System for operating a vehicle

Abstract

The invention relates to a system for operating a vehicle (2) comprising a number of components, wherein at least one operating parameter is provided for each component, wherein a maximum permissible value range is provided for each operating parameter of each component, wherein, for the ongoing operation of the vehicle (2) in which the components interact, at least one diagnostic unit (6, 42, 46) is configured to apply at least one existing diagnostic model for all components, which is configured to use a respective existing value range for each operating parameter of each component, wherein at least one AI unit (4) is configured to generate at least one shadow model and thereby set an alternative value range for at least one operating parameter of at least one component, wherein the at least one AI unit (4) is configured to simulate the at least one shadow model for all components.to determine the operating parameter values ​​for each component from a simulation of at least one shadow model and to verify whether all determined resulting values ​​of the operating parameters of all components (a) lie within the maximum permissible value ranges, or whether at least one determined resulting value of at least one operating parameter of at least one component (b) lies outside the maximum permissible value range of this at least one operating parameter of the at least one component.

Description

[0001] The invention relates to a system for operating a vehicle and a method for operating a vehicle.

[0002] The publication DE 102 35 525 A1 describes a method and a system for monitoring the condition of a vehicle.

[0003] A method for the context-dependent evaluation of a vehicle condition is known from publication DE 10 2021 209 090 A1.

[0004] A method for automatically creating an AI diagnostic model for diagnosing an abnormal condition based on noise and vibration data is shown in publication DE 10 2022 210 195 A1.

[0005] Against this background, the task was to optimize a diagnostic model for a vehicle.

[0006] This problem is solved by a system and a method with the features of the independent claims. Embodiments of the system and the method are described in the dependent claims and the description.

[0007] The system according to the invention is designed to operate a vehicle, e.g., a motor vehicle, which has a number of components. At least one operating parameter is provided for each component.

[0008] It is generally intended that an input value, representing at least one operating parameter, influences the operation of a respective component in the vehicle, and / or that during the operation of the respective component, an output value, representing at least one operating parameter, is generated. For the component, an input value of at least one primary operating parameter can influence a resulting output value of at least one secondary operating parameter. It is possible for at least one primary operating parameter and at least one secondary operating parameter to be the same. It is also possible for at least one primary operating parameter to differ from at least one secondary operating parameter.Furthermore, for each component, depending on its function in the vehicle, a component-specific set of operating parameters is provided, comprising at least one operating parameter. The value range of each operating parameter is limited by a minimum and a maximum value.

[0009] For each operating parameter of each component, a maximum permissible value range is defined, limited by, for example, an absolute minimum permissible value of the respective operating parameter that cannot and / or must not be undercut, and an absolute maximum permissible value of the respective operating parameter that cannot and / or must not be exceeded. The system comprises, as subcomponents, at least one AI unit or artificial intelligence, at least one diagnostic unit, and at least one update unit.

[0010] During vehicle operation, the components interact and influence each other, and it is possible for an input value of at least one primary operating parameter of a first component to influence an output value of at least one secondary operating parameter of a second component. The at least one diagnostic unit is configured to apply at least one existing and / or given diagnostic model to all components, which is configured to use an existing and / or given range of values ​​for each operating parameter of each component.

[0011] Furthermore, the at least one AI unit is trained to generate and / or provide at least one shadow model and to set and / or provide an alternative value range for at least one operating parameter of at least one component. The at least one AI unit (artificial intelligence) is also trained to simulate and calculate the at least one shadow model for all components, and to determine the resulting values ​​for each component from a simulation of the at least one shadow model.The AI ​​unit determines the output values ​​of the operating parameters and verifies whether all determined resulting values, typically output values, of the operating parameters of all components are either (a) within the maximum permissible value ranges, or whether at least one determined resulting value of at least one operating parameter of at least one component is (b) outside the maximum permissible value range of that at least one operating parameter of the at least one component. The AI ​​unit is also configured to discard the at least one shadow model if (b) is true. If, on the other hand, (a) is true for the at least one shadow model, the at least one update unit is configured to provide the at least one shadow model as at least one new available diagnostic model and / or to propose it for use.

[0012] The system, in its configuration, comprises the vehicle or a vehicle system with all components and / or, as at least one main component, a cloud system, a workshop system, and / or a download platform. The vehicle is connected to at least one main component via at least one communication unit of the system and exchanges data and / or information with it, usually wirelessly. Alternatively or additionally, the at least one AI unit, the at least one diagnostic unit, and / or the at least one update unit are located in the vehicle and / or in the at least one main component.

[0013] The at least one diagnostic unit may also include another artificial intelligence or AI unit.

[0014] With the respective diagnostic model, both offboard and onboard faults can be detected during operation of the vehicle, as well as a fault-causing component, whereby the fault may be based on a non-compliance with the value range of an operating parameter, if an output value of at least one operating parameter lies outside the value range provided for this purpose, or, in the case of a configuration, outside the maximum permissible value range.

[0015] The method according to the invention is designed for operating a vehicle comprising a number of components, wherein at least one operating parameter is provided for each component, and wherein a maximum permissible value range is provided and / or considered for each operating parameter of each component. During operation of the vehicle, the components interact, mutually influencing each other and the values ​​of their respective operating parameters. In the method, at least one existing diagnostic model is applied, used, and / or considered for all components, with or from which an existing value range is used and / or applied for each operating parameter of each component. Furthermore, at least one shadow model is generated by at least one control unit, and an alternative value range is set and / or provided for at least one operating parameter of at least one component.Furthermore, the at least one shadow model for all vehicle components is simulated by the at least one AI unit, e.g., by calculation. For each component, the resulting values ​​of the operating parameters are determined from a simulation of the at least one shadow model. It is then checked whether all determined resulting values ​​of the operating parameters of all components (a) lie within the maximum permissible value ranges, or whether at least one determined resulting value of at least one operating parameter of at least one component (b) lies outside the maximum permissible value range of that at least one operating parameter of the at least one component. The at least one shadow model tested and / or verified by simulation by the AI ​​unit is discarded if (b) is true. If (a) is true, the at least one shadow model is provided as a new diagnostic model.

[0016] It is possible that one embodiment of the presented method is carried out with one embodiment of the presented system. The new diagnostic model can be installed in the vehicle and / or its diagnostic unit by means of a software update.

[0017] In this configuration, the at least one AI unit determines, predicts, and / or calculates an expected behavior of the system, a behavior of the at least one existing diagnostic model, and a behavior of the at least one shadow model. Furthermore, the AI ​​unit compares a deviation of the behavior of the at least one diagnostic model from the expected behavior of the system with a deviation of the behavior of the at least one shadow model from the expected behavior of the system. The at least one existing diagnostic model is replaced by the at least one shadow model if the deviation of the behavior of the at least one diagnostic model is greater than the deviation of the behavior of the at least one shadow model.Furthermore, the at least one shadow model is rejected if the deviation in the behavior of the at least one shadow model is at least as large as, or greater than, the deviation in the behavior of the at least one diagnostic model. The validity of each shadow model may be checked by the AI ​​unit and / or update unit, usually during the simulation, where the shadow model is valid if (a) is true and invalid if (b) is true.

[0018] In this configuration, the available value range for each operating parameter of each component of the at least one existing diagnostic model is reduced and / or restricted compared to the maximum permissible value range. Specifically, the minimum value of the value range for each operating parameter of each diagnostic model is set to be greater than the minimum permissible value of the respective operating parameter, and the maximum value of the value range for each operating parameter of each diagnostic model is set to be less than the maximum permissible value of the respective operating parameter, with values ​​of the available value range lying between the minimum permissible value and the maximum permissible value of the respective operating parameter.

[0019] Furthermore, the alternative value range for the at least one operating parameter of the at least one component of the at least one shadow model is extended or increased compared to the existing value range for the at least one operating parameter of the at least one component of the at least one existing diagnostic model, but is still set smaller than the maximum permissible value range for the at least one operating parameter of the at least one component. In this embodiment, the minimum value of the alternative value range for the at least one operating parameter of the at least one component in the at least one shadow model is set larger than the minimum permissible value and smaller than the minimum value of the at least one operating parameter of the at least one existing diagnostic model. Furthermore,The maximum value of the alternative range for the at least one operating parameter of the at least one component in the at least one shadow model is set to be less than the maximum permissible value and greater than the maximum value of the at least one operating parameter of the at least one component in the at least one existing diagnostic model. Values ​​of the alternative range for the at least one operating parameter of the at least one component in the at least one shadow model lie between the minimum permissible value and the maximum permissible value of the at least one operating parameter of the at least one component. Furthermore, the minimum value of the alternative range for the at least one operating parameter of the at least one component in the at least one shadow model lies between the minimum permissible value and the minimum value of the at least one operating parameter of the at least one existing diagnostic model.Furthermore, the maximum value of the alternative value range of the at least one operating parameter of the at least one component in the at least one shadow model lies between the maximum value of the at least one operating parameter of the at least one existing diagnostic model and the maximum permissible value for this.

[0020] If (a) is true, a further embodiment may provide a new extended and / or incremented shadow model with an extended and / or increased alternative value range of the at least one operating parameter of the at least one component, compared to a shadow model already tested and / or verified by the AI ​​unit through simulation, wherein the extended alternative value range is less than the maximum permissible value range and greater than the existing value range of the at least one existing diagnostic model, wherein this extended shadow model is simulated and / or tested by the AI ​​unit, whereby it is checked whether all determined resulting values ​​of the operating parameters of all components (a) lie within or (b) outside the maximum permissible value range of the at least one operating parameter of the at least one component.If (b) is true, a further embodiment may also provide a shadow model that is reduced compared to a shadow model already tested and / or verified by the AI ​​unit through simulation, wherein the reduced alternative value range is smaller than the value range of the shadow model already tested and / or verified and larger than the value range of the at least one available diagnostic model.

[0021] When providing a shadow model from which a new diagnostic model is provided and / or derived, an aging and / or wear process is considered for each component, starting from the initial commissioning of the vehicle. This process depends on the vehicle's operating time and / or age. The respective value range for each operating parameter of each component in the diagnostic model, and / or the alternative value range for at least one operating parameter of at least one component in the shadow model, is set based on the respective operating time and / or age of that component. It can also be taken into account that a component already present in the vehicle has been replaced with a new component, with the new component replacing the existing one.The existing component is replaced, with both components having the same function during vehicle operation and typically located in the same position. The maximum permissible value range of at least one operating parameter of the new component can also be taken into account, which is wider than the maximum permissible value range of the existing component because the existing component is older and has already shown signs of wear in the vehicle. After replacing and / or renewing a component, an extended shadow model can also be generated, in which the value range for at least one operating parameter of the new component is considered to be wider than that of the existing, older component.

[0022] The at least one operating parameter can be, for example, the temperature of a respective component, the electrical voltage applied to a respective component, the electrical current flowing through a respective component, and / or the electrical and / or mechanical power supplied by a respective component. Furthermore, to set a respective value range for the at least one operating parameter of the at least one component for the vehicle described so far (e.g., the first vehicle), information about diagnostic models, shadow models, and / or value ranges from at least one other vehicle (e.g., a second vehicle) can also be considered, whereby the first and the at least one second vehicle can belong to the same fleet.

[0023] At least one component of the vehicle is used during operation, for example, for propulsion. It can be designed as a sensor, an actuator (e.g., a machine or motor), or a control unit.

[0024] The presented method can be performed continuously or permanently, or as needed, for example, after replacing a component. In this process, the at least one existing diagnostic model and its behavior are checked against the expected behavior of the system by the at least one AI unit (e.g., located and / or installed in the vehicle) and replaced by the at least one shadow model deemed valid by the AI ​​unit.

[0025] In this configuration, the diagnosis performed by the at least one diagnostic model is carried out by the at least one diagnostic unit and / or AI unit in the at least one main component, i.e., in the vehicle, in the cloud system, and / or in the workshop system or a workshop if the vehicle is in the workshop. The execution of services related to a diagnosis to be performed or already performed by the at least one diagnostic model can be monitored by or within the at least one diagnostic unit, e.g., continuously.This involves collecting insights into the behavior and / or reliability of results for values ​​from at least one operating range of at least one component, determined using a given diagnostic model. The determined values ​​arise within a chain of interactions between components in the vehicle and / or from interactions between components. This process verifies the values ​​of operating parameters of the components that contribute to a diagnostic result.

[0026] During optimization, the diagnostic models are continuously, or as needed, compared and verified by at least one AI unit against the expected system behavior. If an existing diagnostic model is replaced by a shadow model, the resulting new diagnostic model is optimized and / or improved compared to the existing, outdated diagnostic model, for example, due to its extended value range, for at least one operating parameter of at least one component. If the vehicle's behavior remains unchanged and an existing diagnostic model does not yield any altered results for the values ​​of at least one operating parameter, active verification and / or adjustment of the respective diagnostic model is not strictly necessary, but can still be performed, for example, at regular intervals.

[0027] The at least one shadow model is generated by the at least one AI unit and / or diagnostic unit, e.g., by a diagnostic AI unit with which a diagnosis is carried out by an artificial intelligence, whereby when generating a respective shadow model, the value range of the at least one operating parameter of the at least one component is dynamically adjusted compared to the previously available value range, e.g., increased or extended, whereby new boundary conditions and / or limit values, i.e., the minimum and maximum values ​​provided for a respective value range, are created and / or provided, within which the vehicle can be operated safely.

[0028] It is also possible to consider the dependence of at least one component or its operation within the vehicle on at least one external environmental parameter, such as temperature and / or humidity, within the vehicle's environment. A component designed as a sensor is intended to acquire values ​​of at least one detected and / or measured operating parameter via signals from at least one subcomponent and / or provide these values ​​to at least one control unit. The values ​​and / or signals are adjusted by modifying a respective value range. Individual aging processes of components are also taken into account, with the AI ​​unit continuously monitoring and adjusting forecasts for operating parameter values ​​by providing shadow models.

[0029] To verify the validity of a given shadow model, the AI ​​unit and / or update unit runs it through a simulation using identical information, such as input values, for at least one operating parameter as the at least one existing diagnostic model, and evaluates the results. The at least one AI unit, typically installed and which may be designed and / or referred to as an AI engine or AI machine, can thus generate and / or determine a particularly stable and reliable shadow model during vehicle operation, enabling safe operation of the vehicle system in its current state. Each generated shadow model can be designed and / or referred to as a diagnostic model, usually with minor modifications, if it replaces a previously existing diagnostic model.

[0030] In this implementation, the most reliably generated and / or determined shadow model is transmitted as a candidate for a new diagnostic model from the vehicle and / or workshop system, typically including relevant metadata about a condition, e.g., the operating time and / or age, of at least one component of the vehicle, to the cloud system or cloud backend of a manufacturer of the at least one component and / or the vehicle or an original equipment manufacturer (OEM). There, information about the shadow model is anonymized, consolidated according to various criteria, and passed to an evaluation unit as a subcomponent of the system.

[0031] The AI ​​unit of the cloud system, or rather an optimizer for the diagnostic models, generates specific adjustments for components in generalized diagnostic models for multiple vehicles in the fleet based on the data from the shadow models. From this, optimized diagnostic models are generated, which are actively offered or suggested to the vehicles for download from various data archives (repositories) via a subcomponent configured as an update delivery unit, which can be configured as an update unit configuration. The system includes a trigger as a subcomponent that notifies all vehicles that a new, valid shadow model has been made available, which can also be configured and / or designated as an optimized diagnostic model.Each vehicle downloads the optimized diagnostic model from its assigned data archive and uses the new diagnostic model based on the shadow model after it has been validated by at least one update unit and / or AI unit of the vehicle. This starts a new optimization cycle in which new shadow models are generated and simulated.

[0032] The system and method can generate a diagnostic model for the vehicle from shadow models, ideally avoiding limiting diagnoses due to, for example, poor value ranges. The original equipment manufacturer (OEM) also gains the ability, via the cloud system, to detect age-related, unpredictable changes in components and / or their operating parameters at an early stage.

[0033] For each component, the maximum permissible range of its at least one operating parameter defines the operating parameters, which can be further refined by reducing this range. For example, the component might have a specified temperature range of 0 °C to 40 °C, within which the supplier considers it fully functional. Within a range of -5 °C to 60 °C, it would only be partially functional, and below -5 °C, it would be non-functional. In this case, the supplier would only consider the range of 10 °C to 30 °C as safe and include it in the initial considerations for a diagnostic model to be provided. By generating shadow models, it is possible to expand the temperature range and maximize it as much as possible.Furthermore, shadow models can also account for negative changes to a component due to aging, as the value range for at least one operating parameter of a given shadow model can be restricted or reduced. Dynamic changes to any component in the vehicle can be captured using shadow models.

[0034] Furthermore, electrical operating parameters such as current and voltage can be considered for each component. Value ranges for these parameters can also be individually determined and / or adjusted using shadow models to suit the operation of the respective component within the vehicle. By adapting these value ranges to the vehicle's operation, the driver can be protected from premature warnings resulting from a diagnosis generated by the currently available diagnostic model. Since component aging is also taken into account, its functions can be optimally utilized within the maximum possible value range until the component needs to be replaced.

[0035] The AI ​​unit is designed to determine and maintain the maximum value ranges of operating parameters for the sensors and actuators within the vehicle. The AI ​​unit can predict qualitative and, where necessary, quantitative results for the output values ​​of operating parameters, consider the effects on the entire vehicle, and adjust the value range of an operating parameter for a first component—that is, maximize or minimize it—while adhering to the maximum permissible value range of a second component that interacts with the first. The existing diagnostic model can be replaced and thus adapted by a new, valid shadow model if the output values ​​of the diagnostic unit and the output values ​​of the AI ​​unit differ.

[0036] It is understood that the features mentioned above and those to be explained below can be used not only in the combinations specified, but also in other combinations or on their own, without leaving the scope of the present invention.

[0037] The invention is schematically illustrated with reference to embodiments in the drawing and is described schematically and in detail with reference to the drawing.

[0038] Fig. Figure 1 shows a schematic representation of an embodiment of the system according to the invention for carrying out an embodiment of the method according to the invention.

[0039] The in Fig.1. A schematically represented embodiment of the system comprises a vehicle 2 and, as main components, a workshop system 40, a cloud system 20 of an original manufacturer or original equipment manufacturer (OEM) of the vehicle 2 and / or components of the vehicle 2 and a download platform 60.

[0040] The vehicle 2, or vehicle system, comprises as subcomponents a communication unit 8, a diagnostic unit 6, an AI unit 4 (or artificial intelligence) trained to perform diagnostics, and an update unit 10. Components of the vehicle 2 include several computing units 12a, 12b, 12c and operating units 14a, 14b, 14c, each of which may include at least one sensor and / or actuator as a further component. A maximum permissible value range is defined for each operating parameter of each component, which should not be exceeded or fallen below. Furthermore, the vehicle 2 may have a human-machine interface (MMI, not shown) with a display designed to inform the driver of the vehicle 2 about its operation and the results of vehicle 2 diagnoses.

[0041] The workshop system 40 comprises, as subcomponents, a communication unit 48, an update unit 44, a first diagnostic unit 42, and a second diagnostic unit 46, whereby at least the first diagnostic unit 42 is supported by artificial intelligence (AI) or an AI unit. The cloud system 20, or a cloud backend of the OEM, comprises, as subcomponents, a communication unit 26, an optimizer 22 for diagnostic models, a further update unit configured and / or designated as an update deployment unit 24, and a data archive 28, or repository, for diagnostic models and also for shadow models. The download platform 60 comprises, as at least one subcomponent, a further and / or alternative data archive 62, or repository, for diagnostic models and also for shadow models, and a communication unit not shown further.

[0042] In the embodiment of the method according to the invention, the vehicle 2 and the main components of the system communicate cryptographically via the communication units 8, 26, 48. Furthermore, the vehicle 2 and the main components are protected against tampering. During a diagnostic check of the vehicle 2's operation, the diagnostic unit 6 of the vehicle 2 uses a currently available diagnostic model that considers a range of values ​​for each operating parameter of each component of the vehicle 2, which is limited compared to the respective maximum permissible value range. The components of the vehicle 2 provide or supply values, typically output values, of operating parameters that arise during the operation of the vehicle 2.Errors can be detected and displayed to the driver via the display if at least one output value, usually detected by a sensor, as the value of at least one operating parameter of a component, or even just a single component, lies outside a value range provided for this purpose and available.

[0043] In this embodiment of the method according to the invention, the existing diagnostic module used for diagnosis is optimized by adjusting the value ranges for operating parameters. A shadow model is generated by at least one AI unit 4, and an alternative value range is set for at least one operating parameter of at least one component. The at least one AI unit 4 simulates this shadow model for all components of the vehicle 2. Furthermore, the at least one AI unit 4 determines the resulting values ​​or output values ​​of the operating parameters for each component from a simulation of the shadow model and checks whether all determined resulting values ​​of the operating parameters of all components are (a) within or (b) outside the respective maximum permissible value range. The at least one AI unit 4 rejects the shadow model if (b) is true.If, on the other hand, (a) is true, the at least one shadow model is checked by at least one update unit 10, 44 to determine whether it is valid or not. If it is valid, the shadow model is provided as at least one new available diagnostic model. REFERENCE MARK: 2 vehicles 4 AI units 6 Diagnostic Unit 8 Communication unit 10 update units 12a, 12b, 12c Unit of calculation 14a, 14b, 14c Operating unit 20 Cloud systems 22 optimizers 24 Update Deployment Unit 26 Communication unit 28 Data archive 40 workshop system 42 Diagnostic Unit 44 update units 46 Diagnostic Unit 48 communication unit 60 Download platform 62 Data archive QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] DE 102 35 525 A1

[0002] DE 10 2021 209 090 A1

[0003] DE 10 2022 210 195 A1

[0004]

Claims

[1] System for operating a vehicle (2) comprising a number of components, wherein at least one operating parameter is provided for each component, wherein a maximum permissible range of values ​​is provided for each operating parameter of each component, wherein the system comprises as subcomponents at least one AI unit (4), at least one diagnostic unit (6, 42, 46) and at least one update unit (10, 40), - wherein for an ongoing operation of the vehicle (2) in which the components interact, at least one diagnostic unit (6, 42, 46) is designed to apply at least one existing diagnostic model to all components, which is designed to use an existing range of values ​​for each operating parameter of each component, - wherein the at least one AI unit (4) is configured to generate at least one shadow model and to set an alternative range of values ​​for at least one operating parameter of at least one component, - wherein the at least one AI unit (4) is configured to simulate the at least one shadow model for all components, to determine the values ​​of the operating parameters resulting from a simulation of the at least one shadow model for each component, and to check whether all determined resulting values ​​of the operating parameters of all components (a) are within the maximum permissible value ranges, or whether at least one determined resulting value of at least one operating parameter of at least one component (b) is outside the maximum permissible value range of this at least one operating parameter of the at least one component, - wherein the AI ​​unit (4) is trained to discard the at least one shadow model if (b) is true, and wherein, in the case that (a) is true, the at least one update unit (10, 40) is trained to provide the at least one shadow model as at least one new available diagnostic model. [2] System according to claim 1, comprising the vehicle (2) and as at least one main component a cloud system (20), a workshop system (40) and / or a download platform (60), wherein the vehicle (2) is connected to at least one main component via at least one communication unit (8, 26, 48), and / or wherein the at least one AI unit (4), the at least one diagnostic unit (6, 42, 46) and / or the at least one update unit (10, 40) is arranged in the vehicle (2) and / or in the at least one main component. [3] Method for operating a vehicle (2) comprising a number of components, wherein at least one operating parameter is provided for each component, wherein a maximum permissible range of values ​​is provided for each operating parameter of each component, - wherein for an ongoing operation of the vehicle (2) in which the components interact, at least one existing diagnostic model is applied for all components, with which an existing range of values ​​is used for each operating parameter of each component, - wherein at least one shadow model is generated by at least one AI unit (4) and an alternative value range is set for at least one operating parameter of at least one component, - wherein the at least one AI unit (4) simulates the at least one shadow model for all components, wherein values ​​of the operating parameters resulting from a simulation of the at least one shadow model are determined for each component, checking whether all determined resulting values ​​of the operating parameters of all components (a) are within the maximum permissible value ranges, or whether at least one determined resulting value of at least one operating parameter of at least one component (b) is outside the maximum permissible value range of this at least one operating parameter of the at least one component, - wherein the at least one shadow model is discarded if (b) is true, and wherein, in the event that (a) is true, the at least one shadow model is provided as at least one new available diagnostic model. [4] Method according to claim 3, wherein the at least one AI unit (4) determines an expected behavior of the system, a behavior of the at least one existing diagnostic model and a behavior of the at least one shadow model, wherein a deviation of the behavior of the at least one diagnostic model from the expected behavior of the system is compared with a deviation of the behavior of the at least one shadow model from the expected behavior of the system, wherein the at least one existing diagnostic model is replaced by the at least one shadow model if the deviation of the behavior of the at least one diagnostic model is greater than the deviation of the behavior of the at least one shadow model, wherein the at least one shadow model is discarded if the deviation of the behavior of the at least one shadow model is at least as large as the deviation of the behavior of the at least one diagnostic model. [5] Method according to claim 3 or 4, wherein the respective value range for each operating parameter of each component of the at least one diagnostic model is reduced compared to the maximum permissible value range. [6] Method of one of claims 3 to 5, wherein the alternative value range for the at least one operating parameter of the at least one component of the at least one shadow model is extended compared to the present value range for the at least one operating parameter of the at least one component of the at least one present diagnostic model, but is still set smaller than the maximum permissible value range for the at least one operating parameter of the at least one component. [7] Method according to one of claims 3 to 6, wherein, starting from an initial commissioning of the vehicle (2), an aging process is taken into account for each component, wherein the respective value range for each operating parameter of each component of the present diagnostic model and / or the alternative value range for at least one operating parameter of at least one component of the at least one shadow model is set depending on a respective runtime of the at least one component. [8] Method according to any one of claims 3 to 7, wherein the at least one operating parameter is a temperature of a respective component, an electrical voltage applied to a respective component, an electric current flowing through a respective component and / or an electrical and / or mechanical power supplied by a respective component. [9] Method according to any one of claims 3 to 8, wherein, in order to set a respective value range of the at least one operating parameter of the at least one component for the vehicle (2), information about diagnostic models, shadow models and / or value ranges of at least one further vehicle (2) is also taken into account.

Images