Calibration procedure for ultrasonic parking sensor

Abstract

The present invention relates to a method for calibrating an ultrasonic parking sensor, comprising the following steps: calculating a measurement tuple comprising an arrival azimuth angle and the arrival elevation angle of a received ultrasonic signal, measuring at least one environmental condition, determining a true calibration tuple comprising a true arrival azimuth angle and a true arrival elevation angle of the ultrasonic signal, and determining a calibration function that maps the measurement tuple and the environmental condition to the true calibration tuple.

Description

State of the art

[0001] The present invention relates to a method for calibrating an ultrasonic parking sensor, a method for determining an azimuth angle and an elevation angle of an ultrasonic signal, and an ultrasonic parking sensor.

[0002] Ultrasonic parking sensors play an important role in parking assistance systems. Current ultrasonic parking sensors on the market rely on distance estimation using travel time to detect objects. However, they lack the ability to reliably determine the position of objects in the 3D environment.

[0003] It would be desirable to have an ultrasonic parking sensor that can be calibrated to determine the direction of an ultrasonic signal. Disclosure of the invention

[0004] The method for calibrating an ultrasonic parking sensor with the features of claim 1 of the invention has the advantage that the sensor accuracy can be improved, in particular the detection of an azimuth angle and an elevation angle of a received ultrasonic signal. This can be achieved by reusing the same ultrasonic parking sensor hardware. According to the invention, this is achieved by a method for calibrating an ultrasonic parking sensor, wherein in a first step, a measurement tuple comprising an arrival azimuth angle and the arrival elevation angle of a received ultrasonic signal is calculated. In a further step, at least one environmental condition is measured. Furthermore, a true calibration tuple comprising a true arrival azimuth angle and a true arrival elevation angle of the ultrasonic signal is determined.A calibration function is then determined that maps the measurement tuple and the ambient conditions to the true calibration tuple. This allows for reliable calibration of the ultrasonic parking sensor, and the resulting calibration function can be used to improve the angle measurement accuracy of the ultrasonic parking sensor. The azimuth angle is preferably measured horizontally to a measurement axis perpendicular to the main sensor surface of the ultrasonic parking sensor. The elevation angle is preferably measured vertically to the measurement axis.

[0005] The subclaims show preferred embodiments of the invention.

[0006] The ultrasonic parking sensor used for the calibration method described above is preferably an ultrasonic sensor array with multiple ultrasonic sensor units, and the measurement tuple is calculated by comparing the phase differences of the received ultrasonic signal between the ultrasonic sensor units. Specifically, the ultrasonic sensor array is a 2x2 ultrasonic sensor array with two ultrasonic sensor units per row and column. The use of an integrated ultrasonic sensor array can improve the angle measurement of the ultrasonic parking sensor while maintaining a small installation space.

[0007] Furthermore, it is preferred that the calibration function maps several different measurement tuples and associated environmental conditions to true calibration tuples. By mapping several different measurement tuples to true calibration tuples, the accuracy of the calibration function for incoming ultrasonic signals can be improved. The several different measurement tuples are preferably located in a relevant field of view for a subsequent application. The direction of the measurement tuples corresponds, particularly during operation, to their distribution over a probable distribution of directions of incoming ultrasonic signals.

[0008] The calibration function is preferably stored in a lookup table. A lookup table is an array of data that maps measurement tuples and the measured ambient condition to an output azimuth angle and an output elevation angle, thereby approximating the calibration function. Using a lookup table is easy to implement, enables fast runtime during subsequent use, and allows for mapping complex relationships when determining the calibration function.

[0009] The lookup table is preferably multidimensional. This allows multiple environmental conditions to be included in the calibration function and enables improved output from the calibration function.

[0010] Preferably, the ultrasonic parking sensor array is mounted on a motor, in particular a servo motor, configured to determine the true calibration tuple. This allows the calibration tuple for the corresponding measurement tuple measured by the ultrasonic parking sensor on top of the motor to be determined quickly and reliably.

[0011] Furthermore, the measured environmental condition preferably includes a temperature, a dynamic range, and / or a detected operating condition. The temperature has a significant impact on the propagation time and accuracy of the ultrasonic signal. By incorporating the measured temperature into the calibration function, the accuracy can be improved. The detected operating condition could, for example, be a parking application, where, for example, higher priority could be given to the resolution of the elevation angle, whereas in another application the azimuth angle could be of higher priority, or vice versa. Another detected operating condition could benefit from higher accuracy in some regions of the field of view, thereby further improving the quality of the calibration function.An example of the potential improvement offered by this method is as follows: in some areas, an object might be detected in the field of observation (FOR), which would indicate that an object is present, without any angular information. A further improvement is that more accurate angular information could be obtained as a result of this additional calibration.

[0012] The calibration function is preferably determined using a ranking function, in particular by minimizing the ranking function. The ranking function preferably describes the error between the true calibration tuple and the output of the calibration function, in particular the output azimuth angle and the output elevation angle. The ranking function is preferably a likelihood function. This helps to find an ideal calibration function and thereby improve the quality of its output values.

[0013] The invention also relates to a method for determining an azimuth angle and an elevation angle of an ultrasonic signal. First, the method receives the ultrasonic signal using an ultrasonic parking sensor. The method then calculates a measurement tuple, comprising an arrival azimuth angle and an arrival elevation angle, from the received ultrasonic signal. In a further step, at least one environmental condition is measured. Subsequently, an output azimuth angle and an output elevation angle are determined using a calibration function, wherein the calibration function was determined using one of the inventive methods described above. The output azimuth and elevation angles can then be sent to a control unit for use in a parking assistant.Using the calibration function to determine the output azimuth and elevation angle can help improve the angle measurement accuracy of the ultrasonic parking sensor and enable the ultrasonic parking sensor to determine the position of objects in the 3D environment.

[0014] Furthermore, the invention relates to an ultrasonic parking sensor configured to determine an azimuth angle and an elevation angle of an ultrasonic signal according to the method described above. This improves the accuracy of the angle measurement of the ultrasonic parking sensor, enabling the ultrasonic parking sensor to determine the position of objects in the 3D environment. Short description of the drawings

[0015] A preferred embodiment of the invention will be described in detail below with reference to the accompanying figures. The drawings show: Fig. 1 is a simplified flowchart of a method for calibrating an ultrasonic parking sensor according to a preferred embodiment of the invention, Fig. 2 is a simplified flowchart of a method for determining an azimuth and an elevation angle of an ultrasonic parking sensor according to the preferred embodiment of the invention, and Fig. 3 a simplified representation of an ultrasonic parking sensor according to the preferred embodiment of the invention. Embodiments of the invention

[0016] Preferably, all identical components, elements and / or units in all figures are provided with the same reference symbols.

[0017] Referring to Fig. 1 to 3, a method for calibrating an ultrasonic parking sensor 10, a method for determining an azimuth angle α and an elevation angle β, and the ultrasonic parking sensor 10 according to a preferred embodiment of the invention are described in detail below.

[0018] Fig. Figure 1 shows a flowchart describing the method for calibrating the ultrasonic parking sensor 10. At the beginning of the method, the flowchart shows three independent steps that can be performed simultaneously or sequentially.

[0019] In a first step S1.1, a measurement tuple comprising the arrival azimuth angle α and the arrival elevation angle β of a received ultrasonic signal 12 is calculated. The azimuth angle α and the elevation angle β are preferably measured using an ultrasonic sensor array 15 with a plurality of ultrasonic sensor units 11 (in Fig. 3) is calculated and the phase differences of the received ultrasonic signal 12 between the ultrasonic sensor units 11 are compared.

[0020] In a further step S1.2, at least one environmental condition is measured. The measured environmental condition is preferably a condition that influences the speed of the ultrasonic signal 12, such as the ambient temperature, air pressure, or humidity. The environmental condition can also be a dynamic range and / or a detected operating condition of the ultrasonic parking sensor 10.

[0021] In a further step S1.3, a true calibration tuple comprising a true arrival azimuth angle α and a true arrival elevation angle β of the ultrasonic parking sensor is determined, which corresponds to the calculated measurement tuple.

[0022] The true calibration tuple is preferably determined by a motor on top of which the ultrasonic parking sensor 10 is mounted for calibration. This motor is capable of orienting the ultrasonic parking sensor 10 in space so that the ultrasonic parking sensor 10 receives the ultrasonic signal 12 from different directions. An object 16 from which the ultrasonic signal 12 is reflected back to the ultrasonic parking sensor 10 preferably remains in a fixed position.

[0023] The object 16 from which the ultrasonic signal 12 is reflected back to the ultrasonic parking sensor 10 can preferably also be varied in distance.

[0024] After the measurement tuple is calculated, the ambient condition is measured, and the true calibration tuple is determined in steps S1.1, S1.2, and S1.3. A calibration function is determined in step S2. The calibration function maps the measurement tuple and the ambient condition to the true calibration tuple and can be used during operation of the ultrasonic parking sensor to determine an output tuple comprising an output azimuth angle α and an output elevation angle β based on a measurement tuple and a measured ambient condition.

[0025] Steps S1.1, S1.2 and S1.3 can be performed sequentially or simultaneously.

[0026] Preferably, steps S1.1, S1.2 and S1.3 are performed multiple times so that multiple associated measurement tuples, environmental conditions and true calibration tuples are collected before the calibration function is determined in step S2 based on these multiple values.

[0027] The calibration function is preferably determined using supervised learning, which can help detect the complex interdependencies between the input and output values. The calibration function can also be determined by minimizing a ranking function.

[0028] After the calibration function has been determined, it is stored in a lookup table in step S3. The lookup table is an array of data that maps measurement tuples and the measured ambient condition to the output azimuth angle α and the output elevation angle β, thereby approximating the calibration function.

[0029] The lookup table is preferably stored in a memory of the ultrasonic parking sensor 10, which enables rapid application of the calibration function in real time.

[0030] The lookup table is preferably a multidimensional array that maps the measurement tuple and several environmental conditions, such as temperature, air pressure and humidity, to the output azimuth angle α and the output elevation angle β.

[0031] Fig. 2 shows a flowchart comprising the steps of a method for determining the azimuth angle α and an elevation angle β of an ultrasonic signal 12.

[0032] In a first step 1, the ultrasonic signal 12 is received by an ultrasonic parking sensor 10. The ultrasonic parking sensor 10 is preferably an ultrasonic sensor array 15 with multiple ultrasonic sensor units 11.

[0033] The received ultrasonic signal 12 is preferably an incoming ultrasonic signal 14 based on an outgoing ultrasonic signal 13 emitted by the ultrasonic parking sensor 10 and reflected by an object 16. This allows the ultrasonic parking sensor 10 to measure the time of flight of the ultrasonic signal 12 and determine the distance to the object 15.

[0034] Alternatively, the received incoming ultrasonic signal 14 may be based on a reflected outgoing ultrasonic signal 13 emitted by another ultrasonic parking sensor of the same vehicle.

[0035] Any other ultrasonic sensor could also emit the received ultrasonic signal 12.

[0036] In a second step 2, the measurement tuple, comprising the arrival azimuth angle α and the arrival elevation angle β, is calculated from the received ultrasonic signal 12. The measurement tuple is preferably calculated by comparing the phase differences of the received ultrasonic signal 12 between the ultrasonic sensor units 11 of the ultrasonic sensor array 15.

[0037] Subsequently, in a third step 3, at least one environmental condition is measured. The environmental condition can be a temperature, humidity, air pressure, dynamic range, and / or a detected operating condition.

[0038] Based on the calculated measurement tuple and the measured ambient condition, the method determines the output azimuth angle α and the output elevation angle β in a fourth step 4 using a calibration function. The calibration function is based on a calibration function previously determined by the method for calibrating an ultrasonic parking sensor 10 described above.

[0039] Fig. 3 shows a schematic view of a vehicle 100 with the ultrasonic parking sensor 10. The ultrasonic parking sensor 10 is shown as the ultrasonic sensor array 15 with four ultrasonic sensor units 11 arranged in a 2x2 grid.

[0040] The ultrasonic parking sensor 10 emits the outgoing ultrasonic signal 13 along a main axis XX, which is perpendicular to a main sensor surface. The main sensor surface is preferably perpendicular to a ground surface, so that the main axis XX is horizontal to the ground surface.

[0041] The outgoing ultrasonic signal 13 is reflected by the object 16 and redirected to the ultrasonic parking sensor 10 as the incoming ultrasonic signal 14. When the incoming ultrasonic signal 14 is received by the ultrasonic parking sensor 10, it calculates the arrival azimuth angle α and the arrival elevation angle β.

[0042] The azimuth angle α is measured from the main axis XX in a horizontal plane, with the main axis XX lying within the horizontal plane. The elevation angle β is measured in a vertical plane from the direction of the incoming ultrasonic signal 14 to the horizontal plane.

[0043] The vehicle also includes an environmental sensor 17 that measures at least one environmental condition.

[0044] To improve the calculated measurement tuple, a calibration function is used to determine an output azimuth angle α and elevation angle β based on the measurement tuple and the ambient condition.

Claims

[1] A method for calibrating an ultrasonic parking sensor (10), comprising the following steps: - calculating (S1.1) a measurement tuple comprising an arrival azimuth angle (α) and the arrival elevation angle (β) of a received ultrasonic signal (12), - measuring (S1.2) at least one environmental condition, - determining a true calibration tuple (S1.3) comprising a true arrival azimuth angle (α) and a true arrival elevation angle (β) of the ultrasonic signal (12), and - Determining a calibration function (S2) that maps the measurement tuple and the ambient condition to the true calibration tuple. [2] A method for calibrating an ultrasonic parking sensor (10) according to claim 1, wherein the ultrasonic parking sensor is an ultrasonic sensor array (15) having a plurality of ultrasonic sensor units (11) and the measurement tuple is calculated by comparing the phase differences of the received ultrasonic signal (12) between the ultrasonic sensor units (11). [3] A method for calibrating an ultrasonic parking sensor (10) according to any one of the preceding claims, wherein the calibration function maps a plurality of different measurement tuples and associated environmental conditions to true calibration tuples. [4] A method for calibrating an ultrasonic parking sensor (10) according to any one of the preceding claims, wherein the calibration function is saved in a lookup table (S3). [5] A method for calibrating an ultrasonic parking sensor according to claim 4, wherein the lookup table is multi-dimensional. [6] Method for calibrating an ultrasonic parking sensor according to one of the preceding claims, wherein the ultrasonic sensor array (15) is mounted on a motor, in particular a servo motor, which is configured to determine the true calibration tuple. [7] A method for calibrating an ultrasonic parking sensor according to any one of the preceding claims, wherein the measured ambient condition comprises a temperature, a dynamic range and / or a detected operating condition. [8] A method for calibrating an ultrasonic parking sensor according to any one of the preceding claims, wherein the calibration function is determined using a ranking function. [9] Method for determining an azimuth angle (α) and an elevation angle (β) of an ultrasonic signal (12), comprising the following steps: - receiving (1) the ultrasonic signal (12) with an ultrasonic parking sensor (10), - calculating (2) a measurement tuple comprising an arrival azimuth angle (α) and an arrival altitude angle (β) from the ultrasonic signal (12), - measuring (3) at least one environmental condition and - Determining (4) an output azimuth angle (α) and an output elevation angle (β) using a calibration function, - wherein the calibration function was determined using one of the methods according to one of claims 1 to 8. [10] Ultrasonic parking sensor, arranged to determine an azimuth angle (α) and an elevation angle (β) of an ultrasonic signal (12) according to the method of claim 9.

Images