Location risk management methods, devices, electronic equipment and storage media

Abstract

This disclosure provides a method, apparatus, electronic device, and storage medium for handling location risks. The method includes: acquiring current vehicle data, historical vehicle data, current location data, and historical location data; determining, based on the review time series and historical vehicle data, vehicle data to be combined for each review time within the review time series; determining, based on the review time series and historical location data, location data to be combined for each review time within the review time series; combining the current vehicle data and the vehicle data to be combined to obtain vehicle data to be detected; combining the current location data and the location data to be combined to obtain location data to be detected; inputting the vehicle data to be detected and the location data to be detected into a location risk judgment model to obtain measurement source risk data; and determining the target location risk by combining a location risk factor matrix and the importance coefficients of each location measurement source. This disclosure improves the accuracy of location risk assessment.

Description

Technical Field

[0001] This disclosure relates to the field of vehicle positioning technology, and in particular to a positioning risk handling method, apparatus, electronic device, and storage medium. Background Technology

[0002] Positioning risk has always been one of the core issues in the field of autonomous driving, and a good and stable positioning system is the foundation for the normalized operation of autonomous vehicles.

[0003] Furthermore, location risk is difficult to determine directly based on the exact state of a measurement source or vehicle. This is because location risk is not only related to the performance and algorithm of the measurement source, but also closely related to the vehicle's state. Moreover, the characteristics of various measurement sources are similar, and the collected data is time-series data, exhibiting significant temporal correlation. Therefore, failing to consider these factors when predicting and processing location risk would result in treating all scenarios, all measurement sources, and all time points the same, leading to significant deviations in the predicted location risk and hindering timely and effective follow-up processing based on the prediction results. Summary of the Invention

[0004] To address or at least partially address the aforementioned technical problems, embodiments of this disclosure provide a location risk processing method, apparatus, electronic device, and storage medium. These methods fully integrate vehicle data and location data, consider the temporal relationships between these data, and assess the importance of different fields and different location measurement sources, thereby improving the accuracy of location risk assessment.

[0005] In a first aspect, embodiments of this disclosure provide a method for locating risk processing, the method comprising:

[0006] Acquire current vehicle data, historical vehicle data, current location data, and historical location data; wherein, the current location data includes the measurement data of each location measurement source at the current moment, and the historical location data includes the measurement data of each location measurement source at each historical moment;

[0007] Based on the review time series and the historical vehicle data, determine the vehicle data to be combined corresponding to each review time in the review time series, and determine the location data to be combined corresponding to each review time in the review time series based on the review time series and the historical location data; wherein, the review time series is an arithmetic sequence arranged in ascending order;

[0008] The current vehicle data and the vehicle data to be combined are combined to obtain the vehicle data to be detected, and the current positioning data and the positioning data to be combined are combined to obtain the positioning data to be detected.

[0009] The vehicle data to be detected and the positioning data to be detected are input into a pre-trained positioning risk judgment model to obtain measurement source risk data.

[0010] The target positioning risk is determined based on the measurement source risk data, the preset positioning risk factor matrix, and the importance coefficient of each positioning measurement source.

[0011] Secondly, embodiments of this disclosure also provide a location risk processing device, the device comprising:

[0012] The data acquisition module is used to acquire current vehicle data, historical vehicle data, current positioning data, and historical positioning data; wherein, the current positioning data includes the measurement data of each positioning measurement source at the current moment, and the historical positioning data includes the measurement data of each positioning measurement source at each historical moment;

[0013] The data processing module is used to determine, based on the review time series and the historical vehicle data, the vehicle data to be combined corresponding to each review time in the review time series, and to determine, based on the review time series and the historical positioning data, the positioning data to be combined corresponding to each review time in the review time series; wherein, the review time series is an arithmetic sequence arranged in ascending order;

[0014] The data combination module is used to combine the current vehicle data and the vehicle data to be combined to obtain the vehicle data to be detected, and to combine the current positioning data and the positioning data to be combined to obtain the positioning data to be detected.

[0015] The model calculation module is used to input the vehicle data to be detected and the positioning data to be detected into a pre-trained positioning risk judgment model to obtain measurement source risk data.

[0016] The risk processing module is used to determine the target positioning risk based on the measurement source risk data, the preset positioning risk factor matrix, and the importance coefficient of each positioning measurement source.

[0017] Thirdly, embodiments of this disclosure also provide an electronic device, the electronic device comprising: one or more processors; a storage device for storing one or more programs; and when the one or more programs are executed by the one or more processors, causing the one or more processors to implement the location risk processing method as described above.

[0018] Fourthly, embodiments of this disclosure also provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the location risk processing method as described above.

[0019] This disclosure provides a location risk processing method that acquires current vehicle data, historical vehicle data, current location data, and historical location data. Then, based on the review time series and historical vehicle data, it determines the vehicle data to be combined corresponding to each review time within the review time series, and based on the review time series and historical location data, it determines the location data to be combined corresponding to each review time within the review time series. This allows for temporal organization of the vehicle data and location data, facilitating temporal correlation. Furthermore, the current vehicle data and the vehicle data to be combined are combined to obtain the vehicle data to be detected, and the current location data and the location data to be combined are combined to obtain the location data to be detected. This method, combined with the current vehicle data and current location data at the current moment, facilitates subsequent determination. The current location risk is assessed by inputting the vehicle data and location data to be detected into a pre-trained location risk judgment model to obtain measurement source risk data. This model then analyzes the vehicle and location data, fully considering temporal correlations and the relationships between them, thus improving the accuracy of the measurement source risk data. Based on the measurement source risk data, a pre-set location risk factor matrix, and the importance coefficients of each location measurement source, the target location risk is determined. This process integrates and processes the measurement source risk data according to different fields and the importance of different location measurement sources, achieving a comprehensive combination of vehicle and location data, considering the temporal correlations between these data, and assessing the importance of different fields and different location measurement sources, thereby improving the accuracy of location risk assessment. Attached Figure Description

[0020] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and the originals and elements are not necessarily drawn to scale.

[0021] Figure 1 This is a flowchart of a location risk handling method according to an embodiment of this disclosure;

[0022] Figure 2 This is a schematic diagram of the structure of a location risk processing device according to an embodiment of the present disclosure;

[0023] Figure 3 This is a schematic diagram of the structure of an electronic device according to an embodiment of this disclosure. Detailed Implementation

[0024] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0025] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0026] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0027] Determining and handling positioning risks is a complex issue. The stability and quality of positioning are comprehensively judged by the positioning effects of various measurement sources, vehicle status, and even the vehicle's usage scenario. Therefore, positioning risk is difficult to determine directly based on the exact status of a single measurement source or vehicle. Currently, positioning risk can be obtained by directly using measurement source positioning data as model input. However, this does not consider the special scenarios of autonomous driving, the unique characteristics of positioning data, and its temporal nature, which can affect the accuracy of positioning risk determination. Positioning risk is not only related to the performance of the measurement sources and algorithms but also closely related to the usage scenario and vehicle status. For example, the actual positioning effect may be relatively poor when the vehicle is under an overpass, indoors, or turning. Positioning data is also unique, with certain correlations between measurement sources and similarities in their features. Furthermore, positioning data is time-series data, and positioning risk is also significantly correlated with historical data. Directly using measurement source positioning data as model input does not consider these factors and sets the trust level of each measurement source to be the same, which is questionable. Therefore, directly predicting through the model and treating all measurement sources equally may result in a deviation from the actual positioning risk.

[0028] To address the aforementioned issues, this disclosure provides a location risk processing method that comprehensively considers location data from various measurement sources, temporal variations in location data, vehicle data, and temporal variations in vehicle data, thereby improving the accuracy of location risk processing.

[0029] Figure 1 This is a flowchart illustrating a location risk processing method according to an embodiment of this disclosure. The method can be executed by a location risk processing device, which can be implemented in software and / or hardware and can be configured in an electronic device. Figure 1As shown, the method may specifically include the following steps:

[0030] S110: Obtain current vehicle data, historical vehicle data, current location data, and historical location data.

[0031] The current positioning data includes measurement data from each positioning measurement source at the current moment, while historical positioning data includes measurement data from each positioning measurement source at various historical moments. Current vehicle data includes vehicle data at the current moment, while historical vehicle data includes vehicle data from various historical moments. Vehicle data comprises the vehicle's kinematic parameters (e.g., speed, front wheel deflection angle) and the degree of scene influence on positioning (e.g., whether the vehicle is under a bridge or in a tunnel). Positioning data refers to data acquired directly or indirectly from positioning measurement sources during vehicle operation, including the measurement source status and data transformed from the data collected by the positioning measurement sources, such as whether there are anomalies, whether fusion is required, and confidence levels.

[0032] Specifically, current vehicle data and current location data are collected and stored based on various measurement sources and vehicle sensors installed on the vehicle. Historical vehicle data and historical location data from previous times are retrieved from the storage space for subsequent data processing and analysis.

[0033] Building upon the example above, after acquiring current vehicle data, historical vehicle data, current location data, and historical location data, preprocessing can be performed on the current location data and historical location data. This can be achieved in the following way:

[0034] The number of fields to be merged is determined based on the fields of the measurement data collected by each positioning measurement source;

[0035] Based on the number of fields to be merged, the current location data and each historical location data are expanded into a square matrix.

[0036] The positioning measurement source refers to the measurement source installed on the vehicle for positioning, such as GPS (Global Positioning System), LSLAM (LiDAR Simultaneous Localization and Mapping), and VSLAM (Vision Simultaneous Localization and Mapping). The measurement data fields may include northward data fields, eastward data fields, altitude fields, heading angle fields, pitch angle fields, roll angle fields, and whether data is merged. The merged fields include the fields from the measurement data collected by each positioning measurement source. For example, the measurement data collected by positioning measurement source 1 includes fields 1, 2, and 3; the measurement data collected by positioning measurement source 2 includes fields 1, 3, and 4; and the measurement data collected by positioning measurement source 3 includes fields 4 and 5. In this case, the merged fields include fields 1, 2, 3, 4, and 5. The number of merged fields is greater than the number of positioning measurement sources; as in the example above, the number of merged fields is 5, and the number of positioning measurement sources is 3.

[0037] Specifically, the union of fields from the measurement data collected by each positioning measurement source is used as the merging field, and the number of merging fields is determined. Then, matrices are constructed using the positioning measurement source type as the row identifier and the merging field as the column identifier for both the current positioning data and each historical positioning data point. Finally, the matrices corresponding to the current positioning data and each historical positioning data point are expanded into square matrices according to the number of merging fields, serving as the current positioning data and historical positioning data to be used subsequently.

[0038] For example, if the number of location measurement sources is 3 and the number of merged fields is 5, then the size of the directly obtained current location data and each historical location data is 3×5. It is expanded into a square matrix according to the number of merged fields of 5. Specifically, two rows are expanded under the original current location data and each historical location data. During the expansion process, the data filled is 0, so that its size becomes 5×5.

[0039] S120. Based on the review time series and historical vehicle data, determine the vehicle data to be combined corresponding to each review time in the review time series, and based on the review time series and historical positioning data, determine the positioning data to be combined corresponding to each review time in the review time series.

[0040] The time-lookback sequence is an arithmetic progression arranged in ascending order. Each lookback time is an item in the time-lookback sequence, representing the time difference before the current moment. The current moment and its differences from each lookback time are used as the time points to be combined. Based on these time points, a new sequence can be constructed from largest to smallest. The vehicle data to be combined is the vehicle data obtained by integrating and processing historical vehicle data between any two adjacent time points in the new sequence. The location data to be combined is the location data obtained by integrating and processing historical location data between any two adjacent time points in the new sequence.

[0041] Specifically, the current time and the difference between the current time and each playback time are used as the time points to be combined. Then, a new sequence is constructed based on these time points from largest to smallest. The historical vehicle data between any two adjacent time points in the new sequence are then integrated to obtain the vehicle data to be combined corresponding to each playback time. Similarly, the historical location data between any two adjacent time points in the new sequence are integrated to obtain the location data to be combined corresponding to each playback time. The integration method can be statistical methods such as calculating the mean, maximum, or minimum value, or it can be a data integration method using a pre-established integration model.

[0042] For example, the current time is t, and the time series to be reviewed is [f1, f2, ..., f n The differences between the current time and each of the review times in the time log are taken as the time points to be combined, namely t, t-f1, t-f2, ..., tf. n The new sequence formed by combining the time points from largest to smallest is [t, t-f1, t-f2, ..., tf]. n ], where the historical vehicle data between any two adjacent time points to be combined are [t-f1, t], [t-f2, t-f1], ..., [tf n , tf n-1 Historical vehicle data within the range, where the historical location data between any two adjacent items are [t-f1, t], [t-f2, t-f1], ..., [tf n , tf n-1 Historical location data within [the specified range].

[0043] Based on the above example, the following methods can be used to determine the vehicle data to be combined corresponding to each review time within the review time series, and the location data to be combined corresponding to each review time within the review time series, based on the review time series and historical vehicle data:

[0044] For each playback time within the playback time series, determine the playback period corresponding to that playback time;

[0045] Obtain the vehicle data corresponding to the review period from the historical vehicle data, arrange the vehicle data in chronological order, and determine the median data in the arranged vehicle data as the vehicle data to be combined.

[0046] Retrieve the playback location data corresponding to the playback period from the historical location data, arrange the playback location data in chronological order, and determine the median data in the arranged playback location data as the location data to be combined.

[0047] The lookback period is the time period used for subsequent integration and acquisition of vehicle data and location data to be combined. It can be understood as the time interval between adjacent time points to be combined in the new sequence. If the lookback time is the minimum value in the lookback time series, then the lookback period is the time interval from the difference between the current time and the minimum value to the current time. If the lookback time is any value in the lookback time series other than the minimum value, then the lookback period is the time interval from the difference between the current time and the previous time to the difference between the current time and the lookback time. For example, the current time is t, and the lookback time series is [f1, f2, ..., f...]. n The time points to be combined are t, t-f1, t-f2, ..., tf n The new sequence formed by combining the time points from largest to smallest is [t, t-f1, t-f2, ..., tf]. n The replay time periods are [t-f1, t], [t-f2, t-f1], ..., [tf] n , tf n-1 The vehicle data review section refers to reviewing historical vehicle data within the same time period. The location data review section refers to reviewing historical location data within the same time period.

[0048] Specifically, for each review time within the review time series, following the rules for determining the review period as described above (and not repeated here), each review period can be obtained. Then, for each review period, the corresponding vehicle data can be retrieved from historical vehicle data. This data is then arranged chronologically. The middle vehicle in the arranged data is selected as the vehicle data to be combined. If there are two middle vehicles, their average is used as the vehicle data to be combined. This can be understood as determining the median of the arranged vehicle data as the vehicle data to be combined. Based on this, the vehicle data to be combined corresponding to each review period can be obtained. For each playback period, corresponding playback location data can be obtained from historical location data. This playback location data is then arranged chronologically. The middle playback location data is selected as the location data to be combined. If there are two middle playback location data points, their average is used as the location data to be combined. This can be understood as determining the median of the arranged playback location data as the location data to be combined. Based on this, the location data to be combined corresponding to each playback period can be obtained.

[0049] For example, if the vehicle data for a certain playback period is c1, c2, and c3 (already arranged in chronological order), then the median data c2 is determined as the vehicle data to be combined for that playback period. If the location data for a certain playback period is d1, d2, d3, and d4 (already arranged in chronological order), then the median data (d2+d3) / 2 is determined as the location data to be combined for that playback period.

[0050] S130. Combine the current vehicle data and the vehicle data to be combined to obtain the vehicle data to be detected, and combine the current positioning data and the positioning data to be combined to obtain the positioning data to be detected.

[0051] The vehicle data to be detected is obtained by combining the current vehicle data and the vehicle data to be combined according to the third dimension. For example, the matrix size of the current vehicle data and the vehicle data to be combined is n×n, and the number of matrices is m. Therefore, the vehicle data to be detected is a three-dimensional matrix with a size of n×n×m. The positioning data to be detected is obtained by combining the current positioning data and the positioning data to be combined according to the third dimension. For example, the matrix size of the current positioning data and the positioning data to be combined is 1×d, and the number of matrices is m. Therefore, the positioning data to be detected is a three-dimensional matrix with a size of 1×d×m.

[0052] Specifically, the current vehicle data and the vehicle data to be combined are combined according to the third dimension to obtain three-dimensional vehicle data to be detected. Similarly, the current positioning data and the positioning data to be combined are combined according to the third dimension to obtain three-dimensional positioning data to be detected.

[0053] S140. Input the vehicle data to be detected and the positioning data to be detected into the pre-trained positioning risk judgment model to obtain the measurement source risk data.

[0054] The location risk assessment model is a pre-built and trained model used to process time-related vehicle data and location data to obtain risk data describing the fields of each measurement source and each measurement data. Measurement source risk data is the output data of the location risk assessment model, representing the risk of the fields of the measurement source and measurement data.

[0055] Specifically, the vehicle data and the location data to be detected are input into a pre-trained location risk assessment model. The two data dimensions of vehicle data and location data are comprehensively evaluated. In addition, the model data is analyzed in conjunction with time series changes and measurement data from different measurement sources to obtain the output data of the location risk assessment model, which is the measurement source risk data.

[0056] Based on the above example, the pre-trained location risk assessment model can be trained in the following way:

[0057] Based on sample vehicle data, sample location data, and review time series, multiple sets of test vehicle data and corresponding test location data for each set of test vehicle data are determined.

[0058] For each set of location data to be tested, the location data with the latest time sequence is taken as the target location data. Based on the abnormal thresholds corresponding to each field in the target location data, the target location data is transformed into target data to be tested.

[0059] Using the corresponding vehicle data and location data as independent variables, and the target data corresponding to the location data as the dependent variable, the initial long short-term memory artificial neural network is trained to obtain a pre-trained location risk judgment model.

[0060] The sample vehicle data consists of pre-collected vehicle data, and the sample location data consists of pre-collected location data that is temporally correlated with the sample vehicle data. The test vehicle data serves as the dependent variable of the initial Long Short-Term Memory (LSTM) artificial neural network; its acquisition method is similar to that of the detection vehicle data and will not be elaborated here. Similarly, the test location data serves as the dependent variable of the initial LSTM artificial neural network; its acquisition method is similar to that of the detection location data and will not be elaborated here. The initial LSTM artificial neural network is a LSTM artificial neural network whose model structure and parameters have not been fully adjusted. The target location data is the latest chronologically ordered test location data in each group of test location data. The anomaly threshold is a pre-determined threshold for anomalies in each field of each measurement data from each location measurement source; it can be obtained through experimentation or reasoning and is not limited here. The test target data is the target location data after being converted from 0 to 1 according to the anomaly threshold.

[0061] Specifically, the sample vehicle data and the corresponding sample location data are combined and divided according to the time sequence to obtain multiple sets of test vehicle data and corresponding test location data for each set, which are then used as training data for the training model. For each set of test location data, the latest test location data in the set is taken as the target location data. Then, the target location data is converted from 0 to 1 according to the anomaly thresholds corresponding to each field in the target location data to obtain the test target data. Next, the test vehicle data and test location data with corresponding relationships are used as independent variables of the initial Long Short-Term Memory (LSTM) artificial neural network, and the test target data corresponding to the test location data is used as the dependent variable of the initial LSM artificial neural network. The initial LSM artificial neural network is trained, and when the model training results meet preset requirements, such as convergence or reaching the required number of training iterations, the trained model is obtained as the pre-trained location risk assessment model.

[0062] Based on the above example, the target location data can be transformed into test target data according to the anomaly thresholds corresponding to each field in the target location data in the following way:

[0063] For each field in the target location data, if the field value reaches the corresponding abnormal threshold, the value corresponding to the field in the target data to be tested is determined as the first value; if the field value does not reach the corresponding abnormal threshold, the value corresponding to the field in the target data to be tested is determined as the second value.

[0064] The first value is either 0 or 1, and the second value is either 0 or 1. If the first value is 0, then the second value is 1; if the first value is 1, then the second value is 0.

[0065] Specifically, for each field in the target location data, the field value is compared with the corresponding anomaly threshold. If the field value reaches the anomaly threshold, the value corresponding to the field in the target data to be tested is determined to be the first value. Otherwise, the value corresponding to the field in the target data to be tested is determined to be the second value. Since the first and second values ​​are 0 and 1 respectively, a 0-1 conversion of the target location data can be completed, resulting in the target data to be tested containing 0 / 1 variables.

[0066] S150. Determine the target positioning risk based on the measurement source risk data, the preset positioning risk factor matrix, and the importance coefficient of each positioning measurement source.

[0067] The positioning risk factor matrix is ​​a matrix constructed from the risk factors corresponding to each field of measurement data in the positioning measurement sources. These risk factors describe the importance of the corresponding measurement data fields. Importance coefficients are pre-determined coefficients used to describe the importance of various types of measurement sources. Target positioning risk is a numerical representation of the positioning risk when the vehicle is currently driving using various positioning measurement sources.

[0068] Specifically, by integrating the field dimensions of each field in the measurement source risk data using a preset positioning risk factor matrix, and then further integrating the positioning measurement source dimensions using the importance coefficients of each positioning measurement source, a value describing the current positioning risk of the vehicle can be obtained, which is the target positioning risk.

[0069] Based on the above example, the target location risk can be determined using the measurement source risk data, the preset location risk factor matrix, and the importance coefficients of each location measurement source in the following way:

[0070] Based on the risk data of the measurement sources and the preset positioning risk factor matrix, the positioning risk coefficient corresponding to each positioning measurement source is obtained;

[0071] The target positioning risk is determined based on the positioning risk coefficient and importance coefficient corresponding to each positioning measurement source.

[0072] The positioning risk coefficient is the product of the measurement source risk data corresponding to any positioning measurement source and the preset positioning risk factor matrix, and is used to represent the positioning risk of the positioning measurement source.

[0073] Specifically, multiplying the risk data of the measurement sources with a preset positioning risk factor matrix yields the positioning risk coefficient corresponding to each positioning measurement source. Then, multiplying the positioning risk coefficients and importance coefficients of each corresponding positioning measurement source yields the positioning risk weighted value for each source. Finally, summing these positioning risk weighted values ​​yields the target positioning risk.

[0074] Based on the above example, after determining the target positioning risk, corresponding decisions can be made according to the target positioning risk, specifically:

[0075] If the target location risk reaches the first risk threshold but does not reach the second risk threshold, then an early warning operation will be executed.

[0076] If the target location risk reaches the second risk threshold but does not reach the third risk threshold, then control the vehicle to perform a deceleration operation;

[0077] If the target location risk reaches the third risk threshold, then control the vehicle to perform an emergency stop.

[0078] The first, second, and third risk thresholds are pre-set location risk values ​​used to differentiate the processing of different decisions. The first risk threshold is the minimum location risk value that triggers an early warning operation; the second risk threshold is the minimum location risk value that triggers a deceleration operation; and the third risk threshold is the minimum location risk value that triggers an emergency stop operation. The third risk threshold is greater than the second risk threshold, and the second risk threshold is greater than the first risk threshold. The first, second, and third risk thresholds can be set according to actual needs and are not specifically limited here.

[0079] Specifically, if the target location risk reaches the first risk threshold but not the second risk threshold, it indicates a situation requiring decision-making. However, no vehicle intervention is necessary at this point; simply alerting the user is sufficient. Therefore, a warning is issued to alert the user of the location risk posed by each location measurement source. If the target location risk reaches the second risk threshold but not the third risk threshold, it indicates that the target location risk is affecting driving functions. Therefore, the vehicle needs to be slowed down to avoid an accident. If the target location risk reaches the third risk threshold, it indicates that the target location risk is severely affecting driving functions. Driving should not continue based on the current location measurement data. Therefore, the vehicle needs to be brought to an emergency stop to avoid danger.

[0080] The location risk processing method provided in this embodiment acquires current vehicle data, historical vehicle data, current location data, and historical location data. Then, based on the review time series and historical vehicle data, it determines the vehicle data to be combined corresponding to each review time within the review time series, and based on the review time series and historical location data, it determines the location data to be combined corresponding to each review time within the review time series. This allows for temporal organization of the vehicle and location data, facilitating temporal correlation. Furthermore, the current vehicle data and the vehicle data to be combined are combined to obtain the vehicle data to be detected, and the current location data and the location data to be combined are combined to obtain the location data to be detected. This, combined with the current vehicle data and current location data at the current moment, facilitates subsequent determination of the current vehicle data and current location data. The positioning risk assessment method inputs the vehicle data and positioning data to be detected into a pre-trained positioning risk judgment model to obtain measurement source risk data. This model then analyzes the vehicle and positioning data, fully considering temporal correlations and the relationships between them, thus improving the accuracy of the measurement source risk data. Based on the measurement source risk data, a pre-set positioning risk factor matrix, and the importance coefficients of each positioning measurement source, the target positioning risk is determined. This method integrates and processes the measurement source risk data according to different fields and the importance of different positioning measurement sources, achieving a comprehensive combination of vehicle and positioning data, considering the temporal correlations between these data, and assessing the importance of different fields and different positioning measurement sources, thereby improving the accuracy of positioning risk assessment.

[0081] Figure 2 This is a schematic diagram of the structure of a location risk processing device according to an embodiment of this disclosure. Figure 2 As shown, the device includes: a data acquisition module 210, a data processing module 220, a data combination module 230, a model calculation module 240, and a risk processing module 250.

[0082] The data acquisition module 210 is used to acquire current vehicle data, historical vehicle data, current positioning data, and historical positioning data; wherein the current positioning data includes measurement data from each positioning measurement source at the current moment, and the historical positioning data includes measurement data from each positioning measurement source at each historical moment; the data processing module 220 is used to determine, based on the review time series and the historical vehicle data, the vehicle data to be combined corresponding to each review time in the review time series, and to determine, based on the review time series and the historical positioning data, the positioning data to be combined corresponding to each review time in the review time series; wherein, the... The time series is an arithmetic sequence arranged from smallest to largest; the data combination module 230 is used to combine the current vehicle data and the vehicle data to be combined to obtain the vehicle data to be detected, and to combine the current positioning data and the positioning data to be combined to obtain the positioning data to be detected; the model calculation module 240 is used to input the vehicle data to be detected and the positioning data to be detected into a pre-trained positioning risk judgment model to obtain measurement source risk data; the risk processing module 250 is used to determine the target positioning risk based on the measurement source risk data, the preset positioning risk factor matrix, and the importance coefficient of each positioning measurement source.

[0083] Based on the above example, optionally, the data processing module 220 is further configured to: determine a playback period corresponding to each playback time in the playback time series; obtain playback vehicle data corresponding to the playback period from the historical vehicle data, arrange the playback vehicle data in chronological order, and determine the median data in the arranged playback vehicle data as the vehicle data to be combined; obtain playback location data corresponding to the playback period from the historical location data, arrange the playback location data in chronological order, and determine the median data in the arranged playback location data as the location data to be combined; wherein, if the playback time is the minimum value in the playback time series, then the playback period is the period from the difference between the current time and the minimum value to the current time; if the playback time is any value other than the minimum value in the playback time series, then the playback period is the period from the difference between the current time and the previous time to the difference between the current time and the playback time.

[0084] Based on the above example, optionally, after acquiring the current vehicle data, historical vehicle data, current positioning data, and historical positioning data, the system further includes: a data preprocessing module, used to determine the number of merged fields based on the fields of the measurement data collected by each of the positioning measurement sources; wherein the number of merged fields is greater than the number of positioning measurement sources; and expanding the current positioning data and each of the historical positioning data into a square matrix based on the number of merged fields.

[0085] Based on the above example, optionally, the pre-trained location risk assessment model can be trained using the following model training module:

[0086] Based on the sample vehicle data, sample location data, and the review time series, multiple sets of test vehicle data and test location data corresponding to each set of test vehicle data are determined.

[0087] For each set of location data to be tested, the location data to be tested that is latest in time sequence is taken as the target location data. Based on the abnormal thresholds corresponding to each field in the target location data, the target location data is transformed into target data to be tested.

[0088] Using the corresponding vehicle data and the location data to be tested as independent variables, and the target data to be tested corresponding to the location data to be tested as the dependent variable, the initial long short-term memory artificial neural network is trained to obtain a pre-trained location risk judgment model.

[0089] Based on the above example, optionally, the model training module is further configured to, for each field in the target location data, if the field value of the field reaches the abnormal threshold corresponding to the field, determine the value corresponding to the field in the target data to be tested as a first value; if the field value of the field does not reach the abnormal threshold corresponding to the field, determine the value corresponding to the field in the target data to be tested as a second value; wherein, the first value is 0 or 1, the second value is 0 or 1, if the first value is 0, then the second value is 1, if the first value is 1, then the second value is 0.

[0090] Based on the above example, optionally, the risk processing module 250 is further configured to obtain the positioning risk coefficient corresponding to each of the positioning measurement sources based on the measurement source risk data and the preset positioning risk factor matrix; and to determine the target positioning risk based on the positioning risk coefficient and importance coefficient corresponding to each of the positioning measurement sources.

[0091] Based on the above example, optionally, after determining the target location risk, the system further includes: an operation processing module, configured to: if the target location risk reaches a first risk threshold but does not reach a second risk threshold, perform a warning operation; if the target location risk reaches the second risk threshold but does not reach a third risk threshold, control the vehicle to perform a deceleration operation; if the target location risk reaches the third risk threshold, control the vehicle to perform an emergency stop operation; wherein the third risk threshold is greater than the second risk threshold, and the second risk threshold is greater than the first risk threshold.

[0092] The location risk processing device provided in this disclosure can execute the steps in the location risk processing method provided in this disclosure, and has the execution steps and beneficial effects, which will not be described in detail here.

[0093] Figure 3 This is a schematic diagram of the structure of an electronic device according to an embodiment of this disclosure. See below for details. Figure 3 It shows a schematic diagram of a structure suitable for implementing the electronic device 300 in the embodiments of this disclosure. Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments disclosed herein.

[0094] like Figure 3 As shown, the electronic device 300 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes to implement the methods of the embodiments described herein, based on a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the electronic device 300. The processing device 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0095] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts, thereby implementing the risk location processing method as described above. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 309, or installed from a storage device 308, or installed from a ROM 302. When the computer program is executed by the processing device 301, it performs the functions defined in the methods of embodiments of this disclosure.

[0096] It should be noted that the computer-readable medium described in this disclosure can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this disclosure, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. In this disclosure, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0097] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs that, when executed by the electronic device, cause the electronic device to:

[0098] Acquire current vehicle data, historical vehicle data, current location data, and historical location data; wherein, the current location data includes the measurement data of each location measurement source at the current moment, and the historical location data includes the measurement data of each location measurement source at each historical moment;

[0099] Based on the review time series and the historical vehicle data, determine the vehicle data to be combined corresponding to each review time in the review time series, and determine the location data to be combined corresponding to each review time in the review time series based on the review time series and the historical location data; wherein, the review time series is an arithmetic sequence arranged in ascending order;

[0100] The current vehicle data and the vehicle data to be combined are combined to obtain the vehicle data to be detected, and the current positioning data and the positioning data to be combined are combined to obtain the positioning data to be detected.

[0101] The vehicle data to be detected and the positioning data to be detected are input into a pre-trained positioning risk judgment model to obtain measurement source risk data.

[0102] The target positioning risk is determined based on the measurement source risk data, the preset positioning risk factor matrix, and the importance coefficient of each positioning measurement source.

[0103] Optionally, when one or more of the above-described procedures are executed by the electronic device, the electronic device may also perform other steps described in the above embodiments.

[0104] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0105] The above description is merely a preferred embodiment of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features disclosed in this disclosure that have similar functions.

Claims

1. A method for handling location-based risks, characterized in that, The method includes: Acquire current vehicle data, historical vehicle data, current location data, and historical location data; wherein, the current location data includes the measurement data of each location measurement source at the current moment, and the historical location data includes the measurement data of each location measurement source at each historical moment; Based on the review time series and the historical vehicle data, determine the vehicle data to be combined corresponding to each review time in the review time series, and determine the location data to be combined corresponding to each review time in the review time series based on the review time series and the historical location data; wherein, the review time series is an arithmetic sequence arranged in ascending order; The current vehicle data and the vehicle data to be combined are combined to obtain the vehicle data to be detected, and the current positioning data and the positioning data to be combined are combined to obtain the positioning data to be detected. The vehicle data to be detected and the positioning data to be detected are input into a pre-trained positioning risk judgment model to obtain measurement source risk data. The target positioning risk is determined based on the measurement source risk data, the preset positioning risk factor matrix, and the importance coefficient of each positioning measurement source.

2. The method according to claim 1, characterized in that, The step of determining the vehicle data to be combined corresponding to each review time within the review time series based on the review time series and the historical vehicle data, and determining the location data to be combined corresponding to each review time within the review time series based on the review time series and the historical location data, includes: For each playback time within the playback time series, determine the playback period corresponding to that playback time; Obtain the playback vehicle data corresponding to the playback period from the historical vehicle data, arrange the playback vehicle data in chronological order, and determine the median data in the arranged playback vehicle data as the vehicle data to be combined. Obtain the playback location data corresponding to the playback period from the historical location data, arrange the playback location data in chronological order, and determine the median data in the arranged playback location data as the location data to be combined. Wherein, if the review time is the minimum value in the review time series, then the review period is the period from the difference between the current time and the minimum value to the current time; if the review time is any value in the review time series other than the minimum value, then the review period is the period from the difference between the current time and the time before the review time to the difference between the current time and the review time.

3. The method according to claim 1, characterized in that, After acquiring current vehicle data, historical vehicle data, current location data, and historical location data, the process also includes: The number of fields to be merged is determined based on the fields of the measurement data collected by each of the positioning measurement sources; wherein the number of fields to be merged is greater than the number of positioning measurement sources. Based on the number of merged fields, the current location data and each of the historical location data are expanded into a square matrix.

4. The method according to claim 1, characterized in that, The pre-trained location risk assessment model was trained using the following method: Based on the sample vehicle data, sample location data, and the review time series, multiple sets of test vehicle data and test location data corresponding to each set of test vehicle data are determined. For each set of location data to be tested, the location data to be tested that is latest in time is taken as the target location data. Based on the abnormal thresholds corresponding to each field in the target location data, the target location data is transformed into target data to be tested. Using the corresponding vehicle data and the location data to be tested as independent variables, and the target data to be tested corresponding to the location data to be tested as the dependent variable, the initial long short-term memory artificial neural network is trained to obtain a pre-trained location risk judgment model.

5. The method according to claim 4, characterized in that, The step of converting the target location data into test target data based on the anomaly thresholds corresponding to each field in the target location data includes: For each field in the target location data, if the field value reaches the corresponding abnormal threshold, the value corresponding to the field in the target data to be tested is determined as a first value; if the field value does not reach the corresponding abnormal threshold, the value corresponding to the field in the target data to be tested is determined as a second value; wherein, the first value is 0 or 1, the second value is 0 or 1, if the first value is 0, then the second value is 1, and if the first value is 1, then the second value is 0.

6. The method according to claim 1, characterized in that, The step of determining the target positioning risk based on the measurement source risk data, the preset positioning risk factor matrix, and the importance coefficient of each positioning measurement source includes: Based on the risk data of the measurement sources and the preset positioning risk factor matrix, the positioning risk coefficient corresponding to each of the positioning measurement sources is obtained; The target positioning risk is determined based on the positioning risk coefficient and importance coefficient corresponding to each of the positioning measurement sources.

7. The method according to claim 1, characterized in that, Following the determination of target location risk, the following is also included: If the target location risk reaches the first risk threshold but does not reach the second risk threshold, then an early warning operation is performed; If the target location risk reaches the second risk threshold but does not reach the third risk threshold, then control the vehicle to perform a deceleration operation; If the target location risk reaches the third risk threshold, the vehicle is controlled to perform an emergency stop operation; wherein the third risk threshold is greater than the second risk threshold, and the second risk threshold is greater than the first risk threshold.

8. A location risk handling device, characterized in that, include: The data acquisition module is used to acquire current vehicle data, historical vehicle data, current positioning data, and historical positioning data; wherein, the current positioning data includes the measurement data of each positioning measurement source at the current moment, and the historical positioning data includes the measurement data of each positioning measurement source at each historical moment; The data processing module is used to determine, based on the review time series and the historical vehicle data, the vehicle data to be combined corresponding to each review time in the review time series, and to determine, based on the review time series and the historical positioning data, the positioning data to be combined corresponding to each review time in the review time series; wherein, the review time series is an arithmetic sequence arranged in ascending order; The data combination module is used to combine the current vehicle data and the vehicle data to be combined to obtain the vehicle data to be detected, and to combine the current positioning data and the positioning data to be combined to obtain the positioning data to be detected. The model calculation module is used to input the vehicle data to be detected and the positioning data to be detected into a pre-trained positioning risk judgment model to obtain measurement source risk data. The risk processing module is used to determine the target positioning risk based on the measurement source risk data, the preset positioning risk factor matrix, and the importance coefficient of each positioning measurement source.

9. An electronic device, characterized in that, The electronic device includes: One or more processors; Storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the location risk processing method as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the location risk handling method as described in any one of claims 1-7.

Images