Method for operating a vehicle device and vehicle device

Abstract

The invention relates to a method (10) for operating a vehicle device (12) on the basis of at least one calculated value (16), comprising the following steps: determining the calculated value (16) using an analytical correlation (18) on the basis of at least a first parameter (a) and a second parameter (b) of the vehicle device (12) and an input value (TR) associated with the vehicle device (12); calculating a first parameter (a) by means of a first neural network (20); a second parameter (b) is calculated by means of a second neural network (22). The invention also relates to a vehicle device (12).

Description

[0001] Instruction Manual

[0002] The invention relates to a method for operating a vehicle device according to claim 1. The invention also relates to a vehicle device.

[0003] WO 2020 147 870 A1 describes a method for calculating the temperature of a vehicle device, in particular a drive unit. The temperature at a corresponding temperature detection area is calculated using a temperature model.

[0004] The object of the present invention is to operate a vehicle device more efficiently and effectively. In particular, calculated values, such as the temperature of a vehicle device, should be calculated more accurately and more quickly.

[0005] At least one of these objects is achieved by a method for operating a vehicle device having the features according to claim 1. Thus, the calculated values ​​determined in this way allow a more accurate and reliable assessment of the behavior of the vehicle device. The calculated values ​​can be calculated more accurately. This enables a more efficient and effective operation of the vehicle device.

[0006] The vehicle device may be associated with a vehicle. The vehicle device may be an electric motor, a power electronics device, a thermal management system, a battery and / or a transmission. The vehicle may be an electric vehicle.

[0007] The calculated value may change over time. The calculated value may be determined at time intervals. The time intervals may be predefined time intervals. The time intervals may be constant or variable. The time intervals may be adjusted based on environmental conditions or measured values.

[0008] The first parameter can be a variable. The second parameter can be a variable.

[0009] By applying the first neural network and the second neural network, the first parameter and the second parameter can be calculated even for unknown input variables. The first neural network and / or the second neural network may have at least one input layer, one or more hidden layers, and an output layer.

[0010] The first neural network and / or the second neural network may be a feedforward neural network (FNN), a recurrent neural network (RNN), or a convolutional neural network (CNN).

[0011] In a preferred embodiment of the invention, it is advantageous if the input value is a measured and / or calculated physical quantity of the vehicle device.The input value may be a calculated value determined from a calculated value previously determined in time.

[0012] In a particular embodiment of the invention, it is advantageous if the first neural network and the second neural network are trained separately. This may reduce the complexity and the effort required to train the neural networks. The first neural network and the second neural network may be trained dependently or independently. The first neural network may be trained with training data that is in particular completely different from the training data of the second neural network. The training data of the first neural network and the second neural network may at least partially overlap.

[0013] In a special embodiment of the invention, it is advantageous if the determined calculated value forms an input value for a further determination of the calculated value after the determination in time.The calculated value may be dynamic in time and may therefore be calculated in dependence on time.

[0014] In a particular embodiment of the invention, it is advantageous if the analytical correlation can be represented by an nth order polynomial. The analytical correlation may include addition, subtraction, multiplication and / or division. Often, an exact analytical correlation cannot be specified for one or more vehicle devices, for example because the system is too complex. One possibility is to reduce the complexity by simplifying assumptions, grouping parts / components, etc. This approach provides an approximate analytical correlation. The procedure described herein also applies to this type of analytical correlation. In this context, the term "system" may be understood to mean parts, groups of parts, components and / or individual parts of a vehicle device.

[0015] An advantage of a preferred embodiment of the present invention is that the number of parameters for analyzing the correlation of the first parameter and the second parameter is equal to n + 1. Preferably, each parameter can be calculated by a separate neural network including the first neural network and the second neural network. This can reduce the complexity of the neural network.

[0016] In a preferred embodiment of the present invention, it is provided that the number of neural networks used to determine the calculated value and including the first neural network and the second neural network is equal to n+1. In addition, less than n+1 or n neural networks including the first neural network and the second neural network may be used. For example, one of the parameters for analyzing the correlation may be calculated conventionally. At least one of the parameters for analyzing the correlation may be fixed.

[0017] In an advantageous embodiment of the invention, it is provided that the calculated value indicates a temperature change, the first parameter indicates a heat input, the second parameter indicates a heat dissipation and / or the input value indicates a temperature. The temperature change may be the temperature of the electric motor. The heat input may be caused by electrical energy. The heat dissipation may be conductive, radiative and / or convective.

[0018] Furthermore, at least one of the above-mentioned objects is achieved by a vehicle device having the features of claim 9 .

[0019] In a particular embodiment of the invention, it is advantageous if the vehicle device is a driveline device of a driveline of a vehicle. The driveline device may be configured to propel the vehicle. The driveline device may provide a driving force for the propulsion of a motor vehicle.

[0020] Further advantages and advantageous embodiments of the invention emerge from the description of the figures and from the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The present invention will be described in detail below with reference to the accompanying drawings. In the accompanying drawings:

[0022] Figure 1 A schematic diagram of a method for operating a vehicle device in a specific embodiment of the invention is shown. The method 10 for operating a vehicle device 12 is used, for example, in operating a vehicle having the vehicle device 12. The vehicle device 12 may be a driveline device 14 of the vehicle, in particular an electric motor of the vehicle.

[0023] The operation of the vehicle device 12 depends on the calculated value 16. The calculated value 16 may be a temperature change ΔT of the temperature T of the electric motor. The electric motor may preferably be operated as a function of the temperature change ΔT and thus of the temperature T calculated from the temperature change ΔT.

[0024] The calculated value 16 is determined by the at least one first parameter a and a second parameter b and the input value T R The analysis correlation 18 is calculated. For example, the analysis correlation 18 is given by an n-order polynomial, in particular a first-order polynomial with n=1, as can be expressed by the following formula:

[0025] ΔT=aT R +b

[0026] The number of parameters of the analysis dependency 18 is therefore equal to n+1. The determined calculated value 16 can form an input value for a further determination of the calculated value 16 in time after the determination. as follows:

[0027]

[0028] The first parameter a can represent the heat input P 输入 , and the second parameter b can represent the heat dissipation P 输出 The first parameter a can be constructed to depend on the following input variables:

[0029]

[0030] Where the heat transfer coefficient is α, the contact area involved in the heat transfer is A, and the specific heat capacity of the affected component is C T .

[0031] The second parameter b can be constructed to depend on the following input variables:

[0032]

[0033] where the resistance is R. The resulting heat output is assumed to be:

[0034] P 输入 =I 2 R

[0035] Wherein, the current is I.

[0036] The first parameter a is calculated by the first neural network 20, and the second parameter b is calculated by the second neural network 22. The first neural network 20 and the second neural network 22 are preferably trained independently of each other. The number of neural networks used to determine the calculated value 16 is equal to n+1. The first neural network 20 is trained with the training data a1 to a2. m The second neural network 22 is trained with the training data b1 to b k Conduct training.

[0037] Reference numerals list

[0038] P 输入 Heat input

[0039] P 输出 Heat dissipation

[0040] T R Input value

[0041] ΔT Temperature Change

[0042] aThe first parameter

[0043] bThe second parameter

[0044] 10. Processing

[0045] 12 Vehicle Equipment

[0046] 14 Transmission system

[0047] 16 Calculated value

[0048] 18 Analyze Correlation

[0049] 20. The First Neural Network

[0050] 22 Second Neural Network

Claims

1. A method (10) for operating a vehicle device (12), the method being dependent on at least one calculated value (16) and comprising the following steps: According to at least one first parameter (a) and a second parameter (b) of the vehicle device (12) and an input value (T) associated with the vehicle device (12), R ) to determine said calculated value (16) using analytical correlation (18), calculating the first parameter (a) by a first neural network (20), The second parameter (b) is calculated by a second neural network (22).

2. The method (10) for operating a vehicle device (12) according to claim 1, characterized in that The input value (T R ) is a measured and / or calculated physical quantity of the vehicle device (12).

3. The method (10) for operating a vehicle device (12) according to claim 1 or 2, characterized in that: The first neural network (20) and the second neural network (22) are trained independently of each other.

4. The method (10) for operating a vehicle device (12) according to any one of the preceding claims, characterized in that: The determined calculated value (16) is an input value (T) for further determining the calculated value (16) in time after the determination. R ).

5. The method (10) for operating a vehicle device (12) according to any one of the preceding claims, characterized in that The analytical correlation (18) can be represented by an nth order polynomial.

6. The method (10) for operating a vehicle device (12) according to claim 5, characterized in that The number of parameters of the analytical correlation (18) including the first parameter (a) and the second parameter (b) is equal to n+1.

7. The method (10) for operating a vehicle device (12) according to claim 5 or 6, characterized in that: The number of neural networks used to determine the calculated value (16) and including the first neural network (20) and the second neural network (22) is equal to n+1.

8. The method (10) for operating a vehicle device (12) according to one of the preceding claims, characterized in that The calculated value (16) indicates the temperature change (ΔT), and the first parameter (a) indicates the heat input (P 输入 ), the second parameter (b) indicates the heat dissipation (P 输出 ) and / or the input value (T R ) indicates temperature (T).

9. A vehicle device (12) operable according to the method (10) of the preceding claim.

10. The vehicle device (12) according to claim 9, characterized in that The vehicle device (12) is a drive train device (14) of a drive train of a vehicle.

Images