A cloud-edge-based auto insurance intelligent application method and platform

Abstract

The present invention relates to the technical field of a car insurance cloud platform, and discloses a car insurance intelligent application method based on a cloud edge terminal, the method comprising: collecting vehicle driving data at an accident scene through vehicle sensors and preprocessing to remove vehicle operation periodic data to obtain preprocessed vehicle driving data, compressing the preprocessed vehicle driving data and transmitting it to an edge server; the edge server performs abnormal sensor identification on the received preprocessed vehicle driving data to filter out key sensor operation data; the filtered key sensor operation data is sent to a cloud server, and the cloud server locates a damaged part of the received key sensor operation data; the located damaged part and the collected original vehicle driving data are packaged and uploaded to a car insurance intelligent application platform for automatic declaration, so as to shorten the processing time of car insurance claims and improve claims efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of auto insurance cloud platform, and in particular to an auto insurance intelligent application method and platform based on cloud-edge. Background Art

[0002] With the rapid development of the automobile industry, the number of cars has increased significantly, and the demand for auto insurance claims has also increased. The traditional auto insurance claims process mainly relies on manual operations, which has problems such as low efficiency, high cost and lack of transparency. In recent years, the development of cloud computing, edge computing and artificial intelligence technologies has provided new solutions for intelligent processing of auto insurance. Cloud computing can handle large amounts of data and complex computing tasks by providing large-scale computing and storage resources. Edge computing can process data close to the data source, reduce latency and improve data real-time performance. Artificial intelligence, especially deep learning technology, can demonstrate excellent performance in areas such as image recognition and natural language processing, providing technical support for accident identification and damage assessment in auto insurance claims. Summary of the invention

[0003] In view of this, this solution integrates cloud computing, edge computing and artificial intelligence technologies, and proposes a cloud-edge-based intelligent application method and platform for auto insurance. By introducing cloud-edge-based architecture and artificial intelligence technology, it shortens the processing time of auto insurance claims, reduces manual intervention and improves efficiency.

[0004] To achieve the above purpose, the present invention provides a cloud-edge-based auto insurance intelligent application method, comprising the following steps:

[0005] S1: Collecting vehicle driving data at the accident scene through vehicle sensors and preprocessing to remove vehicle operation periodic data to obtain preprocessed vehicle driving data, compressing and transmitting the preprocessed vehicle driving data to an edge server, wherein data period identification is an implementation method of the preprocessing;

[0006] S2: The edge server performs abnormal sensor identification on the received pre-processed vehicle driving data to filter out key sensor operation data, wherein the node influence calculation based on the sensor association graph is the implementation method of the abnormal sensor identification;

[0007] Identifying abnormal sensors and filtering out key sensor operation data includes the following steps:

[0008] S21: Calculate the correlation coefficient based on the restored sensor time series data to determine the correlation degree between different sensors. The calculation formula is:

[0009] ;

[0010] in:

[0011] Represents the covariance between the time series data restored by the i-th sensor and the j-th sensor;

[0012] and Respectively represent the time series data restored by the i-th and j-th sensors;

[0013] and Respectively represent the standard deviation of the time series data restored by the i-th and j-th sensors;

[0014] S22: construct a sensor association graph according to the calculated association coefficient, represent the sensor as a node in the sensor association graph, use edges to represent the correlation between the sensors, and the weight of the edge is represented as the association coefficient;

[0015] S23: Divide the restored time series data into sliding windows according to the preset window size and calculate the data mean in the sliding window; calculate the deviation between the data mean in the sliding window and the data mean under normal operation; if the deviation is less than the specified threshold, it means that there is no abnormality in the current sensor, otherwise there is an abnormality, and the influence value of the sensor node with abnormality in the sensor association graph is set to MAX, and the influence value of the non-abnormal node is set to 0;

[0016] S24: Based on the sensor association graph and the identified abnormal sensors, use the PageRank algorithm to search for all nodes in the sensor association graph whose node influence values ​​exceed the specified threshold, and add the nodes exceeding the specified threshold as potential abnormal sensor nodes. The node influence calculation formula is:

[0017] ;

[0018] in:

[0019] represents the influence value of sensor i;

[0020] represents the damping factor, and its value is 0.85;

[0021] represents the total number of sensor nodes;

[0022] represents the set of adjacent sensor nodes of sensor i;

[0023] represents the degree of node j;

[0024] S25: taking the potential abnormal sensors and the identified abnormal sensors as key sensors, and screening the operation data of the key sensors from the pre-processed vehicle driving data;

[0025] S3: sending the selected key sensor operation data to the cloud server, and the cloud server locates the damaged part based on the received key sensor operation data, wherein the time sequence diagram model is an implementation method of the damaged part location;

[0026] The cloud server locates the damaged part based on the key sensor operation data received, including the following steps:

[0027] S31: segmenting the operation data of the selected key sensors according to a preset time window size T to obtain subsequences of operation data of different key sensors;

[0028] S32: Extract features from different key sensor operation data subsequences to form feature vectors , the eigenvector Represents the feature vector of the key sensor i in the time window t, wherein the features include the mean value, standard deviation, maximum value and minimum value;

[0029] S33: construct a time sequence diagram model, in which nodes Corresponding time window All key sensor feature vectors within , the edge in the timing graph model Represents a time window and The coupling relationship between the corresponding nodes, the coupling degree calculation formula is:

[0030] ;

[0031] S34: Calculate the coupling degree difference in adjacent time windows, and the calculation formula is:

[0032] ;

[0033] S35: If is less than the preset threshold, and is greater than the preset threshold, then the time window Injury occurs;

[0034] S36: remove non-critical sensor nodes from the sensor association graphs corresponding to all time windows, and only retain the critical sensor association graph composed of critical sensor nodes, using the identified time window of damage as the interval, integrating the critical sensor association graphs before the time window of damage into a standard association graph through graph clustering, integrating the critical sensor association graphs after the time window of damage into an abnormal association graph, and accurately locating the abnormal sensor by calculating the difference matrix between the standard association graph and the abnormal association graph, wherein the graph edit distance is the similarity measurement index of the graph clustering;

[0035] S37: locating the damaged part according to the correspondence between the located abnormal sensor and the detection part of the sensor;

[0036] S4: The located damage parts and the collected original vehicle driving data are packaged and uploaded to the auto insurance intelligent application platform for automatic reporting.

[0037] As a further improvement method of the present invention:

[0038] Optionally, in step S1, the vehicle driving data at the accident scene is collected by a vehicle sensor and preprocessed to remove the vehicle operation periodic data to obtain the preprocessed vehicle driving data, including:

[0039] S11: Perform period fitting on the collected vehicle driving data through Fourier series, and the calculation formula is:

[0040] ;

[0041] in:

[0042] represents the fitted value at time t;

[0043] represents the order of the Fourier series;

[0044] and represents the coefficients of the Fourier series;

[0045] Indicates the cycle length;

[0046] S12: Subtracting the period value obtained by fitting at the corresponding time from the vehicle driving data to eliminate the vehicle operation periodic data and obtain the pre-processed vehicle driving data.

[0047] Optionally, in step S1, encrypting and transmitting the pre-processed vehicle driving data to the edge server includes:

[0048] Fourier transform is used to extract driving information from the preprocessed vehicle driving data, and the extracted signal coefficient information is encrypted and sent to the edge server to reduce the amount of data sent. The specific process includes:

[0049] A11: Perform discrete Fourier transform on the preprocessed time series data to convert the time domain data into the frequency domain to obtain the signals of the time series data at different frequencies. The calculation formula is:

[0050] ;

[0051] in:

[0052] express The data after preprocessing at each moment;

[0053] Indicates frequency Frequency domain signal under conditions;

[0054] Indicates the number of time series data points;

[0055] A12: RC4 encrypts the signal coefficients whose signal amplitudes at different frequencies are greater than a specified threshold, and sends the encrypted data to the edge server;

[0056] A13: Decrypt the encrypted data to restore the signal coefficients, and perform inverse discrete Fourier transform based on the signal coefficients to restore the preprocessed time series data. The calculation formula is:

[0057] ;

[0058] in:

[0059] Represents the frequency domain signal restored according to the signal coefficient;

[0060] Represents the restored time series data.

[0061] Optionally, in step S36, all key sensor association graphs before the damage time window are integrated into a standard association graph through graph clustering, and key sensor association graphs after the damage time window are integrated into an abnormal association graph, including:

[0062] S361: Calculate the similarity between any key sensor association graphs, wherein the graph edit distance is an implementation method of the similarity metric, and the graph edit distance is the minimum number of edit operations required to convert one key sensor association graph into another key sensor association graph, and the edit operation includes adding and deleting edge operations;

[0063] S362: Using graph edit distance as the similarity metric, using the Kmeans algorithm, and setting the cluster center to 1, cluster the adjacency matrices corresponding to the key sensor association graphs before the damage time window, cluster the adjacency matrices corresponding to the key sensor association graphs after the damage time window, perform a mean operation on the adjacency matrices corresponding to the first M key sensor association graphs that are closest to the cluster center, and obtain the adjacency matrices of the standard association graph and the abnormal association graph, respectively, thereby obtaining the corresponding standard association graph and abnormal association graph.

[0064] Optionally, the step S36 of accurately locating the abnormal sensor by calculating a difference matrix between the standard association graph and the abnormal association graph includes:

[0065] A361: Calculate the difference between the adjacency matrix corresponding to the standard association graph and the adjacency matrix corresponding to the abnormal association graph to obtain the difference matrix. The calculation formula is:

[0066] ;

[0067] in:

[0068] is the element of the difference matrix, which represents the difference value of the correlation between the i-th sensor and the j-th sensor between the standard correlation graph and the abnormal correlation graph;

[0069] and denote the correlation coefficient between the i-th sensor and the j-th sensor in the standard correlation diagram and the abnormal correlation diagram respectively;

[0070] A362: Analyze the difference matrix. If the value of the difference matrix element is greater than the preset threshold, there is an abnormality in the two sensors involved in the element.

[0071] The located damaged parts and the collected original vehicle driving data are packaged and uploaded to the auto insurance intelligent application platform for automatic reporting.

[0072] In order to solve the above problems, the present invention provides a cloud-edge-end based auto insurance intelligent application platform, the platform comprising:

[0073] The vehicle driving data collection module is used to collect the vehicle driving data at the accident scene through the vehicle sensor and pre-process it to remove the vehicle operation periodic data, obtain the pre-processed vehicle driving data, and compress the pre-processed vehicle driving data and transmit it to the edge server;

[0074] Key sensor operation data screening module: the edge server performs abnormal sensor identification on the received pre-processed vehicle driving data to screen out key sensor operation data;

[0075] The damage part locating device is used to send the screened key sensor operation data to the cloud server. The cloud server locates the damage part based on the received key sensor operation data, and packages the located damage part and the collected original vehicle driving data and uploads them to the motor vehicle insurance intelligent application platform for automatic reporting.

[0076] In order to solve the above problem, the present invention further provides an electronic device, the electronic device comprising:

[0077] A memory storing at least one instruction;

[0078] A communication interface to realize electronic device communication; and a processor to execute instructions stored in the memory to implement the above-mentioned cloud-edge-based intelligent application method for auto insurance.

[0079] In order to solve the above problems, the present invention also provides a computer-readable storage medium, in which at least one instruction is stored. The at least one instruction is executed by a processor in an electronic device to implement the above-mentioned cloud-edge-based intelligent application method for auto insurance.

[0080] Compared with the prior art, the present invention proposes a cloud-edge-based intelligent application method for auto insurance, which has the following advantages:

[0081] First of all, this solution proposes a cloud-edge-based intelligent application method for auto insurance. This solution combines periodic data elimination, key sensor screening and identification, positioning and other methods to form a complete cloud-edge-based intelligent application process for auto insurance, improve the efficiency of automated auto insurance reporting, and use vehicle sensors to collect accident scene data. It also removes periodic data through preprocessing, reduces the impact of vehicle operation periodic data on the abnormal identification of key sensors, and improves the accuracy of auto insurance processing.

[0082] At the same time, this scheme adopts a node influence calculation method based on the sensor association graph to identify abnormal sensors, which helps to quickly and accurately screen out key sensor operation data, reduce the amount of data, and improve the efficiency of abnormal sensor identification.

[0083] In addition, this solution proposes a method for accurately locating the damage site based on a time-series graph model, which can effectively help determine the specific location of vehicle damage and provide important information for subsequent automated vehicle insurance reporting and processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] Figure 1 A flowchart of a cloud-edge-based intelligent application method for auto insurance provided in one embodiment of the present invention;

[0085] Figure 2A functional module diagram of a cloud-edge-based auto insurance intelligent application platform provided in one embodiment of the present invention;

[0086] Figure 3 A schematic diagram of the structure of an electronic device for implementing a cloud-edge-based intelligent application method for auto insurance provided in one embodiment of the present invention. DETAILED DESCRIPTION

[0087] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings.

[0088] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0089] The embodiment of the present application provides a cloud-edge-based intelligent application method for auto insurance. The execution subject of the cloud-edge-based intelligent application method for auto insurance includes but is not limited to at least one of the electronic devices such as the server and the terminal that can be configured to execute the method provided by the embodiment of the present application. In other words, the cloud-edge-based intelligent application method for auto insurance can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc.

[0090] S1: Collect vehicle driving data at the accident scene through vehicle sensors and preprocess to remove vehicle operation periodic data to obtain preprocessed vehicle driving data, compress the preprocessed vehicle driving data and transmit it to the edge server, wherein data cycle identification is the implementation method of the preprocessing.

[0091] In step S1, the vehicle driving data at the accident scene is collected by the vehicle sensor and pre-processed to remove the vehicle operation periodic data, so as to obtain the pre-processed vehicle driving data, including:

[0092] S11: Perform period fitting on the collected vehicle driving data through Fourier series, and the calculation formula is:

[0093] ;

[0094] in:

[0095] represents the fitted value at time t;

[0096] represents the order of the Fourier series;

[0097] and represents the coefficients of the Fourier series;

[0098] Indicates the cycle length;

[0099] S12: Subtracting the period value obtained by fitting at the corresponding time from the vehicle driving data to eliminate the vehicle operation periodic data and obtain the pre-processed vehicle driving data.

[0100] The step S1 encrypts and transmits the pre-processed vehicle driving data to the edge server, including:

[0101] Fourier transform is used to extract driving information from the preprocessed vehicle driving data, and the extracted signal coefficient information is encrypted and sent to the edge server to reduce the amount of data sent. The specific process includes:

[0102] A11: Perform discrete Fourier transform on the preprocessed time series data to convert the time domain data into the frequency domain to obtain the signals of the time series data at different frequencies. The calculation formula is:

[0103] ;

[0104] in:

[0105] express The data after preprocessing at each moment;

[0106] Indicates frequency Frequency domain signal under conditions;

[0107] Indicates the number of time series data points;

[0108] A12: RC4 encrypts the signal coefficients whose signal amplitudes at different frequencies are greater than a specified threshold, and sends the encrypted data to the edge server;

[0109] A13: Decrypt the encrypted data to restore the signal coefficients, and perform inverse discrete Fourier transform based on the signal coefficients to restore the preprocessed time series data. The calculation formula is:

[0110] ;

[0111] in:

[0112] Represents the frequency domain signal restored according to the signal coefficient;

[0113] Represents the restored time series data.

[0114] S2: The edge server performs abnormal sensor identification on the received pre-processed vehicle driving data to filter out key sensor operation data, wherein node influence calculation based on the sensor association graph is an implementation method of the abnormal sensor identification.

[0115] In step S2, the edge server performs abnormal sensor identification on the received pre-processed vehicle driving data to filter out key sensor operation data, including:

[0116] S21: Calculate the correlation coefficient based on the restored sensor time series data to determine the correlation degree between different sensors. The calculation formula is:

[0117] ;

[0118] in:

[0119] Represents the covariance between the time series data restored by the i-th sensor and the j-th sensor;

[0120] and Respectively represent the time series data restored by the i-th and j-th sensors;

[0121] and Respectively represent the standard deviation of the time series data restored by the i-th and j-th sensors;

[0122] S22: construct a sensor association graph according to the calculated association coefficient, represent the sensor as a node in the sensor association graph, use edges to represent the correlation between the sensors, and the weight of the edge is represented as the association coefficient;

[0123] S23: Divide the restored time series data into sliding windows according to the preset window size and calculate the data mean in the sliding window; calculate the deviation between the data mean in the sliding window and the data mean under normal operation; if the deviation is less than the specified threshold, it means that there is no abnormality in the current sensor, otherwise there is an abnormality, and the influence value of the sensor node with abnormality in the sensor association graph is set to MAX, and the influence value of the non-abnormal node is set to 0;

[0124] S24: Based on the sensor association graph and the identified abnormal sensors, use the PageRank algorithm to search for all nodes in the sensor association graph whose node influence values ​​exceed the specified threshold, and add the nodes exceeding the specified threshold as potential abnormal sensor nodes. The node influence calculation formula is:

[0125] ;

[0126] in:

[0127] represents the influence value of sensor i;

[0128] represents the damping factor, and its value is 0.85;

[0129] represents the total number of sensor nodes;

[0130] represents the set of adjacent sensor nodes of sensor i;

[0131] represents the degree of node j;

[0132] S25: The potential abnormal sensors and the identified abnormal sensors are regarded as key sensors, and the operation data of the key sensors are screened out from the pre-processed vehicle driving data.

[0133] S3: Send the selected key sensor operation data to the cloud server, and the cloud server locates the damaged part based on the received key sensor operation data, wherein the timing diagram model is an implementation method of the damaged part location.

[0134] In step S3, the cloud server locates the damaged part based on the received key sensor operation data, including:

[0135] S31: segmenting the operation data of the selected key sensors according to a preset time window size T to obtain subsequences of operation data of different key sensors;

[0136] S32: Extract features from different key sensor operation data subsequences to form feature vectors , the eigenvector Represents the feature vector of the key sensor i in the time window t, wherein the features include the mean value, standard deviation, maximum value and minimum value;

[0137] S33: construct a time sequence diagram model, in which nodes Corresponding time window All key sensor feature vectors within , the edge in the timing graph model Represents a time window and The coupling relationship between the corresponding nodes, the coupling degree calculation formula is:

[0138] ;

[0139] S34: Calculate the coupling degree difference in adjacent time windows, and the calculation formula is:

[0140] ;

[0141] S35: If is less than the preset threshold, and is greater than the preset threshold, then the time window Injury occurs;

[0142] S36: remove non-critical sensor nodes from the sensor association graphs corresponding to all time windows, and only retain the critical sensor association graph composed of critical sensor nodes, using the identified time window of damage as the interval, integrating the critical sensor association graphs before the time window of damage into a standard association graph through graph clustering, integrating the critical sensor association graphs after the time window of damage into an abnormal association graph, and accurately locating the abnormal sensor by calculating the difference matrix between the standard association graph and the abnormal association graph, wherein the graph edit distance is the similarity measurement index of the graph clustering;

[0143] S37: Locate the damaged part according to the correspondence between the located abnormal sensor and the detection part of the sensor.

[0144] In the step S36, all key sensor association graphs before the damage time window are integrated into a standard association graph through graph clustering, and the key sensor association graphs after the damage time window are integrated into an abnormal association graph, including:

[0145] S361: Calculate the similarity between any key sensor association graphs, wherein the graph edit distance is an implementation method of the similarity metric, and the graph edit distance is the minimum number of edit operations required to convert one key sensor association graph into another key sensor association graph, and the edit operation includes adding and deleting edge operations;

[0146] S362: Using graph edit distance as the similarity metric, using the Kmeans algorithm, and setting the cluster center to 1, cluster the adjacency matrices corresponding to the key sensor association graphs before the damage time window, cluster the adjacency matrices corresponding to the key sensor association graphs after the damage time window, perform a mean operation on the adjacency matrices corresponding to the first M key sensor association graphs that are closest to the cluster center, and obtain the adjacency matrices of the standard association graph and the abnormal association graph, respectively, thereby obtaining the corresponding standard association graph and abnormal association graph.

[0147] In the step S36, the abnormal sensor is accurately located by calculating the difference matrix between the standard association graph and the abnormal association graph, including:

[0148] A361: Calculate the difference between the adjacency matrix corresponding to the standard association graph and the adjacency matrix corresponding to the abnormal association graph to obtain the difference matrix. The calculation formula is:

[0149] ;

[0150] in:

[0151] is the element of the difference matrix, which represents the difference value of the correlation between the i-th sensor and the j-th sensor between the standard correlation graph and the abnormal correlation graph;

[0152] and denote the correlation coefficient between the i-th sensor and the j-th sensor in the standard correlation diagram and the abnormal correlation diagram respectively;

[0153] A362: Analyze the difference matrix. If the value of the difference matrix element is greater than the preset threshold, there is an abnormality in the two sensors involved in the element.

[0154] S4: The located damage parts and the collected original vehicle driving data are packaged and uploaded to the auto insurance intelligent application platform for automatic reporting.

[0155] like Figure 2 As shown, it is a functional module diagram of a cloud-edge-based auto insurance intelligent application platform provided by an embodiment of the present invention, which can implement the cloud-edge-based auto insurance intelligent application method in Example 1.

[0156] The cloud-edge-based auto insurance intelligent application platform 100 of the present invention can be installed in an electronic device. According to the functions implemented, the cloud-edge-based auto insurance intelligent application platform can include a vehicle driving data acquisition module 101, a key sensor operation data screening module 102 and a damage location positioning device 103. The module of the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0157] The vehicle driving data collection module 101 is used to collect the vehicle driving data at the accident scene through the vehicle sensor and pre-process it to remove the vehicle operation periodic data, obtain the pre-processed vehicle driving data, and compress the pre-processed vehicle driving data and transmit it to the edge server;

[0158] The key sensor operation data screening module 102, the edge server performs abnormal sensor identification on the received pre-processed vehicle driving data to screen out the key sensor operation data;

[0159] The damage part locating device 103 is used to send the screened key sensor operation data to the cloud server. The cloud server locates the damage part based on the received key sensor operation data, and packages the located damage part and the collected original vehicle driving data and uploads them to the motor vehicle insurance intelligent application platform for automatic reporting.

[0160] In detail, each module in the cloud-edge-based auto insurance intelligent application platform 100 in the embodiment of the present invention is used in the same manner as above. Figure 1 The technical means are the same as the cloud-edge-based auto insurance intelligent application method described in, and can produce the same technical effects, so I will not go into details here.

[0161] like Figure 3 The figure is a schematic diagram of the structure of an electronic device for implementing a cloud-edge-based intelligent application method for auto insurance provided by an embodiment of the present invention.

[0162] The electronic device 1 may include a processor 10 , a memory 11 , a communication interface 13 and a bus, and may also include a computer program stored in the memory 11 and executable on the processor 10 , such as a program 12 .

[0163] The memory 11 includes at least one type of readable storage medium, and the readable storage medium includes a flash memory, a mobile hard disk, a multimedia card, a card-type memory (e.g., SD or DX memory, etc.), a magnetic memory, a disk, an optical disk, etc. In some embodiments, the memory 11 may be an internal storage unit of the electronic device 1, such as a mobile hard disk of the electronic device 1. In other embodiments, the memory 11 may also be an external storage device of the electronic device 1, such as a plug-in mobile hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device 1. Further, the memory 11 may also include both an internal storage unit of the electronic device 1 and an external storage device. The memory 11 may not only be used to store application software and various types of data installed in the electronic device 1, such as the code of the program 12, etc., but also be used to temporarily store data that has been output or is to be output.

[0164] The processor 10 may be composed of an integrated circuit in some embodiments, for example, a single packaged integrated circuit, or a plurality of packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and a combination of various control chips, etc. The processor 10 is the control core (Control Unit) of the electronic device, and uses various interfaces and lines to connect the various components of the entire electronic device, and executes or executes the programs or modules stored in the memory 11 (such as the program 12 for realizing the intelligent application of auto insurance based on the cloud edge), and calls the data stored in the memory 11 to execute various functions of the electronic device 1 and process data.

[0165] The communication interface 13 may include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between the electronic device 1 and other electronic devices, and to achieve connection and communication between internal components of the electronic device.

[0166] The bus may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The bus is configured to realize connection and communication between the memory 11 and at least one processor 10, etc.

[0167] Figure 3 Only an electronic device with components is shown, and those skilled in the art will understand that Figure 3 The structure shown does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0168] For example, although not shown, the electronic device 1 may also include a power source (such as a battery) for supplying power to various components. Preferably, the power source may be logically connected to the at least one processor 10 through a power management device, so that the power management device can realize functions such as charging management, discharging management, and power consumption management. The power source may also include one or more DC or AC power sources, recharging devices, power failure detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The electronic device 1 may also include a variety of sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be repeated here.

[0169] Optionally, the electronic device 1 may further include a user interface, which may be a display, an input unit (such as a keyboard), or a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, and an OLED (Organic Light-Emitting Diode) touch device. The display may also be appropriately referred to as a display screen or a display unit, which is used to display information processed in the electronic device 1 and to display a visual user interface.

[0170] It should be understood that the embodiment is for illustration only and the scope of the patent application is not limited to this structure.

[0171] It should be noted that the serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments. And the terms "including", "comprising" or any other variants thereof in this article are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "including a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0172] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0173] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A cloud-edge-based intelligent application method for auto insurance, characterized in that: The method comprises the following steps: S1: Collecting vehicle driving data at the accident scene through vehicle sensors and preprocessing to remove vehicle operation periodic data to obtain preprocessed vehicle driving data, compressing and transmitting the preprocessed vehicle driving data to an edge server, wherein data period identification is an implementation method of the preprocessing; S2: The edge server performs abnormal sensor identification on the received pre-processed vehicle driving data to filter out key sensor operation data, wherein the node influence calculation based on the sensor association graph is the implementation method of the abnormal sensor identification; Identifying abnormal sensors and filtering out key sensor operation data includes the following steps: S21: Calculate the correlation coefficient based on the restored sensor time series data to determine the correlation degree between different sensors. The calculation formula is: ； in: Represents the covariance between the time series data restored by the i-th sensor and the j-th sensor; and Respectively represent the time series data restored by the i-th and j-th sensors; and Respectively represent the standard deviation of the time series data restored by the i-th and j-th sensors; S22: construct a sensor association graph according to the calculated association coefficient, represent the sensor as a node in the sensor association graph, use edges to represent the correlation between the sensors, and the weight of the edge is represented as the association coefficient; S23: Divide the restored time series data into sliding windows according to the preset window size and calculate the data mean in the sliding window; calculate the deviation between the data mean in the sliding window and the data mean under normal operation; if the deviation is less than the specified threshold, it means that there is no abnormality in the current sensor, otherwise there is an abnormality, and the influence value of the sensor node with abnormality in the sensor association graph is set to MAX, and the influence value of the non-abnormal node is set to 0; S24: Based on the sensor association graph and the identified abnormal sensors, use the PageRank algorithm to search for all nodes in the sensor association graph whose node influence values ​​exceed the specified threshold, and add the nodes exceeding the specified threshold as potential abnormal sensor nodes. The node influence calculation formula is: ； in: represents the influence value of sensor i; represents the damping factor, and its value is 0.85; represents the total number of sensor nodes; represents the set of adjacent sensor nodes of sensor i; represents the degree of node j; S25: taking the potential abnormal sensors and the identified abnormal sensors as key sensors, and screening the operation data of the key sensors from the pre-processed vehicle driving data; S3: sending the selected key sensor operation data to the cloud server, and the cloud server locates the damaged part based on the received key sensor operation data, wherein the time sequence diagram model is an implementation method of the damaged part location; The cloud server locates the damaged part based on the key sensor operation data received, including the following steps: S31: segmenting the screened operation data of the key sensors according to a preset time window size T to obtain subsequences of operation data of different key sensors; S32: Extract features from different key sensor operation data subsequences to form feature vectors , the eigenvector Represents the feature vector of the key sensor i in the time window t, wherein the features include the mean value, standard deviation, maximum value and minimum value; S33: construct a time sequence diagram model, in which nodes Corresponding time window All key sensor feature vectors within , the edge in the timing graph model Represents a time window and The coupling relationship between the corresponding nodes, the coupling degree calculation formula is: ； S34: Calculate the coupling degree difference in adjacent time windows, and the calculation formula is: ； S35: If is less than the preset threshold, and is greater than the preset threshold, then the time window Injury occurs; S36: remove non-critical sensor nodes from the sensor association graphs corresponding to all time windows, and only retain the critical sensor association graphs composed of critical sensor nodes, using the identified time window of damage as the interval, integrating the critical sensor association graphs before the time window of damage into a standard association graph through graph clustering, integrating the critical sensor association graphs after the time window of damage into an abnormal association graph, and accurately locating the abnormal sensor by calculating the difference matrix between the standard association graph and the abnormal association graph, wherein the graph edit distance is the similarity measurement index of the graph clustering; S37: locating the damaged part according to the correspondence between the located abnormal sensor and the detection part of the sensor; S4: The located damage parts and the collected original vehicle driving data are packaged and uploaded to the auto insurance intelligent application platform for automatic reporting.

2. The cloud-edge-based intelligent application method for auto insurance according to claim 1, characterized in that: In the step S1, the vehicle driving data at the accident scene is collected by the vehicle sensor and pre-processed to remove the vehicle operation periodic data, so as to obtain the pre-processed vehicle driving data, including the following steps: S11: Perform period fitting on the collected vehicle driving data through Fourier series, and the calculation formula is: ； in: represents the fitted value at time t; represents the order of the Fourier series; and represents the coefficients of the Fourier series; Indicates the cycle length; S12: Subtracting the period value obtained by fitting at the corresponding time from the vehicle driving data to eliminate the vehicle operation periodic data and obtain the pre-processed vehicle driving data.

3. The cloud-edge-based intelligent application method for auto insurance according to claim 1, characterized in that: In the step S1, the pre-processed vehicle driving data is encrypted and transmitted to the edge server, including the following steps: Fourier transform is used to extract driving information from the preprocessed vehicle driving data, and the extracted signal coefficient information is encrypted and sent to the edge server to reduce the amount of data sent. The specific process includes: A11: Perform discrete Fourier transform on the preprocessed time series data to convert the time domain data into the frequency domain to obtain the signals of the time series data at different frequencies. The calculation formula is: ； in: express The data after preprocessing at each moment; Indicates frequency Frequency domain signal under conditions; Indicates the number of time series data points; A12: RC4 encrypts the signal coefficients whose signal amplitudes at different frequencies are greater than a specified threshold, and sends the encrypted data to the edge server; A13: Decrypt the encrypted data to restore the signal coefficients, and perform inverse discrete Fourier transform based on the signal coefficients to restore the preprocessed time series data. The calculation formula is: ； in: Represents the frequency domain signal restored according to the signal coefficient; Represents the restored time series data.

4. The cloud-edge-based intelligent application method for auto insurance according to claim 1, characterized in that: In the step S36, all key sensor association graphs before the damage time window are integrated into a standard association graph through graph clustering, and the key sensor association graphs after the damage time window are integrated into an abnormal association graph, including the following steps: S361: Calculate the similarity between any key sensor association graphs, wherein the graph edit distance is an implementation method of the similarity metric, and the graph edit distance is the minimum number of edit operations required to convert one key sensor association graph into another key sensor association graph, and the edit operation includes adding and deleting edge operations; S362: Using graph edit distance as the similarity metric, using the Kmeans algorithm, and setting the cluster center to 1, cluster the adjacency matrices corresponding to the key sensor association graphs before the damage time window, cluster the adjacency matrices corresponding to the key sensor association graphs after the damage time window, perform a mean operation on the adjacency matrices corresponding to the first M key sensor association graphs that are closest to the cluster center, and obtain the adjacency matrices of the standard association graph and the abnormal association graph, respectively, thereby obtaining the corresponding standard association graph and abnormal association graph.

5. The cloud-edge-based auto insurance intelligent application method according to claim 1, characterized in that: In the step S36, the abnormal sensor is accurately located by calculating the difference matrix between the standard association graph and the abnormal association graph, including the following steps: A361: Calculate the difference between the adjacency matrix corresponding to the standard association graph and the adjacency matrix corresponding to the abnormal association graph to obtain the difference matrix. The calculation formula is: ； in: is the element of the difference matrix, which represents the difference value of the correlation between the i-th sensor and the j-th sensor between the standard correlation graph and the abnormal correlation graph; and denote the correlation coefficient between the i-th sensor and the j-th sensor in the standard correlation diagram and the abnormal correlation diagram respectively; A362: Analyze the difference matrix. If the value of the difference matrix element is greater than the preset threshold, there is an abnormality in the two sensors involved in the element.

6. A cloud-edge-based auto insurance intelligent application platform, characterized in that: The platform includes: The vehicle driving data collection module is used to collect the vehicle driving data at the accident scene through the vehicle sensor and pre-process it to remove the vehicle operation periodic data, obtain the pre-processed vehicle driving data, and compress the pre-processed vehicle driving data and transmit it to the edge server; Key sensor operation data screening module: the edge server performs abnormal sensor identification on the received pre-processed vehicle driving data to screen out key sensor operation data; A damage part locating device is used to send the screened key sensor operation data to a cloud server. The cloud server locates the damage part of the received key sensor operation data, and packages the located damage part and the collected original vehicle driving data and uploads them to the auto insurance intelligent application platform for automatic reporting, so as to realize a cloud-edge-based auto insurance intelligent application method as described in any one of claims 1-5.

Images