Method for determining an operating state of a vehicle component

Abstract

One aspect of the present disclosure relates to a method for determining an operational state of a vehicle component of a vehicle. The method comprises providing a first knowledge graph containing information about a plurality of vehicle components of a vehicle, wherein each vehicle component provides one or more signals. The method further comprises determining an operational state of a vehicle component from the plurality of vehicle components, wherein the determining comprises selecting, using the first knowledge graph, a plurality of signals and one or more state parameters required for determining the operational state of the vehicle component. The determining further comprises calculating the one or more state parameters of the vehicle component based on the selected plurality of signals. Finally, the determination of the operational state of the vehicle component comprises determining the operational state of the vehicle component in dependence on the information contained in the first knowledge graph and based on the calculated one or more state parameters.

Description

Technical Field

[0001] This invention relates to techniques for determining the operational status of vehicle components. Related aspects include an electronic vehicle system and a data cloud system. Background Technology

[0002] A typical problem with proactively checking the operational status of vehicle components is the limited amount of diagnostic information available to suppliers, original equipment manufacturers (OEMs), vehicle owners, or repair shops. In some existing methods, data from vehicle signals is often only available within the vehicle and may not be stored for further analysis due to limited storage capacity in electronic vehicle controls. Consequently, in existing methods, when a vehicle component malfunctions or fails, only high-level diagnostic trouble codes (DTCs) are typically generated and stored locally in the fault log. These DTCs may have limited value in terms of the actual nature of the problem. Furthermore, due to the aforementioned limitations, vehicle diagnostics are often only performed at the repair shop (where DTC codes can be read) some time after the problem occurs. In many cases, significant expense and additional expertise are required to pinpoint the cause of the problem. This can result in vehicle components being replaced during vehicle repairs without knowing which component actually caused the problem or why. Moreover, in many cases, the problem remains undetected until it causes further damage to the vehicle component. Summary of the Invention

[0003] A first general aspect of this disclosure relates to a method for determining the operating state of vehicle components of a vehicle. The method includes: providing a first knowledge graph containing information about a plurality of vehicle components of the vehicle, wherein each vehicle component provides one or more signals. The method further includes: determining the operating state of a vehicle component from the plurality of vehicle components, wherein the determination includes: selecting, using the first knowledge graph, a plurality of signals and one or more state parameters required for determining the operating state of the vehicle component. The determination further includes: calculating one or more state parameters of the vehicle component based on the selected plurality of signals. Finally, determining the operating state of the vehicle component includes: determining the operating state of the vehicle component based on information contained in the first knowledge graph and based on the calculated one or more state parameters.

[0004] A second general aspect of this disclosure relates to an electronic vehicle system including a first electronic device and a second electronic device, the electronic vehicle system being designed to perform the methods described according to the first general aspect of this disclosure. The system includes a first electronic device for using a first knowledge graph. The second electronic device is configured to monitor and receive signals from a plurality of vehicle components. The second electronic device is also configured to determine state parameters of the plurality of vehicle components.

[0005] A third general aspect of this disclosure relates to a data cloud system designed to receive information from a second device according to a second general aspect, and to perform the method described in accordance with a first general aspect of this disclosure to update a conceptual model of a second knowledge graph. The data cloud system is also configured to transmit updated information contained in the updated conceptual model of the second knowledge graph to the second device of each of a plurality of vehicles.

[0006] The technologies of the first to third overall aspects may have one or more of the following advantages.

[0007] First, the technology disclosed herein provides the possibility of obtaining, through a vehicle knowledge graph, in a form accessible and callable during vehicle operation, expert knowledge about vehicle components, their potential problems / faults, and related vehicle signal data necessary for calculating state parameters. Therefore, compared to some prior art techniques, this enables detailed vehicle diagnostics of vehicle components at runtime, providing an improved understanding of the operational status of vehicle components compared to on-board diagnostics (“OBD”) known in the prior art.

[0008] Secondly, this technology can help: effectively reduce the cost of diagnostics for vehicle components by proactively providing and controlling the operational status of each component. In cases of deteriorating operational status, the vehicle can proactively notify the vehicle's user or a third party (e.g., via a cloud computing system). This allows for faster problem identification and saves time and money during diagnosis at a repair shop, as diagnostics don't have to be performed based on limited information in the initial period after a problem occurs. Furthermore, this technology can provide an overview of all relevant data regarding the health status of vehicle components before the vehicle actually arrives at the repair shop.

[0009] Third, the technology disclosed herein enables the analysis of the operational status of vehicle components through a cloud vehicle knowledge graph, which stores (e.g., the operational status of multiple vehicles of a particular model or with specific components) and other relevant vehicle data. This technology particularly includes statistical analysis and learning of the cloud vehicle knowledge graph regarding "fleets" (e.g., more than 1000 or more than 10000 vehicles). Through this, the knowledge graph is continuously improved (e.g., identifying which state parameters are needed to evaluate the operational status of vehicle components). In many prior art methods, this data is neither available nor analyzed because vehicles neither transmit operational status data for each vehicle component nor collect this operational status data and other relevant data in the cloud vehicle knowledge graph.

[0010] In this disclosure, some terms are used as follows:

[0011] The term "vehicle" includes any device designed for the transport of passengers and / or goods. A vehicle can be a motor vehicle (e.g., a motor vehicle that is at least partially autonomous / assisted, especially a passenger vehicle (PKW) or freight vehicle (LKW)). However, the vehicle can also be a ship, train, aircraft, or spacecraft.

[0012] The term "vehicle component" is understood to mean any internal component of a vehicle. A vehicle component can be an engine (e.g., a combustion engine, electric motor, hybrid engine, or fuel cell, or a portion of an engine such as a turbocharger), control units (e.g., engine controller), battery packs or other energy storage systems, drivetrain components (e.g., transmission), auxiliary systems (e.g., brake assist, lane keeping assist, parking assist), air conditioning systems, sensors or sensor systems (e.g., camera-based systems, LiDAR systems, RADAR systems, ultrasonic sensor systems), or electronic systems for controlling functions within the interior space. A vehicle component can also be a part of or a combination of several of the aforementioned systems.

[0013] Correspondingly, the term "vehicle parameters" includes a variety of parameters that can play a role in determining the operating state of vehicle components; that is, these parameters can characterize the operating state of one or more vehicle components. A non-exhaustive list of such parameters includes: temperature (e.g., internal or external temperature), pressure (e.g., atmospheric pressure), rotational speed, torque, mass flow, wear parameters, friction parameters, current intensity, voltage, power, efficiency, pressure ratio, speed, acceleration, and so on.

[0014] The term "signal" can include all signals detected for the vehicle during vehicle operation or detected in the past. This refers to signals of vehicle components, generated, for example, by corresponding sensors equipped with or located near those vehicle components.

[0015] In the following text, the term "knowledge graph" refers to all methods of describing data points (e.g., data depicting corresponding signals from vehicle components) at a semantic level; that is, methods in which these data points themselves are fed into a semantic model to make them available. These data points can be interpreted as data entities of concepts, which are defined in the conceptual model using corresponding relationships between them. Furthermore, attributes can be assigned to these concepts. Here, a knowledge graph consists of its conceptual model and the data entities of the concepts defined in that conceptual model. Further elaboration can be found below.

[0016] The term "data cloud system" or "cloud-computing system" refers to infrastructure that makes a system available via a network, such as the Internet. A "data cloud system" typically includes storage space, computing power, and / or application software as services (meaning users can access these resources via the network). In other words, a "data cloud system" is infrastructure provided via a network, and this infrastructure does not necessarily need to exist or be installed on a local system. A "data cloud system" can contain distributed resources (e.g., multiple computer systems in different locations). Here, the provisioning and use of resources in a "data cloud system" are achieved through technical interfaces and protocols, such as via a web browser. Attached Figure Description

[0017] Figure 1 An example of a method 140 for determining the operational status of vehicle components based on a first knowledge graph 10 is illustrated.

[0018] Figure 2 The structure of the first knowledge graph 1 for a vehicle with two vehicle components, Komp A and Komp B, is schematically shown. These two vehicle components generate two signals, S-CA-1, S-CA-2 and S-CB-1, S-CB-2, respectively. The parameter Param-H1 is calculated based on the two signals S-CA-2 and S-CB-2, and this parameter is further used to determine the operating state of vehicle component Komp B.

[0019] Figure 3 The illustration schematically shows possible designs and other aspects of a method 20 for determining the operating state of vehicle components. Detailed Implementation

[0020] First, based on Figure 1 This describes the techniques used to determine the operating state of vehicle components. Then, based on... Figure 2 Let's discuss the exemplary structure of the first knowledge graph 10. Finally, refer to... Figure 3 This disclosure aims to present possible design options and other aspects.

[0021] like Figure 1 As shown, the first general aspect relates to a method 140 for determining the operational state of vehicle components of a vehicle. The method includes: calculating 100 a first knowledge graph 10, which contains information about multiple vehicle components Komp A, Komp B of the vehicle FA (see also...). Figure 2 The name "first knowledge graph" does not necessarily imply the existence of a second knowledge graph. This "first knowledge graph" is only used to distinguish between multiple knowledge graphs in the event of their use. First knowledge graph 10 may, for example, be a local vehicle knowledge graph implemented on a unit of the vehicle diagnostic system or other units of the vehicle (i.e., the vehicle knowledge graph is stored within the vehicle). In other examples, knowledge graph 10 may be stored on a remote system (i.e., outside the vehicle) and connected to the vehicle via an appropriate communication interface (more on this below).

[0022] As described above, the knowledge graph can contain concepts, attributes, and relationships between the corresponding concepts, where each concept, for example, describes a vehicle component (modeling that vehicle component). Thus, in some examples, this first knowledge graph can be used to calculate the state parameters of the vehicle components and / or to determine the operational state of the vehicle components. Furthermore, for example, each of these vehicle components can provide one or more signals S-CA-1, S-CA-2, S-CB-1, S-CB-2. In some examples, as described above, these signals can be generated by corresponding sensors. Figure 2 The example schematically illustrates the structure of a first knowledge graph 10 for a vehicle with two vehicle components, Komp A and Komp B (this structure is greatly simplified relative to many "real-world" scenarios for illustrative purposes). In this example, the two components generate two signals, S-CA-1, S-CA-2 and S-CB-1, S-CB-2, respectively. In other examples, a vehicle may contain more vehicle components (e.g., more than 10 or more than 100 vehicle components), which are depicted in the knowledge graph. Each of these components may, in turn, generate one or more signals (e.g., more than 5 or more than 10 signals). Thus, Figure 2 Only a simplified structure of the first knowledge graph 10 is illustrated, or in some examples, in Figure 2The structure shown may only represent the substructures of the first knowledge graph.

[0023] The technology disclosed herein also includes: determining 140 the operating state of vehicle component Komp B from the plurality of vehicle components Komp A and Komp B. Here, this determination may include: when using a first knowledge graph 10, selecting 110 multiple signals and one or more state parameters Param-H1 required for determining the operating state of the vehicle component. In other words, the first knowledge graph can be queried about which signals and which state parameters are relevant to determining the operating state of the vehicle component (i.e., the vehicle component considered at a specific point in time during the operation of the vehicle FA). Figure 2 In the example, two signals are needed: the second signal S-CA-2 of the first vehicle component Komp A and the second signal S-CB-2 of the second vehicle component Komp B, along with a separate state parameter Param-H1, to determine the operating state of the second vehicle component Komp B. Figure 2 In this document, the single parameter Param-H1 is only disclosed as an example, and in other cases, multiple state parameters may be required to determine the operating state of a vehicle component (e.g., more than 5 or more than 10 state parameters). Alternatively or additionally, only one signal or more than two signals may be needed to determine the operating state of a vehicle component.

[0024] The "determining" step of the method may further include: calculating one or more state parameters Param-H1 for the vehicle component based on the selected plurality of signals. For example, a vehicle diagnostic system can be used to calculate the state parameters. In some examples, one or more signals from the vehicle component may be needed to calculate the state parameters of the same vehicle component. In other examples, one or more signals from a first vehicle component may be needed to calculate the state parameters of a second vehicle component. In this regard, in Figure 2 The diagram may illustrate the aforementioned second signals S-CA-2 and S-CB-2 for these two vehicle components, which are used to calculate the individual state parameter Param-H1. In other examples, the state parameter may depend on multiple signals (e.g., more than 5 or more than 10 state parameters), which are therefore used to calculate the state parameter.

[0025] Finally, the operating state of the vehicle components can be calculated based on information contained in the first knowledge graph 10 and on one or more calculated state parameters Param-H1. For example, the calculation of the operating state can be performed using a vehicle diagnostic system. In this context, the information contained in the first knowledge graph can include relationships between two or more vehicle components Komp A, Komp B from the plurality of vehicle components of the vehicle. For example, this relationship can be based on the association between one or more signals S-CA-2 and S-CB-2 used to determine the operating state of the vehicle components and the state parameter Param-H1. In some examples, this method of the present disclosure can be used to determine the operating state of two or more vehicle components of a vehicle. In other examples, the method can be used to determine the operating state of a larger or smaller number of vehicle components of a vehicle.

[0026] As described above, the first knowledge graph in this technology may include a conceptual model with multiple concepts. In one example, each concept may depict a corresponding vehicle component from the multiple vehicle components. Furthermore, these concepts may be equipped with attributes, and corresponding relationships between these concepts may be defined. Here, concepts may be considered as nodes 11, 12 of knowledge graph 10, and the relationships between these concepts may be considered as edges 13, 14, which connect these concepts (or nodes) to each other. The multiple concepts may, for example, depict a vehicle having all or a portion of these vehicle components. The relationships between these concepts may be: for example, requiring a signal from a first component to calculate state parameters of a second vehicle component; or conversely, requiring a signal from a second component to calculate state parameters of a first vehicle component. In other examples, the relationships between these concepts may indicate: querying the same state parameter from two or more vehicle components. In some examples, the first knowledge graph may include multiple data entities. These data entities may have data points from one or more signals. Additionally or alternatively, data entities may also contain data points from one or more state parameters of each vehicle component. These data entities may also be assigned to corresponding concepts in the conceptual model. Furthermore, the first knowledge graph can be configured to receive one or more signals and one or more state parameters from each vehicle component, and can also be designed to transfer the received signals and state parameters into a conceptual model. Additionally or alternatively, the first knowledge graph can also be designed to transfer the received signals and state parameters into these data entities. In some examples, receiving one or more signals may include generating data points (e.g., time series of data points) from the one or more signals. Additionally or alternatively, data points (e.g., time series of data points) may be generated from the one or more state parameters. In some examples, data points may be generated after the signals (e.g., signals detected by sensors) have been processed accordingly by a vehicle diagnostic system or other units of the vehicle.

[0027] according to Figure 3 The next step of the exemplary method includes storing 210 data points (e.g., time series of the generated data points) with corresponding timestamps, generated from one or more signals from vehicle components, in a first knowledge graph 10. In some examples, these data points may be stored as data entities in the first knowledge graph.

[0028] In some examples of this technology, the state parameters of the vehicle components can be calculated continuously (e.g., via a vehicle diagnostic system) using a first knowledge graph (e.g., more than once a day or more than once an hour while the vehicle is running). In other examples, the state parameters of the vehicle components are calculated at predetermined time points using the first knowledge graph. Furthermore, each state parameter of the vehicle component can be stored in the first knowledge graph (e.g., as a data entity). Here, each state parameter can be equipped with a corresponding timestamp.

[0029] In some examples, the operating state of a vehicle component can be calculated based on one or more current values ​​of the one or more state parameters. In other examples, the operating state of a vehicle component can be calculated based on one or more values ​​of one or more state parameters of the vehicle component stored in a first knowledge graph and / or one or more values ​​of these signals stored in the first knowledge graph. In other words, the knowledge graph can also use historical values ​​of state parameters and / or signals, which are considered for calculating the operating state of the vehicle component. In some examples, the calculation of state parameters can be performed in an event-triggered manner at a specific point in time. In other examples, the calculation of state parameters is performed periodically at specific points in time, given a calculation frequency. For example, the calculation frequency can be provided by the first knowledge graph. In a preferred example, the calculation frequency for each state parameter of the plurality of vehicle components can be determined based on the first knowledge graph. In some examples, the calculation frequency of the first state parameter can be different from the calculation frequency of the second state parameter. For example, this can involve the first and second state parameters of the same vehicle component. In another example, the first and second state parameters can be assigned to different vehicle components. In some examples, the calculation frequency for two or more state parameters can be the same.

[0030] In this disclosure, the determination of the operating state of each of the plurality of vehicle components can be performed. This results in a plurality of operating states. In some examples, the operating state of the vehicle can be determined based on the plurality of operating states. Furthermore, the determination of the operating state of a vehicle component may include classifying the operating state of the vehicle component (or the vehicle) 230. In some examples, the operating state can be classified as faulty (abnormal) or faultless (normal) based on the calculated operating state of the vehicle component. In some examples, the presence of a fault may indicate that the operating state of the vehicle component has recently deteriorated in a predetermined manner; and / or the service life of the vehicle component is nearing its end (e.g., due to wear of vehicle components such as the braking system). In other examples, the classification may include additional and / or additional categories. In some examples, there may be two or more fault categories indicating different types of faults and / or different degrees of severity. Additionally or alternatively, there may be one or more categories indicating an operating state that is still normal, but which is changing in the direction of a fault (i.e., one or more categories predicting the occurrence of a fault). In addition to classification, the determination of the operating state can also be achieved through regression of information contained in the first knowledge graph and based on one or more calculated state parameters.

[0031] In some examples, within this context, the deviation between one or more current values ​​of one or more state parameters of a vehicle component and one or more corresponding values ​​of those state parameters stored in a first knowledge graph can be calculated. For example, when the deviation exceeds a predefined size, the operating state of the vehicle component can be classified as faulty and abnormal. Otherwise, the operating state of the vehicle component can be classified as fault-free and normal. In other examples, one or more corresponding thresholds can be assigned to one or more state parameters of the vehicle component to determine whether the operating state of the vehicle component is normal or abnormal (or whether other fault categories exist). For example, a corresponding threshold can be assigned to each state parameter of the vehicle component. Then, when one or more state parameters of the vehicle component exceed the one or more corresponding thresholds, the operating state of the vehicle component can be classified as faulty and abnormal (or classified as another fault category). Otherwise, the operating state of the vehicle component can be classified as fault-free and normal. In other examples, this criterion can be applied in reverse and includes classifying the operating state of the vehicle component as abnormal when one or more state parameters of the vehicle component fall below the one or more corresponding thresholds, and otherwise classifying the operating state of the vehicle component as normal. In other examples, calculating the operational state of a vehicle component may include calculating an average operational state parameter of the vehicle component, which depends on a weighted sum of all operational state parameters of the vehicle component. In this regard, a corresponding weight may be assigned to each operational state parameter of the vehicle component. In a preferred example, the weights assigned to these operational state parameters are determined using a first knowledge graph.

[0032] In some examples, an average threshold can be assigned to the average operating status parameter. Furthermore, for example, when the average operating status parameter of a vehicle component exceeds the average threshold, the operating status of the vehicle component can be classified as faulty (abnormal); otherwise, the operating status of the vehicle component can be classified as fault-free (normal). In other examples, this criterion can be applied in reverse and includes: when the average operating status parameter of a vehicle component drops below the average threshold, the operating status of the vehicle component is classified as abnormal; otherwise, the operating status of the vehicle component is classified as normal.

[0033] In this disclosure, the first knowledge graph may contain operational state indicators (defined, for example, as additional features in the knowledge graph) for each operational state parameter, representing abnormal operational states of vehicle components, thereby forming multiple operational state indicators. In some examples, each operational state indicator may include a deviation and a predefined magnitude corresponding to the corresponding state parameter Param-H1 of the vehicle component. In other examples, each operational state indicator may include a threshold corresponding to the corresponding state parameter of the vehicle component. In still other examples, each operational state indicator may include an average threshold and an assigned weight corresponding to the corresponding state parameter of the vehicle component.

[0034] In the technology disclosed herein, abnormal operating states of vehicle components can be stored in a first knowledge graph along with corresponding operating state parameters. In some examples, the abnormal operating states of vehicle components may also be timestamped. Furthermore, a reaction can be triggered based on the classified operating states of one or more vehicle components. In some examples, the reaction may include displaying the operating state on a graphical user interface. Additionally or alternatively, the triggered reaction may include providing information about the operating states of one or more vehicle components to a remote system (e.g., for notifying the original equipment manufacturer or auto repair shop) or to the vehicle's unit (e.g., for notifying the driver). In other examples, information about the operating state can be displayed on the user's mobile device. For example, the reaction may include generating an alarm or warning message based on the classified abnormal operating state of a vehicle component as faulty. Additionally or alternatively, in this case, the operation of the vehicle may be stopped or altered. Furthermore, as a reaction to the classified abnormal operating states of vehicle components, an event log file may be generated. In some examples, a notification may also be transmitted to the remote system. When the classified operating state is abnormal, the reaction may include, for example, a notification that the vehicle requires maintenance or repair. Conversely, when the operating status has been classified as normal, the response may include the notification that the vehicle is operating normally. In other examples, the above response may also be implemented as a reaction to identifying other faults mentioned above.

[0035] The technology disclosed herein may further include updating one or more values ​​of the computational frequency in response to an anomaly that classifies the operating state of a vehicle component as a faulty anomaly, when using a first knowledge graph. In some examples, this update may include learning one or more values ​​of the computational frequency via the first knowledge graph. In some examples, this learning may be based on data entities stored in the first knowledge graph. For example, the update may include learning one or more values ​​of the computational frequency via the first knowledge graph based on one or more values ​​of the state parameters of the vehicle component stored in the first knowledge graph. In some examples, stored values ​​of the state parameters may be used for which the operating state of the vehicle component has been classified as a faulty anomaly. Additionally or alternatively, in response to an anomaly that classifies the operating state of a vehicle component as a faulty anomaly, new multiple signals of the plurality of vehicle components required for determining the operating state of the vehicle components may also be determined using the first knowledge graph. In some examples, in response to an anomaly that classifies the operating state of a vehicle component as a faulty anomaly, the number of signals in the new multiple signals required for determining the operating state of the vehicle component may differ from the number of signals in the multiple signals.

[0036] In a next step of the method, one or more learned computation frequencies can be invoked from the first knowledge graph. Furthermore, the operating state of one or more vehicle components from the plurality of vehicle components can be determined at a next time point according to the corresponding invoked computation frequencies from the first knowledge graph. In some examples, determining the operating state at the next time point may include invoking a data entity stored in the first knowledge graph. In other examples, alternatively or additionally, determining the operating state at the next time point may include invoking one or more values ​​of a state parameter stored in the first knowledge graph. In other examples, determining the operating state at the next time point may include invoking 250 new multiple signals required for determining the operating state of the vehicle component. Determining the operating state at the next time point may also include sending a request 260 to the plurality of vehicle components via the first knowledge graph to obtain the new multiple signals. This can provide 270 the new multiple signals required from the plurality of vehicle components for determining the operating state of the vehicle component at the next time point.

[0037] The technology disclosed herein includes transmitting to a remote system the classified operating states (e.g., faulty anomalies) of 300 vehicle components and / or one or more corresponding state parameters that cause the operating state of the vehicle component to be correspondingly determined (e.g., determined to be a faulty anomaly). In some examples, the classified operating states (e.g., faulty anomalies) of the vehicle components and / or the corresponding one or more state parameters may be transmitted to a second knowledge graph. In some examples, the second knowledge graph may be stored in a data cloud system (or other computer systems remote from the vehicle). Thus, the second knowledge graph may also be referred to as a data cloud-vehicle knowledge graph. In some examples, the method may include transmitting to the second knowledge graph the classified operating states (e.g., classified as faulty anomalies) of more than 300 vehicle components or each vehicle component and / or one or more corresponding state parameters that cause the operating state of the corresponding vehicle component to be correspondingly determined (e.g., determined to be a faulty anomaly). In some examples, the corresponding one or more state parameters may be transmitted together with one or more corresponding operating state indicators. Additionally or alternatively, these state parameters may also be transmitted together with a plurality of corresponding signals. The method may further include: transmitting data entities stored in the first knowledge graph to the second knowledge graph.

[0038] The next step may include: storing the transmitted classified operating states and / or one or more status parameters and / or signals of one or more vehicle components of a vehicle, in some cases together with one or more corresponding operating state indicators, as data entities in a second knowledge graph 310. Additionally or alternatively, the status parameters may also be stored as data entities in the second knowledge graph together with multiple corresponding signals. The method may also include: reporting to an operator (e.g., staff at the original equipment manufacturer or in an auto repair shop) or the vehicle driver when the operating state of one or more vehicle components of a vehicle indicates a malfunction 320. In some examples, the report may also include information about the corresponding one or more status parameters, for example, together with one or more corresponding operating state indicators. In one example, the second knowledge graph may include: classified operating states and / or one or more status parameters and / or signals of vehicle components of multiple vehicles. In this way, the second knowledge graph can collect a large amount of "missing data" of various types about a "fleet." For example, such a "fleet" may include 10 or more vehicles, 100 or more vehicles, or 1000 or more vehicles. These vehicles may have certain commonalities. In some cases, these vehicles may be of a specific type or model. Alternatively or additionally, these vehicles may have one or more vehicle components of the same type.

[0039] In the next step, the conceptual model of the second knowledge graph can be updated based on information about the vehicle components of the transmitted plurality of vehicles (e.g., operating states, state parameters, and / or signals). In some examples, updating the conceptual model of the second knowledge graph may include: using statistical analysis of 330 about the information about the vehicle components of the transmitted plurality of vehicles (e.g., operating states, state parameters, and / or signals) to optimize the determination of the operating states of the vehicle components. For example, the conceptual model of the second knowledge graph can be updated based on data entities stored in the second knowledge graph. In this embodiment, updating the conceptual model of the second knowledge graph may include: using statistical analysis about the data entities stored in the second knowledge graph to optimize the determination of the operating states of the vehicle components. In some examples, the use of statistical analysis may include: learning one or more operating state indices corresponding to the operating state parameters. The purpose of doing so may be, for example, to update the values ​​included in the operating state indices of deviation and predefined magnitude. Alternatively or additionally, thresholds and / or average thresholds with assigned weights in the operating state indices of the operating state parameters may also be updated in this way.

[0040] In the next step, the technique may include: determining one or more new state parameters required for determining the operating state of vehicle components based on the updated conceptual model. Additionally or alternatively, multiple new signals required for determining the operating state of vehicle components may also be determined. In some examples, a new computation frequency for one or more state parameters of the multiple vehicle components may be determined based on the updated conceptual model. Furthermore, the conceptual model of the first knowledge graph from the multiple vehicles may be updated based on the updated conceptual model of the second knowledge graph. For example, this disclosure may include: transmitting 340 update information (e.g., via a data cloud system) to the first knowledge graph of each of the multiple vehicles, the update information being contained in the updated conceptual model of the second knowledge graph. Finally, in some examples, the transmitted update information may be stored 350 in the first knowledge graph of each vehicle.

[0041] The second general aspect of this disclosure relates to an electronic system comprising a first electronic device and a second electronic device, the electronic system being designed to perform the methods described according to the first general aspect of this disclosure. The first electronic device may be an electronic device of a vehicle for using a first knowledge graph (e.g., a unit of a vehicle diagnostic system or other independent unit of the vehicle). The second electronic device (e.g., a vehicle diagnostic system) may be used to monitor and receive signals from multiple vehicle components and to calculate state parameters of the multiple vehicle components. The system may have at least one processor, at least one memory (which may contain programs that, when executed, implement the methods of this disclosure), and at least one interface for input and output.

[0042] In the example above, a local first knowledge graph (i.e., stored in the vehicle and describing its use) was used in part. In other examples, the first knowledge graph may be stored on a remote system. For example, the vehicle may transmit multiple signals to the remote system (e.g., periodically or upon request). The steps of this disclosure for determining the operational state of vehicle components can then be performed on that remote system. The determined operational state may be transmitted back by the remote system to the vehicle and / or a third-party location (e.g., to trigger one of the aforementioned responses). In other cases, the first knowledge graph may be stored on a remote system and collaborate with units within the vehicle to determine the operational state of vehicle components.

[0043] In other examples, the knowledge graph stored on a remote system may contain information about multiple vehicles (as described above with respect to the second knowledge graph) and perform the steps of this disclosure to determine the operational status of vehicle components.

[0044] A third general aspect of this disclosure relates to a data cloud system designed to receive information from a second device as described in the second general aspect, and to perform the method described in the first general aspect of this disclosure to update a conceptual model of a second knowledge graph. The data cloud system can also be configured to transmit updated information contained in the updated conceptual model of the second knowledge graph to a second device of each of a plurality of vehicles. In some examples, two or more data cloud systems may also be used in this respect.

[0045] This disclosure also relates to computer programs that perform the methods of this disclosure. This disclosure also relates to computer-readable media and signals that store or encode the computer programs of this disclosure.

Claims

1. A method for determining the operating state of vehicle components of a vehicle, wherein the method includes the following steps: Provide (100) a first knowledge graph (10) containing information about multiple vehicle components (Komp A; Komp B) of a vehicle (FA), wherein each vehicle component provides one or more signals (S-CA-1; S-CA-2; S-CB-1; S-CB-2). Determining (140) the operating state of a vehicle component from the plurality of vehicle components, wherein determining the operating state of a vehicle component from the plurality of vehicle components includes the following steps: When using the first knowledge graph, select (110) multiple signals and one or more state parameters (Param-H1) required to determine the operating state of the vehicle components. (120) Calculate one or more state parameters (Param-H1) of the vehicle component based on the selected multiple signals. The operating state of the vehicle component is determined (130) based on information contained in the first knowledge graph and based on one or more calculated state parameters. in, The method further includes: when using the first knowledge graph, continuously calculating (120) the state parameters of the vehicle component or calculating (120) the state parameters at a specific point in time. The state parameters are calculated at specific time points using a calculation frequency, wherein the calculation frequency is provided by the first knowledge graph. The method further includes: updating one or more values ​​of the computational frequency in response to a classification result of the operating state of the vehicle components, when using the first knowledge graph, wherein the update includes: learning the one or more values ​​of the computational frequency based on data entities stored in the first knowledge graph when using the first knowledge graph.

2. The method according to claim 1, wherein the first knowledge graph comprises: A conceptual model comprising multiple concepts, each of which describes a corresponding vehicle component from the multiple vehicle components, wherein the concepts are equipped with attributes and the corresponding relationships between the concepts are defined; Multiple data entities, each having data points from one or more signals and / or one or more state parameters from each vehicle component, are assigned to corresponding concepts in the conceptual model. The first knowledge graph is configured to receive one or more signals and one or more state parameters of each vehicle component, and is also designed to transfer the received signals and state parameters into the conceptual model and / or into the data entity, wherein receiving the one or more signals further includes generating data points from the one or more signals or from the one or more state parameters.

3. The method of claim 2, wherein receiving the one or more signals further comprises generating a time series of data points from the one or more signals or from the one or more state parameters.

4. The method according to any one of claims 1 to 3, wherein determining the operating state of the vehicle component further includes: Based on the calculated operating state of the vehicle components, the operating state of the vehicle components is classified (230) into abnormal with faults or normal without faults.

5. The method of claim 4, wherein the calculation of the operating state of the vehicle component includes: Calculate the deviation between one or more current values ​​of the one or more state parameters of the vehicle component and one or more values ​​of the one or more state parameters of the vehicle component stored in the first knowledge graph. When the deviation exceeds or falls below a predetermined value, the operating state of the vehicle component is classified as abnormal; otherwise, the operating state of the vehicle component is classified as normal without fault.

6. The method of claim 4, wherein the first knowledge graph includes operational status indicators of abnormal operational states of the vehicle components for each operational status parameter, and wherein a plurality of operational status indicators are thus formed. Each operational status indicator includes a deviation and a predefined magnitude, which correspond to the corresponding status parameter (Param-H1) of the vehicle component.

7. The method according to claim 4, further comprising: The second knowledge graph transmits (300) the classified operating state of the vehicle component and one or more corresponding state parameters that cause the operating state of the vehicle component to be determined as faulty, wherein the one or more corresponding state parameters are transmitted together with one or more corresponding operating state indicators and / or together with multiple corresponding signals.

8. The method of claim 7, wherein the second knowledge graph includes state parameters of vehicle components of a plurality of vehicles transmitted, wherein the transmitted operating states have been classified as faulty anomalies, and wherein the method further includes updating the conceptual model of the second knowledge graph based on the state parameters of the vehicle components of the plurality of vehicles transmitted, wherein updating the conceptual model of the second knowledge graph includes: using (330) statistical analysis of the state parameters of the vehicle components of the plurality of vehicles transmitted to optimize the determination of the operating states of the vehicle components.

9. The method according to claim 8, wherein the method comprises: The concept model of the first knowledge graph from the plurality of vehicles is updated based on the updated concept model of the second knowledge graph. The method further includes transmitting (340) update information to a first knowledge graph from each of the plurality of vehicles, the update information being contained in an updated conceptual model of the second knowledge graph, and storing (350) the transmitted update information in the first knowledge graph of each vehicle.

10. An electronic vehicle system for determining the operating state of vehicle components of a vehicle, the electronic vehicle system being designed to perform the steps of the method according to any one of claims 1 to 9, the system comprising: A first electronic device for using the first knowledge graph; A second electronic device, the second electronic device being configured to: Monitor and receive signals from multiple vehicle components of the vehicle. Calculate the state parameters of the plurality of vehicle components.

11. A data cloud system, said data cloud system being designed for: Information is received from a second electronic device of an electronic vehicle system for determining the operating status of vehicle components, as claimed in claim 10. Perform the steps of the method according to any one of claims 7 to 9 in order to update the conceptual model of the second knowledge graph; A second device transmits updated information contained in the updated conceptual model of the second knowledge graph to each of the plurality of vehicles.

Images