Method, control device and flap arrangement for distinguishing actuations of a triggering device for a flap arrangement of a vehicle, as well as computer program and computer-readable medium

Abstract

The present invention relates to a method, a control device (160) and a flap arrangement (100) for distinguishing actuations of a triggering device (140) for a flap arrangement (100) of a vehicle, as well as a computer program and a computer-readable medium.The method according to the invention comprises receiving first training data sets for intended actuations of the triggering device (140), receiving second training data sets for unintentional actuations of the triggering device (140), determining at least one distinguishing criterion by means of machine learning for distinguishing between intended and unintentional actuations of the triggering device (140) at least partially based on the received first training data sets and at least partially based on the received second training data sets, determining at least one target actuation parameter by means of machine learning at least partially based on the at least one determined distinguishing criterion and providing the at least one determined target actuation parameter.The at least one determined target actuation parameter is designed to indicate an intended actuation of the triggering device (140) and thus to distinguish an intended actuation of the triggering device (140) from an unintentional actuation of the triggering device (140).

Description

[0001] The present invention relates to a method, a control device and a flap arrangement for distinguishing an intended actuation from an unintentional actuation of a triggering device for a flap arrangement of a vehicle as well as a computer program and a computer-readable medium and in particular to a method for distinguishing an intended actuation from an unintentional actuation of the triggering device for the flap arrangement of the vehicle by means of machine learning.

[0002] Sensors are used in numerous devices, such as smartphones, various wearables (e.g., devices worn on the body and / or head), smart headphones, etc. Intelligent sensors (or smart sensors), for example, can be used here, which can preprocess detected sensor signals. Preprocessing of detected sensor signals can be performed, for example, using trained machine learning classifiers, such as neural networks. The machine learning classifiers can be trained using training data covering a broad user spectrum. For example, a machine learning classifier generalized in this way can recognize movement patterns.

[0003] Sensors are also used in vehicles. For example, a variety of sensors, such as triggering devices for vehicle hatches, are known from the prior art. Such triggering devices are conventionally mounted on a movable hatch of the vehicle to allow a user to open the movable hatch. The opening is triggered by an electrical and / or mechanical switch, which emits an electrical switching signal depending on the user's actuation. The known triggering devices are sometimes arranged in a recess in the outer skin of the hatch.

[0004] It is an object of the present invention to at least partially remedy the above-mentioned disadvantages known from the prior art. In particular, it is an object of the present invention to provide a method for efficiently distinguishing between intentional and unintentional actuations of a triggering device for a flap assembly of a vehicle.

[0005] This object is achieved with a method according to claim 1, a control device according to claim 7, a triggering unit according to claim 9, a flap arrangement according to claim 10, a computer program according to claim 11, and a computer-readable medium according to claim 12. Advantageous embodiments are specified in the subclaims.

[0006] The present invention is essentially based on the idea of ​​training a trigger unit for a vehicle's flap assembly using machine learning with training data sets for intentional and unintentional actuations of the triggering device. This allows the machine-learned training data sets to be used to reliably assess how an intentional actuation of the trigger unit differs from an unintentional actuation of the trigger unit. With this knowledge, the operator can then open the flap assembly in the vehicle efficiently and only when there is an intentional desire to open the flap assembly.

[0007] Accordingly, according to a first aspect of the present invention, a method is disclosed for distinguishing an intentional actuation from an unintentional actuation of a triggering device for a flap assembly of a vehicle, which is configured to be openably mounted on a vehicle body. The triggering device is configured to trigger an opening operation of the flap assembly.The method according to the invention comprises receiving first training data sets for intended actuations of the triggering device, receiving second training data sets for unintentional actuations of the triggering device, determining at least one distinguishing criterion by means of machine learning for distinguishing between intended and unintentional actuations of the triggering device at least partially based on the received first training data sets and at least partially based on the received second training data sets, and determining at least one target actuation parameter by means of machine learning at least partially based on the at least one determined distinguishing criterion.The at least one determined target actuation parameter is designed to indicate an intended actuation of the triggering device and thus to distinguish an intended actuation of the triggering device from an unintentional actuation of the triggering device. The method according to the invention further comprises providing the at least one determined target actuation parameter.

[0008] According to the invention, the difference between an intended and an unintentional actuation of the triggering device is learned using machine learning based on the received first and second training data sets during the step of determining at least one distinguishing criterion. Based on this, the at least one target actuation parameter can then be determined in order to quantitatively distinguish an intended from an unintentional actuation of the triggering device.

[0009] In a preferred embodiment of the method according to the invention, the triggering device has a sensor unit with a triggering element and an electrical detection unit designed to detect a movement of the triggering element. The triggering device is designed to generate an actuation signal when the electrical detection unit has detected a movement of the triggering element. It is further preferred that the at least one determined distinguishing criterion comprises an amplitude of the movement of the triggering element and / or a value range of the signal frequency components and / or an acceleration of the triggering element, which can describe the time derivative of the amplitude, and / or a skewness of the actuation signal. The skewness, also called skewness, of the signal is the dimensionless measure of the ratio of the average deviation from the mean divided by the cubic standard deviation.

[0010] In a further advantageous embodiment of the method according to the invention, the at least one determined target actuation parameter defines a value range for the at least one distinguishing criterion, wherein the value range is representative of an intended actuation. The target actuation parameter specifies, in particular, a quantitative distinguishing value by means of which an intended actuation can be distinguished from an unintentional actuation of the triggering device.

[0011] In a further preferred embodiment, the method according to the invention further comprises determining a prioritization of the at least one determined distinguishing criterion, preferably by means of machine learning. The determined prioritization of the at least one determined distinguishing criterion preferably has the order "amplitude of the movement of the trigger element" or "frequency amplitude of the amplitude signal", "acceleration of the trigger element", and "skewness of the actuation signal". This means that the amplitude of the movement of the trigger element is evaluated first, before the acceleration of the trigger element and the skewness of the actuation signal are then evaluated in order to distinguish an intended from an unintentional actuation of the triggering device with the greatest possible probability and preferably with minimal computational effort.

[0012] According to a further aspect of the present invention, a method for determining an intended actuation of a triggering device for a flap assembly of a vehicle is disclosed, which is configured to be openably mounted on a body of the vehicle. The triggering device is configured to generate an actuation signal indicating actuation thereof and to indicate initiation of an opening operation of the flap assembly.The method according to the invention comprises receiving an actuation signal from the triggering device, determining actual actuation parameters of the triggering device based on the received actuation signal, determining an intended actuation of the triggering device if the received actuation parameters substantially correspond to the target actuation parameters provided according to an inventive method of the present invention, and sending a triggering signal which triggers an opening process of the flap arrangement if an intended actuation of the triggering device has been determined.

[0013] The method according to the further aspect thus forms the determination method during the operation of the vehicle in order to then decide, on the basis of the operator operations in comparison with the target operation parameters previously determined by means of machine learning, whether the actuation of the triggering device is intentional or unintentional.

[0014] According to yet another aspect of the present invention, a control device is disclosed which is designed to carry out the steps of the method according to the invention.

[0015] According to an advantageous embodiment, the control device according to the invention has a first control device section for carrying out the step of receiving first training data sets for intended actuations of the triggering device, a second control device section for carrying out the step of receiving second training data sets for unintentional actuations of the triggering device, a third control device section for carrying out the step of determining at least one differentiation criterion by means of machine learning, a fourth control device section for carrying out the step of determining at least one target actuation parameter by means of machine learning and a fifth control device section for carrying out the step of providing the at least one determined target actuation parameter.

[0016] According to yet another aspect of the present invention, a trigger unit for triggering an opening process of a flap assembly for a vehicle is disclosed. The trigger unit according to the invention comprises a trigger device having a sensor unit with a trigger element and an electrical detection unit configured to detect movement of the trigger element, and a control device according to the invention. The trigger device is configured to generate an actuation signal when the sensor device has detected movement of the trigger element.

[0017] According to yet another aspect of the present invention, a flap assembly for a vehicle is disclosed, which has a flap openably mounted on a body of the vehicle and a triggering unit according to the invention.

[0018] According to yet another aspect of the present invention, a computer program is disclosed comprising instructions which, when executed by a computing unit, cause the computing unit to execute a method according to the invention for distinguishing an intended actuation from an unintentional actuation.

[0019] According to yet another aspect of the present invention, a computer-readable medium is disclosed on which the computer program according to the invention is stored.

[0020] Further advantages and features of the present invention will become apparent to those skilled in the art by practicing the teachings described herein and viewing the accompanying drawings in which: Fig. 1 shows a schematic representation of a vehicle with an exemplary triggering unit of a flap arrangement according to the invention, Fig. 2 a schematic representation of an exemplary triggering device of the triggering unit according to the invention of Fig. 1 shows, and Fig. 3 an exemplary flow diagram of a method according to the invention for distinguishing an intended actuation from an unintentional actuation of the triggering device of the triggering unit of the Fig. 1 shows.

[0021] In the context of this disclosure, the term "machine learning" describes the evaluation and investigation of statistical algorithms. These algorithms can learn a specific behavior from data that is available in machine-readable form and contains information about observations or experiences. The behavior is not explicitly programmed; rather, the algorithms learn it directly from the data. Learning from data can also be referred to as "statistical learning." Generally speaking, the goal of "machine learning" is for an algorithm to learn a function from data that subsequently produces a correct output even for unlearned data inputs. So that the algorithm can learn what is "correct," correct output values ​​are already provided in the data used for learning in supervised learning.

[0022] In the context of the present disclosure, the term "training data sets" describes those machine-readable data or data sets that are provided for machine learning. According to the present disclosure, first training data sets are provided that were generated by intentional actuations of the triggering device, and second training data sets are provided that were generated by unintentional actuations of the triggering device. Unintentional actuations can be generated, for example, by exposing the triggering device to the environment, for example, by rain, hail, wind, and / or foreign objects acting on the triggering device.

[0023] The Fig. 1 shows a schematic representation of a vehicle 10 with a flap arrangement 100 according to the invention, which is movably attached to a body 12 of the vehicle 10. The vehicle 10 is in the Fig. 1 is shown as an example of a road vehicle. In other embodiments, the vehicle 10 can be any vehicle, such as an agricultural vehicle, a rail vehicle, a watercraft such as a ship, or an aircraft such as an airplane. Likewise, the flap assembly 100 according to the invention can be any flap assembly attached to the vehicle 10 in an openable and closable manner, such as a front flap, a tailgate, a side door, a charging plug flap, a fuel filler flap, or any other flap that can allow or block access to a vehicle function.

[0024] In the Fig. 1 already indicates that the flap assembly 100 has a triggering unit 130 for triggering an opening process of the flap assembly 100. The triggering unit 130 has a triggering device 140 and a control device 160. The user of the vehicle 10 can express his or her desire to open the flap assembly 100 by actuating the triggering device 140. The triggering device 140 is designed to generate a triggering signal upon actuation and to send it to the control device 160, which in turn is designed to process and analyze the triggering signal. Upon positive detection of an intended actuation of the triggering device 140, an actuator unit (not explicitly shown) designed to electrically open the flap assembly 100 can be controlled.The opening movement of the flap arrangement 100 can be at least partially rotational and / or at least partially translational and / or at least partially pivoting relative to the body 12 of the vehicle 10.

[0025] The trigger unit 130 according to the invention is, as already mentioned, formed from the trigger device 140 and the control device 160. The control device 160 can have several control device sections, such as a first control device section 162, a second control device section 164, a third control device section 166, a fourth control device section 168 and a fifth control device section 169, which will be described with reference to the Fig. 3 will be discussed in more detail below.

[0026] The control device 160 may include a processor and a memory. Alternatively, the control device 160 may be the processor coupled to the memory. The processor may be a central processing unit (CPU). The processor may further be another general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, or the like. The general-purpose processor may be a microprocessor, or the processor may be any conventional processor or the like.

[0027] Memory includes, but is not limited to, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), or portable read-only memory (e.g., CD-ROM). The memory is configured to store associated program instructions and associated data.

[0028] The control device may be a computer or comprise a computer. In the context of the present disclosure, a computer may be understood as any type of logic-implementing entity, which may be hardware, software, firmware, or a combination thereof. Therefore, a computer may be a hard-wired logic circuit or a programmable logic circuit, such as a programmable processor, for example a microprocessor (e.g., a CISC (large instruction set processor) or a RISC (reduced instruction set processor)). A computer may also be software implemented or executed by a processor, for example, any type of computer program, for example, a computer program using virtual machine code, such as Java.Any other manner of implementing the respective functions described in more detail below may be understood as a computer in accordance with an alternative embodiment.

[0029] The Fig. 2 shows a schematic sectional view through a triggering device 140 of a triggering unit 130 according to the invention. Fig. The triggering device 140 shown as an example in Figure 2 comprises a two-part housing unit 142 with a first housing element 141, which is designed to be attached to the vehicle interior skin, and a second housing element 143, which is arranged to be at least partially movable relative to the first housing element 141 from an initial position into a triggering position.

[0030] The triggering device 140 further comprises a sensor unit 144 configured to detect a desire to trigger an opening operation of the flap assembly 100. The triggering device 140 also includes an elastic return device 149 that movably mounts the second housing element 143 relative to the first housing element 141 such that the return device 149 biases the second housing element 143 into an unactuated state of the triggering device 149.

[0031] The sensor unit 144 may include an electrical detection unit 145 that is attached to the first housing element 141 and configured to detect a movement of a trigger element 146 relative to the electrical detection unit 145. The trigger element 146 may, for example, comprise an electrically conductive material and be attached to the second housing element 143.

[0032] In particular, the sensor unit 144 can be based on the inductive or capacitive principle, in which a change in distance between the detection unit 145 and the movably arranged and actuable trigger element 146 leads to a change in capacitance or induction, thus allowing the desire to open the flap arrangement to be determined. Furthermore, it is possible for the sensor unit 144 to be based on a resistive or optical measuring principle. In addition, the sensor unit 144 can have a piezoelectric measuring element designed to determine the force applied by the user of the vehicle 10.

[0033] The trigger element 146 can be mounted on the first housing element 141 by means of a mounting device 148. Alternatively, the trigger element 146 can be rigidly attached to the second housing element 143, for example, to a projection 147 of the second housing element 141, which is designed to move the trigger element 146 toward the detection unit 145 upon actuation of the trigger device 140.

[0034] In the Fig. 3 is an exemplary flowchart of a method according to the invention for distinguishing an intended actuation from an unintentional actuation of the triggering device 140 of the triggering unit 130 of the Fig. 1 shown.

[0035] The procedure of Fig. 3 starts at step 200 and then proceeds to step 210, at which the control device 160, in particular the first control device section 162, receives first training data sets for intended actuations of the triggering device 130. The first training data sets comprise, in particular, recorded data sets for intended actuations of the triggering device 130. For example, the triggering device 130, in particular the second housing element 143 (see Fig. 2) be intentionally actuated to open the flap assembly 100.

[0036] In a subsequent step 220, the control device 160, in particular the second control device section 164, receives second training data sets for unintentional actuations of the triggering device 130. The second training data sets, in particular, comprise recorded data sets for unintentional, such as unwanted, actuations of the triggering device 130. For example, the triggering device 130 can be actuated with raindrops, hailstones, water jets from a vehicle cleaning system, knocking, splashing water, brushes from a car wash, brooms when sweeping the vehicle, and other actions.

[0037] It goes without saying that the first control device section 162 and the second control device section 164 may also be integrally formed as one control device section that receives both the first training data sets and the second training data sets.

[0038] In a subsequent step 230, the control device 160, in particular the third control device section 166, determines at least one distinguishing criterion by means of machine learning to distinguish between intentional and unintentional actuations of the triggering device 130, based at least partially on the first training data sets received in step 210 and at least partially on the second training data sets received in step 220. In step 230, the received first and second training data sets are evaluated, in particular by means of machine learning, to the effect that they are analyzed, in particular, with regard to distinguishing criteria. The distinguishing criteria can be, for example, an amplitude of the movement of the triggering element 146.In the exemplary embodiment shown, the amplitude describes the movement of the trigger element 146 relative to the electrical detection unit 145 and can be specified in millimeters. Preferably, the determined amplitude data are converted over time into the frequency domain. The transformation from the time domain to the frequency domain can be performed, for example, using a fast Fourier transformation.

[0039] Preferably, the actual mechanical amplitude of the movement of the trigger element 146 is recorded over time, thus generating an electrical signal (unit [bit]), the amplitude of which can also be evaluated. Consequently, the amplitude mentioned in the present disclosure can also be the amplitude of the signal. The speed of the trigger element 146 can be determined by time-deriving the amplitude.

[0040] Further examples of the determined distinguishing criterion can be an acceleration of the trigger element 146. The acceleration of the trigger element 146 is proportional to the pressing speed of the second housing element 143 relative to the first housing element 141 and can describe the time derivative of the speed of the trigger element 146. Additionally or alternatively, the determined distinguishing criterion can be the skewness of the actuation signal sent and generated by the sensor unit 144. The skewness, also called skewness, of the signal is the dimensionless measure of the ratio of the average deviation from the mean divided by the cubic standard deviation.

[0041] In a subsequent step 240, the control device 160, in particular the fourth control device section 168, determines at least one target actuation parameter by means of machine learning, at least partially based on the at least one determined distinguishing criterion. The at least one determined target actuation parameter is designed to indicate an intended actuation of the triggering device 140 and thus to distinguish an intended actuation of the triggering device 140 from an unintentional actuation of the triggering device 140. In particular, after determining the at least one distinguishing criterion in step 230, this distinguishing criterion can subsequently be quantified in the form of at least one target actuation parameter such that an intended actuation can be distinguished from an unintentional actuation of the triggering device 140.The at least one target actuation parameter can preferably define a value range for the at least one distinguishing criterion that is representative of an intended actuation. For example, the at least one determined target actuation parameter can indicate a minimum amplitude of the movement of the trigger element 146 and / or a minimum acceleration of the trigger element 146. If the amplitude and / or the acceleration of the trigger element 146 is less than a predetermined amplitude threshold and / or a predetermined acceleration threshold, an unintentional actuation of the triggering device 146 can be diagnosed, so that the flap assembly 100 does not open.

[0042] In a subsequent step 250, the control device 160, in particular the fifth control device section 169, can provide the target actuation parameters determined in step 250. In one exemplary embodiment, these target actuation parameters can be adjusted depending on aging effects of the triggering device 140, such as the mechanics of the triggering device 140. For this purpose, predetermined criteria for the adjustment can be predefined, such as the signal dynamics or the maximum achievable amplitude.

[0043] The Fig.Steps 210 to 250 shown in Figure 2 can represent a learning process. In particular, steps 210 to 250 can serve to learn the discrimination parameters for assessing the actuation signal using a machine learning algorithm on a computer and can therefore also be executed separately from steps 260 to 290 described below. The learning process according to steps 210 to 250 also includes extracting the determined evaluation criteria and creating a machine-readable computer program for execution on a computer or microcontroller.

[0044] Thus, it may be preferred that the learning process carried out in steps 210 to 250 is carried out on a computer or control device provided separately from the vehicle 10 and, as described, is then provided as a computer program to the control device 160.

[0045] In a further exemplary embodiment, it may be preferable to determine the target actuation parameters 250 determined in step 250 for different environmental conditions. For example, the ambient temperature and / or the influence of wind can be taken into account for this purpose.

[0046] After performing step 250, the triggering unit 130, in particular the control device 160, is equipped with the determined and provided target actuation parameters, so that the vehicle 10 can now be used. As already mentioned, steps 260 to 290 can be performed in a separate process, wherein the learning process according to steps 210 to 250 has already been performed and provided previously.

[0047] In a subsequent step 260, after actuation of the trigger unit 130, the control device 160 can receive an actuation signal.

[0048] In a subsequent step 270, the control device 160 can determine actual actuation parameters from the actuation signal received in step 260. For example, the actual amplitude and / or actual frequency components in the actuation signal and / or actual acceleration of the trigger element 146 and / or the actual skewness of the actuation signal can be determined.

[0049] In a subsequent step 280, a comparison can be made between the actual actuation parameters determined in step 270 and the at least one target actuation parameter provided in step 250. If it is determined in step 280 that the actual actuation parameters determined in step 270 are equal to the target actuation parameters provided in step 250, the method proceeds to step 282, where an intended actuation of the triggering device 140 is determined. Consequently, a trigger signal is sent to an actuator unit configured to open the flap assembly 100 before the method is terminated in step 290.

[0050] Alternatively, in step 280, it may be preferable to check whether the determined actual actuation parameter falls below or exceeds the provided target actuation parameter. Depending on whether the actual actuation parameter falls below or exceeds the provided target actuation parameter, a corresponding subsequent step 282 or 284 may be performed. For example, it may be checked whether the actual amplitude of the movement of the trigger element exceeds a predetermined target amplitude threshold. If not, it can be assumed that the actuation occurred unintentionally.

[0051] However, if it is determined in step 280 that the actual actuation parameters determined in step 270 are not equal to the target actuation parameters provided in step 250 or are not within a determined target actuation parameter range, the method proceeds to step 284, where it is determined that there has been an unintentional actuation of the triggering device 140. As a result, no trigger signal is sent, so that the flap assembly 100 does not open before the method is again terminated in step 290.

[0052] According to the invention, a triggering unit 130 can thus be conditioned in advance using suitable first and second training data sets, which are generated empirically, for example, to provide target actuation parameters learned by machine learning, which are then compared with actual actuation parameters generated in the field by the vehicle operator. This can prevent, for example, the flap from opening when the vehicle is driven through a car wash, where water jets may impinge on the triggering unit 130. At the same time, it can be prevented that the flap opens when rain or hail impacts the triggering device 130 from the outside.

Claims

[1] Method for distinguishing an intentional actuation from an unintentional actuation of a triggering device (140) for a flap arrangement (100) of a vehicle, which is designed to be openably attached to a body (12) of a vehicle (10), wherein the triggering device (140) is designed to trigger an opening operation of the flap arrangement (100), the method comprising: - receiving first training data sets for intended actuations of the triggering device (140), - receiving second training data sets for unintentional actuations of the triggering device (140), - Determining at least one distinguishing criterion by means of machine learning for distinguishing between intentional and unintentional actuations of the triggering device (140) at least partially based on the received first training data sets and at least partially based on the received second training data sets, - Determining at least one target actuation parameter by means of machine learning based at least partially on the at least one determined distinguishing criterion, wherein the at least one determined target actuation parameter is designed to indicate an intended actuation of the triggering device (140) and thus to distinguish an intended actuation of the triggering device (140) from an unintentional actuation of the triggering device (140), and - Providing at least one determined target actuation parameter. [2] Method according to claim 1, wherein the triggering device (140) comprises a sensor unit (144) with a triggering element (146) and an electrical detection unit (145) which is designed to detect a movement of the triggering element (146), wherein the triggering device (140) is designed to generate an actuation signal when the electrical detection unit (145) has detected a movement of the triggering element (146), wherein the at least one determined distinguishing criterion comprises an amplitude of the movement of the triggering element (146) and / or a value range of the signal frequency components and / or an acceleration of the triggering element (146) and / or a skewness of the actuation signal. [3] Method according to one of the preceding claims, wherein the at least one determined target actuation parameter defines a value range for the at least one distinguishing criterion, the value range being representative of an intended actuation. [4] Method according to one of the preceding claims, further comprising: - Determining a prioritization of at least one identified distinguishing criterion. [5] Method according to claim 4, wherein the determined prioritization of the at least one determined differentiation criterion comprises the order “amplitude of the movement of the trigger element (112)”, “acceleration of the trigger element (112)” and “skewness of the actuation signal”. [6] Method for determining an intended actuation of a triggering device (130) for a flap arrangement (100) of a vehicle, which is designed to be openably attached to a body (12) of the vehicle (10), wherein the triggering device (130) is designed to generate an actuation signal indicating an actuation thereof and to indicate a triggering of an opening operation of the flap arrangement (100), the method comprising: - receiving an actuation signal from the triggering device (130), - determining actual actuation parameters of the triggering device (130) based on the received actuation signal, - determining an intended actuation of the triggering device (130) if the received actuation parameters substantially correspond to the target actuation parameters provided according to a method of the preceding claims, and - Sending a trigger signal which triggers an opening operation of the flap arrangement (100) when an intended actuation of the triggering device has been detected. [7] Control device (160) adapted to carry out the steps of the method according to any one of the preceding claims. [8] Control device (160) according to claim 7, comprising: - a first control device section (162) for performing the step of receiving first training data sets for intended actuations of the triggering device (130), - a second control device section (164) for performing the step of receiving second training data sets for unintentional actuations of the triggering device (130), - a third control device section (166) for carrying out the step of determining at least one distinguishing criterion by means of machine learning, - a fourth control device section (168) for carrying out the step of determining at least one target actuation parameter by means of machine learning, and - a fifth control device section (169) for carrying out the step of providing the at least one determined target actuation parameter. [9] Triggering unit (130) for triggering an opening operation of a flap arrangement (100) for a vehicle, the triggering unit (130) comprising: - a triggering device (140) comprising a sensor unit (144) with a triggering element (146) and an electrical detection unit (145) designed to detect a movement of the triggering element (146), wherein the triggering device (140) is designed to generate an actuation signal when the electrical detection device (145) has detected a movement of the triggering element (146), and - a control device (160) according to one of claims 7 and 8. [10] Flap arrangement (100) for a vehicle, comprising: - a flap attached to a body (12) of the vehicle (10) in an openable manner, and - a trigger unit according to claim 9. [11] A computer program comprising instructions which, when executed by a computing unit, cause the computing unit to carry out a method according to any one of claims 1 to 6. [12] A computer-readable medium on which the computer program according to claim 11 is stored.

Images