Agricultural work machine with driver assistance system

Abstract

The invention relates to an agricultural machine (1) comprising a driver assistance system (45) which is designed and configured to automatically optimize and adjust working parameters (44, 50a..i) and quality parameters (49, 49a) of the agricultural machine (1), wherein a process model (47) comprising characteristic curve fields (48) is assigned to the driver assistance system (45), such that an optimization method (56) to be implemented by the driver assistance system (45) is designed as a characteristic curve control (57) and the optimization method (56) generates optimized working parameters (44, 50a..i) as a control variable (51) and the driver assistance system (45) adjusts the working parameters (44, 50a..i) depending on the optimized working parameters (44, 50a..i).i) determines the quality parameters (49, 49a) of the agricultural machinery (1), wherein the process model (47) comprises a basic process model (58) and an associated affine output layer (59), wherein the basic process model (58) and the affine output layer (59) form sub-models (60) of the process model (47), and the basic process model (58) is set up and configured to map the relationships of a multitude of process parameters (61) in characteristic fields to generate an optimized quality parameter (49, 49a), and the affine output layer (59) is configured and configured to define the respective quality parameter (49, 49a) as a function of two process parameters (61).

Description

[0001] The invention relates to an agricultural work machine comprising a driver assistance system according to the preamble of claim 1.

[0002] Driver assistance systems that optimize the operation of agricultural machinery, particularly harvesters, and thereby largely relieve the operator of monitoring and adjustment tasks, are already comprehensively described in the prior art. DE 10 2010 017 687 A1 discloses a driver assistance system that determines optimized setting parameters for the working components of an agricultural harvester using a map-based approach. The method disclosed here approximates optimized working parameters in an iterative process. Such systems are well suited for quickly determining optimized working parameters of the harvester under more or less homogeneous harvesting conditions.In rapidly changing harvesting conditions, such methods have the disadvantage that, due to the inertia of the optimization process, a certain settling-in process must be completed until the harvesting machine is operating at an optimized operating point again.

[0003] Prior art has already recognized that the quality of the used map has a significant influence on the respective optimization result. For example, EP 2 687 922 A2 proposed selectively targeting specific operating points of the map located outside the current operating range in order to keep the map updated across a large portion of the overall map. This has the particular effect of adjusting the mathematical relationships governing the map in such a way as to accelerate the determination of optimized operating parameters. However, a disadvantage of such methods is the additional effort required for the targeted control of operating points outside the current operating range.

[0004] To enable a map used to optimize the operating parameters of an agricultural machine, particularly a combine harvester, to quickly determine optimized operating parameters, it is also known from the prior art, according to EP 4 154 699 A1, to assign a so-called control curve to a map. This control curve summarizes the optimal operating points over a large area of ​​the map, and the assistance system controls the parameter optimization process along this optimal curve. This has the particular advantage that the optimal operating points are reached more quickly using this map-based optimization. The disadvantage remains that there is some inertia in the system's response to abrupt changes in harvesting conditions, as the optimization system must first steer towards the new optimal operating point aligned with the control curve.

[0005] Furthermore, DE 10 2023 122 014 discloses an optimization method that can rapidly adapt the optimization of working parameters to changing process conditions, such as abruptly changing harvesting conditions. To achieve this, DE 10 2023 122 014 proposes designing the underlying process model as a dynamic nonlinear process model, comprising a static and a dynamic process model component. This allows the characteristic curve control to react quickly, i.e., dynamically, to changing process conditions, while simultaneously enabling the process model itself to be gradually adapted to these dynamic changes. Such a system is very well suited for reacting quickly to dynamic changes in process conditions.A disadvantage of this method is that the entire stored characteristic map is constantly updated, leading to long dwell times until the process model adaptation is complete. Furthermore, such methods fail to detect recurring, similar process behavior.

[0006] It is therefore an object of the invention to avoid the described disadvantages of the prior art and in particular to create a driver assistance system that adapts quickly to the real harvesting conditions and in particular reacts quickly to abrupt changes in the harvesting conditions and recognizes recurring similar process behavior.

[0007] This problem is solved according to the invention by an agricultural working machine with the characterizing features of claim 1.

[0008] By including a driver assistance system in the agricultural machinery, which is designed and configured to automatically optimize and adjust the working and quality parameters of the agricultural machinery, wherein a process model comprising characteristic curve fields is assigned to the driver assistance system, such that an optimization procedure to be implemented by the driver assistance system is designed as a characteristic curve control, and the optimization procedure generates optimized working parameters as a control variable, and the driver assistance system determines the quality parameters of the agricultural machinery depending on the optimized working parameters, and the process model comprises a basic process model and an affine output layer assigned to it,The process model and the affine output layer form sub-models of the process model. The process model is configured and configured to map the relationships of numerous process parameters in characteristic maps to generate an optimized quality parameter. The affine output layer is configured and configured to define the respective quality parameter as a function of two process parameters. This ensures that the driver assistance system adapts quickly to real harvesting conditions and, in particular, can react quickly to abrupt changes in harvesting conditions. With a small number of available data points, it is difficult to robustly adapt the process model, so a longer time is required before the process model can describe the process, the operation of the agricultural machine, under the current conditions. The affine output layer creates the possibility to...Even with a small number of available data points, the process model can be robustly adapted. The affine output layer enables the process model to describe the process more accurately and quickly, thus achieving good optimization results even with limited data. If sufficient data points are available, as is typically the case in an ongoing harvesting process, a process model that processes significantly more data points—in this case, the basic process model—can be used for optimization. This model ultimately achieves better optimization results due to the more complex relationships it considers.

[0009] In an advantageous embodiment of the invention, the basic process model and the affine output layer can be activated independently of each other, and the activation causes an adjustment of the characteristic curve fields stored in the respective sub-model, so that a high-quality optimization of the working method of the agricultural machine is achieved in a targeted manner depending on available data points / information.

[0010] In an advantageous further development of the invention, it is provided that the adaptation of the process model and its sub-models comprises adaptation phases, wherein the activation of the respective adaptation phase takes place depending on the available process information.

[0011] This ensures that the assistance system can select a suitable process model for optimizing the operation of the agricultural machinery, depending on the available process information.

[0012] In this context, it is advantageous if a first adaptation phase is designed and configured as an initialization phase, wherein in the initialization phase the adaptation of the process model is limited to the adaptation of the sub-model of affine output layers, so that, for example, when entering a crop stand, a sufficiently good optimization result is achieved with the limited process information available. In an advantageous further development of the invention, a second adaptation phase is designed and configured as an adaptation phase, wherein in the adaptation phase the adaptation of the process model is limited to the adaptation of the basic process model, so that with a larger amount of process information available, a more complex adaptation method is selected, which achieves improved optimization results.In this context, it is also advantageous to have a third adaptation phase, the optimal phase, in which the adaptation of the process model is limited to the adaptation of the basic process model. Furthermore, the parameterization of the adaptation procedure in the optimal phase differs from that in the adaptation phase, and this parameterization includes the adaptation rate and the regularization strength. This has the effect that, in a more homogeneous process, such as a harvesting process with a homogeneous crop structure and a steady-state characteristic curve, a process model encompassing complex relationships is accessed, but the adaptation of this complex process model is less extensive due to the more homogeneous conditions.Therefore, in the optimal phase, compared to the adaptation phase, there is a lower rate of adaptation and a lower level of regulation.

[0013] In an advantageous embodiment of the invention, the process information is formed from the set of available data points, the available environmental parameters, and the adaptation state, which is determined by data availability, the excitation of the process by the operating parameters, the convergence behavior of the adaptation method, and the variance of the estimates of certain process parameters in the sub-model "basic process model." The driver assistance system switches between adaptation phases depending on this process information. In this way, comprehensive process information is taken into account, ensuring a high-quality optimization process.

[0014] In this context, it is advantageous if the driver assistance system starts in the initialization phase at the beginning of the harvest and, depending on the amount of available data points, the available environmental parameters, and the adaptation state, which together constitute the process information, switches to the adaptation phase and, upon reaching a quasi-stationary phase of the basic process model, switches to the optimization phase. This ensures that, depending on the available process information, the adaptation phase that provides the best possible model and thus ultimately the best possible optimization result is always selected.Furthermore, by determining the adaptation state, as previously described, by the data availability and / or the excitation of the process by the working parameters and / or the convergence behavior of the adaptation procedure and / or the variance of the estimation of certain process parameters in the sub-model process basic model, it is ensured that the process model represents the process with sufficient accuracy to achieve a high-quality optimization process.

[0015] The complex relationships that play a role in optimizing the operation of an agricultural machine are achieved in an advantageous embodiment of the invention when the basic process model reflects a non-linear, dynamic process behavior.

[0016] In order to achieve a sufficiently good optimization result for the operation of the agricultural machinery even with less available process information, a further advantageous embodiment of the invention provides that the sub-model of affine output layers represents a linear process behavior in such a way that the respective characteristic curve field is stretched or compressed and / or shifted in space.

[0017] A particularly efficient optimization of the working method of the agricultural machine results in a further advantageous embodiment when the quality parameters include one or more of the quality parameters “threshing loss”, “broken grain fraction”, “separation loss”, “cleaning loss”, “threshing load”, “fuel consumption”, “returned grain fraction” and “returned volume”.

[0018] Similarly, an efficient and high-quality optimization process is ensured in a further advantageous embodiment when the process parameters include operating parameters of the agricultural machinery and / or environmental parameters and / or crop parameters. In this context, it is advantageous if the operating parameters of the agricultural machinery include, for example, the threshing drum speed, the concave width, the ground speed, the speeds of the threshing and separating rotors, the position of the so-called closing flaps (if the threshing and separating rotors have such flaps), the fan speed, and the upper and lower sieve widths. A further advantageous development also results when the operating parameters of the agricultural machinery include the crop parameters throughput, straw moisture, and grain moisture.

[0019] Further advantageous embodiments are the subject of further dependent claims and are described below with reference to exemplary embodiments illustrated in several figures. These show: Fig. 1 A schematic view of the agricultural machine with driver assistance system according to the invention Fig. 2 Details of the process model-based optimization of working parameters of the agricultural machinery Fig. 3 Details of the process model-based optimization of working parameters of the agricultural machinery according to the invention

[0020] The in Fig. Figure 1, schematically depicted as a combine harvester 2, incorporates a grain header 3 in its front section, which is connected to the inclined conveyor 4 of the combine harvester 2 in a manner known per se. The crop flow 5 passing through the inclined conveyor 4 is transferred in the upper, rear section of the inclined conveyor 4 to the threshing elements 7 of the combine harvester 2, which are at least partially enclosed at the bottom by a so-called threshing concave 6. A deflecting drum 8 downstream of the threshing elements 7 redirects the crop flow 5 exiting them in the rear section so that it is transferred directly to a separating device 10 designed as a separator rotor assembly 9. It is within the scope of the invention that the separating device 10 can also be designed as a straw walker, which is known per se and therefore not shown.It is also within the scope of the invention that the separating device can be designed with one or two rotors, or that the threshing elements 7 and the separating device 10 are combined into a single- or double-rotor axial flow threshing and separating device. In the separating device 10, the crop flow 5 is conveyed in such a way that freely moving grains 11 contained in the crop flow 5 are separated in the lower region of the separating device 10. Both the grains 11 separated at the threshing concave 6 and in the separating device 10 are fed via the return floor 12 and feed floor 13 to a cleaning device 17 consisting of several sieve levels 14, 15 and a blower 16. The cleaned grain flow 18 is finally transferred to a grain tank 20 by means of elevators 19.

[0021] In the rear section of the separating device 10, a shredding device 23, designed as a straw chopper 22 and enclosed in a funnel-shaped housing 21, is assigned to it in the illustrated embodiment. The straw 24 exiting the separating device 10 in the rear section is fed to the straw chopper 22 from above. Alternatively, the straw 24 can be redirected after the separating device 10 so that it is deposited directly onto the ground 25 in a swath (not shown). At the outlet of the straw chopper 22, the material stream consisting of the shredded straw 24 and the non-grain components separated in the cleaning device 17 is transferred to a material distribution device 26, which discharges the residual material stream 27 in such a way that it is spread out across the ground 25.In the embodiment shown here, the residual material stream 28 separated in the cleaning unit 17 is discharged into the straw chopper 22 by means of a so-called chaff spreader 29 and ultimately conveyed from the combine harvester 12 as a common residual material stream 27 by means of the crop distribution unit 26. In the following, the grain cutter 3, the inclined conveyor 4, the threshing units 7 and the associated threshing concave 6, the separating unit 10, the cleaning unit 17, the elevators 18, the grain tank 20, the straw chopper 22, the crop distribution unit 26, and the chaff spreader 29 are referred to as working elements 30 of the agricultural machine 1.

[0022] Furthermore, the agricultural machine 1 has a vehicle cab 31 in which at least one control and regulating device 33, equipped with a display unit 32, is arranged. This device allows for the automatic control, or manual control by the operator 34 of the agricultural machine 1, of a multitude of processes P, which will be described in more detail below. The control and regulating device 33 communicates with a multitude of sensor systems 36 via a so-called bus system 35 in a manner known per se. Details regarding the structure of the sensor systems 36 are described in detail in DE 101 47 733, the contents of which are hereby incorporated in their entirety into the disclosure of this patent application. Therefore, the structure of the sensor systems 26 will not be described again below.

[0023] Furthermore, it shows Fig. Figure 1 shows a schematic representation of the display unit 32 of the control and regulating device 33 and the computing unit 37 associated with and coupled to the display unit 32. The computing unit 37 is designed to process, in addition to the information 38 generated by the sensor systems 36, external information 39 and information 40 stored in the computing unit 37 itself, such as expert knowledge, into a variety of output signals 41. The output signals 41 are designed to include at least display control signals 42 and working element control signals 43, the former determining the contents of the display unit 32 and the latter causing changes to the various working parameters 44 of the working elements 30 of the agricultural machine 1, with arrow 44 symbolically representing the threshing drum speed.The control and regulating device 33 with its associated display unit 32 and computing unit 37 are part of the driver assistance system 45 according to the invention, which will be described in more detail below.

[0024] According to the invention, a process optimization module 46 is assigned to the driver assistance system 45, wherein the process optimization module 46 is preferably part of the computing unit 37. One or more process models 47 are assigned to the process optimization module 46, which describe the processes P taking place in the agricultural machine 1, which will be described in more detail below. The single or multiple process models 47 are described by characteristic curve fields 48, wherein the relationship between quality parameters 49 and operating parameters 50a..i of the agricultural machine 1 is defined in each characteristic curve field 48.

[0025] The processes P described by the respective process model 47 can be, for example, the threshing process, the separation process, or the cleaning process of the harvested crop 5, to name just a few examples. The quality parameters 49, 49a can be, for example, the quality parameters "threshing loss," "broken grain fraction," "separation loss," "cleaning loss," "threshing load," "fuel consumption," "returned grain fraction," and "returned volume," which are known per se and therefore not described in detail here. The operating parameters 50a..i of the agricultural machinery include, on the one hand, parameters related to the harvested crop flow 5, such as the crop throughput, the layer height of the harvested crop flow 5 detected in the agricultural machinery, and / or the moisture content of the harvested crop flow 5.

[0026] On the other hand, the working parameters 50a..i include parameters relating to the working elements 30 of the agricultural machine, such as the aforementioned threshing drum speed 44 and / or the speed of the blower 16 associated with the cleaning device 17, to name just two examples. The in Fig. 1 The example shown for a characteristic map 48 describing a process model 47 could, for example, describe the quality parameter 49 “separation loss” as a function of the working parameters threshing drum speed 44, 50a and the layer height 50i related to the crop flow 5.

[0027] With such a structure of an agricultural machine 1 designed as a combine harvester 2, comprising the described driver assistance system 45, which is designed and configured to automatically monitor and adjust the working parameters 44, 50a..i and the quality parameters 49 of the combine harvester 2, a driver assistance system 45 is created that supports the operator 34 of the combine harvester 2 in operating the combine harvester 2. This is made possible because a process model 47 comprising characteristic curve fields 48 is assigned to the driver assistance system 45, and the process model 47 defines the respective quality parameter 49 as a function of control variables 51, here the working parameters 50a..i, and the driver assistance system 50 is configured, in a manner to be described in more detail, to determine optimized working parameters 50a..i of the agricultural machine 1 as a function of the respective process model 47.

[0028] To better understand the invention, in Fig. 2. First, general aspects of the driver assistance system 45, known from the prior art, are described in detail. In an optimization step 52, the driver assistance system 45 optimizes the working parameters 44, 50a..i of the agricultural machine 1, designed as a combine harvester 2. The optimization of the working parameters 44, 55a..i depends on a preselected harvesting process strategy 53 that determines the optimization step 52. The harvesting process strategy 53 is generally specified by the operator 34 and, in the simplest case, defines the quality parameters 49 to be achieved, such as maximum permissible grain losses or a certain grain purity 11, to name just two examples of quality criteria 49. The optimized working parameters 44, 55a..i determined in the optimization step 52...The parameters are then transferred to process P and automatically adjusted at the respective working elements 30 of the combine harvester 2, for example, an optimized rotational speed 44 of a threshing drum assigned to the threshing elements 7. The driver assistance system 45 then determines the quality criteria 49a resulting from an adjustment of the working parameters 44, 55a..i.

[0029] Both the optimized work parameters 44, 50a..i and the determined quality parameters 49a are transmitted to the process optimization module 46 in a model adaptation step 54. Depending on the transmitted parameters 44, 50a..i, 49a, the process model 47 is adapted in a known manner if the stored process model 47 and its associated characteristic curve field 48 no longer describe the actual process P and thus the dependencies between the respective quality parameter 49 and the respective control variables 51 with sufficient accuracy. A data preprocessing step 55 typically precedes the model adaptation step 54.

[0030] Fig.Figure 3 now shows details of the driver assistance system 45 according to the invention with the process optimization module 46 associated with it, in which the process model 47 and the characteristic curve fields 48 associated with the process model 47 are stored, wherein the process model 47 describes the relationship between working parameters 50a..i of a working element 30, the control variables 51, and the quality parameters 49, so that the optimization procedure (56) to be implemented by the driver assistance system (45) and which will be explained in more detail below is designed as a characteristic map control (57).

[0031] This general structure enables the driver assistance system 45 to automatically optimize and adjust the working parameters 44, 50a..i and the quality parameters 49,49a of the agricultural machine 1, wherein the optimization procedure 56 generates optimized working parameters 44, 50a..i as a control variable 51 and the driver assistance system 45 determines the quality parameters 49,49a of the agricultural machine 1 depending on the optimized working parameters 44, 50a..i.

[0032] According to the invention, the process model (47) comprises a basic process model (58) and an associated affine output layer (59), wherein the basic process model (58) and the affine output layer (59) form sub-models (60) of the process model (47), and the basic process model (47) is configured and designed to map the relationships of a plurality of process parameters (61) in characteristic maps (48) for the generation of an optimized quality parameter (49, 49a), and the affine output layer (59) is configured and designed to define the respective quality parameter (49, 49a) as a function of two process parameters (61).

[0033] As already described, the working parameters 44 related to the agricultural machine 1, such as the threshing drum speed, the working parameters 50a..i related to the crop, such as the crop throughput, and the quality parameters 49, 49a, such as grain losses, are transferred to the process model 47 assigned to the process optimization module 46, whereby the transfer takes place to the basic process model 58 assigned to the process model 47. The basic process model 58 can be implemented as a neural network 62, preferably as a local model network. In the neural network 62, specific characteristic curve fields 48a..i are stored for various quality parameters 49, 49a, by means of which optimized quality parameters 49, 49a and optimized characteristic curve fields 48a..i are generated, wherein preferably each quality parameter 49 is assigned a characteristic curve field 48a..i. The results from the characteristic curve fields 48a..The derived optimized operating points are transferred according to the invention to the further sub-model 60, the affine output layer 59. The affine output layer 59 defines the quality parameter 49, 49a stored in the respective characteristic map 48 as a function of only two process parameters 61, wherein one of these process parameters 61 forms a multiplier Gglobal, which stretches or compresses the entire characteristic map 48, while the other process parameter 61, Oglabal, causes a spatial displacement 63 of the respective characteristic map 48a..i.

[0034] According to the invention, the basic process model 58 and the affine output layer 59 are further designed such that they can be activated independently of each other and the activation causes an adjustment of the characteristic fields 48 stored in the respective sub-model 61.

[0035] By making the submodel 60, affine output layer 59, dependent solely on two process parameters 61, a low level of complexity is achieved. In contrast, the basic process model 58, due to its large number of parameters and the dynamics it represents, is capable of generating significantly more complex changes to the process model 47. This difference in complexity, along with the ability to adapt both submodels 61 independently, allows for the targeted adaptation of the entire process model 47 to the current data and process situation. If only a few and minimally stimulated data are available (e.g.,At the beginning of the harvesting process in a new field, adapting the basic process model carries the risk of overfitting. This occurs when machine learning methods are trained on known data, but unknown data is used in the application. This can lead to poor optimization results or an inability of the model to accurately predict the untrained data. Due to the small number of parameters in the affine output layer 59, the risk of overfitting is eliminated by adjusting these process parameters 61. Instead, adapting these process parameters 61 allows for rapid and robust adaptation of the process model 47 with minimal and only conditionally excited process data. However, the impact on the model's behavior is limited by the low complexity of the affine output layer 59 submodel 61.Locally varying effects or changing dynamic behavior cannot be modified. Only the model output can be changed by adapting the affine output layer 59 to match the currently acquired data.

[0036] In contrast, adapting the basic process model 58 allows for targeted responses to subtle deviations between the current process model 47 and the behavior observed in the field. Adapting the basic process model 58 enables adjustments to both nonlinear and dynamic process behavior. Furthermore, the model's form allows for targeted responses to the prevailing process excitation in different local areas of the basic process model 58. This means that if good process excitation is present only at the combine harvester's current operating point, this part of the basic process model 58 can be adapted. Other areas of the basic process model 58 remain unaffected.

[0037] Given that the quality of the adaptation, the optimization, of the process model 47, its sub-models 60, and the applied characteristic curve fields 48 depends significantly on the amount of available process information 65, which will be explained in more detail below, the adaptation of the process model 47 and its sub-models 60 comprises adaptation phases 64. The activation of each adaptation phase 64 occurs, as will be explained in more detail below, depending on the available process information 65. In this context, a first adaptation phase 64 is designed and established as an initialization phase 66, in which the adaptation of the process model 47 is limited to the adaptation of the sub-model 60 affine output layer 59. A second adaptation phase 64 is designed and established as an adaptation phase 67, in which the adaptation of the process model 47 is limited to the adaptation of the basic process model 58.A third adaptation phase 64 is designed and set up as optimal phase 68, wherein in optimal phase 68 the adaptation of the process model 47 is limited to the adaptation of the basic process model 58 and the parameterization of the basic process model 58 in optimal phase 68 differs from the parameterization of the basic process model 58 in adaptation phase 67 in a manner to be described in more detail, wherein the parameterization includes the adaptation speed and the regularization strength.

[0038] By operating process model 47 in adaptation phases 64, situation-dependent adaptation of the process model 47 becomes possible. Situation recognition can be performed based on various sources. Firstly, the adapted sub-models 60 and their associated characteristic curve fields 48, as well as the prediction error, can be used as criteria to determine the quality of the entire process model 47 and its sub-models 60. Furthermore, available environmental parameters, such as crop type, crop rotation, weather changes, soil conditions, and agronomic differences related to the crop type, can be used to specify the current harvest situation. With this information about the current situation, appropriate parameterization of the model adaptation can then be performed.This allows both switching between predefined parameter sets and continuous adjustment of the parameters based on the situation. In this way, the driver assistance system can switch between adaptation phases 64 depending on the number of available data points, the available environmental parameters, and the adaptation state, which is determined by data availability, the excitation of the process by the operating parameters, the convergence behavior of the adaptation procedure, and the variance of the estimate of certain process parameters in the submodel "basic process model".

[0039] In each of the described adaptation phases 64, different adaptation algorithms and parameterizations are used for adapting the process model 47 and its submodels 60. In the initialization phase 66, robust adaptation is not possible due to the small amount of data. To nevertheless enable rapid adaptation of the entire process model 47, the adaptation of the affine output layer 59 is used in the initialization phase 66. Upon reaching the adaptation phase 67, which is characterized by the convergence of the parameters of the affine output layer 59, the adaptation switches to the basic process model 58. The nuances of the process behavior can now be learned by adapting the basic process model 58. Upon reaching the optimal phase 68, characterized by a sufficiently accurate process model 47, the adaptation remains with the parameters of the basic process model 58.The now low expected model error means that only small adjustments to the basic process model 58 and thus ultimately to the process model 47 are made through adaptation.

[0040] With regard to the available process information 65, the driver assistance system 45 starts in the initialization phase 66 at the beginning of the harvest. Depending on the amount of available data points, the available environmental parameters, and the adaptation state—determined by data availability, the excitation of the process by the operating parameters, the convergence behavior of the adaptation procedure, and the variance of the estimates of certain process parameters in the sub-model "basic process model"—the system switches to the adaptation phase 67. Upon reaching a quasi-stationary phase of the basic process model 58, the system then switches to the optimal phase 68. Only upon the detection of another large model error 69 (change in weather, field change, crop type change, etc.) is the model adaptation reset to the initialization phase 66, and the procedure is repeated.

[0041] As already described, the complex basic process model 58 represents a non-linear, dynamic process behavior, whereas the sub-model 60 affine output layer 59 represents a linear process behavior in such a way that the respective characteristic curve field 48 is stretched or compressed and / or shifted in space.

[0042] In a manner known per se, the quality parameters 49, 49a can include the “threshing loss”, the “broken grain fraction”, the “separation loss”, the “cleaning loss”, the “threshing load”, the “fuel consumption”, the “return grain fraction” and the “return volume”.

[0043] The process parameters 61 can include, in a manner known per se, the working parameters 44, 50a..i of the agricultural machinery and / or environmental parameters and / or crop parameters.

[0044] Furthermore, the operating parameters 44, 50a..i of the agricultural machine can be, in a manner known per se, the threshing drum speed, the concave width, etc. Also in a manner known per se, the operating parameters 44, 50a..i of the agricultural machine can be the crop parameters throughput, straw moisture, grain moisture. Reference symbol list: 1 Agricultural work machine 2 combine harvesters 3 Grain cutter 4 inclined conveyors 5 Harvested crop flow 6 threshing baskets 7 threshing organ 8 Deflection drum 9 Separator rotor arrangement 10 Separating device 11 grains 12 Return floor 13 Feed tray 14 Sieve level 15 sieve level 16 blowers 17 Cleaning equipment 18 grain stream 19 Elevator 20 grain tank 21 funnel-shaped housing 22 straw choppers 23 Shredding device 24 straw 25 floor 26 Goods distribution device 27 Residual material flow 28 Residual material flow 29 chaff spreaders 30 Working organ 31 Vehicle cabin 32 Display unit 33 Control and regulating device 34 operators 35 bus system 36 sensor system 37 computing units 38 Internal Information 39 External Information 40 Information 41 Output signal 42 Display signal 43 Working organ signal 44 operating parameters 45 Driver assistance systems 46 Process optimization module 47 Process model 48 characteristic curve field 49.49a Quality parameters 50a..i Operating parameters 51 Control variable 51' optimized control variable 52 Optimization step 53 Harvesting process strategy 54 Model adaptation step 55 Data preprocessing step 56 optimization methods 57 Map control 58 Basic Process Model 59 Affiner Output Layer 60 sub-model 61 process parameters 62 Neural network 63 Shift 64 Adaptation phase 65 Process Information 66 Initialization phase 67 Adaptation phase 68 Optimal phase 69 Model errors 70 71 Global process parameters Global process parameters QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] DE 10 2010 017 687 A1

[0002] EP 2 687 922 A2

[0003] EP 4 154 699 A1

[0004] DE 10 2023 122 014

[0005] DE 101 47 733

[0022]

Claims

[1] Agricultural machine (1) comprising a driver assistance system (45) which is designed and configured to automatically optimize and adjust working parameters (44, 50a..i) and quality parameters (49, 49a) of the agricultural machine (1), wherein a process model (47) comprising characteristic curve fields (48) is assigned to the driver assistance system (45), such that an optimization procedure (56) to be implemented by the driver assistance system (45) is designed as a characteristic curve control (57) and the optimization procedure (56) generates optimized working parameters (44, 50a..i) as a control variable (51) and the driver assistance system (45) determines the quality parameters (49, 49a) of the agricultural machine (1) depending on the optimized working parameters (44, 50a..i). characterized by, that the process model (47) comprises a basic process model (58) and an associated affine output layer (59), wherein the basic process model (58) and the affine output layer (59) form sub-models (60) of the process model (47) and the basic process model (58) is set up and configured to map the relationships of a multitude of process parameters (61) in characteristic fields to generate an optimized quality parameter (49, 49a) and the affine output layer (59) is set up and configured to define the respective quality parameter (49, 49a) as a function of two process parameters (61). [2] Agricultural working machine (1) a driver assistance system (45) comprising according to claim 2, characterized by , that the basic process model (58) and the affine output layer (59) can be activated independently of each other and that the activation causes an adjustment of the characteristic curve fields stored in the respective sub-model (60). [3] Agricultural working machine (1) a driver assistance system (45) comprising according to any one of the preceding claims, characterized by , that the adaptation of the process model and its sub-models (60) includes adaptation phases (64), wherein the activation of the respective adaptation phase (64) depends on the available process information (65). [4] Agricultural working machine (1) a driver assistance system (45) comprising according to claim 3, characterized by , that a first adaptation phase (64) is designed and set up as an initialization phase (66), wherein in the initialization phase (66) the adaptation of the process model (47) is limited to the adaptation of the sub-model (60) affine output layer (59). [5] Agricultural working machine (1) a driver assistance system (45) comprising according to claim 3, characterized by, that a second adaptation phase (64) is designed and set up as adaptation phase (67), wherein in adaptation phase (67) the adaptation of the process model (47) is limited to the adaptation of the basic process model (58). [6] Agricultural working machine (1) a driver assistance system (45) comprising according to claim 3, characterized by , that a third adaptation phase (64) is designed and set up as an optimal phase (68), wherein in the optimal phase (68) the adaptation of the process model (47) is limited to the adaptation of the basic process model (58), wherein the parameterization of the adaptation procedure in the optimal phase (68) differs from the parameterization of the adaptation procedure in the adaptation phase (67). [7] Agricultural working machine (1) a driver assistance system (45) comprising according to any one of claims 4 to 6, characterized by, that the driver assistance system (45) switches between the adaptation phases (67) depending on the amount of available data points, the available environmental parameters and the adaptation state. [8] Agricultural working machine (1) a driver assistance system (45) comprising according to claim 7, characterized by , that the driver assistance system (45) starts in the initialization phase (66) at the start of harvesting and switches to the adaptation phase (67) depending on the amount of available data points, the available environmental parameters and the adaptation state and switches to the optimal phase (68) when a quasi-stationary phase of the basic process model (58) is reached. [9] Agricultural working machine (1) a driver assistance system (45) comprising according to any one of the preceding claims, characterized by , that the basic process model (58) represents a non-linear, dynamic process behavior. [10] Agricultural working machine (1) a driver assistance system (45) comprising according to any one of the preceding claims, characterized by , that the submodel (60) affine output layer (59) represents a linear process behavior in such a way that the respective characteristic curve field is stretched or compressed and / or shifted in space. [11] Agricultural working machine (1) a driver assistance system (45) comprising according to any one of the preceding claims, characterized by , that the quality parameter (49, 49a) may include the “threshing loss”, the “broken grain fraction”, the “separation loss”, the “cleaning loss”, the “threshing load”, the “fuel consumption”, the “return grain fraction” and the “return volume”. [12] Agricultural working machine (1) a driver assistance system (45) comprising according to any one of the preceding claims, characterized by, that the process parameters (61) may include working parameters (44, 50a) of the agricultural machinery (1) and / or environmental parameters and / or crop parameters. [13] Agricultural working machine (1) a driver assistance system (45) comprising according to any one of the preceding claims, characterized by , that the working parameters (44, 50 a..i) of the agricultural working machine (1) may be the threshing drum speed, the concave width, etc. [14] Agricultural working machine (1) a driver assistance system (45) comprising according to any one of the preceding claims, characterized by , that the working parameters (44, 50a..i) of the agricultural machinery can be the crop parameters throughput, straw moisture, grain moisture.

Images