Control device and method as well as computer program product for calculating a pitching movement of a vehicle, in particular an autonomous or semi-autonomous vehicle, for example an autonomous or semi-autonomous agricultural machine

Abstract

Control method for calculating a pitching motion in a vehicle powertrain or a vehicle (50), comprising: - Determining several environment image points (22, 24, 26, 28) on a standing surface of the vehicle (50) detected by an environment sensor (20); - Defining an imaginary guide line (30, 32) for the several environment image points (22) by means of a linear regression;and- determining an angle (34, 36) between the auxiliary line (30, 32) and a predefined standing surface (52), wherein the environment image points (22, 24, 26, 28) are located in a predefined area near the vehicle (50), wherein the environment image points (22, 24, 26, 28) are located within a predefined distance (54) from the environment sensor (20), the control procedure further comprising comparing the determined angle (26) with a predefined threshold value, the control procedure further comprising classifying the determined angle (34, 36) as an angle caused by a pitching movement if the determined angle (34, 36) does not exceed the predefined threshold value, the control procedure further comprising classifying the determined angle (34, 36) as an angle caused by an actual slope of the standing surface if the determined angle (34, 36) exceeds a predefined threshold.;

Description

TECHNICAL AREA

[0001] The present invention relates to the field of vehicle powertrains. Furthermore, the present invention relates to vehicles, in particular autonomous or semi-autonomous vehicles, for example, agricultural machinery. Within the scope of the present invention, a control device, a control method, and a computer program product are proposed. TECHNICAL BACKGROUND

[0002] For autonomous and semi-autonomous vehicles, environmental perception plays a central role. This typically sensor-based environmental perception serves to generate and analyze environmental data in order to detect objects in the vehicle's vicinity and the road surface conditions. Based on this data analysis, the vehicle is controlled autonomously or semi-autonomously to avoid collisions with other objects, such as vegetation, buildings, people, and other road users. Another goal of this data analysis is to adapt the vehicle's dynamics to road surface conditions such as width, gradient, and surface roughness.

[0003] In practice, however, environmental sensing is affected by a number of factors. In particular, vehicle vibrations, such as those caused by the engine's inherent vibrations or potholes in the road surface, cause the positioning of the environmental sensors to change. Since these sensors are typically calibrated for a specific installation position, this means that the calibration data is no longer applicable to the situation where the vehicle is vibrating. Consequently, the detection of the road surface or the vehicle's footprint becomes inaccurate.

[0004] The invention is therefore based on the objective of realizing a control device and method for a vehicle powertrain or a vehicle that makes it possible to improve the environmental perception in such a way that errors caused by vibrations of the vehicle are corrected.

[0005] The problem is solved by a control device, a control method and a computer program product according to the independent claims.

[0006] The vehicle in question could be, for example, an autonomously or semi-autonomously driving passenger car or truck, an autonomously or semi-autonomously driving off-road and / or commercial vehicle, or an autonomously or semi-autonomously driving industrial and / or agricultural machine. The vehicle's powertrain is, in particular, an intelligent powertrain.

[0007] The control device can be, for example, an electronic control unit (ECU), an electronic control module (ECM), or a control unit for autonomous driving (e.g., an "autopilot"). The processor used / integrated in the control device can be a central processing unit (CPU) or a graphics processing unit (GPU).

[0008] The control device can be used, for example, in an autonomous or semi-autonomous vehicle, enabling it to operate fully or partially without human driver intervention. The control device can be located inside the vehicle, or outside or partially outside the vehicle. It could, for instance, be an algorithm running on a server or a cloud system. The control device has an interface to output control signals, generated based on the evaluation of environmental data supplied by the ambient sensor, to a mechatronic control unit in the vehicle.

[0009] The environmental sensor is, for example, one or more radar, lidar, and / or ultrasonic sensors. Alternatively, it can be one or more cameras with image sensors, particularly in the form of a CCD or CMOS sensor. The image data generated by the image sensors can consist of luminance data (grayscale values, brightness) or chrominance data (color). The environmental sensor is designed to monitor the vehicle's surroundings and, if necessary, detect objects on the road. For example, the environmental sensor generates sensor data relating to a person or another vehicle approaching the vehicle.

[0010] The environmental sensor can be located on or off the vehicle, for example as part of a central monitoring system. The environmental sensor can communicate with the control unit via wiring or wirelessly to transmit environmental data to the latter.

[0011] The processor determines several environmental image points on the vehicle's footprint, which it extracts from the environmental data generated by the environmental sensor. These environmental image points are pixels located on the vehicle's footprint. For example, they might include radar or lidar points. Essentially, these environmental image points are conceptual graphical points placed on the vehicle's footprint within the environmental image generated by the sensor.

[0012] The processor is also capable of defining an imaginary guideline for the multiple surrounding image points using linear regression or approximation. For example, the guideline could be the graphical representation of a mathematical function that is approximately obtained by fitting the multiple surrounding image points in a coordinate system. Alternatively or additionally, instead of an imaginary guideline, the processor can define an imaginary auxiliary surface or plane in which the multiple surrounding image points lie approximately, for the same purpose.

[0013] In the case of vibrations that cause the sensor position to shift or rotate relative to the vehicle's base, the auxiliary line, surface, or plane represents the base "seen" by the sensor. However, due to the shift or rotation of the sensor position, this does not correspond to the actual base; that is, the surrounding image points do not lie on the actual base, but on a line or surface that forms an angle with the actual base.

[0014] The processor is further designed to determine this angle between the auxiliary line or auxiliary surface / plane and a predefined base surface, which is the actual base surface.

[0015] The present invention is thus able to determine the difference between the surface "seen" by the environmental sensor and the actual surface and to take this into account during object detection. In this way, sensor-based environmental sensing is particularly accurate despite vehicle vibrations.

[0016] Advantageous designs and further developments are specified in the subclaims.

[0017] According to one embodiment, the surrounding image points are located in a predefined area near the vehicle, for example within a predefined distance from the vehicle.

[0018] In this area, the environmental sensor's detection is particularly precise, allowing the environmental pixels to be determined with high accuracy. Additionally or alternatively, the number of environmental pixels and / or the sensor's installation position can be predefined.

[0019] According to a further embodiment, the control method according to the invention additionally includes comparing the determined angle with a predefined threshold value.

[0020] The threshold preferably defines the boundary between a sensor position shift caused by pitching motion and a shift resulting from driving over an incline on the ground surface. If the angle is particularly small, it is highly likely to be due to pitching motion. Larger angle values ​​are more likely to indicate an incline of the ground surface in general. The processor can additionally classify the detected angle as caused by pitching motion, provided the detected angle does not exceed the predefined threshold.

[0021] According to another embodiment, the processor can also correct the surrounding pixels by the determined angle if the determined angle does not exceed the predefined threshold.

[0022] In particular, the processor can shift or move the surrounding image points by the determined angle to align them with the vehicle's actual footprint. This is achieved, for example, by calibrating the image coordinates of the respective surrounding image points. The corrected surrounding image points can then be used for environmental perception, preventing the detection error caused by vibrations from propagating to further object recognition. The calibrated coordinate system can be stored, in particular, in a storage medium of the environmental sensor, the control unit, and / or the vehicle, or alternatively in an external storage medium such as a server or cloud system, so that it can be retrieved as needed.

[0023] According to another embodiment, the processor can interpret the determined angle as an angle caused by a real slope of the base, provided that the determined angle exceeds the predefined threshold.

[0024] Preferably, in this case, the angle is not taken into account in further object recognition. For example, the angle value is changed to zero.

[0025] The computer program product according to the invention is designed to be loaded into a computer's memory and comprises software code sections with which the process steps of the method according to the invention are executed when the computer program product is running on the computer.

[0026] A program is part of the software of a data processing system, such as an evaluation unit or a computer. Software is a collective term for programs and their associated data. The complement to software is hardware. Hardware refers to the mechanical and electronic components of a data processing system. A computer is an evaluation unit.

[0027] Computer program products typically comprise a sequence of instructions that, when the program is loaded, cause the hardware to perform a specific procedure leading to a particular result. When the program in question is used on a computer, the computer program product produces the inventive technical effect described above.

[0028] The computer program product according to the invention is platform independent. That is, it can be executed on any computing platform. Preferably, the computer program product is executed on an evaluation device according to the invention for capturing the vehicle's environment.

[0029] The software code sections are written in any programming language, for example Python.

[0030] Exemplary embodiments are now described by way of example and with reference to the accompanying drawings. These show: Fig. 1 a schematic representation of a control device according to one embodiment; Fig. 2 a schematic representation of a vehicle to which a control device from Fig. 1 cooperating environmental sensor for detecting the vehicle's standing surface is arranged; and Fig. Figure 3 shows a schematic representation of the vehicle, showing a correction of several surrounding image points using linear regression.

[0031] In the figures, identical reference symbols refer to identical or functionally similar reference parts. The relevant reference parts are identified in each figure.

[0032] Fig. Figure 1 shows a schematic representation of a control device 10 according to one embodiment. The control device 10 includes an input interface 12 , a processor 14 and an output interface 16 on.

[0033] The input interface 12 is equipped with an environmental sensor 20 , for example a radar sensor, and is therefore able to obtain data from the environment detection of the environment sensor. 20 for a vehicle 50 (see Fig. 2-3) to receive.

[0034] The processor14 is trained to process the received data in such a way as to detect a pitching movement of the vehicle. 50 to calculate. For this purpose, the processor 14 an imaginary guideline 30 , 32 on, which uses linear regression of several surrounding image points 22 , 24 , 26 , 28 This comes about. This is in Fig. 2 shown in more detail. In the example shown there, the environmental sensor 20 (such as a radar sensor) directly on the vehicle 20 , which is a tractor, arranged. The environmental sensor 20 captured within a field angle 56 the vehicle's surroundings 20 and thus also a standing surface on which the vehicle can stand. 20 is located in Fig. 2 is a predefined standing area represented by a solid line. 52 The actual footprint is shown. At the same time, several initial surrounding image points are also shown. 22(e.g. radar points) in the graphic representation above the predefined base area 52 and several second background image points 26 (e.g. radar points) in the graphic representation below the predefined base area 52 to see the surrounding pixels 22 , 26 are chosen so that they are within a predefined distance 54 from the environmental sensor 20 lie. An imaginary guideline was drawn over the first and second points, respectively. 30 , 32 laid out, which are in line with the predefined standing area 52 each angle 34 , 36 includes.

[0035] The first surrounding image points 22 reflect the readings from the environmental sensor 20 "seen" standing area in case the vehicle 50a nodding motion vertically upwards, or at least with a vertically upward-pointing motion component. The second surrounding image points 26 reflect the readings from the environmental sensor 20 "seen" standing area in case the vehicle 50 a nodding motion vertically downwards, or at least with a vertically downward-pointing motion component.

[0036] The standing area 52 includes a slope 58 , which are connected to the vehicle 20 approaching, flat section at an angle 60 includes. Also when measuring the gradient. 58 Several initial environmental image points can be 24 or several second background image points 28 be determined.

[0037] The processor 14 is also trained to display the surrounding image points 22 , 24 , 26 , 28 about the determined angle 34 , 36to correct in order to eliminate the detection error that is due to the respective nodding movement.

[0038] This is in Fig. Three examples are shown. The first and second surrounding image points. 22 , 24 , 26 , 28 They are moved or relocated, for example with the help of a coordinate transformation, so that they now rest on the predefined base. 52 lie. Thus, the first or second surrounding image points appear. 22 , 24 , 26 , 28 as if they were lying on the actual surface. The same applies to the section with the incline. 58 .

[0039] The processor 14 can also determine the angles 34 , 36 compare with a predefined threshold. In the event that the determined angle 34 , 36If the predefined threshold is not exceeded, the angle will be 34 , 36 as caused by a pitching motion. Otherwise, the angle 34 , 36 classified as being caused by a slope. The output interface 16 is designed to send a control signal, generated based on the calculated pitching movement, to the environment sensor 20 (as in Fig. 1 (shown as an example with a dashed line), the vehicle 50 or another entity (such as a server or cloud system) with which the vehicle 50 and / or the environmental sensor 20 (can communicate) to output. Reference symbol list 10 Control device 12 Input interface 14 processor 16 Output interface 20 Environmental sensor 22, 24 first surrounding image points 26, 28 second surrounding image points 30, 32 Guideline 34, 36 angles 50 vehicles 52 predefined standing areas 54 predefined distance 56 field angle 58 gradient 60° incline

Claims

[1] Control method for calculating a pitching motion in a vehicle powertrain or vehicle (50), comprising: - Determining several environment image points (22, 24, 26, 28) on a vehicle standing surface (50) detected by an environment sensor (20); - Defining an imaginary guideline (30, 32) for the multiple surrounding image points (22) using a linear regression; and - Determining an angle (34, 36) between the auxiliary line (30, 32) and a predefined base (52). [2] Control method according to claim 1, wherein the environment image points (22, 24, 26, 28) are located in a predefined area near the vehicle (50). [3] Control method according to claim 2, wherein the environment image points (22, 24, 26, 28) are located within a predefined distance (54) from the environment sensor (20). [4] Tax procedure according to one of the preceding claims, further comprising comparing the determined angle (26) with a predefined threshold. [5] Control method according to claim 4, further comprising classifying the determined angle (34, 36) as an angle caused by a pitching movement, provided that the determined angle (34, 36) does not exceed the predefined threshold. [6] Control method according to claim 4, further comprising correcting the surrounding image points (22, 24, 26, 28) by the determined angle (34, 36) if the determined angle (34, 36) does not exceed the predefined threshold. [7] Control method according to one of claims 4 to 6, further comprising classifying the determined angle (34, 36) as an angle caused by a real slope of the base surface if the determined angle (34, 36) exceeds the predefined threshold. [8] Control device (10) for calculating a pitching motion in a vehicle powertrain or a vehicle (50), comprising a processor (14) configured to: - to determine several environmental image points (22, 24, 26, 28) on a vehicle standing surface (50) detected by an environmental sensor (20); - to define an imaginary guideline (30, 32) for the multiple surrounding image points (22) using a linear regression; and - to determine an angle (34, 36) between the auxiliary line (30, 32) and a predefined base (52). [9] Vehicle powertrain or vehicle (50), in particular an autonomous or semi-autonomous agricultural machine, comprising a control device (10) according to claim 8. [10] Vehicle powertrain or vehicle (50), further comprising a sensor (20) for environmental sensing. [11] Computer program product stored on a data carrier and configured to perform the control method according to claim 1 when the computer program product is executed by a computer.

Images