Abnormal refueling behavior identification method, device, electronic device and storage medium

Abstract

The present invention discloses a method, device, electronic device and storage medium for identifying abnormal refueling behavior. In the abnormal refueling behavior identification scheme, each time a real-time refueling data is generated, a variety of feature data related to the refueling card used and the vehicle bound to the card are extracted, and these feature data are used to make judgments on various abnormal behaviors related to refueling cashing out for the real-time refueling data. Therefore, the abnormal judgment of this scheme is highly real-time, and this scheme distinguishes various abnormal behaviors of refueling cashing out according to dependence and independence. For abnormal behaviors with independence, only the feature data related to it are used to perform a separate abnormal judgment. For abnormal behaviors with dependence, abnormal judgment is made in a hierarchical manner. Only when the abnormal behavior judgment result of the upper layer is true, the abnormal behavior of the next layer is judged, so that the judgment order of various abnormal behaviors is more refined, which is more in line with the comprehensive identification of various refueling cashing out methods in real life.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method, device, electronic device and storage medium for identifying abnormal refueling behavior. Background Art

[0002] With the promotion of fuel card services, the proportion of payments made via fuel cards has increased annually. Furthermore, petrochemical distributors often utilize fuel cards for promotional activities, offering consumers numerous benefits when using fuel cards, such as rewards, discounts, and points redeemable for gifts. However, while fuel cards have boosted the refined oil business, they have also led to a growing number of instances of gas station employees and drivers cashing out using fuel cards. Currently, there is no control plan for abnormal vehicle refueling behavior. Summary of the Invention

[0003] The purpose of the present invention is to propose a method, device, electronic device and storage medium for identifying abnormal refueling behavior in response to the above-mentioned deficiencies in the prior art. This purpose is achieved through the following technical solutions.

[0004] A first aspect of the present invention provides a method for identifying abnormal refueling behavior, the method comprising:

[0005] Extracting various feature data related to the fuel card and the vehicle bound to the card based on the received real-time fueling data;

[0006] Determining whether the real-time refueling data contains the first type of abnormal refueling behavior using characteristic data associated with the first type of independent abnormal refueling behavior;

[0007] Starting from the first level of the second-category abnormal refueling behavior in the hierarchical table, traversing the second-category abnormal refueling behavior, using feature data related to the currently traversed second-category abnormal refueling behavior, determining whether the real-time refueling data contains the second-category abnormal refueling behavior, wherein the second-category abnormal refueling behaviors of adjacent levels in the hierarchical table are dependent on each other;

[0008] If so, continue traversing the next layer of the second type of abnormal refueling behavior in the hierarchical table, and continue to perform the process of determining whether the real-time refueling data contains the second type of abnormal refueling behavior using the feature data related to the currently traversed second type of abnormal refueling behavior;

[0009] If it does not exist, the traversal ends;

[0010] An early warning is issued for the first type of abnormal refueling behavior and / or the second type of abnormal refueling behavior in the real-time refueling data.

[0011] In some embodiments of the present application, the extraction of various feature data related to the fuel card and the card-bound vehicle based on the received real-time fueling data includes:

[0012] Obtain card binding data, a gas station dimension table, vehicle stop gas station data, a vehicle dimension table, and an enterprise dimension table; and extract various feature data related to the gas card and the card-bound vehicle based on the real-time refueling data and card binding data, the gas station dimension table, vehicle stop gas station data, the vehicle dimension table, and the enterprise dimension table.

[0013] In some embodiments of the present application, the determining whether the real-time refueling data contains the first type of abnormal refueling behavior using feature data associated with the independent first type of abnormal refueling behavior includes:

[0014] Acquire characteristic data related to the first type of abnormal refueling behavior from the multiple characteristic data; input the acquired characteristic data into a prediction model corresponding to the first type of abnormal refueling behavior, and use the input characteristic data to predict whether the real-time refueling data contains the first type of abnormal refueling behavior.

[0015] In some embodiments of the present application, determining whether the real-time refueling data contains the second type of abnormal refueling behavior by utilizing feature data related to the currently traversed second type of abnormal refueling behavior includes:

[0016] Acquire characteristic data related to the second type of abnormal refueling behavior from the multiple characteristic data; input the acquired characteristic data into a prediction model corresponding to the second type of abnormal refueling behavior, and use the input characteristic data by the prediction model to predict whether the real-time refueling data contains the second type of abnormal refueling behavior.

[0017] In some embodiments of the present application, the method further includes a prediction model training process:

[0018] A refueling and vehicle stop wide table is constructed based on historical refueling data, card binding data, vehicle stop gas station data, vehicle dimension table, enterprise dimension table, and gas station dimension table; multiple feature data of the historical refueling data are extracted from the refueling and vehicle stop wide table and added to the training data set, and at least one abnormal behavior label or normal behavior label is labeled for the multiple feature data of the historical refueling data; for the prediction model corresponding to each type of abnormal refueling behavior, the prediction model is trained using the feature data related to the prediction model in the training data set until the prediction model converges.

[0019] In some embodiments of the present application, after issuing an early warning for the first type of abnormal refueling behavior and / or the second type of abnormal refueling behavior in the real-time refueling data, the method further includes:

[0020] When a feedback notification is received, the various feature data of the real-time refueling data are added to the training data set, and the various feature data of the real-time refueling data are marked with abnormal labels or normal behavior labels; for the prediction model corresponding to each type of abnormal refueling behavior, the feature data related to the prediction model in the training data set is used to optimize the training of the prediction model.

[0021] In some embodiments of the present application, the first type of abnormal refueling behavior includes refueling for cash and high-frequency refueling; the second type of abnormal refueling behavior includes refueling with non-card-bound vehicles, refueling with non-corporate vehicles, refueling with non-current vehicle fuel products, refueling products that are not corporate vehicles, refueling amount exceeding the fuel tank capacity of the card-bound vehicle, and abnormal refueling with high fuel consumption.

[0022] A second aspect of the present invention provides a device for identifying abnormal refueling behavior, the device comprising:

[0023] A feature extraction module is used to extract various feature data related to the fuel card and the vehicle bound to the card based on the received real-time fueling data;

[0024] a first identification module, configured to determine whether the real-time refueling data contains the first type of abnormal refueling behavior, using characteristic data associated with the first type of independent abnormal refueling behavior;

[0025] A second identification module is configured to start traversing from the first layer of the second type of abnormal refueling behavior in the hierarchical table, and use feature data related to the currently traversed second type of abnormal refueling behavior to determine whether the real-time refueling data contains the second type of abnormal refueling behavior, where the second type of abnormal refueling behaviors of adjacent layers in the hierarchical table are dependent on each other; if so, continue traversing the next layer of the second type of abnormal refueling behavior in the hierarchical table, and continue to perform the process of determining whether the real-time refueling data contains the second type of abnormal refueling behavior using feature data related to the currently traversed second type of abnormal refueling behavior; if not, terminate the traversal;

[0026] The alarm module is used to issue an early warning for the first type of abnormal refueling behavior and / or the second type of abnormal refueling behavior in the real-time refueling data.

[0027] The third aspect of the present invention provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method described in the first aspect when executing the program.

[0028] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method described in the first aspect when executed by a processor.

[0029] Based on the above-mentioned method and device for identifying abnormal refueling behavior in the first and second aspects, the present invention has at least the following beneficial effects or advantages:

[0030] Each time a piece of real-time refueling data is generated, various feature data related to the fuel card used and the vehicle bound to the card are extracted. These feature data are then used to determine various abnormal behaviors related to refueling cashing out. Therefore, this solution has high real-time anomaly determination. This solution distinguishes various abnormal behaviors of refueling cashing out based on dependence and independence. For independent abnormal behaviors, only the relevant feature data is used to make individual abnormality determinations. For dependent abnormal behaviors, abnormality determinations are made in a hierarchical manner. Only when the abnormal behavior determination result of the previous level is true is the abnormal behavior of the next level determined. This makes the determination order of various abnormal behaviors more refined and more consistent with the comprehensive identification of various refueling cashing out methods in real life. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0032] Figure 1 This is a flow chart of an embodiment of a method for identifying abnormal refueling behavior according to an exemplary embodiment of the present invention;

[0033] Figure 2 1 is a schematic structural diagram of a device for identifying abnormal refueling behavior according to an exemplary embodiment of the present invention;

[0034] Figure 3 FIG1 is a schematic diagram of the hardware structure of an electronic device according to an exemplary embodiment of the present invention;

[0035] Figure 4 The figure is a schematic structural diagram of a storage medium according to an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0036] Exemplary embodiments will be described in detail herein, examples of which are illustrated in the accompanying drawings. In the following description, when referring to the drawings, like numbers in different figures represent like or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present invention, as detailed in the appended claims.

[0037] The terms used in this invention are for the purpose of describing specific embodiments only and are not intended to limit the invention. The singular forms "a," "the," and "the" used in this invention and the appended claims are also intended to include plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0038] It should be understood that although the terms "first," "second," "third," etc. may be used in the present invention to describe various information, such information should not be limited to these terms. These terms are merely used to distinguish information of the same type from one another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the present invention. Depending on the context, the term "if" as used herein may be interpreted as "when," "when," or "in response to determining."

[0039] The existing method for identifying refueling cash-out behavior is to poll at fixed time points, that is, to collect statistics on refueling-related data, including multiple factors, and then perform permutations and combinations based on the factors to obtain suspicious refueling records after calculation.

[0040] However, this fixed-cycle polling calculation has poor real-time warning performance, and the final judgment is based on a value calculated after factor permutations and combinations and the threshold set by the rule. The judgment accuracy will be affected by the threshold set by experience.

[0041] Based on this, the present application proposes a method for identifying abnormal refueling behavior, namely, extracting multiple feature data related to the refueling card and the card-bound vehicle based on the received real-time refueling data, and then using the feature data related to the independent first-type abnormal refueling behavior to determine whether the real-time refueling data has the first-type abnormal refueling behavior, and at the same time starting from the first layer of the second-type abnormal refueling behavior in the hierarchical table, using the feature data related to the currently traversed second-type abnormal refueling behavior, to determine whether the real-time refueling data has the second-type abnormal refueling behavior. The second-type abnormal refueling behaviors of adjacent levels in the hierarchical table are dependent. If so, continue to traverse the next layer of the second-type abnormal refueling behavior in the hierarchical table, and continue to execute the process of using the feature data related to the currently traversed second-type abnormal refueling behavior to determine whether the real-time refueling data has the second-type abnormal refueling behavior. If not, end the traversal, and finally issue an early warning for the first-type abnormal refueling behavior and / or the second-type abnormal refueling behavior in the real-time refueling data.

[0042] The technical effects that can be achieved based on the above description are:

[0043] Each time a piece of real-time refueling data is generated, various feature data related to the fuel card used and the vehicle bound to the card are extracted. These feature data are then used to determine various abnormal behaviors related to refueling cashing out. Therefore, this solution has high real-time anomaly determination. This solution distinguishes various abnormal behaviors of refueling cashing out based on dependence and independence. For independent abnormal behaviors, only the relevant feature data is used to make individual abnormality determinations. For dependent abnormal behaviors, abnormality determinations are made in a hierarchical manner. Only when the abnormal behavior determination result of the previous level is true is the abnormal behavior of the next level determined. This makes the determination order of various abnormal behaviors more refined and more consistent with the comprehensive identification of various refueling cashing out methods in real life.

[0044] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiment of the present application will be clearly and completely described below in conjunction with the drawings in the embodiment of the present application.

[0045] Example 1:

[0046] Figure 1 FIG. 1 is a flow chart of an embodiment of a method for identifying abnormal refueling behavior according to an exemplary embodiment of the present invention. Figure 1 As shown, the abnormal refueling behavior identification method includes the following steps:

[0047] Step 101: extract various feature data related to the fuel card and the vehicle bound to the card based on the received real-time fueling data.

[0048] In this embodiment, real-time refueling data refers to a piece of refueling data generated in real time by a gas station, and this piece of refueling data includes the refueling card number, gas station ID, and refueling event information (such as refueling amount, oil volume, price, oil type, refueling province, city, refueling time, etc.).

[0049] Only a few features can be extracted based on real-time refueling data alone. Multiple types of features can be extracted by combining data from other dimensions. In one possible implementation, card binding data, a gas station dimension table, data on vehicles stopping at gas stations, a vehicle dimension table, and an enterprise dimension table can be obtained. Then, based on the real-time refueling data and card binding data, the gas station dimension table, data on vehicles stopping at gas stations, a vehicle dimension table, and an enterprise dimension table, multiple feature data related to the fuel card and the card-bound vehicle can be extracted.

[0050] It should be noted that the card binding data includes the gas card number, user information, enterprise ID, vehicle ID, etc.; the gas station dimension table records the gas station ID and the corresponding geographic location information (such as longitude and latitude); the vehicle stop gas station data includes vehicle stop event information and vehicle ID; the vehicle dimension table records the network access time, fuel type, mailbox liters, etc.; the enterprise dimension table records the enterprise vehicle ID of each enterprise.

[0051] Exemplarily, the multiple feature data include: the amount of refueling, fuel volume, price, province of refueling, city of refueling and the time interval between the last refueling; the fuel type of the card-bound vehicle; the mileage of the card-bound vehicle; whether the card-bound vehicle is a corporate vehicle; the number of liters of the fuel tank of the card-bound vehicle; the maximum number of liters of the fuel tank of the corporate vehicle; whether the fuel product provided by the gas station is the fuel product in the refueling data; whether the fuel product provided by the gas station is the fuel product used by the card-bound vehicle; whether the fuel product provided by the gas station is the fuel product used by the card-bound vehicle at the gas station within the refueling time range; the amount of refueling, fuel volume, price, province of refueling, city of refueling and the time interval between the last refueling, and the mileage of the card-bound vehicle for each of the previous 9 times; the number of gas stations that are the same as the last 10 times, and the number of fuel products that are the same as the last 10 times; the absolute value of the difference between the average number of liters of refueling in the last time and the last 10 times.

[0052] Those skilled in the art will appreciate that the various characteristic data descriptions given above are merely exemplary descriptions and do not constitute specific limitations on the present application.

[0053] Step 102: Using feature data related to the independent first type of abnormal refueling behavior, determine whether the real-time refueling data contains the first type of abnormal refueling behavior.

[0054] Among them, in real life, people will come up with various ways to cash out by refueling, and each way of cashing out by refueling can be reflected based on data from different aspects. Therefore, this solution proposes the detection of multiple abnormal refueling behaviors, and distinguishes various abnormal refueling behaviors according to dependence and independence. Abnormal refueling behaviors with independence are defined as the first type of abnormal refueling behaviors, and abnormal refueling behaviors with dependence are defined as the second type of abnormal refueling behaviors. In addition, the judgment of each abnormal refueling behavior is based on the characteristic data related to it for separate abnormal judgment to improve the judgment accuracy.

[0055] In one possible implementation, feature data related to the first type of abnormal refueling behavior is obtained from a variety of feature data, and the obtained feature data is input into a prediction model corresponding to the first type of abnormal refueling behavior. The prediction model uses the input feature data to predict whether the real-time refueling data contains the first type of abnormal refueling behavior.

[0056] The feature data related to the first type of abnormal refueling behavior may be obtained by determining a correlation coefficient between the feature data and the first type of abnormal refueling behavior, and selecting feature data having a correlation coefficient higher than a certain threshold.

[0057] It should be noted that before inputting the feature data into the prediction model, the feature data may be converted into a one-hot vector, and the one-hot vector may be further normalized to obtain data suitable for inputting into the prediction model.

[0058] In specific implementation, the independent first-category abnormal refueling behavior includes two types: refueling cash-out behavior (refueling cards are frequently used to refuel at a small number of gas stations, which may be a gas station employee using his own gas card to cash out, or a middleman refueling cash-out behavior) and high-frequency refueling behavior (the time interval between recent refueling is very small), and each first-category abnormal refueling behavior corresponds to a prediction model.

[0059] Step 103: Start traversing from the second type of abnormal refueling behavior in the first layer of the hierarchical table.

[0060] Among them, the second type of abnormal refueling behavior of adjacent levels in the hierarchical table is dependent, that is, only when the abnormal refueling behavior of the previous level is established will the abnormal refueling behavior of the next level appear.

[0061] Exemplarily, the second type of abnormal refueling behavior with dependence includes refueling of non-card-bound vehicles, refueling of non-enterprise vehicles, fuel products that are not current vehicles, fuel products that are not enterprise vehicles, refueling amount exceeding the fuel tank capacity of the card-bound vehicle (single refueling amount exceeds the fuel tank capacity or multiple refueling but the total refueling amount minus the theoretical mileage exceeds the fuel tank capacity), abnormal refueling with high fuel consumption (the fuel consumption per kilometer of the vehicle from the last refueling to this refueling is much greater than the average fuel consumption), and each second type of abnormal refueling behavior corresponds to a prediction model.

[0062] The second category of abnormal refueling behavior is categorized as follows: refueling of a non-card-linked vehicle -> refueling of a non-company vehicle, refueling of a fuel type other than the current vehicle's -> refueling of a fuel type other than the company's -> refueling of a fuel amount exceeding the card-linked vehicle's fuel tank capacity | abnormal high fuel consumption. The first tier includes refueling of non-card-linked vehicles, and the same applies to refueling of vehicles with abnormal fuel consumption.

[0063] Step 104: Determine whether the real-time refueling data contains the second type of abnormal refueling behavior using the feature data related to the currently traversed second type of abnormal refueling behavior.

[0064] Based on the prediction implementation described in the above step 103, in step 104, feature data related to the second type of abnormal refueling behavior can also be obtained from multiple feature data, and the obtained feature data can be input into the prediction model corresponding to the second type of abnormal refueling behavior. The prediction model uses the input feature data to predict whether the real-time refueling data contains the second type of abnormal refueling behavior.

[0065] The characteristic data related to the second type of abnormal refueling behavior may be obtained by determining the correlation coefficient between the characteristic data and the first type of abnormal refueling behavior, and selecting the characteristic data with a correlation coefficient higher than a certain threshold.

[0066] It should be noted that before inputting the feature data into the prediction model, the feature data may be converted into a one-hot vector, and the one-hot vector may be further normalized to obtain data suitable for inputting into the prediction model.

[0067] Step 105: If it exists, when the currently traversed second-type abnormal refueling behavior is not the last layer of the hierarchical table, continue to traverse the next layer of the second-type abnormal refueling behavior in the hierarchical table and continue to execute the process of step 104.

[0068] Step 106: If the second type of abnormal refueling behavior does not exist or the currently traversed layer is the last layer of the layer table, then the traversal of the layer table ends.

[0069] Among them, when the second type of abnormal refueling behavior currently traversed does not hold true, the prediction of the subsequent abnormal refueling behavior in the hierarchical table is directly terminated.

[0070] It is worth noting that the prediction process of the first type of abnormal refueling behavior in step 102 and the second type of abnormal refueling behavior in steps 103 to 106 can be performed in parallel and independently.

[0071] It should be noted that before executing steps 102 to 106 above, it is necessary to pre-train the prediction model corresponding to each abnormal refueling behavior. The training process includes: constructing a refueling and vehicle stop wide table based on historical refueling data, card binding data, vehicle stop gas station data, vehicle dimension table, enterprise dimension table, and gas station dimension table, and then extracting various feature data of historical refueling data from the refueling and vehicle stop wide table, and marking the various feature data of historical refueling data with at least one abnormal behavior label or normal behavior label, and then for the prediction model corresponding to each type of abnormal refueling behavior, use the feature data related to the prediction model in the training data set to train the prediction model until the prediction model converges.

[0072] Among them, a binary classification model, such as a logistic regression probability estimation model, can be used for the selection of the prediction model. It can be understood by those skilled in the art that the model convergence condition can be that the accuracy, recall rate and loss value of the model all meet the corresponding threshold conditions.

[0073] In specific implementation, the refueling and vehicle stop wide table is a database table containing multiple fields. For the construction process of the refueling and vehicle stop wide table, the fuel card number is used to associate the historical refueling data and the card binding data to generate Table A, and Table A is grouped and counted according to the enterprise ID to obtain the vehicle ID set for each enterprise ID, thereby generating Table B; Table A and Table B are associated by the enterprise ID, and then the vehicle stop gas station data is associated within the time range of the vehicle stop at the gas station through the enterprise ID, geographic location information and refueling time, and then the refueling event data, card binding information, enterprise vehicle set and enterprise vehicle set at the same gas station at the refueling time are aggregated according to the fuel card number and refueling time. Then, the vehicle dimension table, enterprise dimension table and gas station dimension table are associated, and finally the refueling event, stop event, vehicle information, refueling station information, enterprise information, enterprise vehicle information set and enterprise vehicle set at the gas station at the same time are obtained as the refueling and vehicle stop wide table.

[0074] Step 107: issuing an early warning for the first type of abnormal refueling behavior and / or the second type of abnormal refueling behavior in the real-time refueling data.

[0075] It should be noted that after executing step 107, the relevant salesperson will make manual judgments and processes after receiving the abnormal refueling behavior warning, and will feedback the erroneous results of the judgment. Therefore, when receiving the feedback notification, the training data set is expanded by adding multiple feature data of the real-time refueling data to the training data set, and marking the multiple feature data of the real-time refueling data with abnormal labels or normal behavior labels. Then, for the prediction model corresponding to each type of abnormal refueling behavior, the prediction model is optimized and trained using the feature data related to the prediction model in the expanded training data set.

[0076] Among them, through the feedback of manual judgment of erroneous results, the amount of abnormal data in the training data set can be expanded, so that the prediction model can be continuously optimized and upgraded, making the prediction accuracy of the prediction model higher.

[0077] At this point, the above is completed Figure 1The identification process shown extracts various feature data related to the fuel card used and the vehicle bound to the card for each piece of real-time refueling data. These feature data are then used to determine various abnormal behaviors related to refueling cashing out. Therefore, this solution offers high real-time anomaly determination. This solution distinguishes various abnormal behaviors related to refueling cashing out based on dependency and independence. For independent abnormal behaviors, only the relevant feature data is used to make individual anomaly determinations. For dependent abnormal behaviors, anomaly determinations are made in a hierarchical manner. Only when the abnormal behavior determination result for the previous level is true is the abnormal behavior at the next level determined. This refines the order in which various abnormal behaviors are determined, making it more consistent with the comprehensive identification of various refueling cashing out methods in real life.

[0078] Corresponding to the aforementioned embodiment of the method for identifying abnormal refueling behavior, the present invention also provides an embodiment of a device for identifying abnormal refueling behavior.

[0079] Figure 2 This is a schematic diagram of a structure of an abnormal refueling behavior recognition device according to an exemplary embodiment of the present invention. The device is used to execute the abnormal refueling behavior recognition method provided in any of the above embodiments, such as Figure 2 As shown, the abnormal refueling behavior identification device includes:

[0080] A feature extraction module 210 is configured to extract various feature data related to the fuel card and the vehicle bound to the fuel card based on the received real-time fueling data;

[0081] The first identification module 220 is configured to determine whether the real-time refueling data contains the first type of abnormal refueling behavior using feature data associated with the first type of independent abnormal refueling behavior;

[0082] The second identification module 230 is configured to start traversing from the first layer of the second-category abnormal refueling behavior in the hierarchical table, and use feature data related to the currently traversed second-category abnormal refueling behavior to determine whether the real-time refueling data contains the second-category abnormal refueling behavior, where the second-category abnormal refueling behaviors of adjacent layers in the hierarchical table are dependent on each other; if so, continue traversing the next layer of the second-category abnormal refueling behavior in the hierarchical table, and continue to use feature data related to the currently traversed second-category abnormal refueling behavior to determine whether the real-time refueling data contains the second-category abnormal refueling behavior; if not, terminate the traversal;

[0083] The alarm module 240 is used to issue an early warning for the first type of abnormal refueling behavior and / or the second type of abnormal refueling behavior in the real-time refueling data.

[0084] The implementation process of the functions and effects of each unit in the above-mentioned device is specifically described in the implementation process of the corresponding steps in the above-mentioned method, and will not be repeated here.

[0085] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to the partial description of the method embodiments. The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present invention. A person of ordinary skill in the art can understand and implement the present invention without inventive work.

[0086] An embodiment of the present invention further provides an electronic device corresponding to the abnormal refueling behavior identification method provided in the above embodiment, so as to execute the above abnormal refueling behavior identification method.

[0087] Figure 3 This is a hardware structure diagram of an electronic device according to an exemplary embodiment of the present invention. The electronic device includes: a communication interface 601, a processor 602, a memory 603, and a bus 604. Communication interface 601, processor 602, and memory 603 communicate with each other via bus 604. Processor 602 executes the abnormal refueling behavior identification method described above by reading and executing machine-executable instructions corresponding to the control logic of the abnormal refueling behavior identification method in memory 603. The details of this method are described in the above embodiments and will not be repeated here.

[0088] The memory 603 mentioned in the present invention can be any electronic, magnetic, optical or other physical storage device, and can contain stored information, such as executable instructions, data, etc. Specifically, the memory 603 can be RAM (Random Access Memory), flash memory, a storage drive (such as a hard disk drive), any type of storage disk (such as an optical disk, DVD, etc.), or similar storage media, or a combination thereof. The communication connection between the system network element and at least one other network element is achieved through at least one communication interface 601 (which can be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. can be used.

[0089] The bus 604 may be an ISA bus, a PCI bus, or an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The memory 603 is used to store programs, and the processor 602 executes the programs after receiving an execution instruction.

[0090] The processor 602 may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by an integrated logic circuit of hardware in the processor 602 or by instructions in the form of software. The above-mentioned processor 602 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The various methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in conjunction with the embodiments of the present application can be directly embodied as being executed by a hardware decoding processor, or can be executed by a combination of hardware and software modules in the decoding processor.

[0091] The electronic device provided in the embodiment of the present application and the method for identifying abnormal refueling behavior provided in the embodiment of the present application are based on the same inventive concept and have the same beneficial effects as the methods adopted, operated or implemented by them.

[0092] The present application also provides a computer-readable storage medium corresponding to the abnormal refueling behavior identification method provided in the above embodiment. Figure 4 As shown, the computer-readable storage medium is a CD 30 on which a computer program (ie, a program product) is stored. When the computer program is run by a processor, the abnormal refueling behavior identification method provided by any of the aforementioned embodiments will be executed.

[0093] It should be noted that examples of the computer-readable storage medium may also include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical or magnetic storage media, which are not listed here one by one.

[0094] The computer-readable storage medium provided in the above-mentioned embodiment of the present application and the abnormal refueling behavior identification method provided in the embodiment of the present application are based on the same inventive concept and have the same beneficial effects as the method adopted, run or implemented by the application program stored therein.

[0095] Other embodiments of the present invention will readily occur to those skilled in the art after considering the specification and practicing the invention disclosed herein. The present invention is intended to cover any variations, uses, or adaptations of the present invention that follow from the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, with the true scope and spirit of the invention being indicated by the following claims.

[0096] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0097] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for identifying abnormal refueling behavior, characterized in that: The method comprises: Extracting various feature data related to the fuel card and the vehicle bound to the card based on the received real-time fueling data; Determining whether the real-time refueling data contains the first type of abnormal refueling behavior using characteristic data associated with the first type of independent abnormal refueling behavior; Starting from the first level of the second-category abnormal refueling behavior in the hierarchical table, traversing the second-category abnormal refueling behavior, using feature data related to the currently traversed second-category abnormal refueling behavior, determining whether the real-time refueling data contains the second-category abnormal refueling behavior, wherein the second-category abnormal refueling behaviors of adjacent levels in the hierarchical table are dependent on each other; If so, if the currently traversed second-category abnormal refueling behavior is not the last layer of the hierarchical table, the next layer of the second-category abnormal refueling behavior in the hierarchical table is traversed, and the process of determining whether the real-time refueling data contains the second-category abnormal refueling behavior is continued using the feature data related to the currently traversed second-category abnormal refueling behavior; If it does not exist, the traversal ends; An early warning is issued for the first type of abnormal refueling behavior and / or the second type of abnormal refueling behavior in the real-time refueling data.

2. The method according to claim 1, characterized in that The method extracts various feature data related to the fuel card and the vehicle bound to the fuel card based on the received real-time fueling data, including: Obtain card binding data, gas station dimension table, vehicle stop gas station data, vehicle dimension table, and enterprise dimension table; Based on the real-time refueling data and card binding data, the gas station dimension table, the vehicle stop gas station data, the vehicle dimension table, and the enterprise dimension table, various feature data related to the refueling card and the card-bound vehicle are extracted.

3. The method according to claim 1, characterized in that The determining whether the real-time refueling data contains the first type of abnormal refueling behavior using the feature data related to the independent first type of abnormal refueling behavior includes: Acquire characteristic data related to the first type of abnormal refueling behavior from the multiple characteristic data; The acquired feature data is input into a prediction model corresponding to the first type of abnormal refueling behavior, and the prediction model uses the input feature data to predict whether the real-time refueling data contains the first type of abnormal refueling behavior.

4. The method according to claim 3, characterized in that The determining whether the real-time refueling data contains the second type of abnormal refueling behavior by utilizing the characteristic data related to the currently traversed second type of abnormal refueling behavior includes: Acquiring characteristic data related to the second type of abnormal refueling behavior from the multiple characteristic data; The acquired feature data is input into a prediction model corresponding to the second type of abnormal refueling behavior, and the prediction model uses the input feature data to predict whether the real-time refueling data contains the second type of abnormal refueling behavior.

5. The method according to claim 4, characterized in that The method also includes a prediction model training process: Build a refueling and vehicle stop wide table based on historical refueling data, card binding data, vehicle stop gas station data, vehicle dimension table, enterprise dimension table, and gas station dimension table; Extracting multiple feature data of the historical refueling data from the refueling and vehicle stop wide table and adding them to the training data set, and labeling the multiple feature data of the historical refueling data with at least one abnormal behavior label or a normal behavior label; For the prediction model corresponding to each type of abnormal refueling behavior, the prediction model is trained using feature data related to the prediction model in the training data set until the prediction model converges.

6. The method according to claim 4, characterized in that After issuing an early warning for the first type of abnormal refueling behavior and / or the second type of abnormal refueling behavior in the real-time refueling data, the method further includes: Upon receiving the feedback notification, the multiple feature data of the real-time refueling data are added to the training data set, and abnormal labels or normal behavior labels are marked for the multiple feature data of the real-time refueling data; For the prediction model corresponding to each type of abnormal refueling behavior, the prediction model is optimized and trained using the feature data related to the prediction model in the training data set.

7. The method according to any one of claims 1 to 6, characterized in that The first type of abnormal refueling behavior includes refueling for cash and high-frequency refueling; The second type of abnormal refueling behavior includes refueling of non-card-bound vehicles, refueling of non-corporate vehicles, refueling of fuel products that are not used by the current vehicle, refueling of fuel products that are not used by the corporate vehicle, refueling amount exceeding the fuel tank capacity of the card-bound vehicle, and abnormal refueling with high fuel consumption.

8. A device for identifying abnormal refueling behavior, characterized in that: The device comprises: A feature extraction module is used to extract various feature data related to the fuel card and the vehicle bound to the card based on the received real-time fueling data; a first identification module, configured to determine whether the real-time refueling data contains the first type of abnormal refueling behavior, using characteristic data associated with the first type of independent abnormal refueling behavior; A second identification module is configured to start traversing from the first layer of the second type of abnormal refueling behavior in the hierarchical table, and use feature data related to the currently traversed second type of abnormal refueling behavior to determine whether the real-time refueling data contains the second type of abnormal refueling behavior, where the second type of abnormal refueling behaviors of adjacent layers in the hierarchical table are dependent on each other; if so, continue traversing the next layer of the second type of abnormal refueling behavior in the hierarchical table, and continue to perform the process of determining whether the real-time refueling data contains the second type of abnormal refueling behavior using feature data related to the currently traversed second type of abnormal refueling behavior; if not, terminate the traversal; The alarm module is used to issue an early warning for the first type of abnormal refueling behavior and / or the second type of abnormal refueling behavior in the real-time refueling data.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images