Method for determining the installation position of an environment-monitoring sensor of a motor vehicle, computing device, computer program and electronically readable data carrier

Abstract

Method for determining the installation position of an environment-monitoring sensor (1, 1a, 1b), in particular a radar sensor or lidar sensor, measuring by reflection from objects to be detected, in a motor vehicle, comprising the following steps: - Providing a geometric first model of at least one component of the motor vehicle on which the environmental sensor (1, 1a, 1b) is to be installed, - Providing a second geometric model in at least one detection position relative to the first geometric model in a computational space, wherein the second model describes the surface of an object (2) to be detected, - for several test installation positions of the environmental sensor (1, 1a, 1b) on the first model in the computational space, determination of a surface orientation that enables a maximum sensor signal of the environmental sensor (1, 1a, 1b) in the test installation position, describing a visual normal (8) of the environmental sensor (1, 1a, 1b) for at least one surface section (7) of the second model as a function of the test installation position and an evaluation information for the surface section (7) as a function of a comparison of a surface normal of the surface section (7) with the visual normal (8), - Display of the evaluation information and / or an evaluation of the test installation positions based on their evaluation information by determining an evaluation value for each test installation position, wherein the size of the surface sections (7) for which an evaluation information has been determined is weighted on the basis of the evaluation information to obtain a surface integral as at least a part of the evaluation value.

Description

[0001] The invention relates to a method for determining the installation position of an environment-monitoring sensor, in particular a radar sensor or lidar sensor, which measures by reflection from objects to be detected, in a motor vehicle. The invention also relates to a computing device, a computer program, and an electronically readable data carrier.

[0002] Environmental sensors are common in modern vehicles, including not only land vehicles (cars) but also marine vessels and / or aircraft, to gather information about the vehicle's surroundings. In particular, environmental sensors whose physical measurement principle is based on the reflection of a transmitted signal from an object and the measurement of the reflected received signal are frequently used. Classic examples of such sensors are radar sensors and lidar sensors, as well as infrared cameras with their own illumination (PMDs).

[0003] To ensure that the corresponding environmental sensors can reliably detect objects in the vicinity of the vehicle, their positioning—that is, determining their installation position, including their orientation and orientation—requires careful attention. These environmental sensors are typically mounted on vehicle components, specifically on their surfaces, thus defining a specific detection range. It is crucial to select the installation position to maximize object detection reliability. Therefore, determining suitable installation positions as early as possible in the vehicle design and development process is particularly important.

[0004] In this context, it is common practice to calculate the so-called retroreflective behavior, for example in the form of the RCS (Radar Cross Section), using high-frequency full-wave simulation. This can involve, for instance, determining a surface integral of the current distribution from the electromagnetic excitation. It is also possible to use ray tracing, in which a quasi-optical virtual beam is tracked as it strikes the object, is reflected, and potentially returns to the receiver. From this, specific RCS values ​​can be determined. However, such high-frequency full-wave simulations are extremely complex, so that, for example, an assessment of an installation location can take several days, depending on the granularity of the object and the accuracy of the calculation.

[0005] DE 10 2011 010 861 A1 relates to a method for optimizing radomes for a motor vehicle radar sensor. It proposes an analysis method for the qualitative and quantitative analysis of the radar frequency-dependent transmission and / or reflection behavior of a multitude of paint and plastic components from a group of paint and plastic components used in the manufacture of radome-forming motor vehicle components consisting of paint and plastic layers. Subsequently, a simulation method is performed to create a model for a radome using the analysis results of the paint and plastic components, and a radome model optimized with respect to its radar frequency-dependent transmission behavior is selected. Thus, the focus is on the structural and mechanical integration of radomes in front of a radar sensor to prevent reflections.

[0006] DE 10 2004 058 703 A1 discloses an arrangement and a method for determining an arrangement of sensors on a motor vehicle. The fields of view of the optical sensors are intended to form overlapping areas for detecting obstacles. A critical detection area of ​​the arrangement is provided with a virtual grid, onto whose cells the respective fields of view of the sensors are mapped. The grid is used to create an objective function, which incorporates at least one predefined condition for the best possible resolution of the obstacles and is optimized to determine an arrangement of the sensors. The aim is to optimize the overlapping areas and effective ranges of the sensors.

[0007] DE 10 2016 200 806 A1 relates to an adapter for mounting a laser measuring device on a curved surface and a method for installing a sensor on a curved surface. The adapter has a first side with a receptacle designed to receive the laser measuring device and a second side, opposite the first side, which includes an annular support for the curved surface. In a method for installing a sensor at a mounting location on a curved section of an outer surface of a vehicle, wherein a measuring axis of the installed sensor is defined by the normal of the outer surface at the mounting location, the vehicle is parked in front of a wall, a possible mounting location on the outer surface is selected, and reference points on the wall are determined.Using the determined reference points, a yaw angle and / or a pitch angle is determined, the angles are compared with specified angles and, if at least one angle deviates from the respective specified angle by more than a specified tolerance, the procedure is repeated for a new possible installation location.

[0008] DE 10 2017 213 214 A1 concerns a sensor product, a simulator, and various methods. It deals with driver assistance systems that use sensor data from different sensors to create an environmental representation, for example through sensor fusion, which is then evaluated by the driver assistance system. The objective is to strive for reusability and simplification in the design of driver assistance systems. Currently, the methods for fusing measured values ​​or sensor data into an internal representation are established based on the developers' experience and engineering expertise. In this context, the aim is to understand which phenomena lead to which measured values, so that conclusions can then be drawn about the phenomena from the measured values.It is proposed to provide a sensor model, which describes the hardware and / or physical properties of the sensor, separately from an environment model, and to derive sensor data through simulation. These results can also be used in the design of driver assistance systems.

[0009] DE 10 2017 213 214 A1 relates to a method for modeling a motor vehicle sensor in a virtual test environment by defining and using a sensor support, a raycast distribution shape, a group of raycast properties, a raycast reflection factor and a raycast echo.

[0010] US 8,417,490 B1 discloses a method for providing an integrated software development environment for the design, verification, and validation of advanced vehicle safety systems. The system enables the development of automotive software on a host computer using a collection of computer programs that are executed concurrently as processes and synchronized by a central process. The disclosed software uses separate synchronized processes that allow the generation of signals from different sources, either through a simulation running on the host computer or from actual sensor and data bus signals coming from and going to the actual vehicle hardware, which is connected in real time to its bus counterparts in the host computer.The methods serve to provide algorithm prototyping, analysis and testing through an integrated framework for dynamic data modeling and application development.

[0011] The invention is therefore based on the objective of providing a simple and quick method for finding suitable installation positions for environmental sensors on a motor vehicle.

[0012] To solve this problem, a method of the type mentioned at the outset comprises, according to the invention, the features of claim 1. The following steps are provided: - Providing a geometric first model of at least one component of the motor vehicle on which the environmental sensor is to be installed, - Providing a second geometric model in at least one detection position relative to the first geometric model in a computational space, wherein the second model describes a surface of an object to be detected, - for several test installation positions of the environmental sensor on the first model in the computational space, determination of a surface orientation that enables a maximum sensor signal from the environmental sensor in the test installation position, describing the visual normal of the environmental sensor for at least one surface section of the second model as a function of the test installation position, and determination of evaluation information for the surface section as a function of a comparison of a surface normal of the surface section with the visual normal. - Display of the evaluation information and / or an evaluation of the test installation positions based on their evaluation information by determining an evaluation value for each test installation position.

[0013] Within the scope of the present invention, as is customary for such installation tasks, the term "position" is used as a collective term for the actual position (e.g., Cartesian coordinates x, y, z) and the orientation (angular position). This means that an installation position defines both the precise installation location and the installation orientation in which the environmental sensor is installed. The same applies to the detection position, which specifies the relative position and orientation of the second model to the first model in the virtual computational space, i.e., of a conceptual virtual object to the virtual component of the motor vehicle and thus to the environmental sensor installed on this component. The method, particularly with regard to calculations in the computational space, is performed at least partially, preferably completely, by a computing device.This means that the first model, the second model, and the computational space are to be understood as data constructs used as computational tools.

[0014] For environmental sensors that measure by reflection from objects to be detected, there are essentially two configurations: one for a monostatic measurement method and one for a bistatic (or more generally, multistatic) measurement method. In monostatic measurement methods, the transmitter and receiver are located at the same location, so object reflections must be directed back in the direction of arrival for optimal detectability. Therefore, the line of sight must be chosen along the direction of transmission of the (virtual) environmental sensor's signal. In bistatic measurement methods, the transmitter and receiver of the environmental sensor are located at different locations. Therefore, object reflections must be perpendicular to the distance normal between the transmitter and receiver for the highest detection reliability; this distance normal forms the line of sight midway between the transmitter and receiver components.An optimally detectable surface must run perpendicular to the respective viewing normals.

[0015] It should be noted here that, depending on the type of measurement operation of the environmental sensor, a distribution of visual normals or a field of visual normals covering the sensor's detection range can also be determined. For example, if the environmental sensor transmits the signal simultaneously in a fan-shaped and / or cone-shaped detection area, there is usually a main measurement direction, which, in monostatic methods, forms the central visual normal, with the visual normals away from the main measurement direction being slightly tilted. Multiple visual normals can also be determined for a scanning environmental sensor that emits the measurement signal sequentially in different directions, again, for example, as a distribution of visual normals and / or a field of visual normals, whereby the resolution can be selected depending on the surface sections to be considered and their size.

[0016] A fundamental idea of ​​the present invention is therefore to perform a geometric analysis by examining the extent to which surface normals on the virtual object to be detected, described by the second model, correspond to the visual normals in specific detection positions. This describes how well the virtual object to be detected would be detected by the environmental sensor. In this way, a quick and efficient selection of an installation position for a radar or lidar sensor is possible for the best possible detection of objects within the detection range, without the need for a complex high-frequency simulation. By comparing surface orientations, the surface areas that contribute most to reflection and thus detection when excited can be identified.Because both the visual display of the evaluation information, for example, as will be explained in more detail, in the form of a color-coded surface reproduction, and the evaluation value make it possible to select the most suitable installation position from several test installation positions.

[0017] The inventive method can be universally applied to any objects and / or motor vehicles to be detected. Furthermore, its use beyond ground-based motor vehicles, particularly automobiles, is conceivable for ships and / or aircraft, generally sea and / or aircraft.

[0018] The present invention offers a particular advantage in exploiting the fact that CAD (Computer-Aided Design) programs are frequently used in a very early development phase of a motor vehicle to select the installation positions of environmental sensors. This means, however, that the motor vehicle and its components are already available as suitable initial models in the CAD software. Since other objects, in this case objects to be detected, are often also available as models in the CAD software or can be easily introduced, a preferred embodiment of the present invention provides that the computational space is provided by CAD software.

[0019] In this context, it is particularly advantageous if the determination of the surface normals of at least one surface section and / or the comparison with the visual normal is carried out using a demolding analysis function of the CAD software. Demolding analysis is already known in the art for use in injection molding tools to ensure that a workpiece can be removed from the mold after the injection process. If the surfaces are not correctly aligned, the workpiece cannot be removed (non-destructively) from the mold. Incorrectly aligned surfaces can even prevent the workpiece from being manufactured at all. Examples of such problems include undercuts and the like.

[0020] According to the present invention, it is now proposed to use a function that is already frequently provided in CAD software, namely the demolding analysis function, within the scope of the present invention by searching for surface components, i.e., surface elements, on the object to be detected, represented by the second model, whose surface normal corresponds to the viewing normal of the environmental sensor or deviates only slightly from it, for which corresponding tolerance ranges can be defined, as will be explained in more detail below.

[0021] A further advantageous embodiment of the present invention provides that, for displaying the evaluation information, a rendered representation of the second model is generated, wherein the evaluated surface sections are displayed differently according to their evaluation information, in particular by different colors. For example, different colors can be used to indicate angular ranges in which a deviation of the surface normal from the visual normal, as determined during comparison, lies. For example, green can be used in a deviation angle range of + / - 5°, yellow in the deviation angle range of + / - 5° to + / - 10°, and pink in the deviation angle range of + / - 10° to + / - 15°, with all deviation angle ranges greater than + / - 15° being marked in red as exclusion areas.In this way, an observer can very quickly identify which surface areas of the object to be detected are actually suitable for detection, specifically, for example, monostatic or bistatic reflections. In other words, coloring can be done based on defined deviation angle intervals of the deviation angles determined in comparison.

[0022] In a further development of the present invention, it can be provided that, for determining the evaluation information, deviation ranges, in particular tolerance ranges describing the detection capability of the environmental sensor in the event of deviations, are used, each of which is assigned a value of the evaluation information, whereby it is checked whether a deviation detected during the comparison lies within a deviation range. The deviation can be determined in particular as a deviation angle, as already explained. For example, it is known for many environmental sensors to what extent a meaningful residual reflection still exists when the surface normal deviates from the visual normal, which can optionally also be verified or determined by experiments. This results in certain tolerance or...Reliability ranges exist, whereby, for example, a radar sensor can be considered to have very good reflection within a deviation angle range of + / - 5°, acceptable reflection within a deviation angle range of + / - 10°, poor reflection within a deviation angle range of + / - 15°, and no meaningfully measurable reflection at larger deviation angles. For example, corresponding numerical values ​​can be assigned to the individual deviation angle ranges, such as 0-5°, 5-10°, 10-15°, and >15°, which can then be translated into corresponding colors when color-coding the display.

[0023] According to the invention, the size of the surface sections for which evaluation information has been determined is weighted according to the evaluation information and summed to obtain a surface integral as at least a part of the evaluation value. For example, in the case of deviation ranges, each of which is assigned a value of the evaluation information (and optionally a color), this value of the evaluation information can be used for weighting the corresponding surface section. In the case of a color-coded representation, it can therefore be said, for example, that the surface elements or surface sections of the different colors are summed, with weighting information being applied depending on the value of the evaluation information or color. Advantageously, in this context, it can be provided that the surface integral is normalized to the total surface area of ​​the object described by the second model.Furthermore, it is conceivable, as is common in electrical engineering, to specify this standardized measure, which forms the evaluation value, as a logarithmic measure, i.e., as an area measure of detection in dB, or as a detection probability in percent. This facilitates the comparison of different installation positions on or in the vehicle and, in particular, as will be discussed in more detail below, also simplifies the definition of an objective function for automatic, optimized installation position detection.

[0024] A particularly advantageous embodiment of the present invention provides that an optimal installation position is determined using an optimization procedure that varies the test installation position and maximizes the evaluation value. In other words, an optimization algorithm is used that aims to maximize the evaluation value, thus finding an installation position optimized with regard to the detection capability of the environmental sensor. Accordingly, to find the best detection performance, the environmental sensor is virtually tilted, shifted, and / or rotated in its installation position. Depending on the shape of the object to be detected, as described by the second model, such a variation of the installation position can also result in a variation of the detection rate, and thus also of the evaluation value.For example, certain surface sections with good detection performance may be added, while others may be less effective or contribute less to the detection. It should be noted that the height of the environmental sensor relative to the object being detected can also lead to significant differences in detection.

[0025] This results in a classic optimization problem: finding the best mounting position for object detection along multiple axes and coordinates. In a specific implementation, the objective function of the optimization procedure can be formulated to maximize the evaluation value determined as a weighted surface integral or using one or more such surface integrals. It is also possible, of course, to logarithmize the evaluation value or express it as a percentage, as previously explained, particularly after normalization. In general terms, the goal of the optimization procedure described here is to find a maximum of reflective surfaces when the rotation, tilt, and translation—in short, the mounting position—of the environmental sensor is changed.

[0026] The positioning of the second model in the detection position relative to the first model can preferably be automated, for example, based on specifications oriented towards real-world physical conditions, but manual positioning is also possible in principle. With automated specifications, entire relative trajectories, i.e., successive detection positions, can be defined for the second model in relation to specific objects to be detected. This is particularly relevant when the second model represents another road user, especially an overtaking and / or overtaken vehicle, and the like. In addition to other road users, other objects to be detected can, of course, be used as further road users if the functionality of the environmental sensor is intended to allow this, for example, static obstacles such as trees, houses, walls, guardrails, and the like.

[0027] When manually positioning the second model in the computational space, it can be advantageous to display the detection range of the environmental sensor in at least one test installation position within a representation of the computational space. Particularly when using CAD software, it can be possible to display the field of view, and thus the detection range, of the environmental sensor relative to its current installation position, so that the object to be detected, modeled by the second model, can be positioned within the detection range. CAD software also offers suitable functions for defining and displaying the detection range. In this context, it can also be advantageous to determine a cumulative detection range for a group of installation positions and / or a range of installation positions and use it for the manual and / or automatic positioning of the second model in the computational space.For example, if, in an optimization procedure, shoring positions in a specific shoring position area or in a specific group of shoring positions are to be examined, it may be useful to consider a cumulative detection area.

[0028] Advantageously, the visual normal can be defined as at least one surface normal of a front surface of a model of the environmental sensor and / or of a surface portion of the first model corresponding to the surface of the environmental sensor in its installation position. For example, it is easily possible to determine normals on surfaces in CAD software, particularly as a central visual normal, for instance, in the case of a cone-shaped emission of the transmitted signal. The CAD software can, for example, provide a corresponding function that, upon selection of the front surface of the (virtual) environmental sensor or the partial surface of the first model corresponding to its installation position (which could, for example, be color-coded), displays the visual normal.

[0029] A particular advantage is that the first model can be assigned spatially resolved installation space information that describes the available installation space on the component. This installation space information is then used as an additional evaluation criterion, especially in the optimization process. Since the same installation spaces for an environmental sensor do not always exist at different locations on vehicle components, optimization with respect to installation space is possible, particularly within the framework of an optimization process. This is provided that installation space information is available in the corresponding optimization algorithm, specifically assigned to different areas of the first model that are potential installation positions. The optimization task, and thus the objective function, can then be formulated not only to find the optimal installation position but also to utilize the installation space in the best possible way.Appropriate weightings can also be applied for this purpose.

[0030] Advantageously, the evaluation information is determined for different detection positions and / or secondary models, whereby the evaluation information and / or evaluation values ​​are weighted differently for different detection positions and / or secondary models when evaluating a test installation position. It can therefore be intended, on the one hand, to incorporate the detection of different objects, represented as different secondary models, with different weightings into an overall evaluation for selecting an optimal installation position, particularly within the framework of an optimization process. For example, a radar sensor, as an environmental sensor, should preferably detect other vehicles, but does not necessarily have to detect pedestrians – depending on the functional requirements. A lidar sensor, on the other hand, should primarily detect pedestrians and small obstacles.This can be addressed by weighting the objects to be detected as described by the second models. However, using different detection positions for the objects to be detected also proves extremely useful. For example, if another vehicle is to be tested and evaluated for detectability in a test installation position, the vehicle can be positioned relative to the first model at various angles – rear, side, front, etc. – to assess detectability in different positions, which are based on physical reality, i.e., actual traffic situations. Weighting can also be applied here if necessary. In particular, as already explained, a predetermined number of second models, detection positions, and, if applicable, detection trajectories can be defined.It should be noted that a detection trajectory, i.e., a sequence of detection positions to be used, can also describe, for example, a rotation of the object described by the second model to the motor vehicle, i.e., the first model.

[0031] As previously explained, an environmental sensor designed for monostatic or bistatic measurement can be considered. If a bistatic or even multistatic environmental sensor is used for detection, the corresponding visual standards necessary for bistatic or multistatic object detection by the environmental sensor must be applied. Therefore, the method can universally evaluate monostatic, bistatic, or multistatic object detection methods.

[0032] The invention further relates to a computing device configured to carry out the method according to the invention. For this purpose, the computing device expediently comprises at least one processor and at least one storage medium. For example, CAD software can be stored in the storage medium and executed by the processor to implement at least partial steps of the method according to the invention. Corresponding functional units can be used to implement the individual steps. The computing device can be part of a development system for motor vehicles, in particular a design system, which can be used, for example, in an early development phase. All aspects relating to the method according to the invention can be applied analogously to the computing device according to the invention, with which the aforementioned advantages can therefore also be obtained.

[0033] A computer program according to the invention can, for example, be loaded into a storage medium of a computing device and includes program means for carrying out the steps of a method according to the invention when the computer program is executed on the computing device. The computer program can be stored on an electronically readable data carrier according to the invention, which thus comprises electronic control information that includes at least one computer program according to the invention and is designed such that, when the data carrier is used in a computing device, the steps of a method according to the invention are carried out. The electronically readable data carrier according to the invention can, in particular, be a non-transient data carrier, for example, a CD-ROM.

[0034] Further advantages and details of the present invention will become apparent from the exemplary embodiments described below and from the drawings. These show: Fig. 1 a sketch of the monostatic measurement method, Fig. 2 a sketch of the bistastatic measurement method, Fig. 3 a flowchart of an embodiment of the method according to the invention, Fig. 4 a first colored representation of a second model, Fig. 5 a second colored representation of a second model, Fig. 6 a third colored representation of a second model, and Fig. 7 a computing device according to the invention.

[0035] Fig. Figure 1 illustrates the visual normal in the monostatic measurement method with an environmental sensor 1 (only indicated here) and an object 2 to be detected (also indicated here). The environmental sensor could be, for example, a radar sensor or a lidar sensor. Signals 3 and 4 emitted by the environmental sensor 1 strike the surface of the object 2 and are reflected there, particularly according to the rule "angle of incidence = angle of reflection". Corresponding reflected signals are generated, as shown by arrows 5 and 6. Since, in this case, both the transmitter and the receiver of the environmental sensor 1 are located at the same point, the signals reflected back in the same direction (arrow 5) from the surface section 7 are received by the environmental sensor 1, allowing a reflection to be measured there. However, the reflected signals shown in line 6 are not received.The visual normal 8 (indicated by a dashed line) corresponds here to the direction of the transmitted signal. If it coincides with the surface normal of surface section 7, the corresponding back reflection occurs. It should be noted, of course, that a certain tolerance range usually exists, for example, + / - 5°, within which the surface normal can deviate from the visual normal 8 in order to still ensure good detectability.

[0036] Fig. Figure 2 shows the case of a bistastatic measurement method, here with a two-part environmental sensor 1a, 1b, where sub-sensor 1a is the transmitter and sub-sensor 1b is the receiver. The visual normal 8 clearly deviates from the direction of the transmitted signal 3, which is received as a reflected signal by sub-sensor 1b, as it forms the angle bisector between the transmitted signal 3 and the received signal 5. If this corresponds to the surface normal, maximum detection performance is achieved.

[0037] Fig. Figure 3 shows a flowchart of an embodiment of the method according to the invention, in which an optimal installation position for an environment sensor 1, 1a, 1b, for example a radar sensor or a lidar sensor, is to be determined on a component, for example a body part, of a motor vehicle, and thus in particular an installation position, i.e. installation position and installation orientation, in which the best possible detection performance for objects 2 to be detected is given.

[0038] Ultimately, it is irrelevant whether the embodiment described here is used for a monostatic measurement method, a bistatic measurement method, or even a multistatic measurement method.

[0039] In step S1, a computational space is defined for a specific test installation position. This space is initially derived from a first model of the component, which, based on the test installation position, also defines a virtual environmental sensor and its detection range, as well as its line-of-sight normal distribution given a cone-shaped radiation pattern of the transmitted signal. Within this detection range, a second model, describing the surface of an object 2 to be detected, is then positioned. Both the second model, particularly based on the object 2 to be detected, and the corresponding detection position, for example, as part of a sequence of detection positions (detection trajectory), can be predefined, especially with regard to the intended application of the environmental sensor.For example, if the environment sensor is to detect other motor vehicles next to the vehicle itself as a radar sensor, corresponding second models and detection positions can be provided, for example a sequence of detection positions in the sense of a detection trajectory, which is intended to depict that the other motor vehicle as object 2 to be detected passes by the vehicle itself.

[0040] The computational space is provided by CAD software, which is generally known in the prior art. This is because the embodiment described here is implemented at an early stage in the development of the motor vehicle, a stage at which such CAD software is frequently used anyway.

[0041] In step S2, the visual normals 8 in the virtual installation position of the environmental sensor 1, 1a, 1b within the computation space are first determined, in particular depending on the size and division of the surface sections 7 to be considered of the virtual surface of the object 2 to be detected defined by the second model as a visual normal field with conical emission.

[0042] In step S3, a demolding analysis function of the CAD software is used to determine the surface normals in the portion of the virtual surface of the second model contained within the detection area according to the installation position and to compare them with the corresponding visual normal 8. Various deviation ranges, in this case deviation angle ranges, are defined, to which different values ​​of an evaluation information are assigned. Depending on the deviation angle range in which the angle between the surface normal and the visual normal lies, the corresponding evaluation information is assigned to the surface section 7.For example, angular deviations in the range of + / - 5° can fall within a deviation angle range for very good detection capability, deviation angles between 5° and 10° within a deviation angle range for medium reflection and thus detection capability, and deviation angles between 10° and 15° within a range for poor detection capability, while deviation angles in a deviation angle range >15° (in magnitude) cannot be assigned to any meaningfully detectable reflection. Furthermore, colors are assigned to the respective deviation angle ranges or values ​​of the evaluation information to create an explanatory representation for a user within the CAD software, which is optional within the scope of the procedure described here and with regard to the... Fig. 4-6 will be explained in more detail.

[0043] The Fig. Figures 4-6 show colored surface areas (represented with different hatching patterns) according to the evaluation information. Figure 13 shows a second model of another vehicle, for example, a side section in Fig. 4, a rear section in Fig. 5 and a front section in Fig. 6. For example, area 9 can be colored green for excellent reflection properties, such as a deviation of the surface normal from the visual normal 8 within a deviation angle range of -5° to +5°. Areas 10 can be colored yellow, for example, for an angular deviation of 5° and 10°, areas 11 can be colored pink, for example, for an angular deviation of 10° to 15°, and areas 12 can be colored red, for example, for an angular deviation greater than 15°. Areas 12 can thus be understood as exclusion areas where no significant reflection is to be expected.

[0044] Returning to Fig. Step S4 checks whether further evaluation information should be determined for additional detection positions of the current second model and / or other second models, and thus in particular for other detectable objects 2. If so, the computation space is adjusted accordingly in step S1, and the process is carried out as described above in steps S2 and S3.

[0045] Once all evaluation information has been determined for a test installation position, an evaluation value is calculated for this test installation position in step S5. For this purpose, the surface sections assigned to the different deviation angle ranges are weighted according to their corresponding evaluation information and combined into a weighted surface integral for each detection position and every second model. These individual evaluation values ​​can then be combined into an overall evaluation value, whereby different detection positions and / or different second models, which are then assigned to different detectable objects 2, can also be weighted differently.

[0046] In step S6, it becomes clear that the determination of the evaluation value for a test installation position is integrated into an optimization procedure, implemented by an optimization algorithm, after an objective function is evaluated. This objective function attempts to maximize the evaluation value and thus the weighted surface integrals. However, further evaluation criteria can also be included in the objective function if, for example, different test installation positions are assigned installation space information, which might have a higher value the more installation space is available or the more preferred the installation space is for the environmental sensor. Ultimately, installation space properties of the vehicle component can also be considered within the objective function.

[0047] In step S7, it is checked whether a termination condition for the optimization process has been met, for example, if the deviation of the objective function value from the previous value is less than a threshold and / or a maximum number of optimization steps have been performed. If the termination criterion is not met, the test installation position is varied in step S8, for which the optimization algorithm may contain a corresponding change instruction. With the changed test installation position, the process continues from step S1 and a new overall evaluation value is derived.

[0048] If the abort criterion in step S7 is met, the current test installation position is output as the installation position to be used for the environmental sensor 1, 1a, 1b.

[0049] In other embodiments, variations of the method according to the invention are also conceivable, for example, partially manual intermediate steps and the like. In particular, it is possible, for example, to carry out the optimization with regard to a suitable installation position manually, by, for example, selecting installation positions by hand and representations as in the Fig. 4-6 are generated so that a user can assess the quality of the individual installation positions.

[0050] Fig.Figure 7 shows a schematic diagram of a computing device 14 according to the invention, which is part of a development system 15 for motor vehicles. The computing device 14 comprises a storage medium 16 and at least one processor 17. It is designed to carry out the method according to the invention. In particular, the CAD software 18 used can be stored in the storage medium 16, as well as data 19 relating to various first and second models.

[0051] The computing unit 14 may, in particular, include an evaluation unit for carrying out steps S1-S5 and an optimization unit for higher-level optimization (steps S6, S7 and S8). A display device and an input device are expediently assigned to the computing unit 14.

Claims

[1] Method for determining the installation position of an environment-monitoring sensor (1, 1a, 1b), in particular a radar sensor or lidar sensor, measuring by reflection from objects to be detected, in a motor vehicle, comprising the following steps: - Providing a geometric first model of at least one component of the motor vehicle on which the environmental sensor (1, 1a, 1b) is to be installed, - Providing a second geometric model in at least one detection position relative to the first geometric model in a computational space, wherein the second model describes the surface of an object (2) to be detected, - for several test installation positions of the environmental sensor (1, 1a, 1b) on the first model in the computational space, determination of a surface orientation that enables a maximum sensor signal of the environmental sensor (1, 1a, 1b) in the test installation position, describing a visual normal (8) of the environmental sensor (1, 1a, 1b) for at least one surface section (7) of the second model as a function of the test installation position and an evaluation information for the surface section (7) as a function of a comparison of a surface normal of the surface section (7) with the visual normal (8), - Display of the evaluation information and / or an evaluation of the test installation positions based on their evaluation information by determining an evaluation value for each test installation position, wherein the size of the surface sections (7) for which an evaluation information was determined is weighted on the basis of the evaluation information to obtain a surface integral as at least a part of the evaluation value. [2] Method according to claim 1, characterized by , that the computation space is provided by CAD software (18). [3] Method according to claim 2, characterized by , that the determination of the surface normals of at least one surface section (7) and / or the comparison with the visual normal (8) is carried out using a demolding analysis function of the CAD software. [4] Method according to any of the preceding claims, characterized by, that to display the evaluation information a rendered representation of the second model is generated, wherein the evaluated surface sections (7) are displayed differently according to their evaluation information, in particular colored differently. [5] Method according to any of the preceding claims, characterized by , that to determine the evaluation information deviation ranges, in particular the detection capability of the environmental sensor (1, 1a, 1b) in case of deviations, tolerance ranges describing deviations, to which a value of the evaluation information is assigned, are used, whereby it is checked whether a deviation found in the comparison lies within a deviation range. [6] Method according to any of the preceding claims, characterized by , that the surface integral is normalized to a total surface area of ​​the object (2) described by the second model. [7] Method according to any of the preceding claims, characterized by , that an optimal installation position is determined in an optimization procedure that varies the test installation position and maximizes the evaluation value. [8] Method according to any of the preceding claims, characterized by , that for manual positioning of the second model in the computation space the detection range of the environmental sensor (1, 1a, 1b) is displayed in a representation of the computation space in at least one test installation position. [9] Method according to any of the preceding claims, characterized by , that the first model is assigned spatially resolved installation space information that describes the available installation space at the component, whereby the installation space information is used as a further evaluation criterion in the evaluation, especially in the optimization procedure. [10] Method according to any of the preceding claims, characterized by, that the determination of the evaluation information is carried out for different detection positions and / or second models, whereby when evaluating a test installation position, a different weighting of the evaluation information and / or evaluation values ​​is carried out for different detection positions and / or second models. [11] Method according to any of the preceding claims, characterized by , that the environment sensor (1, 1a, 1b) measures monostatically or bistastatically. [12] Computing device (14) configured to perform a method according to any of the preceding claims. [13] Computer program which performs the steps of a method according to any one of claims 1 to 11 when executed on a computing device (14). [14] Electronically readable data carrier on which a computer program according to claim 13 is stored.

Images