System tolerance determination of templates for batch products

Abstract

Disclosed is a computer-implemented method for determining a tolerance region around a target parameter sample of a parameterized simulation model of a product, in particular a steer-by-wire system and / or a highly automated driving steering system, in an ODD (Operational Design Domain), the method includes determining a tolerance region around a target parameter sample in the ODD based on a maximum allowable distance and a difference metric, where the difference metric defines a parameterized simulation model allocated to the target parameter sample and a distance between each parameterized simulation model.

Description

Background Technology

[0001] Mass-produced products, such as steering systems (especially high-volume products), exhibit variations in product parameters (e.g., friction, elasticity, and / or inertia) due to manufacturing tolerances and errors. Furthermore, mass-produced products experience additional parameter variations due to aging (e.g., wear and / or environmental influences). The range and combinations of values ​​for these parameter variations in mass-produced products in practice constitute their so-called "operational design domain" (ODD).

[0002] Due to parameter variations throughout the ODD and simplifications during model formation, each simulation model for (batch) products will exhibit deviations between modeled and actual behavior. These simulation models, including model biases and / or model uncertainties throughout the ODD, form the basis for simulation-based product approval. The characterization of model uncertainties required for this purpose is currently typically based on selected product prototypes, as a complete characterization is usually too complex. In the following text, the target version (i.e., the ideal version) of the prototype is referred to as the product template, and its physical realization is called the product sample. Typically, the product template used to characterize uncertainties is selected based on expert opinion.

[0003] However, due to limitations in manufacturing precision, it is impossible to precisely replicate the chosen (ideal) product template in practice, resulting in slightly different system behaviors between the product template and the associated product sample. Typically, the acceptable parameter tolerances used to achieve the product template are defined through expert knowledge. Alternatively, one can seek to manufacture and / or post-process the product sample as accurately as possible. However, this does not preclude setting and / or manufacturing overly lenient parameter tolerances, which could lead to significant deviations in system behavior between the product template and the product sample. On the other hand, it also does not preclude setting and / or manufacturing overly stringent parameter tolerances, whereby the system behavior of the product template and the product sample may be virtually identical, but this could result in unnecessarily high manufacturing costs.

[0004] The problem to be solved by this disclosure may be, for example, providing a method to determine how strong the deviation between a product sample and a pre-given product template is, so as to still be able to fully realize the system behavior of the product template.

[0005] Compared to traditional steering systems, steer-by-wire (SbW) steering systems and / or highly automated driving (HAD) steering systems have more stringent regulatory requirements for product approval. To ensure that real-world testing and trial costs do not increase significantly more than for traditional steering systems due to the stricter approval requirements for (mass-produced) SbW and HAD steering systems, the industry is focusing on simulation-based approval processes. For this simulation-based approval, it is crucial that the validated and verified simulation model of the steering system has known model uncertainties.

[0006] Within the scope of the company's internal simulation-based approval process for SbW and / or HAD steering systems, model uncertainties should be characterized based on a selection of product samples.

[0007] Therefore, the upstream problem that needs to be addressed may be to systematically select associated product templates using model-based criteria, such that the product templates represent the entire ODD of a (large) batch steering system with quantifiable residual uncertainties.

[0008] Currently, there is no systematic method to determine the parameter tolerances that are meaningful for realizing the product template, thereby ensuring that the quantitative differences in system behavior between the product template and the product sample are negligible, thus avoiding tolerances that are too strict or too loose.

[0009] In systems theory, various metrics are known to quantify the differences between two systems (systems and products can be equated below) and thus compare them. The gap metric, ν-gap metric, and L2 metric are explained below.

[0010] The gap metric uses a scalar in the real number interval [0,1] to quantify the difference in the uncontrolled (open-loop) input / output behavior of two systems P1 and P2 relative to their stability and performance properties in controlled operation (closed-loop). A metric result close to 0 means that the two systems are very similar, and any controller that can stabilize P1 can also stabilize system P2 with similar controlled performance. A metric result of 0 means that the systems under consideration, P1 and P2, behave exactly the same. On the other hand, a metric result close to or equal to 1 indicates that systems P1 and P2 are very different. Furthermore, the gap metric can describe the robust stability of closed-loop control loops with model uncertainties. An explicit controller design is required to evaluate the gap metric. For a detailed definition and properties of the gap metric, see Chapter 17 of Kemin Zhou and John C. Doyle's "Essentials of Robust Control" (1st edition, Pearson, 1997, ISBN: 9780135258332).

[0011] The theoretical formulation and meaning of the ν-gap metric are very similar to those of the gap metric, but the definitions of the two metrics are quite different. To evaluate the ν-gap metric, controller design is not required; instead, a turns-by-turns study of the systems to be compared, P1 and P2, is necessary. For the definition and properties of the ν-gap metric, see Chapter 17 of Kemin Zhou and John C. Doyle's "Essentials of Robust Control" (1st edition, Pearson, 1997, ISBN: 9780135258332), or Glenn Vinnicombe's "Frequencydomain uncertainty and the graph topology" (IEEE Transactions on Automatic Control, Vol. 38, No. 9, pp. 1371-1383, September 1993, DOI: 10.1109 / 9.237648).

[0012] The definition of the L2 metric corresponds to the ν-gap metric, but does not include turns analysis. Therefore, while the two metrics are similar in principle and meaning, the theoretical persuasiveness of the L2 metric is lower. For a detailed definition and properties of the L2 metric, please refer to Chapter 17 of Kemin Zhou and John C. Doyle's "Essentials of Robust Control" (1st edition, Pearson, 1997, ISBN: 9780135258332).

[0013] The three proposed difference measures all have the following properties: These metrics, based on uncontrolled input / output behavior, quantify the differences in stability and performance attributes between two systems under controlled operation.

[0014] - The measurement results can be used to determine whether two systems P1 and P2 are similar enough that P1 can be regarded as a representative of P2.

[0015] - The measurement result can also be interpreted as the (system theory) distance between the compared systems.

[0016] - For all systems to be compared, the following rule applies: L2 result ≤ ν - gap result ≤ gap result. Summary of the Invention

[0017] A first general aspect of this disclosure relates to a computer-implemented method for determining a tolerance region around a sample of target parameters of a parameterizable simulation model of a product in an Operational Design Domain (ODD). The product may be, for example, a steer-by-wire system and / or a steering system for highly automated driving.

[0018] The method includes determining a tolerance region around the target parameter sample in the ODD based on the maximum permissible distance and a difference metric, wherein the difference metric defines the distance between the parameterized simulation model assigned to the target parameter sample and each parameterized simulation model.

[0019] The second general aspect of this disclosure relates to a computer system configured to perform a computer-implemented method, according to the first general aspect (or its implementation), for determining a tolerance region around a target parameter sample of a parameterizable simulation model of a product in an operational design domain (ODD).

[0020] The third general aspect of this disclosure relates to a computer program configured to perform a computer-implemented method, according to the first general aspect (or its implementation), for determining a tolerance region around a target parameter sample of a parameterizable simulation model of a product in an operational design domain (ODD).

[0021] The fourth general aspect of this disclosure relates to a computer-readable medium or signal that stores and / or contains a computer program according to the third general aspect (or its implementation).

[0022] Using the method presented here according to the first general aspect (or its implementation), tolerance regions can be systematically determined separately for the realization (i.e., physical manufacturing) of a product template on a model-based standard, thereby matching the systematic behaviors of the product template and the product sample, except for permissible quantitative differences. Specifically, parameter tolerances for the product template can be determined. The product template to be manufactured is assigned to a target parameter sample, thus becoming the target version (i.e., the ideal version) of the prototype. In the realization of the product template, the product sample can now be constructed as a prototype, with its parameter sample located within the tolerance region surrounding the target parameter sample.

[0023] In particular, the following advantages compared to the prior art can be achieved by the method according to the first general aspect (or its implementation): - A systematic approach for determining individual parameter tolerances (i.e., the allowable range of values ​​for one or more parameters) for implementing a product template; - It can calculate the individual tolerance range (i.e., the allowable range for all parameters) for all product templates. -Quantifiable differences in systematic behavior between product templates and product samples due to mismatched parameters throughout the ODD; - Search for the possibility of local product parameters or combinations of product parameters that have a particularly strong influence on the system behavior of the product template near the product template.

[0024] The method proposed here according to the first general aspect (or its implementation) can be used to develop real (i.e., physical) products (e.g., steer-by-wire (SbW) systems) during the design phase and / or system development phase (i.e., after the design phase).

[0025] For example, during the design phase or system development, this invention can be used to determine individual tolerance regions for the (product) parameters that realize all product templates (e.g., characterize model uncertainties), while preserving acceptable differences in system behavior between product templates and product samples. Specifically, individual parameter tolerances for all product templates to be realized can be calculated, with acceptable differences in system behavior maintained separately under these tolerances. This is crucial because prototypes whose parameters correspond perfectly to target parameter samples are difficult or can only be manufactured and / or modified (e.g., gap-filling) at considerable expense.

[0026] Furthermore, for example, during the design phase, by means of intermediate results from the method according to the first general aspect (or its implementation) and additional mathematical studies, local product parameters and / or combinations of product parameters that have a particularly strong influence on the system behavior near the product template can be found.

[0027] The target parameter sample can be one of a large number of representative parameter samples in the ODD, that is, one of a large number of parameter samples representing the entire ODD with quantifiable residual uncertainties (in batches). The realized product template (more precisely, the product sample) is the representative product sample. On the other hand, the target parameter sample does not have to be a representative parameter sample; that is, the target parameter sample can be arbitrarily or for other reasons (e.g., as a limit template) pre-given to achieve the target.

[0028] Based on representative product samples, uncertainties between the actual product behavior and its modeled behavior throughout the entire ODD can be systematically characterized in further system development. These uncertainties are caused by parameter variations and simplifications during modeling. Subsequently, the representative product samples, including the characterized model uncertainties, can be used for product control development and / or product approval.

[0029] The method presented herein, according to the first general aspect (or its implementation), can be applied in particular to the scope of processes used for approving SbW steering systems and HAD steering systems. In this case, implementing a product template representing the entire ODD can be used for the verification and / or validation of the steering system.

[0030] The method presented here according to the first general aspect (or its implementation) can be performed entirely or partially in an analytical manner. This avoids sampling a large number of parameter samples around the target parameter sample. In implementations, the tolerance region can sometimes be precisely determined. Attached Figure Description

[0031] Figure 1 An exemplary implementation of a computer-based method is illustrated, which is used to determine a tolerance region around a sample of target parameters of a parameterizable simulation model of a product (particularly a steer-by-wire system and / or a steering system for highly automated driving) in an operational design domain (ODD).

[0032] Figure 2 An exemplary implementation of the method is shown.

[0033] Figure 3a It shows Figure 3b The exemplary ODD in the example contains a large number of exemplary and arbitrary parametric simulation models.

[0034] Figure 3b An exemplary ODD with a sample of target parameters surrounded by a tolerance region is shown. Detailed Implementation

[0035] The method 100 presented in this disclosure aims to determine a tolerance region 30 around a target parameter sample 10 of a parameterizable simulation model of a product (particularly a steer-by-wire system and / or a steering system for highly automated driving) in the Operational Design Domain (ODD). Alternatively or additionally, method 100 can also be used to manufacture a product template in the form of a product sample that implements the product template.

[0036] Method 100 is performed entirely or partially analytically, provided that the parameterizable simulation model exists in analytical form.

[0037] First, a computer-implemented method 100 is disclosed for determining a tolerance region 30 around a sample 10 of target parameters of a parameterizable simulation model of a product in an Operational Design Domain (ODD). This product can be, in particular, a batch product, i.e., one that can be mass-produced. Method 100 is also applicable to non-batch or small-batch products, and is particularly suitable for situations where a large number of the same type of product should be manufactured, but these products may vary (e.g., due to production and / or material reasons). A large number of the same type of product can include, for example, >1e5 products per year, >5e5 products per year, or >1e6 products per year.

[0038] For example, the product could be a steer-by-wire system. Alternatively or additionally, the product could be a steering system for automated driving, particularly highly automated driving.

[0039] For example, Figure 1As shown, method 100 includes determining a tolerance region 30 around the target parameter sample 10 in 140 ODDs based on the maximum permissible distance and a difference metric, wherein the difference metric defines the parameterized simulation model assigned to the target parameter sample 10 (which is always the same for the corresponding parameterized simulation model). Figure 2 The distance between the template PUM and each parameterized simulation model.

[0040] Difference metrics can be based on gap metrics, ν-gap metrics, and / or L2 metrics. Specifically, difference metrics can be gap metrics, ν-gap metrics, or L2 metrics. Alternatively, difference metrics can be based on a combination of gap metrics, ν-gap metrics, and / or L2 metrics. A difference metric can output a quantitative measure of the difference between a pair of parametric simulation models, i.e., a quantitative measure of how similar or dissimilar the two parametric simulation models in the pair are. In this respect, difference metrics can also be called similarity metrics or comparative metrics. The quantitative measure output by the difference metric can be called distance. If the two parametric simulation models in a pair are similar, the distance may be small. In particular, if the two parametric simulation models in a pair are identical (i.e., maximally similar), the distance can be zero. Conversely, if the two parametric simulation models in a pair are dissimilar, the distance can be large.

[0041] Figures 3a-3b An exemplary result is shown for determining the tolerance region 30 around the target parameter sample 10.

[0042] The target parameter sample 10 and tolerance region 30 can define the product template to be manufactured, which will also be referred to as the target product template below.

[0043] For example, Figure 1 As illustrated schematically as an option, method 100 may further include: outputting 150 requirements for the production of a product template to be manufactured based on target parameter sample 10 and tolerance region 30, wherein the requirement is met if (e.g., all) parameters of the product template to be manufactured are within tolerance region 30. Figure 2 In an exemplary implementation, step 150 is referred to as “converting the tolerance region into a separate parameter tolerance”.

[0044] Based on this requirement, production (e.g., prototyping) can manufacture product samples. This requirement may include, or may be, an algorithm that checks whether one or more parameters of the product template to be manufactured (i.e., the product sample) are within a tolerance region 30 surrounding the target parameter sample 10. For example, such an algorithm may be useful when the tolerance region is high-dimensional and deviates significantly from the product space (i.e., the direct product with intervals).

[0045] For example, the tolerance region 30 in the 140 ODD can be determined as follows: within the tolerance region 30, all distances between the parameterized simulation model and the target parameter sample 10 are less than the maximum permissible distance. Alternatively, the tolerance region 30 in the 140 ODD can also be determined as follows: within the tolerance region 30, all distances between the parameterized simulation model and the target parameter sample 10 are less than or equal to the maximum permissible distance. For example, it is also conceivable that the tolerance region 30 in the 140 ODD can be determined as follows: within the tolerance region 30, all distances between the parameterized simulation model and the target parameter sample 10 are at most approximately the maximum permissible distance.

[0046] Furthermore, the tolerance region 30 in the 140 ODD can be determined in such a way that the tolerance region 30 is precisely or approximately maximized. For example, the tolerance region 30 can be maximized based on the Lebesgue metric. The tolerance region can be (but does not necessarily have to be) a region in a mathematical sense.

[0047] Furthermore, the tolerance region 30 in the 140 ODD can be determined as follows: the tolerance region 30 is a connected set in the ODD. Alternatively or additionally, the tolerance region 30 can be a convex set in the ODD. In particular, the tolerance region can be a convex connected set in the ODD.

[0048] For example, the tolerance region 30 can be a manifold, a polyhedron (especially a convex polyhedron), a hyperelliptic, or a hypercube. A hypercube is particularly advantageous because its parameters and tolerances are independent of each other. If this is not the case, the tolerance region 30 can be defined as an (arbitrary) manifold. To reduce the memory requirements of such a manifold, approximating it with a polyhedron or hyperelliptic may be meaningful.

[0049] An ODD can be defined through numerical and / or analytical descriptions. A parameterized simulation model assigned to the target parameter sample 10 can be defined by evaluating a parameterizable analytical simulation model on the target parameter sample 10 in the ODD. For example... Figure 1 As illustrated schematically and as an option, method 100 may further include: forming a parameterized simulation model 120 assigned to the target parameter sample 10, wherein a parameterizable analytical simulation model is evaluated for the target parameter sample 10 in the ODD.

[0050] The corresponding parametric simulation model can be defined by evaluating parameter samples in the product's ODD (Optical Design Datasheet). For example... Figure 1 As illustrated and as an option, method 100 may further include: forming a corresponding parameterized simulation model 121, wherein a parameterizable analytical simulation model is evaluated for (arbitrary) parameter samples in the ODD.

[0051] Method 100 or step 140 may further include transforming the parameterizable analytical simulation model to obtain an associated parameterized simulation model for any ODD sample (e.g., target parameter sample 10, or parameter samples of the corresponding parameterized simulation model). Figure 2 In an exemplary implementation, this step is referred to as “converting PAM to PUM at each ODD point”, where PAM represents a parameterizable analytical simulation model, PUM represents a parameterized simulation model of the parameterizable analytical simulation model, and ODD points represent (arbitrary, corresponding) parameter samples in the ODD.

[0052] Method 100 or step 140 may further include establishing a difference measure function for any pair consisting of the parameterized simulation model assigned to the target parameter sample 10 and the corresponding parameterized simulation model in the entire ODD (or a portion thereof). Such a function may be a concatenation of difference measures on two transformed parameterizable analytical simulation models. Figure 2 In an exemplary implementation, this step is called “establishing a metric function between all PUMs and the template PUM”, where each metric function represents a difference metric function.

[0053] exist Figure 2 In an exemplary implementation, the tolerance region around the target parameter sample 10 in the 140 ODD is actually determined based on the maximum permissible distance and the difference metric and is referred to as the “calculated tolerance region”.

[0054] The target parameter sample 10 can be one of a large number of target parameter samples representing the ODD. In this case, the parameter sample 20 of the corresponding parameterized simulation model can be located within the representative range of the target parameter sample 10, that is, limited to a portion of the ODD.

[0055] For example, Figure 1 As illustrated and as an option, method 100 may also include identifying one or more parameters within the 160 tolerance region 30 that have a significant impact on the product. This can be done, for example, through sensitivity analysis. This can improve the understanding of the product and its behavior. This is particularly useful in the case of high-dimensional ODDs, as it is no longer possible to directly examine the tolerance region 30. This can improve the understanding of the system within the tolerance region. These results can be taken into account when implementing the target product template.

[0056] Method 100 may include determining a tolerance region around each of the 140 target parameter samples 10 in the ODD. Thus, in this case, an associated tolerance region can be determined around each target parameter sample in the ODD.

[0057] Method 100 may include setting a maximum permissible distance. Figure 2 In the exemplary embodiment shown, this step is referred to as "setting the maximum permissible distance". In the case of a large number of target parameter samples, this setting can be performed globally (i.e., for all target parameters) or individually for target parameter samples within a large number of target parameter samples.

[0058] As already discussed, tolerance region 30 can be used to identify one or more product samples whose parameters lie within a tolerance region surrounding a target parameter sample. As an alternative to or supplement to the possibilities of output 150 already discussed, method 100 may include comparing a parameter sample of a real product sample stored in a database with the tolerance region, particularly based on similarity relationships. If the similarity is unsatisfactory, one or more instructions may be output specifying how a sufficiently satisfactory product template can be manufactured from the real product template (e.g., through assembly, gap filling, etc.). This product template may be, but is not necessarily, a representative product template.

[0059] Product samples that realize the product template to be manufactured can be used to design the control system of the product. The design of the product's control system can be based, in particular, on a large number of product samples, each realizing the product template to be manufactured. Specifically, if the product template to be manufactured represents an ODD (Original Design Specification), the design and / or layout of the product's robust control system can be improved. In particular, this allows for the design of the product's control system with exceptional robustness and reliability. This, in particular, can enhance the product's safety.

[0060] Furthermore, one or more requirements for the product can be checked based on product samples that realize the product template to be manufactured. Checking one or more requirements for the product can be done particularly well based on a large number of product samples that each realize the product template to be manufactured. This allows for robust and reliable product design. In particular, it can also improve product safety.

[0061] Figure 2 An exemplary implementation of method 100 is shown. In this case, for example, the following steps can be performed sequentially: - "Convert PAM to PUM at each ODD point" - "Establish a metric function between all PUMs and the template PUM" - "Set the maximum allowable distance" - "Calculate tolerance area" 140 - "Convert the tolerance region into a separate parametric tolerance" 150.

[0062] Figure 3a It shows Figure 3bThe exemplary ODD contains exemplary and arbitrary large numbers of parametric simulation models (more precisely: their parameter samples 20, represented in points). These large numbers of parametric simulation models are for illustrative purposes only.

[0063] Figure 3b An exemplary ODD is shown, in which target parameter sample 10 (represented by a cross) is surrounded by tolerance region 30.

[0064] In this example, the ODD is a two-dimensional product space, meaning it is expanded by two parameters (one in the x-direction and the other in the y-direction), each of which can take values ​​within an interval. Typically, the ODD does not have to be a product space; it can also be an (arbitrary) manifold. The dimension of the ODD can be arbitrary. Here, tolerance region 30 is an exemplary rectangle.

[0065] For this example, each parameter sample 20 in the ODD can be represented in 3D (i.e., in the z-direction orthogonal to the page plane) (more precisely, its parameterized simulation model, some of which are in...). Figure 3a The distance between the target parameter sample 10 (represented by a point in the diagram) and the parameterized simulation model (based on a difference metric) is shown. Conversely, in 2D, this is represented by four contour lines (e.g., 40) relative to the variable in the z-direction. The four solid lines each form a contour line, meaning they have the same z-value. In addition to the contour lines, the boundaries (dashed lines) of the influence range of the corresponding representative parameterized simulation model are also shown.

[0066] For example, in Figures 3a-3b In this process, another target parameter sample can be located to the left of the boundary (dashed line) of the influence range, or the tolerance area around the other target parameter sample can be determined according to method 100.

[0067] The present invention also discloses a computer system configured to perform a computer-implemented method 100 for determining a tolerance region 30 around a target parameter sample 10 of a parameterizable simulation model of a product in an operational design domain (ODD). The computer system may include a processor and / or working memory.

[0068] The present invention also discloses a computer program configured to perform a computer-implemented method 100 for determining a tolerance region 30 around a target parameter sample 10 of a parameterizable simulation model of a product in an operational design domain (ODD). The computer program may exist, for example, in an interpretable or compilable form. The computer program may be loaded (even partially loaded) into the RAM of a computer (e.g., a computer system) for execution, for example, as a bit sequence or byte sequence.

[0069] In addition, a computer-readable medium or signal for storing and / or containing the computer program is also disclosed. The medium may, for example, include RAM, ROM, EPROM, HDD, SSD, etc., wherein / the signal is stored thereon.

Claims

1. A computer-implemented method (100) for determining a tolerance region (30) around a parameterizable simulation model of a product, particularly a steer-by-wire system and / or a steering system for highly automated driving, in an operational design domain (ODD), the method comprising: -Based on the maximum permissible distance and the difference measure, determine the tolerance region (30) around the target parameter sample (10) in the ODD (140), wherein the difference measure defines the distance between the parameterized simulation model assigned to the target parameter sample (10) and each parameterized simulation model.

2. The method (100) according to claim 1, wherein, The target parameter sample (10) and the tolerance region (30) define the product template to be manufactured.

3. The method (100) according to claim 2, comprising: - The production requirements of the product template to be manufactured are based on the target parameter sample (10) and the tolerance region (30) output (150), wherein if the parameters of the product template to be manufactured are within the tolerance region (30), the requirements are met; Optionally, the requirement includes an algorithm that can check whether the parameters of the product template to be manufactured are within the tolerance region (30).

4. The method (100) according to any one of the preceding claims, wherein, The difference measure is based on gap measure, ν-gap measure and / or L2 measure.

5. The method (100) according to any one of the preceding claims, wherein, The tolerance region (30) in the ODD (140) is determined in such a way that each distance from the parameterized simulation model assigned to the target parameter sample (10) in the tolerance region (30) is less than the maximum permissible distance.

6. The method (100) according to claim 5, wherein, The tolerance region (30) in the ODD (140) is determined in such a way that the tolerance region (30) is maximized precisely or approximately, optionally wherein the tolerance region (30) is maximized based on the Lebesgue metric.

7. The method (100) according to any one of the preceding claims, wherein, The tolerance region (30) is a polyhedron, a hyperelliptic or a hypercube, wherein the polyhedron is in particular a convex polyhedron.

8. The method (100) according to any one of the preceding claims, wherein, The parameterized simulation model assigned to the target parameter sample (10) is defined by evaluating a parameterizable analytical simulation model of the target parameter sample (10) in the ODD.

9. The method (100) according to any one of the preceding claims, wherein, The corresponding parametric simulation model is defined by evaluating a parametric analytical simulation model based on parameter samples in the product's ODD.

10. The method (100) according to any one of the preceding claims, wherein, The target parameter sample (10) is one of a large number of target parameter samples representing ODD.

11. The method (100) according to claim 10, wherein, The parameter sample (20) of the corresponding parameterized simulation model is within the representative range of the target parameter sample (10).

12. The method (100) according to any one of the preceding claims, comprising: - Identify (160) one or more parameters in the tolerance region (30) that have a significant impact on the product.

13. A computer system configured to perform a computer-implemented method (100) according to any of the preceding claims for determining a tolerance region (30) around a target parameter sample (10) of a parameterizable simulation model of a product in an operational design domain (ODD).

14. A computer program configured to perform a computer-implemented method (100) for determining a tolerance region (30) around a target parameter sample (10) of a parameterizable simulation model of a product in an operational design domain (ODD) according to any one of claims 1 to 12.

15. A computer-readable medium or signal that stores and / or contains a computer program according to claim 14.

Images