Ultrasonic sensor unit and vehicle

Abstract

The invention relates to an ultrasonic sensor unit comprising a first ultrasonic element, which is provided for transmitting and / or receiving at least one ultrasonic wave, and a logic unit, which is provided for generating an echo signal and a trace signal on the basis of the at least one transmitted and / or received ultrasonic wave, wherein the logic unit is configured to predict at least one further ultrasonic wave, further echo signal by means of a Kalman filter on the basis of the echo signal and / or the trace signal, and wherein the logic unit is configured to predict at least one further ultrasonic wave, further echo signal by means of a Kalman filter on the basis of the echo signal and / or the trace signal by means of a Kuhn-Mongolian algorithm. The further ultrasonic wave, the further echo signal and / or the further trace signal are assigned to the transmitted ultrasonic wave, the echo signal and / or the trace signal. The invention also relates to a vehicle having a corresponding ultrasonic sensor unit.

Description

Technical Field

[0001] This invention relates to an ultrasonic sensor unit and a vehicle. Background Technology

[0002] Currently, various solutions utilizing ultrasonic signals exist in the automotive field. Due to the increasing number of ultrasonic sensors in vehicles and the rising demands for quality and reliability in ultrasonic sensing devices, the need for innovative and robust ultrasonic sensor units continues to grow.

[0003] In the automotive industry, the continuous pursuit of weight reduction to lower fuel consumption and increasingly fierce competition have created cost pressures, leading to a growing demand for cheaper and more efficient vehicle components. Summary of the Invention

[0004] The ultrasonic sensor unit according to the present invention has the following advantages compared with known ultrasonic sensor units: significantly improved robustness of ultrasonic measurements and easier adaptation to changes in the environment of the ultrasonic sensor unit. It is a highly efficient tracking system that enables simplified and systematic processing of sensor data, resulting in more accurate and reliable predictions and assignments. This approach improves the robustness of the system and makes it more adaptable to changes in the environment or the number of tracked objects.

[0005] According to the present invention, this is achieved by the following: an ultrasonic sensor unit having a first ultrasonic element and a logic unit, wherein the first ultrasonic element is configured to transmit and / or receive at least one ultrasonic wave, wherein the logic unit is configured to generate an echo signal and a spuren signal based on at least one transmitted and / or received ultrasonic wave, wherein the logic unit is configured to predict at least one additional ultrasonic wave and an additional echo signal for the additional spuren signal based on the echo signal and / or the spuren signal using a Kalman filter, wherein the logic unit is configured to assign the additional ultrasonic wave, the additional echo signal, and / or the additional spuren signal to the transmitted ultrasonic wave, the echo signal, and / or the spuren signal using a Kuhn-Munkres-Algorithmus algorithm.

[0006] In other words, using an ultrasonic sensor unit, a data document combining a Kalman filter and a den Hungarian algorithm or a Kuhn-Monkres algorithm can be implemented to detect echo data and correlate them within the ultrasonic system. This can be achieved through five main steps: prediction, correlation, update, initialization, and analysis. Preferably, the prediction step uses a Kalman filter to predict measurements for the next time interval (e.g., a frame). In the correlation step, a den Hungarian algorithm or a Kuhn-Monkres algorithm can be used to correlate the predicted data with current measurements within the current time interval, for example, within a matching window. In the update step, the filtering parameters of the Kalman filter can then be updated, particularly simultaneously with an observer using covariance, to provide a matching window and other desired parameters for the next measurement cycle. Due to the imbalance in the number of entries in the trace and in the arriving echoes, the trace can be validated without correlation with new echoes, especially to free up storage space for new data. For cases where echoes cannot be correlated with existing traces, a new trace can be initialized, which can generate new historical records. Preferably, the echo signal can be one or more data entries based on ultrasound waves that have been transmitted and received by the same ultrasonic element, or it can be one or more data entries relating to ultrasound waves that have been transmitted by a first ultrasonic element and received by a second ultrasonic element. More preferably, the trace signal can be, in particular, a trace or similar signal, which can have multiple data entries.

[0007] Preferred extensions of the invention are shown below.

[0008] Preferably, the logic unit is configured to update the Kalman filter based on the associated additional ultrasonic wave, the associated additional echo signal and / or the associated additional trace signal, in order to construct a matching window and at least one adapted parameter for the additional measurement.

[0009] One advantage of this implementation is that it can identify erroneous or similar measurements and exclude them for further processing. Preferably, the matching window includes a range within which the measurement value of the next additional measurement can fall. A Kalman filter can be used to estimate the system state by combining system measurements with state predictions. Kalman filters are particularly useful in systems with noisy or incomplete measurements, and in situations where it is uncertain how the system evolves over time. In this context, the primary application of a Kalman filter is to predict the distance of the next echo based on a set of stored echoes in a trace.

[0010] More preferably, the number of entries in the assigned other echo signals is not equal to the number of entries in the assigned other trace signals, wherein the logic unit is configured to assign a new trace signal to the entries of the assigned other echo signals if an entry of an echo signal cannot be assigned to the assigned other trace signal.

[0011] The advantage of this implementation is that the ultrasonic sensor unit makes the following independent decisions: whether the received or measured ultrasonic value belongs to a track, or whether a new track should be constructed.

[0012] The "assignment" step can be used to accurately assign an echo to each trace, and vice versa. After the prediction step, each trace is assigned a predicted value for the next echo distance. The prediction can then be compared with the arriving echo, and the difference can be calculated. Based on this difference, the cost of each pair can be calculated. Assuming there are M traces on the x-axis and N echoes on the y-axis that need to be assigned to each trace, the cost w(x,y) of each pair (trace, echo) can be determined. Preferably, all assignments can be performed to minimize the total cost of the cost table W(x,y).

[0013] For example, a trace cannot be assigned to an echo, and vice versa. This could mean that the trace has not been updated with new information during this measurement period. This leads to two challenges—first, a local minimum may not be the global minimum of cost; second, an asymmetric matrix may exist.

[0014] To solve this minimum-cost assignment problem, the Kuhn-Moncres algorithm (also known as the Hungarian algorithm) can be used, which has a polynomial time complexity. This algorithm works by iteratively finding a set of augmenting paths in a weighted bipartite graph.

[0015] Preferably, the cost matrix of the Hungarian algorithm is pre-prepared using complexity. Preferably, the minima is transformed into a maximalmatrix: W(x,y) = -W(x,y). The cost matrix is ​​balanced by filling rows or columns with standard cost limits, such as the maximum invalid cost defined as a positive number. The result of this step is a balanced NxN matrix with the covered costs. The algorithm operates by finding maximum consistency in a bipartite graph, where one set of nodes x represents the current echo and the other set of nodes y represents the trace.

[0016] The assignment process described above continues until all nodes are assigned, meaning the number of assignments may correspond to the size of the cost matrix. When this occurs, it is preferable to find the path with the minimum total cost, where each element in a row is assigned exactly to an element in its corresponding column. This ensures that each echo and each trace is accurately assigned to a pair (Gegenstück). Since the cost matrix is ​​balanced at the start of the algorithm, there may be echoes that cannot be associated with a trace, because the number of traces is less than the number of echoes, and vice versa.

[0017] Preferably, the logic unit has at least one observer of a Kalman filter, wherein the observer of the Kalman filter is configured to adapt at least one output value of the Kalman filter when the output value of the Kalman filter deviates from a predetermined amount from the ultrasonic wave, echo signal and / or trace signal.

[0018] One advantage of this implementation is that the results of the Kalman filter can be further improved by means of an observer, thereby further improving the accuracy of the entire system.

[0019] In practical applications of Kalman filters, it is difficult to determine and adapt the accuracy of the calibration in real time, and it is also difficult to analyze the system performance based on the calibration of recorded measurements. To address this issue, an observer can be used to monitor the state of the system. The observer can analyze the filter's output and compare it with actual measurements. If deviations in the filter's estimation can be identified, the observer can perform necessary adjustments to the filter's parameters to improve its accuracy and reliability.

[0020] If the filter works correctly, the measured noise is preferably white noise with a mean of zero. Therefore, the consistency of the filter can be verified by applying a confidence check to the posterior error covariance and a confidence check to the innovationskovarianz covariance.

[0021] Preferably, the elements of the innovation covariance matrix can be used to check the consistency of the Kalman filter's output. If the innovation is too large relative to the innovation covariance matrix, this may indicate that the filter is not working correctly and may require adaptation. Furthermore, the innovation covariance matrix is ​​a good metric for adapting the association window used in the association step. Preferably, the first element of the innovation covariance matrix can be used to filter out irrelevant data, ensuring that only important and relevant data are considered for the association step. This improves the accuracy and efficiency of the algorithm.

[0022] Preferably, the logic unit is configured to update at least one operating parameter of the observer based on an associated additional ultrasonic signal, an associated additional echo signal, and / or an associated additional trace signal.

[0023] When an echo is assigned to a track, it is preferably added to the track's stored echo list. This may trigger an update of the Kalman filter used for that track. Updating the Kalman filter may include integrating the predicted state with the newly measured state, i.e., the updated echo distance and its slope or first time derivative, which may be derived by adding the new echo to the track. Additionally, the state control is preferably updated by calculating the second derivative of the last echo distance. The updated state estimate and covariance matrix are then preferably used as the basis for the next prediction step in the Kalman filtering algorithm.

[0024] One advantage of this implementation is that the accuracy of the observer can be further improved, and in particular, a synergistic effect can be generated between the observer and the Kalman filter, which can further improve the matching accuracy and thus further improve the measurement accuracy of the sensor unit.

[0025] Preferably, updating at least one operating parameter of the observer includes updating the covariance matrix.

[0026] The advantage of this implementation is that multiple components can be considered to improve the operating parameters of the observer.

[0027] More preferably, the logic unit is configured to calculate the difference between the echo signal or trace signal and another echo signal or another trace signal, wherein the logic unit is configured to determine the value of each value pair for the other echo signal and the other trace signal based on the difference.

[0028] One advantage of this implementation is that the comparability of ultrasonic sensor values ​​can be improved by means of the constructed difference.

[0029] Preferably, the logic unit is configured to minimize the total value using the Kuhn-Moncres algorithm, wherein the total value is the sum of multiple values ​​for each value pair, and wherein the logic unit is configured to assign additional ultrasound, additional echo signals and / or additional trace signals to the ultrasound, echo signals or trace signals based on the total value and the Kuhn-Moncres algorithm.

[0030] The advantage of this implementation is that the stroke with the minimum total cost can be determined, which ensures that each echo and each trace is associated with exactly one mating part (cost).

[0031] Preferably, the ultrasonic sensor unit has a second ultrasonic sensor element, wherein the logic unit is configured to assign the second echo signal detected by the second ultrasonic sensor element to another trace signal using the Kuhn-Moncres algorithm.

[0032] One advantage of this embodiment is that multiple ultrasonic sensors are used (e.g., the multiple ultrasonic sensors can be arranged around the vehicle) to further improve the accuracy of the ultrasonic sensor unit.

[0033] Another aspect of the invention relates to a vehicle having an ultrasonic sensor unit as described above and below. Attached Figure Description

[0034] Embodiments of the present invention will now be described in detail with reference to the accompanying drawings. The drawings show:

[0035] Figure 1 An ultrasonic sensor unit according to one embodiment is shown.

[0036] Figures 2 to 4 A schematic diagram is shown to illustrate the operation of an ultrasonic sensor unit according to one embodiment.

[0037] Figure 5 and Figure 6 A block diagram is shown to illustrate the operation of an ultrasonic sensor unit according to one embodiment.

[0038] Figure 7 A simulated view of a vehicle equipped with an ultrasonic sensor unit according to one embodiment is shown.

[0039] Figure 8 and Figure 9 A schematic diagram illustrating the operation of an ultrasonic sensor unit according to one embodiment is shown, and

[0040] Figure 10 A vehicle according to one embodiment is shown.

[0041] Preferably, in all the drawings, all the same elements, units and / or steps are given the same reference numerals. Detailed Implementation

[0042] Figure 1An ultrasonic sensor unit 10 according to one embodiment is shown. The ultrasonic sensor unit 10 has a first ultrasonic element 12 and a logic unit 14, wherein the first ultrasonic element 12 is configured to transmit and / or receive at least one ultrasonic wave 16, wherein the logic unit 14 is configured to generate an echo signal and a trace signal based on at least one transmitted and / or received ultrasonic wave 16, wherein the logic unit 14 is configured to predict at least one additional ultrasonic wave, additional echo signal and / or additional trace signal based on the echo signal and trace signal using a Kalman filter, wherein the logic unit 14 is configured to assign the additional ultrasonic wave, additional echo signal and / or additional trace signal to the transmitted ultrasonic wave, echo signal and / or trace signal using a Kuhn-Moncres algorithm.

[0043] like Figure 1 As shown, the ultrasonic sensor unit is preferably arranged in the vehicle 100. Preferably, the ultrasonic sensor unit may include a logic unit 14, which may be constructed separately or integrally with the ultrasonic element 12. More preferably, the ultrasonic sensor unit 10 may include a second ultrasonic element 18, which in particular can receive the ultrasonic waves 16 emitted by the first ultrasonic element 12. Figure 1 As shown, the ultrasonic elements respectively construct the observation area 200, and in particular, in the segment 204 monitored, for example, by the first ultrasonic element 12, obstacles 206 are identified. More preferably, obstacles 206 can also be identified in a second segment 204 monitored, for example, by the second ultrasonic sensor element 18. For example, obstacle 206 could also be a tree 208.

[0044] Figure 2 A schematic diagram 300 is shown to illustrate the operation of an ultrasonic sensor unit 10 according to one embodiment. Schematic diagram 300 preferably has a first axis 302 and a second axis 304. (As shown...) Figure 2 As can be seen in the diagram, multiple traces 306 are drawn in schematic diagram 300. For example, corresponding ultrasonic measurements can be assigned to the first trace 308, the second trace 310, or the third trace 312.

[0045] Figure 3 A schematic diagram 320 is shown to illustrate the operation of an ultrasonic sensor unit 10 according to one embodiment. Preferably, schematic diagram 320 has a first axis 322 and a second axis 324. As can be seen in schematic diagram 320, the ultrasonic measurements can be summarized in particular as curve 326, a second curve 328, and a third curve 330.

[0046] Figure 4A schematic diagram 350 is shown to illustrate the operation of an ultrasonic sensor unit 10 according to one embodiment. The schematic diagram 350 preferably has a first axis 352 and a second axis 354. Preferably, a plurality of measurement values ​​362 can be plotted on the schematic diagram. Preferably, the ultrasonic sensor unit 10 is configured to construct a first trace 356 based on the plurality of measurement values ​​362. More preferably, the ultrasonic sensor unit 10 can also detect a second trace 358 and a third trace 360.

[0047] Figure 5 A block diagram 400 is shown illustrating the operation of an ultrasonic sensor unit 10 according to one embodiment. Preferably, the diagram 400 has two inputs, specifically an echo signal input 401 and a trace signal input 402. Preferably, algorithm 404 can take over further processing of the input signals in converter 406. Preferably, a Kalman prediction step 408 can be applied to the input signal 406. Based on the prediction step 408, a matching step can be performed, particularly with the aid of Hungarian algorithm 410, which can utilize the input signal 406. More preferably, a Kalman filter update step 412 can be performed, which can take into account the result of Hungarian algorithm 410 and the input signal 406.

[0048] Figure 6 A block diagram 450 is shown illustrating the operation of an ultrasonic sensor unit 10 according to one embodiment. Preferably, a new measurement value 452 may exist for the ultrasonic sensor unit 10. In step 454, the new value 452 may be extended from the previous value 453. The combination of data in step 456, especially with the echo 460, may be used in the prediction step 458. Based on the updated values ​​464 of the echo and the trace, these values ​​may be fed into the assignment process 464 in step 462. Based on this, a conflict resolution sub-step 482 may be performed in the association 464. Preferably, step 482 may include multiple sub-steps 484, 486, 488, 490 so that multiple echo values ​​can be assigned to each trace. Verification 468 may be performed by means of trace 466. More preferably, update 470 may be performed by means of trace and echo 478. Initialization 472 may be performed based on association 464 by means of echo 480. More preferably, the observer of the Kalman filter 474 can be applied based on update 470. The results of verification 468, update 470, and initialization 472 can be combined in step 476 to enable the construction and storage of the previous trace data 453, thereby completing the loop.

[0049] Figure 7A simulated view 500 is shown to illustrate the operation of an ultrasonic sensor unit 10 according to one embodiment. As shown in the simulated view 500, a vehicle 502 can emit ultrasonic waves 506, wherein, in particular, an obstacle 504 is simulated.

[0050] Figure 8 A schematic diagram is shown to illustrate the operation of an ultrasonic sensor unit 10 according to one embodiment. Schematic diagram 510 has a first axis 512 and a second axis 514, the first axis specifically including the distance of the echo, and the second axis including a time feature. As shown in schematic diagram 510, a first curve 516 and a second curve 518 can be shown, wherein the ultrasonic sensor unit 10 is configured to add new measurements to the corresponding curves based on a Kalman filter and a Hungarian algorithm.

[0051] Figure 9 A schematic diagram 520 is shown to illustrate the operation of an ultrasonic sensor unit 10 according to one embodiment. The schematic diagram 520 has a first axis 522 and a second axis 524. Multiple measurement values ​​530 are plotted in the schematic diagram 520, wherein the ultrasonic sensor unit 10 is configured to identify a first curve 526 and a second curve 528.

[0052] Figure 10 A vehicle 100 according to one embodiment is shown. The vehicle 100 preferably has an ultrasonic sensor unit 10 as described above and below.

Claims

1. The ultrasonic sensor unit (10) has: - First ultrasonic element (12), -Logic unit (14), - wherein the first ultrasonic element (12) is configured to emit and / or receive at least one ultrasonic wave (16), -in, The logic unit (14) is configured to generate an echo signal and a trace signal based on at least one emitted and / or received ultrasonic wave (16). - Wherein, the logic unit (14) is configured to predict at least one additional ultrasonic wave, additional echo signal and / or additional trace signal by means of a Kalman filter based on the echo signal and / or the trace signal. - wherein the logic unit (14) is configured to, by means of the Kuhn-Moncres algorithm, assign the additional ultrasonic wave, the additional echo signal and / or the additional trace signal to the emitted ultrasonic wave, the echo signal and / or the trace signal.

2. The ultrasonic sensor unit (10) according to claim 1, - wherein the logic unit (14) is configured to update the Kalman filter based on the associated additional ultrasound, the associated additional echo signal and / or the associated additional trace signal in order to construct a matching window and at least one adapted parameter for the additional measurement.

3. The ultrasonic sensor unit (10) according to claim 2, -in, The number of entries in the other associated echo signal is not equal to the number of entries in the other associated trace signal. -The logic unit (14) is configured to assign the entry of the other echo signal to a new trace signal if the entry of the echo signal cannot be assigned to the other trace signal to which it is assigned.

4. The ultrasonic sensor unit (10) according to any one of claims 2 to 3, -in, The logic unit (14) has at least one observer of the Kalman filter. -in, The observer of the Kalman filter is configured to adapt at least one output value of the Kalman filter when the output value of the Kalman filter deviates from the ultrasonic wave (16), the echo signal and / or the trace signal by a predetermined amount.

5. The ultrasonic sensor unit (10) according to claim 4, - wherein the logic unit (14) is configured to update at least one operating parameter of the observer based on the associated additional ultrasonic wave, the associated additional echo signal and / or the associated additional trace signal.

6. The ultrasonic sensor unit (10) according to claim 5, -Wherein, updating at least one operating parameter of the observer includes updating the covariance matrix.

7. The ultrasonic sensor unit (10) according to any one of the preceding claims, - Wherein, the logic unit (14) is configured to calculate the difference between the echo signal or the trace signal and the other echo signal or the other trace signal. -in, The logic unit (14) is configured to determine the value of each value pair for the additional echo signal and the additional trace signal based on the difference.

8. The ultrasonic sensor unit (10) according to claim 7, - Wherein, the logic unit (14) is configured to minimize the total value using the Kuhn-Moncres algorithm, -in, The total value is the sum of multiple values ​​used for each value pair. - wherein the logic unit (14) is configured to assign the additional ultrasonic wave, the additional echo signal and / or the additional trace signal to the ultrasonic wave, the echo signal or the trace signal based on the total value and the Kuhn-Moncres algorithm.

9. The ultrasonic sensor unit (10) according to any one of the preceding claims, -The ultrasonic sensor unit (10) has a second ultrasonic sensor element (18), wherein, The logic unit (14) is configured to assign the second echo signal detected by the second ultrasonic sensor element (18) to the additional trace signal by means of the Kuhn-Moncres algorithm.

10. A vehicle (100) having an ultrasonic sensor unit (10) according to any one of the preceding claims.

Images