accesses reference data (9), wherein the reference data (9) are agronomical data, in particular ground data, field crop data, yield data and / or area data, geographical data and / or weather data, wherein the reference data (9) are stored in the database (3) and / or in a reference database (8) which is independent of the database and the processing device (2) is connected to the reference database (8) for the transfer of data.

3. The method according to claim 1 or claim 2, characterized in that in order to generate the digital image (11) of the autonomous agricultural working machine (6), the processing device (2) accesses data (13) of the autonomous agricultural working machine (6), in particular geometrical data, operating status data and / or configurational data of the autonomous agricultural working machine (6).

4. The method according to one of claims 1 to 3, characterized by the following method step: - generating, by means of the processing device (2), a digital image (12) of a working assembly (7) which can be adapted to the autonomous agricultural working machine and storing the digital image (12) of the working assembly (7) in the database (3).

5. The method according to claim 4, characterized in that data are derived from the digital image (12) of the working assembly (7) which are fed to the mathematical model of the at least one working step as input variables of the simulation.

6. The method according to claim 4 or claim 5, characterized in that in order to generate the digital image (12) of the working assembly (7), the processing device (2) accesses data (14) of the working assembly (7), in particular geometrical data, operating status data and / or configurational data of the working assembly (7).

7. The method according to one of claims 1 to 6, characterized in that at least one optimization objective (15), in particular an optimization objective (15) relating to the conservation of the ground, the throughput or the processing time, is specified in a user-specific manner, wherein the at least one optimization objective (15) is fed to the mathematical model of the at least one working step as a boundary condition of the simulation.

8. The method according to one of claims 1 to 7, characterized in that the instruction for implementing the working step comprises the following: - a specification which defines a route to be followed by the autonomous agricultural working machine (6); - a specification which defines a control of working assemblies of the autonomous agricultural working machine (6); and / or - a specification which defines a control of at least one working assembly (7) which is adapted to the autonomous agricultural working machine (6).

9. The method according to one of claims 1 to 8, characterized in that an instruction for preparing the autonomous agricultural working machine (6) with a view to implementing the working step is generated as the output variable of the simulation, wherein the instruction is transmitted to a person who executes the instruction for preparing the autonomous agricultural working machine (6).

10. The method according to claim 9, characterized in that the instruction for preparing the autonomous agricultural working machine (6) comprises the following: - a specification which defines a ballast loading of the autonomous agricultural working machine (6); - a specification which defines a working assembly (7) to be adapted to the autonomous agricultural working machine (6); and / or - a specification which defines operating resources to be supplied to the autonomous agricultural working machine (7) and / or to the working assembly (7) to be adapted to the autonomous agricultural working machine (6).

11. The method according to one of claims 1 to 10, characterized in that the control device of the autonomous agricultural working machine (6) processes the real time data in a manner such that the instruction for the implementation of the working step by the autonomous agricultural working machine (6) transmitted by the processing device (2) is automatically adapted by the control device of the autonomous agricultural working machine (6) as a function of the determined operating parameters and / or environmental parameters.

12. The method according to one of claims 1 to 11, characterized by the following method step: - simulating, by means of the processing device (2), the agricultural working procedure with a sequence of working steps wherein an adapted digital image (10) of the agricultural area to be processed is generated as an output variable of the simulation of a working step of the sequence of working steps, the digital image being stored in the database (3), wherein digital images (11) of a plurality of autonomous agricultural working machines (6) are stored in the database (3), wherein data are derived from the adapted digital images (10) of the agricultural area to be processed and from the digital images (11) of the plurality of autonomous agricultural working machines (6), the data being fed as input variables to a respective mathematical model of a working step used for the simulation of the working step, and wherein an instruction for the implementation of the working step by at least one of the plurality of autonomous agricultural working machines (6) is generated as an output variable of each simulation of a working step of the sequence of working steps.

13. The method according to one of claims 1 to 12, characterized in that the simulation of the working step is graphically depicted by means of a display device (4) connected to the processing device (2).

14. A computer program product comprising commands which, during the execution of the program by a processing device (2), cause it to execute the method according to one of claims 1 to 13 in an autonomous agricultural working machine.

15. Use of a computer program product according to claim 14, characterized in that the execution of the program by the processing device (2) is enabled when an authorisation is present.