Vehicle and road cloud system data anomaly detection method, control device, vehicle and program product

Abstract

The invention discloses a vehicle and road cloud system data anomaly detection method, a control device, a vehicle and a program product. The method comprises the steps of dynamically collecting a vehicle and road cloud system operation data set; inputting the vehicle and road cloud system operation data set to an anomaly detection integration platform, performing data anomaly detection through the anomaly detection integration platform, and outputting an anomaly detection result corresponding to each anomaly detection item; and the anomaly detection integration platform comprises a hierarchical detection system corresponding to each anomaly detection item. According to the invention, automatic data anomaly detection of the vehicle-road cloud system can be realized based on the anomaly detection integrated platform, the problems of cross-domain anomaly relevance deficiency, long debugging period and difficulty in fault positioning of the vehicle-road cloud system are solved, the data anomaly detection efficiency and accuracy of the vehicle-road cloud system are improved, and the method can be widely applied to the technical field of data detection.

Description

Technical Field

[0001] This application relates to the field of data detection technology, and in particular to a method, control device, vehicle, and program product for detecting data anomalies in a vehicle-road-cloud system. Background Technology

[0002] The vehicle-road-cloud integrated system achieves traffic environment perception, real-time decision-making, and global optimization through the collaboration of vehicle terminals (OBU), roadside units (RSU), edge computing nodes, and a cloud platform. The system involves multi-source heterogeneous data (such as sensor data, V2X communication messages, high-precision maps, and cloud computing logs), and its anomaly detection mainly relies on the following technologies: 1) Single-point detection: Independent analysis of vehicle, roadside, or cloud data, lacking cross-level correlation (e.g., abnormal roadside camera data is not linked to vehicle decision-making logic). 2) Static rule engine: Alarms are triggered based on fixed thresholds (such as communication delay > 100ms, GPS positioning deviation > 2m), which cannot adapt to dynamic traffic scenarios (such as sudden congestion, severe weather); 3) Traditional machine learning models: use isolation forests (pre-configured rule-based forests) or statistical methods (3σ principle) to detect outliers, but they are not good at fusion analysis of multimodal data (images + time series + text).

[0003] Therefore, the vehicle-road-cloud system suffers from low efficiency and accuracy in anomaly detection, insufficient dynamic adaptability in anomaly detection, and long debugging cycle and difficulty in fault location. Summary of the Invention

[0004] The main objective of this application is to propose a method, control device, vehicle, and program product for detecting data anomalies in a vehicle-road-cloud system, which can realize automated anomaly detection of data in the vehicle-road-cloud system and improve the efficiency and accuracy of anomaly detection.

[0005] On the one hand, this application proposes a method for detecting data anomalies in a vehicle-road-cloud system, the method comprising the following steps: Dynamically collect operational datasets from the vehicle-road-cloud system; The vehicle-road-cloud system's operational dataset is input into the anomaly detection integration platform, which performs data anomaly detection and outputs the anomaly detection results corresponding to each anomaly detection item. The anomaly detection integration platform includes a hierarchical detection system corresponding to each anomaly detection item.

[0006] In some embodiments, the dynamically collected vehicle-road-cloud system operation dataset specifically includes: The system dynamically collects the module operation datasets corresponding to each system deployment module in the vehicle-road-cloud system; the system deployment modules include the vehicle terminal processing module, the roadside facility operation module, and the system cloud processing module. Perform data preprocessing on the runtime datasets of each module to determine the preprocessed runtime datasets of each module. The vehicle-road-cloud system's operating dataset is determined based on the preprocessed data datasets of each module.

[0007] In some embodiments, the method further includes: Construct the hierarchical detection system for each of the aforementioned anomaly detection items and the associations between each of the aforementioned anomaly detection items; Based on each of the aforementioned anomaly detection items and the corresponding hierarchical detection system, an anomaly detection integration platform is established. The anomaly detection integration platform is managed, including updating each anomaly detection item and dynamically updating the hierarchical detection system corresponding to each anomaly detection item.

[0008] In some embodiments, the hierarchical detection system includes a rule filtering unit and a model analysis unit. The step of inputting the vehicle-road cloud system's runtime dataset into an anomaly detection integration platform, performing data anomaly detection through the anomaly detection integration platform, and outputting the anomaly detection results corresponding to each anomaly detection item specifically includes: The vehicle-road-cloud system operation dataset is input into the hierarchical detection system corresponding to each of the anomaly detection items; For each of the aforementioned anomaly detection items, the rule filtering unit performs rule filtering based on pre-configured rule conditions to extract the corresponding target data subset from the vehicle-road-cloud system's operational dataset; For each of the aforementioned anomaly detection items, the corresponding target data subset is input to the model analysis unit. Using a pre-configured anomaly detection model, anomaly analysis is performed on the target data subset, and the abnormal operation data corresponding to the anomaly detection item and the target anomaly type corresponding to the abnormal operation data are output from the target data subset.

[0009] In some embodiments, the step of performing rule filtering based on pre-configured rule conditions through the rule filtering unit for each of the anomaly detection items, and extracting the corresponding target data subset from the vehicle-road-cloud system's operational dataset, specifically includes: The rule filtering unit extracts a subset of operational data corresponding to the set of fields to be detected from the vehicle-road-cloud system's operational dataset; the subset of operational data includes multiple fields to be detected. According to the rules, each of the data fields to be detected contained in the running data subset is matched, and the target data subset that meets the rules is output. The rules include threshold comparison, logical operation and time window constraint.

[0010] In some embodiments, for each of the anomaly detection items, the corresponding target data subset is input to the model analysis unit, and anomaly analysis is performed on the target data subset using a pre-configured anomaly detection model. The result is the output of the abnormal operation data corresponding to the anomaly detection item in the target data subset and the target anomaly type corresponding to the abnormal operation data. Specifically, this includes: Obtain the anomaly detection model, which includes at least one of a time-series prediction model, a clustering model, and a classification model; Using the anomaly detection model, anomaly probability calculation is performed on each running data in the target data subset according to multiple preset anomaly types. When the anomaly probability of any preset anomaly type corresponding to the running data is greater than a preset probability threshold, the running data is output as the abnormal running data. The target anomaly type corresponding to the abnormal running data is determined from the multiple preset anomaly types, wherein the target anomaly type is the preset anomaly type with the highest anomaly probability.

[0011] In some embodiments, managing the anomaly detection integration platform specifically includes: For each of the hierarchical detection systems, the rule conditions and the set of field types to be detected in the rule filtering unit are dynamically updated, wherein the update of the rule conditions is based on historical anomaly detection results and user-defined input. For each of the hierarchical detection systems, the anomaly detection model corresponding to the model analysis unit is dynamically updated. The update of the anomaly detection model is based on the historical anomaly detection results and the user's model parameter update operation. The information of each anomaly detection item is managed, including updating the anomaly detection item description information and the associated information of the hierarchical detection system.

[0012] On the other hand, embodiments of this application propose a data anomaly detection device for a vehicle-road-cloud system, the device comprising: The first module is used to dynamically collect the operational dataset of the vehicle-road-cloud system. The second module is used to input the vehicle-road-cloud system operation dataset into the anomaly detection integration platform, perform data anomaly detection through the anomaly detection integration platform, and output the anomaly detection results corresponding to each anomaly detection item; the anomaly detection integration platform includes a hierarchical detection system corresponding to each anomaly detection item.

[0013] On the other hand, embodiments of this application propose an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-mentioned vehicle-road-cloud system data anomaly detection method.

[0014] On the other hand, embodiments of this application propose a computer program product, including a computer program that, when executed by a processor, implements the above-described method for detecting abnormal data in a vehicle-road-cloud system.

[0015] The embodiments of this application include at least the following beneficial effects: This application provides a method, control device, vehicle, and program product for detecting data anomalies in a vehicle-road-cloud system. It dynamically collects operational datasets from the vehicle-road-cloud system, inputs these datasets to an anomaly detection integration platform, performs data anomaly detection through the platform, and outputs the anomaly detection results corresponding to each anomaly detection item. The anomaly detection integration platform includes a hierarchical detection system corresponding to each anomaly detection item. This application can achieve automated anomaly detection of vehicle-road-cloud systems based on the anomaly detection integration platform, solving the problems of lack of cross-domain anomaly correlation, long debugging cycles, and difficulty in fault location in vehicle-road-cloud systems, thereby improving the efficiency and accuracy of data anomaly detection in vehicle-road-cloud systems. Attached Figure Description

[0016] Figure 1 This is a flowchart of a method for detecting data anomalies in a vehicle-road-cloud system provided in an embodiment of this application; Figure 2 This is a flowchart of step S101 in the embodiments of this application; Figure 3 This is a flowchart of step S102 in the embodiments of this application; Figure 4 This is a flowchart of step S302 in the embodiments of this application; Figure 5 This is a flowchart of step S303 in the embodiments of this application; Figure 6 This is another flowchart of a method for detecting data anomalies in a vehicle-road-cloud system provided in an embodiment of this application; Figure 7 This is a schematic diagram of the structure of the anomaly detection integration platform in this application embodiment; Figure 8 This is a flowchart of step S603 in the embodiments of this application; Figure 9 This is a schematic diagram of the structure of a vehicle-road-cloud system data anomaly detection device provided in an embodiment of this application; Figure 10 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit it. In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with those of this application; they are merely examples of apparatuses and methods consistent with some aspects of the embodiments of this application as detailed in the appended claims.

[0018] It is understood that the terms “first,” “second,” etc., used in this application may be used herein to describe various concepts, but unless otherwise stated, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another. For example, without departing from the scope of the embodiments of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the words “if,” “when,” or “in response to a determination” as used herein may be interpreted as “when…” or “when…” or “in response to a determination.”

[0019] As used in this application, the terms "at least one", "multiple", "each", "any", etc., "at least one" includes one, two or more, "multiple" includes two or more, "each" refers to each of the corresponding multiples, and "any" refers to any one of the multiples.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0021] It should be noted that in all specific embodiments of this application, when processing data related to user identity or characteristics, such as user information, user behavior data, user historical data, and user location information, user permission or consent is obtained first. Furthermore, the collection, use, and processing of this data comply with relevant laws, regulations, and standards. In addition, when embodiments of this application require access to sensitive personal information of users, separate permission or consent from the user is obtained through pop-ups or redirection to confirmation pages. Only after obtaining the user's separate permission or consent is the necessary user-related data required for the proper functioning of these embodiments acquired.

[0022] Reference Figure 1 , Figure 1This is an optional flowchart of a method for detecting data anomalies in a vehicle-road-cloud system provided in an embodiment of this application. The method may include, but is not limited to, steps S101 to S102: Step S101: Dynamically collect the operational dataset of the vehicle-road-cloud system; Step S102: Input the vehicle-road-cloud system running dataset into the anomaly detection integration platform, perform data anomaly detection through the anomaly detection integration platform, and output the anomaly detection results corresponding to each anomaly detection item; the anomaly detection integration platform includes a hierarchical detection system corresponding to each anomaly detection item.

[0023] In some embodiments, refer to Figure 2 Step S101 may include, but is not limited to, steps S201 to S203: Step S201: Dynamically collect the module operation datasets corresponding to each system deployment module in the vehicle-road-cloud system; the system deployment modules include the vehicle terminal processing module, the roadside facility operation module, and the system cloud processing module; Step S202: Perform data preprocessing on the running datasets of each module to determine the preprocessed running datasets of each module. The data preprocessing includes data cleaning and data formatting. Step S203: Determine the vehicle-road-cloud system running dataset based on the preprocessed module running datasets of each data.

[0024] In some embodiments, exemplarily, the vehicle terminal processing module collects operational data from the vehicle terminal, including vehicle status data (including vehicle sensor data and vehicle CAN bus data, etc.), vehicle-to-everything (V2X) communication data (including V2X messages and communication link status data, etc.), and environmental perception data (including surrounding vehicle / pedestrian trajectory prediction data and local high-precision map matching results (such as lane-level positioning deviations, etc.). The roadside facility operation module collects traffic environment perception data (including roadside sensor data and traffic facility status data, etc.) and edge computing data (including vehicle aggregation information and local...). Event detection results (such as traffic accidents and illegal parking) and communication and equipment status (including communication logs between roadside facilities and vehicles / cloud, as well as edge server resource utilization) are collected through cloud processing modules. Global traffic data (including aggregated vehicle trajectory data of the entire road network, dynamic updates of high-precision maps, and centralized control strategies for traffic lights) and system operation data (including microservice call chain logs, database operation records, and cloud resource monitoring indicators) and cross-domain fusion data (including multi-source abnormal event correlation analysis and consistency verification results between vehicle-side and roadside data (such as positioning deviation alarms)) are collected.

[0025] In some embodiments, the hierarchical detection system includes a rule filtering unit and a model analysis unit, referring to... Figure 3Step S102 may include, but is not limited to, steps S301 to S303: Step S301: Input the vehicle-road-cloud system running dataset into the hierarchical detection system corresponding to each anomaly detection item; Step S302: For each anomaly detection item, the rule filtering unit performs rule filtering according to the pre-configured rule conditions to extract the corresponding target data subset from the vehicle-road-cloud system running dataset; Step S303: For each anomaly detection item, the corresponding target data subset is input into the model analysis unit. Using the pre-configured anomaly detection model, anomaly analysis is performed on the target data subset, and the abnormal running data corresponding to the anomaly detection item and the target anomaly type corresponding to the abnormal running data are output from the target data subset.

[0026] In some embodiments, anomaly detection items may include, but are not limited to, vehicle-side anomaly detection items, roadside facility anomaly detection items, cloud-based anomaly detection items, cross-domain related anomaly detection items, and dynamic environment adaptability detection items. Among them, vehicle-side anomaly detection items include vehicle sensor data anomaly detection items (such as missing or distorted LiDAR / camera data, IMU (Inertial Measurement Unit) data mutation, and GPS positioning drift, etc.) and vehicle communication anomaly detection items (such as V2X communication packet loss rate anomaly detection, message delay detection, and message tampering or forgery detection, etc.).

[0027] The roadside facility anomaly detection items include roadside sensing equipment anomaly detection items (such as camera image freezing or sudden resolution drop, millimeter-wave radar target loss, and abnormal roadside meteorological sensor data) and edge computing node anomaly detection items (such as edge server CPU / memory overload, local decision delay, and data reporting failure to the cloud). The cloud anomaly detection items include cloud service anomaly detection items (such as microservice call chain breakage, database response timeout, and cloud resource elastic scaling failure) and global traffic decision anomaly detection items (such as traffic light control command conflict, high-precision map update delay, and unreasonable vehicle route planning results).

[0028] Cross-domain anomaly detection items include vehicle-road cooperative anomaly detection items (such as mismatch between vehicle-side perceived target ID and roadside perceived target ID, and anomaly detection of deviation between vehicle positioning and roadside calibration coordinate system), road-cloud cooperative anomaly detection items (such as discontinuous timestamps of roadside data reported to the cloud, and anomaly detection of differences in confidence between edge preprocessing results and cloud fusion results), and vehicle-cloud cooperative anomaly detection items (such as conflicts between cloud instructions received by the vehicle and the current state, and inconsistencies between the vehicle's historical trajectory and the trajectory stored in the cloud).

[0029] The dynamic environment adaptability detection items include traffic flow change scenario detection items (such as sudden congestion causing a surge in V2X message density, triggering message dropping, and traffic light control not responding in time when emergency vehicles have priority passage) and severe weather scenario detection items (such as heavy rain causing simultaneous distortion of multiple vehicle sensor data, and blurry images from roadside cameras, but no abnormalities in radar data).

[0030] By setting up multi-level and multi-dimensional anomaly detection items, a collaborative analysis mechanism for vehicle-side, roadside, and cloud data is established, effectively solving the problem of missing cross-domain anomaly correlation and improving the accuracy and comprehensiveness of anomaly detection.

[0031] In some embodiments, refer to Figure 4 Step S302 may include, but is not limited to, steps S401 to S402: Step S401: Through the rule filtering unit, extract the subset of running data corresponding to the set of fields to be detected from the vehicle-road-cloud system running dataset; the subset of running data includes multiple data fields to be detected; Step S402: According to the rule conditions, match each data field to be detected contained in the running data subset, and output the target data subset that meets the rule conditions.

[0032] In some embodiments, the rule filtering unit identifies the field type of each data field in the vehicle-road-cloud system operation dataset. The field type is used to reflect the business attributes of the data field (including device status, time series, and identity). Based on the field type of each data field and the set of field types to be detected (including multiple field types to be detected), data filtering is performed to extract multiple data fields to be detected that match the set of field types to be detected from all data fields in the vehicle-road-cloud system operation dataset. Based on the multiple data fields to be detected, the corresponding subset of operation data is determined.

[0033] In some embodiments, anomaly detection is performed on a subset of running data using pre-configured rules and conditions. Target data subsets that meet the rules and conditions are then selected from the running data subset. These rules and conditions include threshold comparison, logical operations, and time window constraints. Specifically, threshold comparison compares the data field to be detected with a predefined threshold to achieve single-point anomaly detection. Logical operations combine multiple sub-conditions containing the data field to be detected using logical operators to achieve complex combined condition judgment. Time window constraints statistically analyze the data state of the data field to be detected within a preset time range to achieve time-series correlation detection.

[0034] The rules and conditions support dynamic updates to adapt to changes in data distribution, enabling high-precision extraction of target data, reducing the waste of computing resources in downstream modules (such as the model analysis unit mentioned above), and avoiding data filtering that relies on fixed rules, which would lead to insufficient filtering accuracy.

[0035] In some embodiments, refer to Figure 5 Step S303 may include, but is not limited to, steps S501 to S502: Step S501: Obtain the anomaly detection model; Step S502: Using an anomaly detection model, calculate the anomaly probability of each running data in the target data subset based on multiple preset anomaly types. When the anomaly probability of any preset anomaly type corresponding to the running data is greater than a preset probability threshold, output the running data as abnormal running data. Determine the target anomaly type corresponding to the abnormal running data from multiple preset anomaly types, where the target anomaly type is the preset anomaly type with the highest anomaly probability.

[0036] In some embodiments, the anomaly detection model includes at least one of a time-series prediction model, a clustering model, and a classification model, wherein the time-series prediction model is used to detect the degree of deviation between the data value and the predicted value; the clustering model is used to identify outliers in the target data subset; and the classification model is used to determine the anomaly type label.

[0037] In some embodiments, an anomaly type mapping table is constructed, which includes each preset anomaly type and a preset probability threshold corresponding to each preset anomaly type. The anomaly detection model calculates the anomaly probability of each running data in the target data subset according to the anomaly type mapping table. When the anomaly probability of any preset anomaly type corresponding to the running data is found to be greater than the preset probability threshold, the running data is marked as abnormal running data, and the preset anomaly type with the highest anomaly probability is determined as the target anomaly type corresponding to the running data. Optionally, the target anomaly type may include, but is not limited to, instantaneous anomalies, continuous anomalies, and periodic anomalies corresponding to anomaly detection items. Instantaneous anomalies correspond to a single burst of data deviation, continuous anomalies correspond to data exceeding the standard within a continuous time window, and periodic anomalies correspond to repetitive deviation patterns at fixed intervals.

[0038] In some embodiments, refer to Figure 6 The above method may also include, but is not limited to, steps S601 to S603: Step S601: Construct a hierarchical detection system for each anomaly detection item and the associations between each anomaly detection item; Step S602: Establish an anomaly detection integrated platform based on each anomaly detection item and the corresponding hierarchical detection system. Step S603: Manage the anomaly detection integration platform, including updating each anomaly detection item and dynamically updating the hierarchical detection system corresponding to each anomaly detection item.

[0039] In some embodiments, refer to Figure 7 Assuming there are anomaly detection items 1 to N, construct hierarchical detection systems corresponding to anomaly detection items 1 to N respectively. For example, anomaly detection item 1 is associated with hierarchical detection system 1. Hierarchical detection system 1 includes rule filtering unit 1 and model analysis unit 1. Based on each anomaly detection item and the hierarchical detection system associated with each anomaly detection item, establish a corresponding anomaly detection integration platform.

[0040] In some embodiments, refer to Figure 8 Step S603 may include, but is not limited to, steps S701 to S703: Step S701: For each layered detection system, dynamically update the corresponding rule conditions and the set of field types to be detected in the rule filtering unit. The update of the rule conditions is based on historical anomaly detection results and user-defined input. Step S702: For each layered detection system, dynamically update the corresponding anomaly detection model in the model analysis unit. The update of the anomaly detection model is based on historical anomaly detection results and user model parameter update operations. Step S703: Manage the information of each anomaly detection item, including updating the description information of the anomaly detection item and the associated information of the hierarchical detection system.

[0041] In some embodiments, the description information of the anomaly detection item includes dynamic adjustment mechanisms, detection triggering conditions, detection objects, and anomaly handling suggestions, etc., and the association information of the hierarchical detection system includes the identifier of the associated hierarchical detection system and an overview of the hierarchical detection method, etc.

[0042] In some embodiments, based on historical anomaly detection results, the rule conditions in the rule filtering unit and the corresponding anomaly detection model in the model analysis unit are dynamically updated to optimize the rule conditions and anomaly detection model, so that the rule conditions and anomaly detection model can continuously adapt to changes in data distribution and improve the efficiency and accuracy of anomaly detection.

[0043] In some embodiments, optionally, the user determines the target rule sub-condition from the rule conditions, performs a user-defined input operation, updates the relevant parameters of the target rule sub-condition, determines the updated target rule sub-condition, and dynamically updates the rule conditions accordingly based on the updated target rule sub-condition.

[0044] Reference Figure 9 , Figure 9This is an optional structural diagram of a vehicle-road-cloud system data anomaly detection device provided in an embodiment of this application. The device is used to implement the aforementioned vehicle-road-cloud system data anomaly detection method and may include: The first module is used to dynamically collect the operational dataset of the vehicle-road-cloud system. The second module is used to input the vehicle-road-cloud system's operational dataset into the anomaly detection integration platform, perform data anomaly detection through the anomaly detection integration platform, and output the anomaly detection results corresponding to each anomaly detection item; the anomaly detection integration platform includes a hierarchical detection system corresponding to each anomaly detection item.

[0045] It is understood that the content of the above method embodiments is applicable to the present device embodiments. The specific functions implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0046] This invention also provides an electronic device, including a memory, a processor, and a program stored in the memory and executable on the processor. When the program is executed by the processor, it implements the vehicle-road-cloud system data anomaly detection method of the above embodiments.

[0047] It is understood that the content of the above method embodiments is applicable to the present device embodiments. The specific functions implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0048] Please see Figure 10 , Figure 10 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes: The processor 901 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory 902 can be implemented as a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 902 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 902 and is called and executed by the processor 901 to execute the vehicle-road-cloud system data anomaly detection method of the embodiments of this application. The input / output interface 903 is used to implement information input and output; The communication interface 904 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 905 transmits information between various components of the device (e.g., processor 901, memory 902, input / output interface 903, and communication interface 904); The processor 901, memory 902, input / output interface 903, and communication interface 904 are connected to each other within the device via bus 905.

[0049] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the control processor, and these remote memories can be connected to the control device via a network.

[0050] The non-transient software program and instructions required to implement the vehicle-road-cloud system data anomaly detection method of the above embodiments are stored in memory. When executed by the processor, the vehicle-road-cloud system data anomaly detection method of the above embodiments is executed. For example, executing... Figure 2 Method steps S201 to S203, Figure 3 Method steps S301 to S303, Figure 5 The method steps S501 to S502, etc.

[0051] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0052] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0053] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0054] Furthermore, one embodiment of the present invention also provides a computer program product, including a computer program or computer instructions, which are stored in a computer-readable storage medium. A processor of a computer device reads the computer program or computer instructions from the computer-readable storage medium and executes the computer program or computer instructions, causing the computer device to perform the aforementioned vehicle-road-cloud system data anomaly detection method. Exemplarily, the above-described method is executed... Figures 1 to 6 as well as Figure 8 The methods and steps in the text.

[0055] It is worth noting that, since the computer program product of this embodiment can execute the vehicle-road-cloud system data anomaly detection method of any of the above embodiments, the specific implementation method and technical effect of the computer program product of this embodiment can refer to the specific implementation method and technical effect of the vehicle-road-cloud system data anomaly detection method of any of the above embodiments.

[0056] It will be understood by those skilled in the art that all or some of the steps and systems in the methods disclosed above can be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, as is known to those skilled in the art, communication media typically include computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0057] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

Claims

1. A method for detecting data anomalies in a vehicle-road-cloud system, characterized in that, The method includes the following steps: Dynamically collect operational datasets from the vehicle-road-cloud system; The vehicle-road-cloud system's operational dataset is input into the anomaly detection integration platform, which performs data anomaly detection and outputs the anomaly detection results corresponding to each anomaly detection item. The anomaly detection integration platform includes a hierarchical detection system corresponding to each anomaly detection item.

2. The method for detecting data anomalies in a vehicle-road-cloud system according to claim 1, characterized in that, The dynamically collected vehicle-road-cloud system's operational dataset specifically includes: The system dynamically collects the module operation datasets corresponding to each system deployment module in the vehicle-road-cloud system; the system deployment modules include the vehicle terminal processing module, the roadside facility operation module, and the system cloud processing module. Perform data preprocessing on the runtime datasets of each module to determine the preprocessed runtime datasets of each module. The vehicle-road-cloud system's operating dataset is determined based on the preprocessed data datasets of each module.

3. The method for detecting data anomalies in a vehicle-road-cloud system according to claim 1, characterized in that, The method further includes: Construct the hierarchical detection system for each of the aforementioned anomaly detection items and the associations between each of the aforementioned anomaly detection items; Based on each of the aforementioned anomaly detection items and the corresponding hierarchical detection system, an anomaly detection integration platform is established. The anomaly detection integration platform is managed, including updating each anomaly detection item and dynamically updating the hierarchical detection system corresponding to each anomaly detection item.

4. The method for detecting data anomalies in a vehicle-road-cloud system according to claim 3, characterized in that, The hierarchical detection system includes a rule filtering unit and a model analysis unit. The system inputs the vehicle-road-cloud system's runtime dataset to an anomaly detection integration platform, performs data anomaly detection through the platform, and outputs the anomaly detection results corresponding to each anomaly detection item. Specifically, this includes: The vehicle-road-cloud system operation dataset is input into the hierarchical detection system corresponding to each of the anomaly detection items; For each of the aforementioned anomaly detection items, the rule filtering unit performs rule filtering based on pre-configured rule conditions to extract the corresponding target data subset from the vehicle-road-cloud system's operational dataset; For each of the aforementioned anomaly detection items, the corresponding target data subset is input to the model analysis unit. Using a pre-configured anomaly detection model, anomaly analysis is performed on the target data subset, and the abnormal operation data corresponding to the anomaly detection item and the target anomaly type corresponding to the abnormal operation data are output from the target data subset.

5. The method for detecting data anomalies in a vehicle-road-cloud system according to claim 4, characterized in that, For each of the aforementioned anomaly detection items, the rule filtering unit performs rule filtering based on pre-configured rule conditions to extract the corresponding target data subset from the vehicle-road-cloud system's operational dataset. Specifically, this includes: The rule filtering unit extracts a subset of operational data corresponding to the set of fields to be detected from the vehicle-road-cloud system's operational dataset; the subset of operational data includes multiple fields to be detected. According to the rules, each of the data fields to be detected contained in the running data subset is matched, and the target data subset that meets the rules is output. The rules include threshold comparison, logical operation and time window constraint.

6. The method for detecting data anomalies in a vehicle-road-cloud system according to claim 4, characterized in that, For each of the aforementioned anomaly detection items, the corresponding target data subset is input to the model analysis unit. Using a pre-configured anomaly detection model, anomaly analysis is performed on the target data subset, and the abnormal operational data corresponding to the anomaly detection item and the target anomaly type corresponding to the abnormal operational data are output from the target data subset. Specifically, this includes: Obtain the anomaly detection model, which includes at least one of a time-series prediction model, a clustering model, and a classification model; Using the anomaly detection model, anomaly probability calculation is performed on each running data in the target data subset according to multiple preset anomaly types. When the anomaly probability of any preset anomaly type corresponding to the running data is greater than a preset probability threshold, the running data is output as the abnormal running data. The target anomaly type corresponding to the abnormal running data is determined from the multiple preset anomaly types, wherein the target anomaly type is the preset anomaly type with the highest anomaly probability.

7. The method for detecting data anomalies in a vehicle-road-cloud system according to claim 5, characterized in that, The management of the anomaly detection integration platform specifically includes: For each of the hierarchical detection systems, the rule conditions and the set of field types to be detected in the rule filtering unit are dynamically updated, wherein the update of the rule conditions is based on historical anomaly detection results and user-defined input. For each of the hierarchical detection systems, the anomaly detection model corresponding to the model analysis unit is dynamically updated. The update of the anomaly detection model is based on the historical anomaly detection results and the user's model parameter update operation. The information of each anomaly detection item is managed, including updating the anomaly detection item description information and the associated information of the hierarchical detection system.

8. A data anomaly detection device for a vehicle-road-cloud system, characterized in that, The device includes: The first module is used to dynamically collect the operational dataset of the vehicle-road-cloud system. The second module is used to input the vehicle-road-cloud system operation dataset into the anomaly detection integration platform, perform data anomaly detection through the anomaly detection integration platform, and output the anomaly detection results corresponding to each anomaly detection item; the anomaly detection integration platform includes a hierarchical detection system corresponding to each anomaly detection item.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-mentioned vehicle-road-cloud system data anomaly detection method.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the vehicle-road-cloud system data anomaly detection method as described in any one of claims 1 to 7.

Images