A method for correcting a matrix, computer program code, and a storage medium

Abstract

The invention relates to a method for correcting a matrix 10 which describes a vehicle component of a vehicle. It comprises the steps of reading the matrix 10 via an interface of an electronic computi

Description

FIELD OF THE INVENTION

[0001] The present invention relates to the field of automobiles. More specifically, the present invention relates to a method for correcting a matrix which describes or characterizes a vehicle component, for example, a motor of a vehicle. Furthermore, the present invention relates to a corresponding computer program code and a computer-readable storage medium. BACKGROUND INFORMATION

[0002] To generate torque on an electric engine or an electric motor, a 2D control table may be used that provides information as to how much current to inject based on different conditions, such as the flux, which consists of voltage and speed, and the demanded torque. Collected from testing, the flux data and / or a flux table that may be determined by data from the electric motor may be condensed into a two-dimensional control table. This control table could be represented as a matrix.

[0003] Correcting a matrix is applicable in various settings. Whether one is correcting digital media or scaling and manipulating a matrix to reach a desired goal. A systematic way to scale the matrix with smoothness and non-disruptively would be helpful. SUMMARY OF THE INVENTION

[0004] There are two main challenges for scaling or correcting the matrix. First is a non-uniform scaling factor across multiple overlapping regions. One might want to increase or decrease the element of the matrix by a multiple, and these regions can overlap with one another making it difficult to decide which direction and value of factor to choose. The second challenge is smoothing out the neighboring area once the factor is supplied to the targeted point. This smoothing needs to be carried out in a delicate matter to ensure the resolution or the pattern is not disrupted.

[0005] It is an object of the present invention to provide a method, a computer program, and a computer-readable storage medium, by which a smooth matrix correction is provided so that if the matrix is used for developing an electric engine, calibrating of a torque accuracy is possible.

[0006] The subject matters according to the independent patent claims solve this object. Further embodiments are shown in the description, the dependent claims, and the drawings.

[0007] A first aspect of the present invention relates to a method for correcting or smoothing, or in particular, scaling a matrix, which describes a vehicle component of a vehicle, wherein the vehicle component may be an engine of the vehicle. The method according to the first aspect of the invention comprises the following steps:

[0008] In a first step, reading the matrix via an interface of an electronic computing device and detecting the matrix elements of the matrix, in particular, by the computing device. In a second step, performing a cluster analysis or clustering of the matrix, wherein at least a portion on the matrix elements of the matrix are clustered together in at least one cluster. In a third step, performing a linear regression for non-clustered components, in particular matrix elements, of each of the at least one cluster, thereby generating a, in particular an average, value for each cluster. In a fourth step, scanning the values of the cluster using a Gaussian function. In a fifth step, building a correction matrix that includes the scaled values and is provided at the interface.

[0009] The electronic computing device may include electronic means, for example, processors, circuits, in particular integrated circuits, and further electronic means for performing at least a part of the steps of the inventive method. The interface may build a data and / or a communication interface, by which data can be transferred from or to the electronic computing device.

[0010] With the method it is possible to correct a matrix in a systematic way to scale the matrix with smoothness in a non-disruptive way. This method may be used in many applications as image processing, where the matrix may be used for correcting digital media such as a pixelated image. The original intention of developing this method is to correct electric motor characterization matrices to improve the torque accuracy.

[0011] As said correcting of a matrix with non-uniform factors may be a complex problem to solve especially when regions are overlapping, the method provided according to the intention may partition the region into clusters and linearly regress the scaling value, allowing for a simpler and more systematic approach. The Gaussian function is used to apply factors, while preserving the characteristic of the original matrix.

[0012] The correction matrix could be used for calibrating and characterizing a motor or electric engine. The current process of calibrating and characterizing the motor to achieve acceptable torque accuracy is a time-consuming procedure that may take about three or four weeks of continuous work. Also, there is no systematic way. The development time heavily depends on the engineer who is working on it.

[0013] The method according to the invention solves this problem by using machine learning algorithms and statistical tools. It provides a more systematic way to solve the problem with a reduced development time.

[0014] According to an embodiment of the present invention, the matrix is a n-dimensional and the cluster analysis is performed before the matrix elements n-1 dimensionally.

[0015] According to another embodiment of the present invention an OPTICS algorithm is used for the cluster analysis. OPTICS stands for “ordering points to identify the clustering structure” and is an algorithm for finding density-based cluster and spatial data.

[0016] In yet another embodiment of the present invention as the linear regression ordinary least squares method is used. Ordinary least squares or OLS is a type of linear least square method for choosing near unknown parameters in a linear regression model by the principle of least squares: minimizing the sum of the squares of the difference between the observed dependent value in the input data set and the output of the function of the independent variable.

[0017] In still another embodiment of the present invention a motor, in particular an electric motor, is specified as the vehicle component.

[0018] In another embodiment of the present invention a current and / or voltage and / or engine speed and / or torque are detected as the matrix elements, in particular, by a corresponding sensor.

[0019] In a further embodiment of the present invention the matrix comprises a two-dimensional or 2D control table, by means of which a ratio of current and torque is detected for a certain motor state.

[0020] In yet another embodiment of the present invention a torque accuracy is set for the motor by means of the correction matrix.

[0021] In particular, the method is a computer-implemented method. Therefore, a second aspect of the invention relates to a computer program product comprising program code means for performing a method according to the first aspect.

[0022] A third aspect of the invention relates to a computer-readable storage medium comprising at least the computer program code according to the second aspect of the invention.

[0023] Advantages and advantageous embodiments of the first aspect of the present invention are to be regarded as advantages and advantageous embodiments of the second and / or third aspect of the present invention and vice versa.

[0024] Further advantages, features, and details of the invention derive from the following description of preferred embodiments as well as from the drawing. The features and feature combinations previously mentioned in the description as well as the features and feature combinations mentioned in the following description of the figures and / or shown in the figures alone can be employed not only in the respectively indicated combination but also in any other combination or taken alone without leaving the scope of the invention. BRIEF DESCRIPTION OF THE DRAWING

[0025] The novel features and characteristic of the disclosure are set forth in the appended claims. The accompanying drawings, which are incorporated in and constitute a part of this disclosure, illustrate exemplary embodiments and together with the description, serve to explain the disclosed principles. The same numbers are used throughout the figures to reference like features and components. Some embodiments of system and / or methods in accordance with embodiments of the present subject matter are now described below, by way of example only, and with reference to the accompanying figures.

[0026] The drawing shows in:

[0027] Fig. 1 a schematic diagram of a method for correcting a matrix; and

[0028] Fig. 2 a schematic diagram of clustered matrix elements derived by the method according to Fig. 1.

[0029] In the figures the same elements or elements having the same function are indicated by the same reference signs. DETAILED DESCRIPTION

[0030] In the present document, the word "exemplary" is used herein to mean "serving as an example, instance, or illustration". Any embodiment or implementation of the present subject matter described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments.

[0031] While the disclosure is susceptible to various modifications and alternative forms, specific embodiments thereof have been shown by way of example in the drawing and will be described in detail below. It should be understood, however, that it is not intended to limit the disclosure to the particular forms disclosed, but on the contrary, the disclosure is to cover all modifications, equivalents, and alternatives falling within the scope of the disclosure.

[0032] The terms “comprises”, “comprising”, or any other variations thereof, are intended to cover a non-exclusive inclusion so that a setup, device, or method that comprises a list of components or steps does not include only those components or steps but may include other components or steps not expressly listed or inherent to such setup or device or method. In other words, one or more elements in a system or apparatus preceded by “comprises” or “comprise” does not or do not, without more constraints, preclude the existence of other elements or additional elements in the system or method.

[0033] In the following detailed description of the embodiment of the disclosure, reference is made to the accompanying drawing that forms part hereof, and in which is shown by way of illustration a specific embodiment in which the disclosure may be practiced. This embodiment is described in sufficient detail to enable those skilled in the art to practice the disclosure, and it is to be understood that other embodiments may be utilized and that changes may be made without departing from the scope of the present disclosure. The following description is, therefore, not to be taken in a limiting sense.

[0034] Fig. 1 shows a method for correcting, in particular, a smoothing out or scanning of a matrix 10, which describes a vehicle component of a vehicle. The matrix 10 may contain matrix elements like an uncorrected data set 12 and correcting factors 14.

[0035] In a first step of the method reading the matrix 10 via an interface of an electronic computing device (not shown here) and detecting the matrix elements of the matrix 10 is performed.

[0036] In a second step of the shown method a cluster analysis 16 of the matrix 10 is performed, wherein at least a portion of the matrix elements of the matrix 10 are clustered together in at least one cluster 18. In the example, there are n clusters 18. The cluster analysis 16 may be implemented using an OPTICS algorithm. OPTICS stands for “ordering points to identify the clustering structure” and is an algorithm for finding densitybased cluster 18 and spatial data.

[0037] In a third step a linear regression 20 for non-clustered components is performed for each of the at least one cluster 18. Thereby generating a quasi-average value for each cluster 18.

[0038] In a fourth step scaling the values of the clusters 18 used in a Gaussian function 22 is performed.

[0039] In a fifth step of the method building a correction matrix 24, in particular, corrected data set, which includes the scaled values as performed, and the correction matrix 24 could be provided at the interface of the electronic computing device.

[0040] Correcting a matrix 10 is applicable in various settings including correcting digital media or scaling and manipulating a matrix 10 to reach a desired goal, the method discussed here offers a schematic way to scale the matrix 10 with smoothness and in a non-disruptive way.

[0041] One might want to increase or decrease the element of the matrix 10 by a multiple, and these regions can overlap with one another making it difficult to decide which direction and value of factor to choose. The second challenge is smoothing out the neighboring area once the factor is applied to the targeted point. The smoothing needs to be carried out in a delicate matter to ensure the resolution or the pattern is not disrupted.

[0042] Clustering is an unsupervised machine-learning approach used to group neighboring points together based on their distance density. In this method, ordering points to identify the cluster 18 structure or an OPTICS algorithm is chosen, in particular, over its close relative, named density paced spatial clustering of applications with noise (DBSCAN), due to OPTICS benefits of performing well under the varied density per data set. This omits the need to define the DBSCAN parameter, epsilon.

[0043] After running the clustering algorithm of performing the cluster analysis 16, each of the points is grouped into its cluster 18, labelled with numbers as shown in Fig. 2.

[0044] Once the points are clustered, x (Id) and y (Iq) values are averaged and saved as the center. Up to this point we have only focused on where each point lies in a two-dimensional plane, for example x and y values. Now for each cluster group or cluster 18, the said values are modeled with linear regression 20, specifically an ordinary least squares (OLS), condensing all the g values in that cluster 18 into a single value. In other words, the matrix 10 could be n-dimensional and the cluster analysis 16 is performed for the matrix elements n-1 dimensionally.

[0045] The Gaussian function 22 is used to apply a factor z0 and slowly smooth out the neighboring points from the center x0 and yo as illustrated with equation 1: z(x, y) — (¾ ~ 0? 2o?'

[0046] The center value in x (Id) axis is x0, the center value in y (Iq) axis is yo, and the initial scaling factor is zo. Euler’s number is e, and the standard deviation is o.

[0047] The function of equation 1 allows surrounding points to be also scaled, preventing the whole data set from losing its pattern. The whole matrix 10 is iterated through this formula.

[0048] Correcting the matrix 10 with non-uniform factors can be a complex program to solve, especially when the regions are overlapping. The method shown here partions the region into clusters 18 and uses linear regression 20 to scale values, allowing for a more systematic approach. The Gaussian function 22 is used to apply source factors, all by considering the characteristics of the matrix 10.

[0049] As the vehicle component, a motor, in particular an electric motor or electric engine, could be specified or described. Therefore, the matrix elements could be detected via sensor and could characterize a current and / or a voltage and / or an engine speed and / or a torque of said motor or engine. The matrix 10 may comprise a 2D control table, by means of which the ratio of current and torque is detected for certain motor states. Therefore, the method, or in particular the correction matrix 24, could be used to set a torque accuracy for the motor. Reference Signs 10 matrix 12 uncorrected data set 14 correcting factors 16 cluster analysis 18 cluster 20 linear regression 22 Gaussian function 24 correction matrix

Claims

1. A method for correcting a matrix (10) which describes a vehicle component of a vehicle, comprising the steps:reading the matrix (10) via an interface of an electronic computing device and detecting the matrix elements of the matrix (10);performing a cluster analysis (16) of the matrix (10), wherein at least a portion of the matrix elements of the matrix (10) are clustered together in at least one cluster (18);performing a linear regression (20) for non-clustered components of each of the at least one cluster (18), thereby generating a value for each cluster (18);scaling the value of the clusters (18) using a Gaussian function (22); and building a correction matrix (24) that includes the scaled values and is provided at the interface.

2. The method according to claim 1, characterized in thatthe matrix (10) is n-dimensional and the cluster analysis (16) is performed for the matrix elements n-1 dimensionally.

3. The method according to claim 1 or 2, characterized in thatan OPTICS algorithm is used for the cluster analysis (16).

4. The method according to any one of claims 1 to 3, characterized in thatordinary least squares are used for the linear regression (20).

5. The method according to any one of the preceding claims, characterized in thata motor, in particular an electric motor, is specified as the vehicle component.

6. The method according to claim 5, characterized in thata current and / or voltage and / or engine speed and / or torque are detected as the matrix (10) elements, in particular by a sensor.

7. The method according to claim 5 or 6, characterized in thatthe matrix (10) comprises a 2D control table, by which the ratio of current and torque is detected for certain motor states.

8. The method according to any one of the preceding claims, characterized in thata torque accuracy is set for the motor by the correction matrix (24).

9. A computer program code comprising program code means for performing a method according to any one of the claims 1 to 4.

10. A computer-readable storage medium comprising the computer program code according to claim 5.12

Images