Method and device for determining a contacted sub-region of a flat component of a motor vehicle

Abstract

A method for determining a contacted sub-region among a plurality of defined sub-regions (11-19) of a flat component (1) of a motor vehicle (100), said method being carried out using at least one ultrasonic sensor (2) which is concealedly attached to the flat component (1), has the steps of: using (S10) the ultrasonic sensor (2) to capture a measurement signal curve (6) which indicates an ultrasonic signal received by the ultrasonic sensor (2) in response to the emission of an ultrasonic signal; and classifying (S20) the measurement signal curve (6) or a signal curve derived from the measurement signal curve in order to determine whether the flat component (1) is being contacted and, if so, at which sub-region of the plurality of defined sub-regions (11-19) the flat component is being contacted.

Description

[0001] METHOD AND DEVICE FOR DETERMINING A CONTACTED PART OF A SURFACE COMPONENT OF A MOTOR VEHICLE

[0002] The present invention relates to the field of motor vehicle sensor technology and more specifically to a method and a device for determining a touched partial area among several predefined partial areas of a planar component of a motor vehicle.

[0003] For numerous applications, it is desirable to provide the on-board electronics of a vehicle with information about when and where the vehicle's exterior is being touched. Examples of such applications include vandalism prevention, but also convenience features like a door opening assistant that opens or closes a vehicle door upon touch, and keyless entry systems.

[0004] US Patent 6,191,682 B1 discloses a keyless entry system with a sensor for unlocking the door locks of a vehicle. The sensor, preferably in the form of a piezoelectric crystal, is attached at a discrete location on a hidden surface of the vehicle's body and electrically connected to an electronic signal processing module that includes an output relay. The output relay is electrically connected to the power control relay of the door lock, the sensor being designed to allow access to a locked vehicle by forcefully tapping on the outer surface of the body at the precise location and within the limited area where the hidden sensor is discreetly attached.

[0005] US patent 2006 / 0232378 A1 discloses a piezoelectric sensor that is activated by touching a door handle and triggers a signal to open a lock of a door locking device.

[0006] Modern motor vehicles are equipped with ultrasonic sensors that use them to determine the distance to obstacles in the vehicle's vicinity by emitting an ultrasonic signal into the surroundings and receiving a reflected signal. The sensors then analyze signal peaks in the reflected signal to determine the distance. By triangulating multiple measurements, the position of the obstacles can also be determined. This information about obstacles can be provided to a driving or parking assistance system, which can be active during slow-speed driving (for example, at speeds between 0 and 50 km / h).

[0007] Against this background, the present invention aims to improve the detection of contact with a flat component of a parked motor vehicle.

[0008] The problem is solved, according to a first aspect, by a method for determining a contacted sub-area among several predefined sub-areas of a planar component of a motor vehicle, which is carried out using at least one ultrasonic sensor concealed on the planar component. The method comprises: using the ultrasonic sensor to detect a measurement signal waveform that indicates an ultrasonic signal received by the ultrasonic sensor in response to the emission of an ultrasonic signal; and classifying the measurement signal waveform, or a signal waveform derived from the measurement signal waveform, to determine whether and at which of the several predefined sub-areas the planar component is in contact.

[0009] A fundamental concept of the present invention is that the ultrasonic signal received by the ultrasonic sensor contains not only reflections from the vehicle's surroundings but also a structure-borne sound component resulting from a portion of the emitted ultrasonic signal propagating and reflected within the flat component. The inventors have recognized that changes in the structure-borne sound component of the received ultrasonic signal can be used to determine whether and where the flat component on which the ultrasonic sensor is mounted is being touched. Specifically, depending on the position (in which of the sub-areas) of the flat component being touched, different measurement signal profiles emerge. These profiles can be distinguished from one another based on characteristic features and thus classified into several classes, each corresponding to one of the predefined sub-areas.For example, depending on the specific sub-area affected, different peak values ​​or peak positions may occur, or different phase shifts may result. Therefore, an algorithmic classification of the measurement signal waveform or a signal waveform derived from it is conceivable. However, it is also possible to use artificial intelligence, such as a trained neural network, for the classification.

[0010] Accordingly, the proposed method advantageously allows for the detection of contact with a flat component of the vehicle and its position without requiring an additional sensor to be installed in the area of ​​the contact point on the flat component. This advantageously enables the detection of contacts even in positions where, for structural or aesthetic reasons (e.g., structural reinforcement struts, decorative panels, and the like running along the inside of the flat component), a separate sensor cannot be installed. Furthermore, the material costs, particularly the number of necessary sensors and associated wiring, are significantly reduced, since the ultrasonic sensors already provided as part of the driving or parking assistance system for obstacle detection can also be used for contact detection.In principle, it is sufficient if only a single ultrasonic sensor is present per flat component to detect touches at any position of the flat component; however, a higher number of ultrasonic sensors per component can optionally be used, which can further improve the precision of the classification.

[0011] Accordingly, precise touch detection can advantageously be achieved with the fewest possible components at any location on a flat component of a motor vehicle. The motor vehicle can be, for example, a passenger car or a commercial vehicle such as a truck, a bus, or a tractor.

[0012] The planar component of the motor vehicle can in particular be a component of the outer shell, such as a sheet metal component, especially an outer sheet, a door sheet and the like.

[0013] However, it is also conceivable that the flat component is made of a different material, such as glass, for example, a display glass of a touch panel or a pane of glass. In the latter case, the ultrasonic sensor would then be concealed in an area of ​​the glass component (covered by a casing).

[0014] The predefined sub-areas of the planar component can be specified in advance for the specific planar component as needed. For example, a first predefined sub-area could be an area of ​​a door handle, and a second predefined sub-area could be an area outside of a door handle. The predefined sub-areas can form an irregular or regular grid that completely covers the planar component, or they can cover only selected areas of interest. The person skilled in the art will thus be guided by the functionality to be provided in response to contact with the planar component, on the one hand, and by the granularity of classification achievable by the proposed method, on the other, in order to determine a suitable number of predefined sub-areas of appropriate size, depending on the application.

[0015] The ultrasonic sensor can, for example, comprise an ultrasonic membrane, a piezoelectric element attached to the inside of the ultrasonic membrane, and a control circuit for the piezoelectric element. Controlled by a control device of the vehicle, the control circuit can set the piezoelectric element, and via the piezoelectric element the ultrasonic membrane, into vibration, thereby emitting ultrasonic signals into the surroundings of the vehicle, as well as into the flat component.When a reflected signal component of such an emitted ultrasound signal from the environment (ambient sound) and / or from the planar component (structure-borne sound) reaches the ultrasound membrane again, this reflected signal component excites the ultrasound membrane and, via the ultrasound membrane, the piezoelectric element to vibrations, which are converted by the piezoelectric element into an electrical signal, so that the control device can detect a measurement signal profile using the control circuit.

[0016] Concealed mounting on the flat component means, in particular, that the flat component does not have an opening for the ultrasonic membrane of the ultrasonic sensor, but rather that the ultrasonic membrane of the ultrasonic sensor rests against or is pressed against an inner surface of the flat component. It is also conceivable that the piezoelectric element is mounted directly on the inner surface of the flat component and that a section of the flat component itself is used as the ultrasonic membrane.

[0017] With such a concealed mounting, ultrasound is inevitably emitted not only into the vehicle's surroundings but also into the flat component itself, and reflected structure-borne sound from the component is received by the ultrasonic sensor. This means that, due to the concealed mounting of the ultrasonic sensor, the received ultrasonic signal, and therefore the measurement signal, can also contain a structure-borne sound component. Traditionally, attempts have been made to dampen and suppress this component using a butyl material; however, there is a growing trend to reduce the damping and to computationally filter out the structure-borne sound component during obstacle detection in slow-moving vehicles.This advantageously makes the structure-borne sound component accessible for the proposed method when the vehicle is stationary and the proposed functionality for detecting contact with the planar component is to be implemented. A "signal waveform," such as the measurement signal waveform or the reference signal waveform discussed later, is understood to be a processable and analyzable representation of the corresponding signal over a predefined time period. For example, the respective signal waveform could be a sampling of the signal at a suitable sampling frequency, perhaps in the range of a few megahertz.

[0018] The signal waveform derived from the measurement signal waveform may be a signal waveform obtained through signal processing of the measurement signal waveform.

[0019] To reduce the processing load, the signal waveform derived from the measurement signal waveform can, for example, be a downsampled version of the measurement signal waveform, based on a resolution of approximately 50 to 200, preferably approximately 100 ps per sample. It is also conceivable to downmix the respective signal, for example, in an IQ mixer implemented as a DSP; in this case, the signal waveform derived from the measurement signal waveform can, for example, include the downmixed i-signal waveform and the downmixed q-signal waveform within the specified time interval.

[0020] The respective signal waveform can exist as a signal, for example as an analog or digital signal, i.e. as a signal in transit, or the respective signal waveform can exist as static data (a sequence of samples stored in working memory or the like).

[0021] According to one embodiment, the flat component is an outer sheet metal panel of the motor vehicle, preferably a door panel of the motor vehicle.

[0022] Accordingly, the information about the touched portion of the outer sheet metal can be used, for example, for vandalism prevention, door opening assistance, or keyless entry. According to one embodiment, the received ultrasonic signal includes a structure-borne sound component.

[0023] A structure-borne sound component is, in particular, a signal component resulting from ultrasound that propagates within the planar component, is reflected within the planar component, and, after the ultrasound signal is emitted, travels back to the ultrasound sensor and is received there.

[0024] This means that when the ultrasonic sensor is concealed on the flat component, a butyl damping material or similar can either be omitted or designed in such a way that structure-borne sound components reflected from the flat component can still penetrate to the ultrasonic membrane of the ultrasonic sensor. For example, a damping pad with a diameter of only about 10-15 cm can be used.

[0025] Accordingly, the structure-borne sound component in the received ultrasonic signal is advantageously available for detecting contact with the planar component according to the proposed method. If, on the other hand, the same ultrasonic sensor is to be used for obstacle detection in the vicinity of the vehicle, the structure-borne sound component can be computationally filtered out.

[0026] According to a further embodiment, the measurement signal waveform and the signal waveform derived from the measurement signal waveform include at least amplitude information and phase information about the received ultrasound signal.

[0027] Accordingly, phase information, in addition to amplitude information, can also be used for classification. This advantageously enables a particularly precise and fine-grained classification into a larger number of different sub-areas of the planar component. It should be noted that a downmixed IQ signal is an example of a signal that includes both amplitude and phase information.

[0028] According to a further embodiment, the proposed method further comprises: subtracting a reference signal waveform from the acquired measurement signal waveform to obtain a differential signal waveform, which is used and classified as the signal waveform derived from the measurement signal waveform.

[0029] The reference signal profile is, in particular, a signal profile that is indicative of structure-borne sound, as it occurs in response to the emission of the ultrasound signal in the planar component when the planar component is not touched.

[0030] Accordingly, the differential signal curve, which is formed by calculating the difference between the measurement signal curve and the reference signal curve, is in particular a signal curve derived from the measurement signal that is indicative of changes to the structure-borne sound of the planar component in response to contact with the planar component.

[0031] Thus, it is advantageous to first determine, based on the differential signal waveform, whether the planar component is being touched. This is the case, for example, if the amplitudes of the differential signal waveform describe a predetermined threshold value, or if the energy content of the differential signal waveform (integral of the amplitudes over a predetermined time range, excluding a dead time at the beginning of the measurement) exceeds a predefined threshold value. The differential signal waveform can then be classified, if necessary, to determine the position of the contact.

[0032] It should be noted that the raw measurement signal waveform can be subtracted from a raw reference signal waveform; however, it is also conceivable to subtract a downsampled or IQ-transformed or otherwise derived reference signal waveform from a correspondingly downsampled or IQ-transformed or otherwise derived measurement signal waveform.

[0033] According to another embodiment, the reference signal profile indicates an ultrasound signal received by the ultrasound sensor in response to the emission of an ultrasound signal and is captured in a stationary state of the received ultrasound signal.

[0034] For example, the reference signal profile can be recorded at the factory under laboratory conditions, taking care that the flat component is not touched and that no reflections from the surroundings of the flat component are allowed to return to the ultrasonic sensor.

[0035] However, it is also intended that during the vehicle's operational period, stationary states of the ultrasound signal (the measurement signal waveform) are detected, and the reference signal waveform is regularly updated when such a stationary state is detected. For example, if the vehicle is parked next to a post, the received ultrasound signal contains both a structure-borne sound component and a static, stationary ambient sound component resulting from the post (which is stationary due to the parking situation). If a stationary state of the ultrasound signal is established—a state in which peak values ​​of the amplitudes, positions, and / or integrals over the amplitudes of the measurement signal waveform, or similar parameters, remain constant for a predetermined period—a new reference signal waveform can be acquired and defined for further use.Furthermore, by subtracting the newly defined reference signal waveform from the measurement signal waveform, both the always present structure-borne sound component in the planar component and the ambient sound component resulting from the post can be eliminated, and advantageously only an additional structure-borne sound component resulting from contact with the planar component remains in the differential signal waveform. According to a further embodiment, the classification comprises: providing at least the measurement signal waveform or the signal waveform derived from the measurement signal waveform as input data to a trained neural network; and obtaining from the trained neural network an indication of which of the several predefined sub-areas is being touched.

[0036] Accordingly, the effort required to manually configure an algorithmic classification of the measurement signal profile or the signal profile derived from it can be advantageously avoided, and artificial intelligence can be used to automatically learn the assignment of different measurement signal profiles (derived signal profiles) to different touched sub-areas of the planar component.

[0037] The training of the neural network can, for example, be carried out under laboratory conditions, whereby the neural network is provided with training data sets, each comprising a signal waveform measured under laboratory conditions or a signal waveform derived therefrom as training input data and a specification of the sub-area touched during the measurement as training output data, and the parameters of the neurons of the neural network can each be adapted to the training output data according to a suitable training algorithm.

[0038] This advantageously allows for the efficient automated creation of classifiers for each flat component.

[0039] According to another embodiment, the classification is further based on a door-closed information or a door-open information.

[0040] In particular, the input data to the trained neural network can also include door-closed or door-open information. If the planar component is, for example, a car door panel, the damping of the door panel due to contact with a door seal when closed can affect the structure-borne sound component of the received ultrasonic signal. By taking the door-open or door-closed information into account during classification, correct identification of the touched area is advantageously enabled for both open and closed doors.

[0041] According to another embodiment, classification is also carried out using temperature information.

[0042] In particular, the input data to the trained neural network can also include temperature information.

[0043] The inventors recognized that the measurement signal profile, or the differential signal profile derived from it, which results from contacting specific areas of a flat component, can be temperature-dependent. Accordingly, by taking the current temperature into account during the classification process, the accuracy of determining the contacted area can be ensured even when the flat component is exposed to large temperature fluctuations, as is often the case with the outer panels of a motor vehicle in climates with variable weather or large day-night temperature differences.

[0044] The temperature information can be obtained, for example, using an outside temperature sensor of the vehicle.

[0045] According to a further embodiment, the proposed method also includes: triggering an action depending on which of the several predefined sub-areas has been touched. The triggered action depends on the application and may, for example, include activating a camera that records an area in which a person touching the predefined sub-area is located in order to secure evidence (vandalism prevention), or may include causing an exterior door of the vehicle to open or close automatically (door opening assistance, for example, if the flat component is a door panel of the vehicle and the predefined area corresponds to a door handle), or locking or unlocking the vehicle (keyless entry), and the like.

[0046] According to a further embodiment, the proposed method also includes: generating a door opening signal or a door closing signal in response to a determination that one of the several predefined sub-areas associated with a door handle is touched.

[0047] The area associated with the door handle may be an area of ​​the door panel in which an actual door handle opening is formed, or it may be an area of ​​the door panel in which a door handle would conventionally be located on motor vehicles and which is therefore habitually touched by users of the motor vehicle in order to open the door.

[0048] It is understood that a door opening signal is generated when the door is currently closed, and a door closing signal is generated when the door is currently open.

[0049] Accordingly, a door opening and closing assistance system that can be operated by simple touch can be developed without the need for a separate sensor at the position of the door handle.

[0050] According to a further embodiment, the proposed method further comprises: recording a sequence in which several of the several predefined sub-areas of one or more planar components of the motor vehicle are touched in temporal proximity to one another; comparing the recorded sequence with a predefined sequence; and generating an unlock signal if the comparison reveals a match. "In temporal proximity" can, for example, mean that no more than a predetermined maximum time period may elapse between any two touches; otherwise, the recorded sequence is deleted and the recording is started again from the beginning.

[0051] Accordingly, a code lock functionality can advantageously be provided, in which the user touches several predefined areas on the outer skin of the motor vehicle (which need not be visually identifiable) arranged in a predefined, secret sequence in order to unlock the motor vehicle.

[0052] The predefined sequence can be factory-set or can be specified by the vehicle user in a learning procedure.

[0053] "Unlocking" can include unlocking previously locked doors, but can alternatively or additionally also include releasing the motor vehicle's engine for starting with a start button.

[0054] According to a second aspect, a computer program product is proposed which includes commands that, when the program is executed by a control device connectable to an ultrasonic sensor, cause it to execute the procedure described above.

[0055] A computer program product, such as a computer program tool, can be provided or delivered, for example, as a storage medium such as a memory card, USB stick, CD-ROM, DVD, EEPROM, or as a downloadable file from a server on a network. This can be done, for example, in a wireless communication network by transmitting a corresponding file containing the computer program product or the computer program tool. According to a third aspect, a control device is proposed that is configured to determine a contacted sub-area among several predefined sub-areas of a planar component of a motor vehicle using at least one ultrasonic sensor concealed on the planar component.The control device comprises: a first unit configured to use the ultrasonic sensor to detect a measurement signal profile indicating an ultrasonic signal received by the ultrasonic sensor in response to the emission of an ultrasonic signal; and a second unit configured to classify the measurement signal profile or a signal profile derived from the measurement signal profile in order to determine whether and at which of the several predefined sub-areas the planar component is touched.

[0056] The respective unit can be implemented in hardware and / or software. In a hardware implementation, the respective unit can, for example, be a computer or a microprocessor. In a software implementation, the respective unit can be a computer program, a function, a routine, an algorithm, a part of program code, or an executable object. Furthermore, each of the aforementioned units can also be part of a higher-level control system of the vehicle, such as a central electronic control unit and / or an engine control unit (ECU).

[0057] According to a fourth aspect, a motor vehicle is proposed which has: a planar component; an ultrasonic sensor which is concealed on the planar component; and the control device according to the third aspect.

[0058] The embodiments, features, and advantages described for the proposed method of the first aspect apply accordingly to the proposed computer program product of the second aspect, the proposed control device of the third aspect, and the proposed motor vehicle of the fourth aspect. Further possible implementations of the invention also include combinations of features or embodiments described previously or subsequently with respect to the exemplary embodiments, even if not explicitly mentioned. In such cases, the person skilled in the art will also add individual aspects as improvements or additions to the respective basic form of the invention.

[0059] Further advantageous embodiments and aspects of the invention are the subject of the dependent claims and the exemplary embodiments of the invention described below. The invention will be explained in more detail below with reference to preferred embodiments and the accompanying figures.

[0060] Fig. 1 shows a flat component of a motor vehicle according to a first embodiment;

[0061] Fig. 2 shows an exploded view to illustrate means for concealed mounting of an ultrasonic sensor according to the first embodiment;

[0062] Fig. 3 illustrates a functional configuration of a control device of a motor vehicle according to the first embodiment;

[0063] Fig. 4 illustrates steps of a method according to the first embodiment;

[0064] Fig. 5 shows a plot of a measurement signal profile for a non-touched flat component;

[0065] Fig. 6 shows a plot of a measurement signal profile upon contact with a first sub-area of ​​the planar component; Fig. 7 shows a plot of a measurement signal profile upon contact with a second sub-area of ​​the planar component;

[0066] Fig. 8 shows a motor vehicle according to a second embodiment; and

[0067] Fig. 9 illustrates a trained neural network according to a third embodiment.

[0068] In the figures, identical or functionally equivalent elements have been given the same reference symbols, unless otherwise indicated.

[0069] Fig. 1 shows a flat component 1 of a motor vehicle (100 in Fig. 8) according to a first embodiment. The flat component 1 is, for example, a door panel. The flat component 1 is conceptually divided into several sub-areas 11-19, the contact of which is to be detected. Several ultrasonic sensors 21-23, collectively or pars pro toto also designated by reference numeral 2, are concealed on the flat component 1.

[0070] It should be noted that the ultrasonic sensors 21-23 are arranged in concealed positions behind the door panel 1, where this is structurally feasible and expedient for the primary purpose of the ultrasonic sensors 21-23: measuring the vehicle's surroundings (100 in Fig. 8). Due to structural stiffening elements, decorative panels, and the like located behind the door panel 1, it may be difficult or impossible to arrange a sensor, such as an ultrasonic sensor 2 or a pressure-sensitive sensor, behind each of the areas of interest 11-19. Thus, as shown in Fig. 1, no ultrasonic sensors are arranged behind the areas 11, 12, 13, 15, 17, and 19. Nevertheless, with the proposed method, as now described, contact with even those areas 11-19 behind which no ultrasonic sensor is arranged can be detected.Figure 2 shows an exploded view illustrating means for the concealed mounting of an ultrasonic sensor 2 according to the first embodiment. The ultrasonic sensor 2, with an ultrasonic membrane (not shown) pointing downwards to the left in the figure, is inserted into a holder 4. The underside of the holder 4 has an adhesive layer by which the holder 4 is bonded to a butyl damping pad 3, which has an opening in the center for the ultrasonic membrane of the ultrasonic sensor 2. The butyl damping pad is bonded to the inside of the door panel (1 in Figure 1), which is not shown in Figure 2. A cap 5 is placed on the ultrasonic sensor 2 and holds it in the holder. In particular, this arrangement presses the ultrasonic membrane of the ultrasonic sensor 2 against the inside of the door panel (1 in Figure 1) and acoustically couples it to the door panel.

[0071] The butyl damping pad 3 is designed to dampen, but not completely suppress, the transmission of structure-borne sound into the door panel (1 in Fig. 1) when an ultrasonic signal is emitted by the ultrasonic sensor 2, as well as the feedback of reflected structure-borne sound from the door panel (1 in Fig. 1) to the ultrasonic diaphragm of the ultrasonic sensor 2. For example, the butyl damping pad 3 can have a diameter of 15 cm or less, particularly 10 cm. Accordingly, ultrasonic signals received by the ultrasonic sensor 2 also include a structure-borne sound component. If the ultrasonic sensor 2 is used to measure obstacles in the vicinity of the motor vehicle (100 in Fig. 8), this interfering structure-borne sound component can be computationally filtered out. In contrast, the structure-borne sound component can be advantageously used for the detection of touches according to the method described below.

[0072] Fig. 3 illustrates a functional configuration of a control device 8 of the motor vehicle (100 in Fig. 8) according to the first embodiment. The control device 8 comprises a first functional unit 81 and a second functional unit 82. The functional units 81 and 82 can be implemented in hardware or in software if the control device 8 executes a computer program that causes the control device to carry out the proposed method. It is understood that the control device 8 is connected to the respective ultrasonic sensor 2 via a bus or wirelessly.

[0073] Fig. 4 illustrates steps of a method according to the first embodiment.

[0074] Reference is made to Fig. 1, Fig. 3 and Fig. 4.

[0075] In step S10, the first unit 81 of the control device 81 activates one of the ultrasonic sensors 2 and causes it to emit an ultrasonic signal and then acquire a measurement signal profile (6 in Figs. 5 to 7) that indicates an ultrasonic signal received by the ultrasonic sensor 2 in response to the emission of the ultrasonic signal. Such an ultrasonic signal received in response to the emission of the ultrasonic signal comprises, in particular, signal components of the emitted ultrasonic signal reflected in the vicinity of the motor vehicle (100 in Fig. 8) as well as signal components of the ultrasonic signal coupled into the door panel 1 that are reflected within the door panel 1 (at its edges, creases, etc.) as structure-borne sound components. The structure-borne sound component will be considered primarily in the following.

[0076] It should be noted that, for the sake of clarity, the description here and in the following assumes that the control device 8 uses any selected ultrasonic sensor 2, for example, ultrasonic sensor 21, to emit an ultrasonic signal and to receive its reflected signal components and record them as a measurement signal waveform (direct measurement). However, those skilled in the art understand that a first ultrasonic sensor 2, e.g., ultrasonic sensor 22, can also emit a signal and another ultrasonic sensor 2, e.g., ultrasonic sensor 23, can receive its reflected signal components (indirect measurement). Those skilled in the art also understand that both direct and indirect measurements can be combined, and, for example, all of the ultrasonic sensors 2 can each emit an ultrasonic signal and all of the ultrasonic sensors 2 can record a measurement signal waveform with reflected signal components.Here, the multiple ultrasonic sensors 2 can either emit a signal sequentially, or all ultrasonic sensors 2 can emit simultaneously but at slightly different frequencies, so that the signals of the ultrasonic sensors 2 remain distinguishable. Some or all of the signals obtained in this way through direct and indirect measurements can be processed in the same manner as the measurement signal profile described below, which was acquired with a single ultrasonic sensor 2. The more measurement signal profiles are considered during classification, the more accurate the determination of the affected of the several sub-areas 1 1 -19 can be.

[0077] Fig. 5 shows a plot of a measurement signal waveform 6 acquired in step S10 when none of the sub-areas (cf. 11-19 in Fig. 1) of a (not shown) planar component of a prototype manufactured by the inventors is touched. In Figs. 5-7, the distance d to a reflection point, which is proportional to the signal propagation time, is plotted along the horizontal d-axis, and the amplitude is plotted along the vertical a-axis.

[0078] In the present embodiment, the measurement signal waveform 6 is a measurement signal waveform 6 downmixed by an IQ mixer, comprising an in-phase signal (i-signal) 61 and a quadrature signal (q-signal) 62. As such, the measurement signal waveform 6 includes both phase information and amplitude information about the received ultrasound signal.

[0079] With an untouched planar component, as shown in Fig. 5, two further peaks are observed in the measurement signal 6 after a transient phase at small distances, at a first distance d1 and a second distance d2. These are reflections at the edges of the planar component or the like. Here, the i-signal 61 and the q-signal d2 are out of phase: at distance d1, the q-signal 62 has a positive peak and the i-signal 61 a negative peak, and at distance d2, the i-signal 61 has a positive peak and the q-signal 62 a negative peak. The amplitudes of the peaks of the i-signal 61 and the q-signal 62 are approximately equal. Fig. 6 shows a plot of a measurement signal 6 that was acquired while one of the sub-areas of the planar component of the prototype was touched. As can be seen from a comparison of Fig.As can be seen in Figure 5, the overall picture is not fundamentally different, but the contact causes a phase shift. The i-signal 61 and q-signal 62 are now in phase and each show a positive peak at the first distance d1 and a negative peak at the second distance d2. Furthermore, the amplitudes of the peaks of the i-signal 61 are significantly lower than the amplitudes of the peaks of the q-signal 62.

[0080] Fig. 7 shows a plot of a measurement signal 6, which was acquired while a second of the sub-areas of the prototype's planar component was being touched. As can be seen from a comparison of Fig. 7 with Fig. 5 and Fig. 6, touching the second sub-area results in a different phase shift than touching the first sub-area. The i-signal 61 and the q-signal 62 are again in phase, but have reversed their sign; now the peaks at the first distance d1 are negative and at the second distance d2 are positive. Furthermore, the amplitudes of the peaks of the i-signal 61 are now significantly higher than the amplitudes of the peaks of the q-signal 62.

[0081] Thus, the experiments carried out with the prototype qualitatively demonstrated that it is possible to determine, by means of a classification of the recorded measurement signal profiles, for example rule-based or algorithmic, which of several predefined sub-areas of a planar component is touched.

[0082] With further reference to Figures 1, 3, 4, and 5-7, in step S20 the second unit 82 of the control device 8 classifies the measurement signal profile 6 acquired in step S10 in order to determine whether and at which of the several predefined sub-areas 1 1-19 the door panel 1 is in contact. According to the first embodiment, the second unit 82 applies a rule-based classification method. The parameters for the rule-based classification method were previously determined by laboratory measurements on a prototype of the

[0083] The door panel 1 is determined. Accordingly, the method of the first embodiment can determine which of several sub-areas 1 1-19 of a door panel 1 is touched, without requiring a separate sensor in each of the sub-areas 1 1-19. In particular, no separate pressure-sensitive sensors need to be provided for touch detection; instead, the ultrasonic sensors 2 installed on a vehicle (100 in Fig. 8) with driving or parking assistance for detecting obstacles in the vicinity of the vehicle (100 in Fig. 8) can simply be used. Depending on whether the vehicle (100 in Fig. 8) is stationary or moving, either the environmental detection or the touch detection proposed according to the first embodiment can be performed.

[0084] Depending on which of the several predefined sub-areas 11-19 is touched according to the determination carried out in step S20, the control device 8 can trigger an action. This will now be explained in more detail using a second embodiment.

[0085] Fig. 8 shows a motor vehicle 100 according to a second embodiment. The second embodiment is based on the first embodiment, and what is described therein also applies to the second embodiment. The second embodiment illustrates applications for touch detection and the actions performed in response.

[0086] The motor vehicle 100 has an outer skin 101, which is formed from several flat components, including a front door panel 102 and a rear door panel 103. A sub-area 12 of several predefined sub-areas of the front door panel 102 is a sub-area associated with a front door handle. This could mean, for example, that a door handle is located in sub-area 12, or it could mean that a conventional motor vehicle would have a door handle there. Accordingly, a sub-area 13 of several predefined sub-areas of the rear door panel 103 is a sub-area associated with a rear door handle. The ultrasonic sensors 23, 24, 25 are concealed on the front door panel 102, and the ultrasonic sensors 21 and 22 are concealed on the rear door panel 103. Another ultrasonic sensor 26 – not used here – is attached to a fender of the motor vehicle 100.

[0087] The motor vehicle also has the control device 8, which can be structured as shown in Fig. 3.

[0088] According to the second embodiment, the proposed method is executed in steps S10 and S20 for each of the doors. If, in step S20, it is detected that the section 12 of the front door is being touched, the control device 8 generates a door opening signal. This door opening signal causes a door motor (not shown) of the front door to open the front door. Similarly, if, in step S20, it is detected that the section 13 of the rear door is being touched, a door opening signal is generated, which causes a door motor (not shown) of the rear door to open the rear door.

[0089] When the front door or the rear door is open and the corresponding sub-area 12 or 13 of the respective door panel 102, 103 is touched again, the control device / generates a door closing signal which causes the respective door motor to close the respective door.

[0090] It should be noted that the inventors have recognized that the shape of the measurement signal curve 6 (Fig. 5-7) to be classified in step S20 (Fig. 4) depends on the damping properties of the door panel 1 (Fig. 1), which differ depending on whether the door is open or closed, since the closed door, for example, rests against a door sealing rubber and is damped by it.

[0091] Therefore, according to the second embodiment, the classification in step S20 (Fig. 4) can also be based on either a door-closed or a door-open information. That is, depending on whether the door is closed or open, different rules can be used to classify the measurement signal waveform 6 (Figs. 5-7).

[0092] Although not shown in Fig. 8, for both the front door panel 102 and the rear door panel 103, not only the shown sub-areas 12 and 13, respectively, can be predefined, but for each of the front door panel 102 and the rear door panel 103, nine sub-areas 11-19, as shown in Fig. 1, can be defined in the manner of a keypad. Reference is now made to Figs. 1, 4, and 8.

[0093] According to a further development of the second embodiment, the method according to the first embodiment is executed cyclically, and the control device 8 records a sequence in which different sections 11-19 of the rear door panel 103 and / or the front door panel 102 are touched in temporal association. That is, the control device 8 registers and records each touch, with the record being deleted or reset if no further touch occurs for a predetermined period of time. After each further touch, the control device 8 compares the recorded sequence with a predefined sequence.If the recorded sequence matches the predefined sequence, the control device 8 generates an unlocking signal, which causes the doors of the motor vehicle 100 to unlock and / or the engine of the motor vehicle 100 to be enabled for starting with a (not shown) start button. In this way, a flat component, such as the front door panel 102 or the rear door panel 103, can be used as a keypad for entering a predefined, secret sequence, thus enabling keyless entry.

[0094] In the further training, touching the sub-areas 12, 13 (Fig. 8) associated with the door handles can have no effect as long as the motor vehicle 100 has not yet been unlocked by "entering" (touching) the secret sequence. Fig. 9 illustrates a trained neural network 7 according to a third embodiment. Reference is made to Figs. 9, 8 and 1.

[0095] The third embodiment is based on the first embodiment, the second embodiment, or one of its further developments. That is, in the third embodiment as well, the contact of one of several partial areas 1 1-19 on a flat component 1 of a motor vehicle 100 is determined according to the method illustrated in Fig. 4, and optionally an action, such as opening or closing a door or locking or unlocking the motor vehicle 100, is performed.

[0096] The third embodiment differs from the first and second embodiments in that in step S20 (Fig. 4) the classification of the measurement signal profile 6 is not performed algorithmically or rule-based, but using the trained neural network 7.

[0097] The neural network 7 can be implemented in the control device 8. More precisely, in the control device 8 of the motor vehicle 100, a neural network 7 specifically trained for each of the planar components 102, 103 of the motor vehicle can be implemented.

[0098] The trained neural network 7 from Fig. 9 is a feedforward network comprising an input layer 71, at least one hidden layer 72, and an output layer 73. Samples 601-60n of the input signal waveform 6 are provided as input data to the neurons of the input layer 71. In response to the input, the neurons of the input layer 71 activate some of the neurons of the hidden layer 72, which in turn activate one or more neurons of the output layer 73. At the output layer 73, the trained neural network 7 provides, in response to the samples 601-60n, a signal in the form of several binary touch indications 111, 112, ..., 11m.Here, the touch indication 11 1 is a value of logical TRUE or numeric one if the first sub-area 11 is touched, and FALSE or zero otherwise; the touch indication 112 is logical TRUE or numeric one if the second sub-area 12 is touched, and so on. The last of the touch indications 11 m is logical TRUE or numeric 1 if none of the sub-areas 1 1 to 19 are touched.

[0099] To obtain such a trained neural network 7, an untrained neural network 7 can be used as a starting point. A large number of training datasets consisting of input data (samples 601-6nn) and output data (1 11 - 11 m) can then be used to train the neural network 7 by iteratively adjusting the parameters of the individual neurons of the individual layers 71-73. This training enables the neural network 7 to generate the output data of the training dataset for the input data of each training dataset. The training datasets can be easily obtained by touching the corresponding planar component 1 (a prototype thereof), taking a measurement signal profile 6 measured according to step S10 (Fig. 4) as input data, and generating touch indications 1 11 ..11 m corresponding to the touched sub-area or sub-areas 1 1-19.

[0100] The inventors have already achieved usable results for a planar component 1 with four sub-areas 11-14 using a trained network consisting of only three layers 71 to 73 in combination with 250 neurons in the input layer 71 and five neurons in the output layer 73.It is understood, however, that this structure of a neural network 7 is purely exemplary and other structures of neural networks are also conceivable; in particular, more than one hidden layer 72 can be used, the granularity of the sampling can be increased or decreased (thus, more or fewer than 250 neurons at the input layer 71 are conceivable), the number of neurons at the output layer can be adapted to any desired number of predefined sub-regions 11-19, and it is also conceivable to use fundamentally differently structured neural networks for classification, such as a convolutional neural network (CNN) or a recurrent neural network (RNN). According to the third embodiment, a neural network 7 can be used to perform the classification and determination of the touched sub-region 1 1-19.In this way, it is advantageously not necessary to have an expert manually develop a set of rules for a rule-based or algorithmic classification for each individual planar component 1 from measured signal profiles 6. Rather, the neural network 7, through training with signal profiles 6 measured on the respective prototype, can itself recognize the potentially difficult-to-formalize regularities by which different signal profiles 6 can be assigned to the different affected sub-areas 11-19.

[0101] This advantage becomes particularly apparent when several additional factors besides the samples 601-6nn of the measurement signal 6 are included in the classification. For example, if, as described for the second embodiment, door-open or door-closed information is to be considered, this can easily be provided at another neuron of the input layer 71. Alternatively or additionally, it is also conceivable to include temperature information measured by an outdoor temperature sensor in the classification, so that the classification is also based on the temperature information. The temperature information can be encoded as a floating-point number or in multiple bits, and the single floating-point number or the multiple bits can be provided at one or more additional neurons of the input layer 71.In this way, the neural network 7 only needs to be trained with measurement signal profiles 6 obtained at several different temperatures and the corresponding touch indications 11 1 -1 1 m and can then use the insights gained in the classification when a temperature information indicates a temperature different from the temperatures used in the training.

[0102] Although the present invention has been described using exemplary embodiments, it is modifiable in many ways. Only door panels have been described as the planar components 1, 102, 103, but this is not a limitation. The keypad functionality for generating the unlocking signal could also be implemented on other planar components of the outer skin 101 of the motor vehicle 100, for example, on a hood or a fender. It is also conceivable to apply the proposed technical teaching to planar components that are not sheet metal, for example, to a glass pane, such as a window pane of the vehicle 100. The concealed ultrasonic sensor 2 could be provided in an area of ​​the window pane hidden by an A-, B-, or S-pillar.

[0103] The number of predefined sub-areas 11-19 is not limited to four or nine, and any suitable number of sub-areas 11-19 can be predefined.

[0104] For the sake of clarity, the exemplary embodiments described how the acquired measurement signal waveform 6, or a downmixed form of the acquired measurement signal waveform 6 in an IQ mixer, is classified. However, to more clearly highlight the differences for classification depending on the affected sub-area 11-19, a previously acquired reference signal waveform is first subtracted from the acquired measurement signal waveform 6 or its downmixed form, and the result of this subtraction is used as the signal waveform derived from the measurement signal waveform 6 for classification purposes. The reference signal waveform may have been determined statically under laboratory conditions and be fixed.

[0105] Furthermore, the considerations presented so far implicitly assumed that the measurement signal profile 6, when the vehicle 100 is stationary, essentially only comprises the structure-borne sound component. However, it is conceivable, for example, that the vehicle 100 is parked next to an obstacle. In this case, all measurement signal profiles 6 measured in step S10 would also indicate an ambient sound component of the received ultrasound signal, resulting from reflections of the emitted portion of the ultrasound signal off the obstacle. This ambient sound component could complicate or even prevent classification. Therefore, it is also planned that the reference signal profile will be updated regularly, specifically whenever a steady state of the received ultrasound signal or the recorded measurement signal profile 6 is established.This means that if no significant changes to the measurement signal profile 6 are measured over a predetermined period of time, the measurement signal profile 6 measured in this way can be defined as the new reference signal profile. If a measurement signal profile 6 indicating a reflection from an obstacle in the vicinity of the parked vehicle 100 is defined as the new reference signal profile, the ambient sound component resulting from the obstacle is subsequently "subtracted away," and the signal profile derived from the respective measurement signal profile 6 by subtracting the reference signal profile is provided to the classification step S20. This signal profile then contains only any structure-borne sound component originating from contact with the planar component 1 and no longer contains ambient sound or structure-borne sound occurring even without contact.

[0106] The actions to be triggered were described as a door opening assistance (opening or closing a door of the vehicle 100). However, it could also be, for example, a camera of the vehicle 100 that could be activated, which, for the purpose of vandalism prevention, precisely records an area near the touched sub-area 1 1-19.

[0107] REFERENCE MARK LIST

[0108] 1 flat component, door panel

[0109] 2 ultrasonic sensors

[0110] 3 Butyl damping pads

[0111] 4 holders with adhesive tape

[0112] 5 caps

[0113] 6 Measurement signal waveform

[0114] 7 trained neural networks

[0115] 8 Control device

[0116] 81 first unit

[0117] 82 second unit

[0118] 61 i-Signal

[0119] 62q signal

[0120] 71 Input neuron layer

[0121] 72 hidden neuronal layers

[0122] 73 Exit neuron layer

[0123] I I-19 Sub-area

[0124] 21-26 Ultrasonic sensor

[0125] 100 motor vehicles

[0126] 101 Outer skin

[0127] 102 Front door panel, flat component

[0128] 103 Rear door panel, flat component

[0129] III -11 m Contact indication

[0130] 601 -6nn Sampling a Amplitude d Time period / Distance

[0131] S10-S20 Procedure steps

Claims

PATENT CLAIMS 1. Method for determining a touched sub-area among several predefined sub-areas (11-19) of a planar component (1) of a motor vehicle (100), which is designed using at least one ultrasonic sensor (2) which is concealed on the planar component (1), comprising: Use (S10) of the ultrasonic sensor (2) to record a measurement signal waveform (6) which indicates an ultrasonic signal received by the ultrasonic sensor (2) in response to the emission of an ultrasonic signal; Classifying (S20) the measurement signal waveform (6) or a signal waveform derived from the measurement signal waveform to determine whether and at which of the several predefined sub-areas (11-19) the planar component (1) is touched.

2. Method according to claim 1, characterized in that the planar component (1 ) is an outer sheet metal of the motor vehicle (100), preferably a door sheet metal of the motor vehicle (100).

3. Method according to one of the preceding claims, characterized in that the received ultrasound signal comprises a structure-borne sound component.

4. Method according to one of the preceding claims, characterized in that the measurement signal profile (6) or the signal profile derived from the measurement signal profile comprises at least amplitude information and phase information about the received ultrasound signal.

5. A method according to any one of the preceding claims, further comprising: Subtracting a reference signal waveform from the acquired measurement signal waveform (6) to obtain a difference signal waveform which is used and classified as the signal waveform derived from the measurement signal waveform (6).

6. Method according to claim 5, characterized in that the reference signal profile indicates an ultrasound signal received by the ultrasound sensor (2) in response to the emission of an ultrasound signal and is recorded in a stationary state of the received ultrasound signal.

7. Method according to any of the preceding claims, characterized in that the classification comprises: Providing at least the measurement signal waveform (6) or the signal waveform derived from the measurement signal waveform as input data to a trained neural network (7); and Erlangen, from the trained neural network (7), an indication of whether and which of the several predefined sub-areas (11-19) is touched.

8. Method according to one of the preceding claims, characterized in that the classification (S20) is further based on a door-closed information or a door-open information.

9. Method according to one of the preceding claims, characterized in that the classification (S20) is further carried out on the basis of temperature information.

10. A method according to any one of the preceding claims, further comprising: Triggering an action depending on which of the several predefined sub-areas (11-19) has been touched.

11. Method according to claim 10, further comprising: Generating a door opening signal or a door closing signal in response to a determination that one of the several predefined sub-areas (11-19) associated with a door handle is touched.

12. A method according to any one of the preceding claims, further comprising: Recording a sequence in which several of the several predefined sub-areas (11-19) of one or more planar components (1) of the motor vehicle (100) are touched in temporal relation to each other; and Comparing the recorded sequence with a predefined sequence; and generating an unlock signal if the comparison reveals a match.

13. Computer program product comprising commands which, when the program is executed by a control device (8) connectable to an ultrasonic sensor (2), cause it to execute the method according to one of claims 1 - 12.

14. Control device (8), configured for determining a touched partial area among several predefined partial areas (11-19) of a planar component (1) of a motor vehicle (100) using at least one ultrasonic sensor (2) concealed on the planar component (1), comprising: a first unit (81) configured to use the ultrasonic sensor (2) to detect a measurement signal waveform (6) indicating an ultrasonic signal received by the ultrasonic sensor (2) in response to the emission of an ultrasonic signal; and a second unit (82) configured to classify the measurement signal waveform (6) or a signal waveform derived from the measurement signal waveform in order to determine whether and at which of the several predefined partial areas (11-19) the planar component (1) is touched.

15. Motor vehicle (100) comprising: a planar component (1); an ultrasonic sensor (2) concealed on the planar component (1); and the control device (8) according to claim 14.

Images