FIELD CHOPPERS AND OPERATING PROCEDURES FOR THEM

Description

[0001] The present invention relates to a forage harvester, an operating method for it and a computer program product that can be used in the forage harvester or for carrying out the method.

[0002] When operating a forage harvester, a balance must be struck between the requirements of breaking down the nutrients contained in the harvested crop so that they can be fully and efficiently utilized by livestock fed with the chopped crop, and enabling time- and energy-efficient harvesting operations. On the one hand, grains contained in the harvested crop should be chopped finely enough for easy digestion; on the other hand, as little time and drive energy as possible should be used for this chopping process.

[0003] To quantify the suitability of chopped corn as animal feed, the Corn Silage Processing Score, abbreviated as CSPS, has been defined. The CSPS is defined as a function of the size distribution of fiber and starch particles in feed material and is conventionally measured by sieving methods performed on the finished feed.

[0004] From DE 10 2020 123 526 A1, a forage harvester is known in which operating parameters relevant to the harvesting result, such as chopper drum speed, cracker gap width, or speed difference of the cracker rollers, are linked to a processing quality by means of a characteristic map, so that suitable values ​​of these operating parameters can be determined and set for a given processing quality based on the characteristic map. However, a definition of the processing quality is not given.

[0005] DE 10 2018 104 286 describes a forage harvester in which the variable to be set using a characteristic map is the compactability of the chopped material.

[0006] One object of the invention is to provide a self-propelled forage harvester that produces nutritionally valuable chopped material with minimal time and energy input.

[0007] According to one aspect of the invention, the problem is solved by providing a self-propelled forage harvester with a chopping drum driven at a variable speed for chopping crop material, a pair of cracker rollers with variable gap width arranged behind the chopping drum on a path of the crop material through the forage harvester for crushing kernels in the chopped crop material, and a control unit for controlling the speed and the gap width, in which the control unit is part of an automatic setting system controlled by a characteristic map, wherein the characteristic map describes a Corn Silage Processing Score, CSPS, as a function of at least the speed and the gap width.

[0008] The control unit can be operated to adjust the speed and gap width according to a target value of the CSPS, based on the characteristic map.

[0009] If the control unit detects a significant deviation between the target and actual values ​​of the CSPS, an operating parameter must be adjusted to reduce the deviation. Due to the high inertia of the chopping drum, adjusting its rotational speed is energy-intensive and should be avoided, even if it could be set arbitrarily. In practice, however, the drum speed is usually directly linked to the motor speed, so that if the motor is to be operated at the optimum of its speed-torque characteristic, the drum speed is fixed. Therefore, it is preferable for the control unit to first attempt to counteract the deviation by changing the gap width in a direction for which a reduction in the deviation is expected based on the characteristic map, and to only change the rotational speed if the gap width cannot be changed further because it has reached the limit of its adjustment range.

[0010] Measuring the CSPS using conventional sieving tests is too time-consuming to allow for adjustments to the forage harvester's settings based on the results. Since the properties of the harvested crop can change during harvesting, for example due to varying soil conditions, adjusting the forage harvester's settings in this way is hardly practical. Therefore, it is preferable to use an image processing unit connected to a camera positioned along the crop's path behind the cracker roller pair to determine the actual CSPS value from images of the crop provided by the camera.

[0011] The image processing unit can be configured, in particular, to determine at least one dimension of the particles depicted in the images and, based on this dimension, to decide which size fraction the particle belongs to for determining the actual CSPS value. This eliminates the need for mechanical sieving, as the sieve fraction into which the harvested material particles would have ended up is determined computationally. In this way, an actual CSPS value can be determined with such a short delay, on the order of seconds or fractions of a second, that it can be assumed that the resulting corrections to the gap width and rotational speed are still appropriate for the harvested material being processed at the time of the correction.

[0012] The CSPS value achieved by the automatic adjustment system during operation depends not only on rotational speed and gap width, but also on the characteristics of the crop, which can vary depending on the variety, location, and climatic conditions during growth. Therefore, a map that once accurately predicted the CSPS value at a given location may not necessarily do so at a different location or even at the same location in a different year. For this reason, the control unit should be able to optimize the map based on deviations between a CSPS value predicted by the map for a given rotational speed and gap width pair and a CSPS value actually measured for that pair. This ensures a satisfactory CSPS prediction even for different locations or harvest years.

[0013] To ensure that only significant measurements are included in the optimization, it can be stipulated that only a CSPS value measured during a quasi-stationary state of the forage harvester is used for optimization. The forage harvester is considered to be in a quasi-stationary state when, over a given time period longer than the dwell time of the crop in the harvester, changes in rotational speed, gap width, and CSPS value are smaller than a predefined relative range of variation. This means that the currently measurable CSPS value can be assumed to be sufficiently accurate to match the value achieved by the chopping drum and cracker rollers of the forage harvester at the same time. The predefined relative range of variation can, for example, be less than or equal to 10% or less than or equal to 5%.

[0014] The characteristic map can be defined by a multi-parameter function of at least the rotational speed and the gap width, as well as a set of values ​​for the parameters of this function. This allows the characteristic map to be updated by changing the parameter values ​​to adapt to altered crop characteristics, based on a small number of value pairs consisting of an expected CSPS value for given rotational speed and gap width values, and an actual CSPS value measured at these values.

[0015] An optimization unit can be set up to optimize the set of parameter values ​​based on a data set of value pairs of rotational speed and gap width and CSPS values ​​measured for these value pairs, with the aim of minimizing a deviation between an actual CSPS value and a CSPS value expected from the characteristic map.

[0016] The data set can consist solely of CSPS values ​​collected during the operation of the forage harvester and the corresponding pairs of values ​​for rotational speed and gap width; preferably, however, it also includes pre-stored data that is available before the start of operation. By incorporating this data into the optimization of the parameter set, the statistical basis of the optimization can be broadened, and sharp fluctuations in parameter values ​​determined during optimization, especially in the initial phase of operation, can be avoided.

[0017] On the other hand, it must be ensured that the pre-stored data does not permanently distort the optimization result if it becomes outdated. For this purpose, it can be provided that the value pairs of the data set define a point grid within a two-dimensional adjustment range defined by the values ​​that the rotational speed and gap width can assume during operation. Furthermore, if a measured CSPS value exists for a value pair of rotational speed (n HT) and gap width (CC) that does not correspond to a point of the grid, the nearest point of the grid to this value pair is determined, and the CSPS value of the data set assigned to this point is overwritten with the measured CSPS value.

[0018] The multi-parameter function is preferably a second-order polynomial with the rotational speed and the slit width as variables. Higher-order polynomials are also possible, but the time required to fit the parameter values ​​increases with the order of the polynomial, since the number of pairs of expected and measured CSPS values ​​needed for fitting the parameter values ​​increases significantly with the order of the polynomial.

[0019] To prevent old value pairs that no longer correspond to the current characteristics of the harvested crop from distorting the optimization result, the weighting with which pairs of measured and expected CSPS values ​​are included in the optimization should decrease with the age of the pairs.

[0020] If the number of available pairs of measured and expected values ​​of the CSPS is smaller than the number of parameters of the multi-parameter function, this does not mean that optimization is impossible, but merely that there can be a large number of sets of parameter values ​​for which the difference between the measured and expected value vanishes. One of these sets can certainly be used for optimization.

[0021] According to a second aspect, the task is solved by a method for operating a forage harvester, in particular a forage harvester as described above, with the following steps: Specifying a multi-parameter function of at least the rotational speed of a chopping drum and the gap width of a cracker gap, as well as an original set of parameter values ​​that define an original characteristic map of the CSPS as a function of at least the rotational speed and the gap width; determining an actual value of the CSPS for actual values ​​of the rotational speed and the gap width; replacing the original set of parameter values ​​with an optimized set, wherein the parameter values ​​of the optimized set are chosen such that the deviation between the actual value of the CSPS and a set expected from the characteristic map is smaller when using the optimized parameter values ​​than when using the original parameter values.

[0022] It goes without saying that the characteristic curve can depend not only on the rotational speed and the gap width, but also on other operating parameters of the forage harvester, such as a speed difference between the rollers of the cracker gap, and that parameters of the multi-parameter function linked to these other operating parameters can also be subject to optimization. However, here too, it is desirable to keep the number of operating parameters manageable in order to obtain an optimization result quickly.

[0023] The invention further relates to a computer program product comprising executable instructions which, when executed by a computer, enable the computer to operate as at least a control unit, optionally also as an optimization unit and / or image processing unit, in a forage harvester as described above or to execute the method described above.

[0024] Further features and advantages of the invention will become apparent from the following description of exemplary embodiments with reference to the accompanying figures. These show: Fig. 1 shows a side view of a forage harvester according to the invention; Fig. 2 shows a section of an image captured by a camera of the forage harvester; Fig. 3 shows an exemplary characteristic map; and Fig. 4 shows a diagram of a data set that forms the basis for the calculation of the characteristic map by an on-board computer of the forage harvester.

[0025] Fig. 1Figure 1 shows a partially transparent side view of a forage harvester 1. The forage harvester has a rigid body 2 to which a header 3, in this case a corn header, is attached and can be adjusted for height. The crop cut and collected by the header 3 is compacted by a pair of intake rollers 4 and fed into a chopping unit, where a rotating chopping drum 5, in conjunction with a stationary cutting edge 6, chops the crop. The length of the resulting chopped particles is proportional to the ratio of the speeds of the intake rollers 4 and the chopping drum 5.

[0026] As the crop passes through the forage harvester 1, a pair of cracker rollers 7 follow the chopping mechanism in a manner known per se. These rollers are driven at different peripheral speeds to crush any grains contained in the crop. The width of the gap between the cracker rollers 7 is adjustable and determines the size of the grain fragments obtained at the exit of the gap.

[0027] A secondary accelerator 8 following the cracker rollers 7 serves to accelerate the chopped material to a speed required to pass through a discharge spout 9.

[0028] A camera 10 is positioned downstream of the cracker rollers 7 along the path of the harvested material to capture images of the chopped material. It can be positioned upstream of the post-accelerator 8, contrary to the figure shown. Positioning it downstream, such as at the discharge spout 9, has the advantage that the particle density in the stream of chopped material is reduced there, making individual particles easier to identify in the images and reducing the likelihood of particles overlapping and therefore making it impossible to reliably assess their size.

[0029] The camera 10 is connected to an image processing unit 12 in an on-board computer 11 of the forage harvester 1. The image processing unit 12 is programmed to identify grains and grain fragments in the images of the camera 10. Fig. 2The result of such an identification is shown: Each pixel of the image is assigned one of two values, depending on whether the pixel shows a grain or grain fragment or not. Fig. 2 Grains and grain fragments are depicted in white against a black background.

[0030] In a subsequent step, the image processing unit 12 determines for each identified grain or grain fragment as described in Fig 2An example illustrates a direction L of its smallest dimension d and, using a known imaging scale of camera 10, the magnitude of this dimension d, in order to determine, based on the result, whether the grain or grain fragment could have passed through a sieve with a mesh size prescribed for the experimental determination of the CSPS. Thus, all grains or grain fragments in an image or a series of successive images are each assigned to one of several size fractions, and the thickness of each fraction is calculated based on the mass of the particles assigned to it. The assignment of a CSPS measurement to the thicknesses thus obtained follows the same rules as if the fractions had been obtained in the conventional manner by physical sieving and then weighed, and therefore need not be explained further here.

[0031] Even though it is not exactly known how the CSPS value achieved during chopping depends on the rotational speed n HT of the chopping drum 5 and the width CC of the gap between the cracker rollers 7, it can be assumed that it can be approximated by a Taylor series in two dimensions, i.e. that CSPS CC n HT = p 0 + p 1 CC + p 2 n HT + p 3 CC 2 + p 4 CC n HT + p 5 n _ HT 2 to a good approximation, where p0, ..., p5 are empirical parameters whose values ​​can change depending on the properties of the crop. Parameter values ​​p0*<, ..., p5* determined by the manufacturer for normal or supposedly normal crops can therefore be stored in the on-board computer 11 in order to find and set suitable values ​​of CC and nHT for a desired CPPS value of the crop at the beginning of a harvesting operation; however, it must be assumed that the CPPS value actually achieved with the values ​​of CC and nHT found in this way will deviate from the desired value and that CC and nHT will have to be adjusted based on the observed deviation in order to achieve the desired value. This adjustment is the task of an optimization unit 13 of the on-board computer 11; a control unit 14 controls the speed of a motor driving the chopping drum and the width of a cracker gap between the cracker rollers 7.In other words, the control unit 14 together with the chopping drum and the cracker gap as crop treatment means and the camera 10 as a sensor forms an automatic setting device within the meaning of EP 3 542 610 B1.

[0032] A procedure for this is described with reference to Fig. 3 described. Fig. 3 This shows an exemplary characteristic map CF for the CSPS, calculated by the on-board computer 11 as a function of the rotational speed n HT and the gap width CC using the formula above, based on the parameter values ​​p 0 *< , ..., p 5 * specified by the manufacturer. For good processing, the CSPS of the harvested crop should reach at least 70%; this value is therefore set as the target value of the CSPS, and for the start of a harvesting operation, a pair of values ​​P 1 = (n HT , CC) of rotational speed and gap width is set, for which CSPS(n HT , CC) = 70%.

[0033] Since the crop processed in the relevant harvesting operation differs from that on which the parameter values ​​p 0 *, ..., p 5 * were determined, it is to be expected that the measured value CSPS(P 1) determined by the image processing unit 12, which is obtained when the forage harvester is operated with the setting P 1, will deviate from the expected 70%, and will be, for example, at 65%. Fig. 3 represented as rhombus C 1 .

[0034] To still achieve the desired CSPS value, the on-board computer 11 reduces the gap width CC. For this purpose, it can calculate the difference between the target value (70%) and the actual value (65%) of the CSPS, add it to the target value to obtain a new target value (75%), and determine a new setting P2 where CSPS(P2) = 75%. If such a setting P2 exists that differs from P1 only in the gap width CC, then it is selected and set; if necessary, nHT is also adjusted. It is expected that the resulting CSPS value C2 will be considerably closer to the target value of 70% than C1.

[0035] Any remaining difference between the target and actual value (measured value) can be further reduced by determining and setting a new P3 value as described above. A resulting CSPS value C3 is then measured; this provides three data points that allow the recalculation of parameters p0, p1, and p3, which describe the dependence of the CSPS value on the gap width CC, for the current crop. By replacing the original values ​​in the formula above with the newly obtained values, an updated characteristic map is produced that more accurately reflects the current relationship between CSPS and CC than the original one.

[0036] If more than three CSPS measurements are available at the same rotational speed n HT and different gap widths CC, it is generally no longer possible to find parameter values ​​p 0 , p 1 , p 3 that exactly reproduce the measured values; in this case, the on-board computer 11 selects as parameter values ​​p 0 , p 1 , p 3 those that reproduce the measured values ​​with the minimum squared deviation. Calculation methods for this, also known as least-squares fit, are well known.

[0037] The fact that the measured values ​​C1, C2, C3 do not allow for an update of the other parameters p2, p4, p5, which describe the influence of the rotational speed nHT, is not a disadvantage as long as the rotational speed nHT does not need to be changed. However, as soon as this happens, measured values ​​from the CSPS will inevitably be obtained, which allow conclusions to be drawn about these parameters.

[0038] In fact, the largely independent nature of the parameters p₂, p₄, p₅ from the measured values ​​C₁, C₂, C₃ recorded at the same rotational speed allows for a simplification of the structure of the control unit 14: Instead of a single control unit for controlling both the rotational speed nₐHT and the gap width CC, two control units can be provided, one for the gap width CC and the other for the rotational speed nₐHT. As long as the rotational speed nₐHT is kept constant, only the control unit for the gap width can optimize its assigned parameters p₀, p₁, p₃, treating the remaining parameters p₂, p₄, p₅ as external information to be considered when determining the gap width.If a different speed value is set later, the speed control unit only needs to adjust parameter p2 to adapt the characteristic map to the CSPS value observed for this changed speed. p0, p1, and p3 represent external information that is taken into account but cannot be changed by the speed control unit. If CSPS measurement results are available for more than two speed values, the speed control unit can also adjust p4 and p5 to the measurement results, but it remains unaffected by parameters p0, p1, and p3. Since the gap width control unit receives the adjusted parameter values ​​p2, p4, and p5 as external information, it can correctly incorporate them into any subsequent adjustment.The weighting with which individual measurements are included in this calculation can be set to decrease with the age of the measurements, so that new measurements, which are more likely to be representative of the current processing characteristics of the harvested crop, influence the result more strongly than older ones.

[0039] Adjusting the characteristic map does not require waiting for three or more measured values ​​as described above. Equivalent to the procedure described above for determining the corrected setting P2, the parameter p0 can be corrected immediately upon receiving the measured value C1 so that the formula correctly predicts the measured value C1. In the case described above, this would mean reducing p0 by 5% and then determining P2 using the corrected formula.

[0040] In the method described above, the calculation of parameters p0, ..., p5 is based entirely on CSPS measurements taken during operation of the forage harvester, as soon as the number of measurements is sufficient to perform the necessary calculations. Therefore, each individual measurement has a significant impact on the optimization result, and even a single measurement error can lead to the calculation of an incorrect parameter set. If this incorrect parameter set is used to control rotational speed and gap width, it will produce a CSPS value that deviates considerably from the expected value.

[0041] A parameter set may be considered faulty, for example, if the CSPS value calculated using the parameter values ​​decreases with increasing gap width CC or increasing rotational speed n HT. To prevent such events, among others, the control unit 14 can be configured to recognize as incorrect an optimization result in which, compared to a previous result, one of the parameters p 0 , ..., p 5 has changed its sign, or in which the change in one of the parameters exceeds a permissible upper limit. When such an incorrect result occurs, the control unit 14 can continue to use the set of parameter values ​​p 0 , ..., p 5 obtained in the previous optimization and discard any measurements acquired since the calculation of this set; another possibility is to completely discard the CSPS measurements obtained during the current operation of the forage harvester and convert them to the parameter values ​​p 0 *< , ... specified by the manufacturer., p 5 * to return and begin the procedure described above again.

[0042] According to a preferred training method, the manufacturer does not specify a set of parameter values ​​p 0 *< , ..., p 5 *, but rather a data set of value pairs of the rotational speed n HT and the gap width CC and CSPS values ​​measured for these value pairs. Fig. 4 An example of such a dataset is shown as a three-dimensional diagram. The adjustment ranges of the gap width CC and the rotational speed n HT are preferably divided into equally sized intervals of m1 and m2, respectively. Fig. 4m1 = m2 = 4, realistic values ​​are up to 20, M1 = 10 or M2 = 10 is preferred. The two-dimensional adjustment range spanned by the gap width CC and the rotational speed nHT is thus divided into n1*n2 sub-intervals. If the gap width CC and the rotational speed nHT are each adjustable in an interval [0,1] (in arbitrary units), then the midpoints of the sub-intervals form a regular grid at coordinate values ​​0.5 / m1, 1.5 / m1, ..., (m1 - 0.5) / m1 for the gap width and 0.5 / m2, 1.5 / m2, ..., (m2 - 0.5) / m2 for the rotational speed. For each of these sub-intervals, a maximum of one initial measurement value recorded by the manufacturer is stored, which is stored in Fig. 4 the height of a cuboid at the location of this subinterval is determined. There can also be subintervals for which no initial measurement value is stored, in Fig. 4 represented by a missing cuboid at the location of the subinterval.

[0043] A set of original parameter values ​​p 0 *, ..., p 5 * derived from this data set can be stored in the control unit 14 or derived from the data set as described above at the beginning of each use of the forage harvester.

[0044] When the forage harvester reaches a steady state during operation, i.e., when changes in gap width CC, rotational speed n HT, and CSPS value remain within an interval of, for example, + / - 5% of their initial value during a given time period, the control unit 14 identifies the sub-interval of the setting range in which the current pair of values ​​for gap width CC and rotational speed n HT lies and overwrites the initial measurement value assigned to this sub-interval using the CSPS value measured in this steady state. In the simplest case, the initial measurement value can simply be overwritten with the new measurement value; however, it is also conceivable to first calculate an expected CSPS value for the current pair of values ​​for gap width CC and rotational speed n HT from the stored initial measurement values, determine the difference between this and the measured CSPS value, and then correct the CSPS value stored for the sub-interval by this difference.Based on the modified stored CSPS values, an updated set of parameters p 0 , ..., p 5 is then calculated using the same procedure as before.

[0045] While in the first described configuration the parameters p0, ..., p5 were completely recalculated based on the CSPS measurements obtained during operation as soon as the number of measurements allowed, in the configuration considered here the initial measurements of a subinterval remain relevant for the parameter calculation as long as no new measurement is available for that subinterval. This reduces the impact that a single erroneous measurement can have on the resulting parameter set. This also reduces the probability that the calculation will produce a clearly incorrect result. If this does occur, the same mitigation measures as described above apply. Reference sign

[0046] 1 Forage harvester 2 Body 3 Header 4 Infeed roller 5 Chopping drum 6 Cutting edge 7 Cracker roller 8 Post-accelerator 9 Discharge spout 10 Camera 11 On-board computer 12 Image processing unit 13 Optimization unit 14 Control unit

Claims

1. Self-propelled forage harvester (1) comprising a cutterhead (5) driven at a variable rotational speed (nHT) for chopping harvested crop, a cracker roller pair (7) arranged behind the cutterhead (5) on a path of the harvested crop through the forage harvester (1) and having a variable gap width (CC) for comminuting grains in the chopped harvested crop, and a control unit (14) for controlling the rotational speed and the gap width; the control unit (14) being part of an automatic setting machine controlled on the basis of a characteristic map (CF), characterized in that the characteristic map (CF) describes a corn silage processing score, CSPS, depending at least on the rotational speed (nHT) and the gap width (CC).

2. Self-propelled forage harvester according to Claim 1, wherein the control unit (14) is configured to set the rotational speed (nHT) and the gap width (CC) appropriately with respect to a target value of the CSPS on the basis of the characteristic map (CF).

3. Self-propelled forage harvester according to Claim 2, wherein the control unit (14) is configured, in the case of a significant deviation between the target and actual values of the CSPS, firstly to change the gap width (CC) in a direction for which a reduction of the deviation is to be expected based on the characteristic map (CF), and to change the rotational speed (nHT) only if the gap width (CC) has reached a limit of its setting range.

4. Self-propelled forage harvester according to Claim 1, 2 or 3, wherein an image processing unit (12) is connected to a camera (10) arranged behind the cracker roller pair (7) on the path of the harvested crop, in order to determine an actual value of the CSPS from images of the harvested crop supplied by the camera (10).

5. Self-propelled forage harvester according to Claim 4, wherein the image processing unit (12) is configured to determine at least one dimension (d) of particles depicted in the images and, on the basis of the dimension, to decide about the association of the particle with a size fraction relevant for the determination of the actual value of the CSPS.

6. Self-propelled forage harvester according to any of the preceding claims, wherein the control unit (14) is configured to optimize the characteristic map (CF) on the basis of deviations between a CSPS value predicted by the characteristic map (CF) for a value pair of rotational speed (nHT) and gap width (CC) and a CSPS value measured for the value pair.

7. Self-propelled forage harvester according to Claim 6, wherein the control unit (14) is configured to use a measured CSPS value for the optimization of the characteristic map only if said value was measured during a quasi-steady state of the forage harvester, a quasi-steady state being characterized in that in a specified period of time which is longer a dwell time of the harvested crop in the forage harvester, changes in the rotational speed (nHT), the gap width (CC) and the CSPS value are smaller than a specified relative fluctuation range.

8. Self-propelled forage harvester according to any of the preceding claims, wherein the characteristic map (CF) is defined by a multi-parameter function of at least the rotational speed (nHT) and the gap width (CC) and also a set of values (p0, ..., p5) of the parameters of this function, and an optimization unit (13) is configured, based on a dataset of value pairs of the rotational speed (nHT) and the gap width (CC) and CSPS values measured for these value pairs, to optimize the set of parameter values (p0, ..., p5) with the aim of minimizing a deviation between an actual value (C1, ...) of the CSPS and a value of the CSPS expected on the basis of the characteristic map.

9. Self-propelled forage harvester according to Claim 8, characterized in that the values that can be assumed by the rotational speed (nHT) and the gap width (CC) during operation define a two-dimensional setting range, in that the value pairs of the dataset define a point grid in the two-dimensional setting range, and in that, if there is a measured CSPS value for a value pair of rotational speed (nHT) and gap width (CC) that does not correspond to any point of the grid, the nearest neighbouring point of the grid with respect to this value pair is determined and the CSPS value of the dataset assigned to this point is overwritten on the basis of the measured CSPS value.

10. Self-propelled forage harvester according to Claim 9, characterized in that the multi-parameter function is a second-order polynomial in rotational speed and gap width.

11. Self-propelled forage harvester according to either of Claims 9 and 10, characterized in that the optimization unit is configured to perform a first optimization before the number of pairs of measured and expected values of the CSPS corresponds to the number of parameters.

12. Method for operating a forage harvester, in particular a forage harvester according to any of Claims 1 to 11, comprising the steps of: - specifying a multi-parameter function (CSPS(nHT, CC)) of at least the rotational speed (nHT) of a cutterhead (5) and the gap width (CC) of a cracker gap and also an original set (p0*, ..., p5*) of parameter values which define an original characteristic map (CF) of the corn silage processing score, CSPS, depending on at least the rotational speed and the gap width; - determining an actual value (C1, ...) of the CSPS for actual values (P1, ...) of the rotational speed and the gap width; - replacing the original set of parameter values (p0*, ..., p5*) with an optimized set (p0, ..., p5), the parameter values of the optimized set being chosen such that the deviation between the actual value (C1, ...) of the CSPS and a value expected on the basis of the characteristic map is smaller when using the optimized parameter values than when using the original parameter values.

13. Computer program product comprising executable instructions which, when they are executed by a computer, enable the computer to operate at least as a control unit, optionally also as an optimization unit and / or image processing unit, in a forage harvester according to any of Claims 1 to 11, or to execute the method according to Claim 12.