Vehicle rental state recognition method, device and equipment and medium

Abstract

The invention relates to the technical field of vehicle identification, and discloses a vehicle lease state identification method, device and equipment and a medium, and the method comprises the steps: obtaining the positioning information of a target vehicle in a preset time period, determining a plurality of parking points of the target vehicle according to the positioning information through a clustering algorithm, and carrying out the timestamp marking of the plurality of parking points, obtaining time distribution characteristics; determining the staying dispersion among the plurality of staying points, and comparing the staying dispersion with a preset dispersion threshold; if the staying dispersion is greater than the preset dispersion threshold, determining a repeated staying point value among the plurality of staying points according to the time distribution characteristics; and if the repeated stop point data is greater than a preset threshold value, analyzing the use frequency information of the target vehicle and the user behavior data through a deep learning model, and outputting a lease state recognition result of the target vehicle. The invention aims to improve the recognition accuracy and recognition efficiency of the vehicle rental state.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle identification, and particularly to a method, device, equipment and medium for identifying the rental status of a vehicle. Background Art

[0002] In the financial field, the automobile financial leasing business, as an important asset financing method, has developed rapidly in recent years. However, with the expansion of the business scale, the risk control challenges faced by financial institutions have become increasingly severe, especially the identification problem of whether the financed vehicle is used on the short-term car rental platform.

[0003] Currently, the industry mainly relies on manual investigations after customer overdue or accident to identify risks. This method has significant limitations. First, the manual investigation is inefficient and difficult to meet the real-time monitoring requirements of large-scale vehicle data, thus increasing the operating cost and risk control difficulty of identifying the rental status of vehicles. Second, the existing risk control means rely on a single channel (such as the information provided by customers or third-party reports), lacking the comprehensive analysis ability of multi-dimensional data and being difficult to comprehensively and accurately evaluate the actual use situation of vehicles. That is to say, the related technologies cannot accurately and quickly identify the rental status of vehicles. Summary of the Invention

[0004] The present invention provides a method, device, computer equipment and medium for identifying the rental status of a vehicle to solve the technical problem that the rental status of a vehicle cannot be accurately and quickly identified in the related technologies.

[0005] In a first aspect, a method for identifying the rental status of a vehicle based on is provided, including:

[0006] Obtain the positioning information of a target vehicle within a preset time period, and determine a number of stop points of the target vehicle according to the positioning information through a clustering algorithm, and perform timestamp marking on the number of stop points to obtain a time distribution feature;

[0007] Determine the stop dispersion between the number of stop points, and compare the stop dispersion with a preset dispersion threshold;

[0008] If the stop dispersion is greater than the preset dispersion threshold, determine the repeated stop point value between the number of stop points according to the time distribution feature;

[0009] If the repeated stop point data is greater than a preset threshold, analyze the usage frequency information and user behavior data of the target vehicle through a deep learning model, and output the identification result of the rental status of the target vehicle.

[0010] In a second aspect, a device for identifying the rental status of a vehicle is provided, including:

[0011] An acquisition module, configured to acquire the positioning information of a target vehicle within a preset time period, determine a plurality of stop points of the target vehicle according to the positioning information through a clustering algorithm, and perform timestamp marking on the plurality of stop points to obtain a time distribution feature;

[0012] A determination module, configured to determine the stop dispersion between the plurality of stop points, and compare the stop dispersion with a preset dispersion threshold;

[0013] The determination module is further configured to, if the stop dispersion is greater than the preset dispersion threshold, determine the repeated stop point value between the plurality of stop points according to the time distribution feature;

[0014] An analysis module, configured to, if the repeated stop point data is greater than a preset threshold, analyze the usage frequency information and user behavior data of the target vehicle through a deep learning model, and output a recognition result of the rental status of the target vehicle.

[0015] In a third aspect, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the above vehicle rental status recognition method are implemented.

[0016] In a fourth aspect, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above vehicle rental status recognition method are implemented.

[0017] In the solution implemented by the above vehicle rental status recognition method, device, computer device, and storage medium, the positioning information of a target vehicle within a preset time period can be acquired, a plurality of stop points of the target vehicle can be determined according to the positioning information through a clustering algorithm, and timestamp marking can be performed on the plurality of stop points to obtain a time distribution feature. Further, the stop dispersion between the plurality of stop points can be determined, and the stop dispersion can be compared with a preset dispersion threshold. If the stop dispersion is greater than the preset dispersion threshold, the repeated stop point value between the plurality of stop points can be determined according to the time distribution feature. If the repeated stop point data is greater than a preset threshold, the usage frequency information and user behavior data of the target vehicle can be analyzed through a deep learning model, and a recognition result of the rental status of the target vehicle can be output. The present invention aims to acquire the positioning information of a target vehicle within a preset time period, analyze the stop points and mark timestamps in combination with a clustering algorithm, further calculate the stop dispersion and repeated stop points, and finally analyze user behavior data in combination with a deep learning model, so as to achieve accurate rental status recognition. This method improves the automation and intelligence level of rental management, can efficiently and accurately identify the usage frequency and status of vehicles, and provides strong support for vehicle scheduling and operation decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0019] Figure 1 is a schematic diagram of an application environment of a vehicle rental status recognition method in an embodiment of the present invention;

[0020] Figure 2 is a schematic flowchart of a vehicle rental status recognition method in an embodiment of the present invention;

[0021] Figure 3 is Figure 1 a schematic flowchart of a specific implementation manner of step S10 in;

[0022] Figure 4 is a schematic structural diagram of a vehicle rental status recognition device in an embodiment of the present invention;

[0023] Figure 5 is a schematic structural diagram of a computer device in an embodiment of the present invention;

[0024] Figure 6 is another schematic structural diagram of a computer device in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0025] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0026] The vehicle rental status recognition method provided by the embodiments of the present invention can be applied, for example, in Figure 1In the application environment, the client communicates with the server through the network. The server can obtain the positioning information of the target vehicle within a preset time period through the client, and determine several stop points of the target vehicle according to the positioning information through a clustering algorithm, and mark time stamps for several of the stop points to obtain a time distribution feature; determine the stop dispersion between several of the stop points, and compare the stop dispersion with a preset dispersion threshold; if the stop dispersion is greater than the preset dispersion threshold, determine the repeated stop point value between several of the stop points according to the time distribution feature; if the repeated stop point data is greater than a preset threshold, analyze the usage frequency information and user behavior data of the target vehicle through a deep learning model, and output the rental status recognition result of the target vehicle. In the present invention, by obtaining the positioning information of the target vehicle within a preset time period, analyzing the stop points and marking time stamps in combination with a clustering algorithm, further calculating the stop dispersion and repeated stop points, and finally analyzing user behavior data in combination with a deep learning model, accurate rental status recognition is achieved. This method improves the automation and intelligence level of rental management, can efficiently and accurately identify the usage frequency and status of vehicles, and provides strong support for vehicle scheduling and operation decision-making. Among them, the client can be, but is not limited to, various personal computers, laptop computers, smart phones, tablet computers, and portable wearable devices. The server can be implemented by an independent server or a server cluster composed of multiple servers. The present invention will be described in detail below through specific embodiments.

[0027] Please refer to Figure 2 as shown in Figure 2 a flowchart of a method for recognizing the rental status of a vehicle provided by an embodiment of the present invention, including the following steps:

[0028] S10: Obtain the positioning information of the target vehicle within a preset time period, and determine several stop points of the target vehicle according to the positioning information through a clustering algorithm, and mark time stamps for several of the stop points to obtain a time distribution feature.

[0029] The method for recognizing the rental status of a vehicle provided by the present invention can be applied to various application scenarios, such as risk assessment, customer behavior analysis, fraud detection, etc.

[0030] Exemplarily, the positioning information of the target vehicle within a specific time period can be obtained. This positioning information can be obtained through GPS or other geographic information systems, which is not limited in this application. Next, a clustering algorithm can be used to analyze the positioning points. Among them, the clustering algorithm can usually group the data so that the data points within the same group have a high similarity to each other, while having a low similarity to the data points in other groups. In the above embodiment, the clustering algorithm can identify the locations where the vehicle stays according to the movement trajectory of the vehicle within the time period. Each stay point is a clustering center, representing the location where the vehicle stays for a relatively long time.

[0031] To further analyze each stay point, a timestamp will be added to each stay point after clustering to record the time period when the stay point appears. The timestamp marking helps to clarify the duration of each stay point and the staying pattern of the vehicle at this location. Through the above steps, the time distribution characteristics of the stay points can be obtained. For example, it can be known the stay frequency at a certain location, the length of the stay time, etc. These information can reveal the rules of vehicle behavior.

[0032] In some embodiments, in step S10, the positioning information includes a number of positioning points. The steps of determining a number of stay points of the target vehicle according to the positioning information by the clustering algorithm are as follows:

[0033] S11: Analyze a number of the positioning points according to a preset distance threshold by the clustering algorithm to obtain an analysis result.

[0034] S12: If the analysis result shows that the distance between the positioning points is greater than the preset distance threshold, then determine the positioning points as the stay points.

[0035] S13: If the analysis result shows that the distance between the positioning points is less than or equal to the preset distance threshold, then determine the center point between the positioning points as the stay point.

[0036] For steps S11 - S13, the stay points can be identified and determined through the clustering algorithm and the distance threshold. Among them, the clustering algorithm is an algorithm that divides data into different groups by analyzing the similarity or distance between data points. Here, the positioning points (such as GPS coordinates) can be clustered according to their relative distances. The preset distance threshold is used to represent the predefined distance standard for judging which positioning points can be grouped into one category. If the distance between the positioning points is less than or equal to this threshold, they will be considered to belong to the same group. Otherwise, they will be considered not to be in the same group. Further, the clustering algorithm will output an analysis result showing which positioning points have a distance greater than or less than the preset distance threshold.

[0037] Exemplarily, if the distance between certain positioning points is greater than a preset distance threshold, then they do not belong to the same clustering group, which means that these positioning points are not concentrated at the same location. In this case, the positioning points themselves are regarded as stop points, indicating that the activity or stay at this location has been completed, which may be a certain static behavior (such as parking, staying in a certain area).

[0038] Exemplarily, if the distance between positioning points is less than or equal to the preset distance threshold, then these positioning points are regarded as belonging to the same clustering group, that is to say, these positioning points are very likely to be active or staying in the same area. The center point of these positioned points after clustering will be determined as the stop point, representing the core position of this area.

[0039] S20: Determine the stop dispersion between several of the stop points, and compare the stop dispersion with a preset dispersion threshold.

[0040] It should be noted that the stop dispersion is an index used to describe the dispersion degree or spread degree of the distribution of multiple stop points. Briefly speaking, the stop dispersion measures whether these stop points are concentrated or dispersed in space. If the stop points are very close, the dispersion is low; if the stop points are widely distributed, the dispersion is higher.

[0041] Exemplarily, the stop dispersion can be obtained by calculating the distance difference between stop points. For example, the average distance and standard deviation of stop points can be calculated, or other statistical methods (such as variance) can be used to quantify the dispersion degree of these stop points.

[0042] In addition, the preset dispersion threshold is a pre-set standard used to judge whether the current distribution of stop points meets the expected concentration or dispersion degree. By comparing with the preset dispersion threshold, it can be judged whether the distribution of stop points is normal. For example, if the dispersion is too high, it may indicate that some stop points do not meet the expectations, or the activity range is too dispersed, and further analysis may be required.

[0043] Exemplarily, by comparing the actually calculated stop dispersion with the preset dispersion threshold, the following judgments can be made: If the dispersion is less than the preset dispersion threshold, it means that the stop points are relatively concentrated, indicating that the stop behavior in this area is relatively concentrated or the activity is relatively stable. If the dispersion is greater than the preset dispersion threshold, it may mean that the stop points are relatively dispersed, and it may be necessary to further analyze whether there are abnormalities in these dispersed stop points, or whether there are multiple stop areas.

[0044] In the above embodiments, by calculating the dispersion between a number of stop points (i.e., the difference in their spatial distribution), and comparing this dispersion with a preset dispersion threshold, it is thus determined whether the distribution of these stop points meets the expectations. If the dispersion is too large, it may indicate that the activity range is too scattered and further analysis or measures need to be taken.

[0045] In S20, that is, in determining the stop dispersion between a number of the stop points, it specifically includes the following steps: determining the average center point of the stop points; calculating the Euclidean distance between each stop point and the average center point; and calculating the stop dispersion between the stop points according to the Euclidean distance by the following formula: where D is the stop dispersion between the stop points, n is the number of the stop points; and d is the Euclidean distance.

[0046] In S20, that is, in comparing the stop dispersion with a preset dispersion threshold, it specifically includes the following steps: normalizing the stop dispersion to obtain the normalized stop dispersion; and comparing the normalized stop dispersion with the preset dispersion threshold.

[0047] It should be noted that normalization is the process of converting data into a standard range, and a common range is [0, 1]. The purpose of normalization is to eliminate the influence brought by different data magnitudes, which is particularly important when comparing data with different measurement units or different data scales.

[0048] Exemplarily, in the above steps, the stop dispersion may have relatively large values or relatively small values, and after normalization, all the stop dispersion values will be adjusted to the same standard range (such as [0, 1]). For example, if the dispersion range of a certain group of stop points is between 0 and 100, after normalization, the dispersion is converted into a value between [0, 1], which can facilitate direct comparison with the preset threshold.

[0049] Exemplarily, the stop dispersion value after normalization is within a standard range and can be directly compared with the preset dispersion threshold. The preset dispersion threshold is usually also normalized to the same range (for example, [0, 1]), which enables convenient comparison. If the normalized stop dispersion is greater than the preset dispersion threshold: it indicates that the distribution of the stop points is relatively scattered, and it may be necessary to pay attention to whether the activities in this area are normal. If the normalized stop dispersion is less than or equal to the preset dispersion threshold: it indicates that the stop points are relatively concentrated and meet the expectations.

[0050] In the above embodiments, in order to standardize the calculation result of the residence dispersion so as to be able to effectively compare it with a preset dispersion threshold. Through this processing, it is possible to more accurately determine whether the distribution of the residence points conforms to the expected concentration or dispersion standard. The comparison between the normalized residence dispersion and the preset dispersion threshold can automatically identify the spatial distribution characteristics of the residence points, and then make corresponding judgments.

[0051] S30: If the residence dispersion is greater than the preset dispersion threshold, determine the repeated residence point value between several of the residence points according to the time distribution characteristics.

[0052] If, in S30, the normalized residence dispersion is greater than the preset dispersion threshold, it indicates that the residence points are relatively dispersed in space, which may be caused by multiple residence activities or areas.

[0053] It should be noted that repeated residence points refer to the situation where, despite being spatially dispersed, there are still multiple residence behaviors or residence points that are repeated in time. Although these residence points may be somewhat dispersed in space, they may be continuous, recurring, or occur within similar time periods in terms of time.

[0054] If the residence points have obvious periodicity or regularity in time (for example, within a certain period of time, there are multiple stays in a certain area), these residence points may represent different stages or different time points of the same residence activity.

[0055] Based on the time distribution characteristics, some algorithms can be used to determine whether multiple residence points belong to the same residence behavior. For example, if the time interval between a certain residence point and another residence point in time is very short (possibly meeting a set time window), they may be regarded as repeated points of the same stay. Or, if multiple residence points appear within the same time period but have a large difference in spatial distribution, they may be marked as repeated residence points, that is, the manifestations of the same activity at different locations.

[0056] In the above embodiments, further analysis is performed on the residence points with relatively dispersed spatial distribution (dispersion greater than the threshold), and the time distribution characteristics of the residence points are used to determine whether these residence points are repeated stays of the same residence activity or behavior. In this way, the system can identify residence points that are repeated in time even though they are spatially dispersed, so as to more accurately understand the essence of the residence activity. This helps to distinguish between truly different activity areas and multiple residence points of the same activity, and further optimize the analysis results.

[0057] In some embodiments of the present invention, in S30, that is, determining the repeated stay point values between several of the stay points according to the time distribution characteristics, specifically includes the following steps: segmenting the stay points according to the time distribution characteristics to divide several time windows; determining the occurrence frequency information of the stay points in each of the time windows to obtain the repeated occurrence times of the stay points; performing weighted calculation on the repeated occurrence times corresponding to each of the stay points to obtain the repeated stay point values.

[0058] Exemplarily, the time distribution characteristics indicate that the occurrence of stay points is not only affected by the spatial position but also closely related to time. To analyze the regularity of stay points in time, it is first necessary to divide the entire analysis period into several time windows according to the time dimension. These time windows are usually of a fixed length (such as every hour, every day, or according to the actual required time granularity). For example, assuming the time window length is 1 hour, then the stay points can be divided by hour in chronological order to analyze the distribution of stay points within each hour. The division of time windows can better analyze the time regularity of stay points, identify whether certain stay points appear repeatedly within a short period, and thus contribute to the subsequent analysis of the repeatability of stay points.

[0059] Exemplarily, if a certain stay point appears multiple times within this time window (i.e., stays multiple times within this time period), then the occurrence frequency of this stay point is relatively high. For the stay points within each time window, the number of times this point appears within the window will be recorded. The repeated occurrence times can reflect the intensity or frequency of the stay activity of this stay point within this window. For example, assuming the analysis is carried out according to a 1-hour time window, if a certain stay point appears 3 times within a certain hour, then the repeated occurrence times of this stay point within this hour is 3.

[0060] Exemplarily, the repeated occurrence times in different time windows may have different weights. For example, some time windows may be more critical than other time periods (such as peak hours or specific activity periods), so higher weights may be assigned to the repeated occurrence times within these windows. Among them, weighting can be achieved in some ways, such as setting weights using the relative importance of time periods, or adjusting the weight values according to the frequency of stay points and time intervals. After performing weighted calculation on the repeated occurrence times of each stay point in different time windows, a repeated stay point value is finally obtained. This value represents the repeatability or concentration of the stay point within the entire time period, reflecting whether a certain stay point appears frequently in multiple time windows.

[0061] S40: If the repeated stay point data is greater than a preset threshold, then analyze the usage frequency information and user behavior data of the target vehicle through a deep learning model, and output the rental status recognition result of the target vehicle.

[0062] Exemplarily, if the repeated stop point data is greater than a preset threshold, the usage frequency information of the target vehicle and the user behavior data are analyzed through a deep learning model. Through the analysis of the deep learning model, the rental status recognition result of the target vehicle can be finally output. This result may include multiple states, such as: being rented: the vehicle is currently being rented; idle: the vehicle has not been rented for a long time and may be in an idle state; high-frequency usage: the vehicle is frequently rented and has a high usage frequency, and may need maintenance or scheduling.

[0063] The step of S40 is based on the data obtained from the previous step (repeated stop point analysis), and further analyzes the usage frequency and user behavior of the target vehicle through a deep learning model to determine the rental status of the vehicle. This process automatically identifies whether the vehicle is being rented and provides relevant intelligent decision-making support. Through such analysis, the operator can understand the usage status of the vehicle in real time and optimize scheduling, maintenance, and resource allocation.

[0064] In some embodiments, the above method further includes: obtaining a training data set and a pre-trained model, where the training data set includes historical usage frequency information and historical user behavior data of a number of vehicles; performing label annotation on the training data set to obtain an annotation result, where the annotation result includes the rental status corresponding to the historical usage frequency information and the historical user behavior data; training the pre-trained model through the training data set and the annotation result to obtain the deep learning model.

[0065] Specifically, the annotation result corresponding to the training data set can be used as the label of this group of input data, and then each group of training sets with labels is input into the pre-trained model for supervised learning. When the training end condition is met, such as when the number of training times reaches the number threshold or the output accuracy of the model reaches the accuracy threshold, the training ends, and the trained deep learning model is obtained.

[0066] On the basis of the above embodiments, after obtaining the deep learning model, it further includes: performing iterative training on the natural language processing model based on the training data set and the annotation result to extract data features, and calculating to obtain a loss function; using a preset method to perform iterative training on the loss function with the aim of reducing the value of the loss function until the expected threshold is met; obtaining the iterated deep learning model based on the loss function after iterative training.

[0067] It can be understood that in order to train a deep learning model with higher accuracy, the deep learning model can be repeatedly iteratively trained to continuously reduce the loss function until the loss function meets the expected threshold requirements.

[0068] It should be noted that the present application does not limit the above preset method and the expected threshold. For example, the preset method can be a gradient descent algorithm, a batch gradient descent algorithm, a stochastic gradient descent algorithm, etc. The present application takes the gradient descent algorithm as an example for illustration.

[0069] The purpose of the gradient descent algorithm is to find the minimum value of the loss function or converge to the minimum value through an iterative manner. Geometrically speaking, in the gradient descent algorithm, at the place where the function changes and increases the fastest, along the direction opposite to the vector, the gradient decreases the fastest, so it is easier to find the minimum value of the function. Based on this, in the embodiments of the present application, the gradient descent algorithm can be used to repeatedly iterate and train the deep learning model to continuously reduce the loss function, thereby reducing the error of the calculation result.

[0070] It can be seen that in the above solution, by obtaining the positioning information of the target vehicle within a preset time period, analyzing the stop points and marking time stamps in combination with the clustering algorithm, further calculating the stop dispersion and repeated stop points, and finally analyzing the user behavior data in combination with the deep learning model, accurate rental status recognition can be achieved. This method improves the automation and intelligence level of rental management, can efficiently and accurately identify the usage frequency and status of vehicles, and provides strong support for vehicle scheduling and operation decision-making.

[0071] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean the order of execution is prior or subsequent. The execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0072] In one embodiment, a rental status recognition device for a vehicle is provided. The rental status recognition device for the vehicle corresponds one-to-one with the rental status recognition method for the vehicle in the above embodiment. As Figure 4 shown, the rental status recognition device for the vehicle includes an acquisition module 101, a determination module 102, and an analysis module 103. The detailed descriptions of each functional module are as follows:

[0073] The acquisition module 101 is configured to acquire the positioning information of the target vehicle within a preset time period, and determine a plurality of stop points of the target vehicle according to the positioning information through a clustering algorithm, and perform time stamp marking on the plurality of stop points to obtain a time distribution feature;

[0074] The determination module 102 is configured to determine the stop dispersion between the plurality of stop points, and compare the stop dispersion with a preset dispersion threshold;

[0075] The determination module 102 is further configured to, if the stop dispersion is greater than the preset dispersion threshold, determine the repeated stop point value between the plurality of stop points according to the time distribution feature;

[0076] An analysis module 103, configured to analyze the usage frequency information and user behavior data of the target vehicle through a deep learning model if the repeated stop point data is greater than a preset threshold, and output an identification result of the rental status of the target vehicle.

[0077] In one embodiment, the acquisition module 101 is specifically configured to:

[0078] Analyze a plurality of the positioning points according to a preset distance threshold through a clustering algorithm to obtain an analysis result;

[0079] If the analysis result is that the distance between the positioning points is greater than the preset distance threshold, determine the positioning points as the stop points;

[0080] If the analysis result is that the distance between the positioning points is less than or equal to the preset distance threshold, determine the center point between the positioning points as the stop point.

[0081] In one embodiment, the determination module 102 is specifically configured to:

[0082] Determine the average center point of the stop points;

[0083] Calculate the Euclidean distance between each of the stop points and the average center point;

[0084] According to the Euclidean distance, calculate the stop dispersion between the stop points through the following formula:

[0085]

[0086] where D is the stop dispersion between the stop points, n is the number of the stop points; d is the Euclidean distance.

[0087] In one embodiment, the determination module 102 is further configured to:

[0088] Normalize the stop dispersion to obtain a normalized stop dispersion;

[0089] Compare the normalized stop dispersion with the preset dispersion threshold.

[0090] In one embodiment, the determination module 102 is further configured to:

[0091] Segment the stop points according to the time distribution feature to divide a plurality of time windows;

[0092] Determine the occurrence frequency information of the stop points in each of the time windows to obtain the repeated occurrence times of the stop points;

[0093] Perform a weighted calculation on the recurrence times corresponding to each of the said stop points to obtain the value of the repeated stop points.

[0094] In one embodiment, the acquisition module 101 is further configured to:

[0095] Acquire a training data set and a pre-trained model, wherein the training data set includes historical usage frequency information and historical user behavior data of a number of vehicles;

[0096] Perform label annotation on the training data set to obtain an annotation result, wherein the annotation result includes the rental status corresponding to the historical usage frequency information and the historical user behavior data;

[0097] Train the pre-trained model with the training data set and the annotation result to obtain the deep learning model.

[0098] In one embodiment, the acquisition module 101 is further configured to:

[0099] Iteratively train the natural language processing model based on the training data set and the annotation result to extract data features, and calculate a loss function;

[0100] Iteratively train the loss function by using a preset method for the purpose of reducing the value of the loss function until the expected threshold is met;

[0101] Based on the loss function after iterative training, obtain the iterated deep learning model.

[0102] The present invention provides a rental status recognition device for a vehicle. By acquiring the positioning information of a target vehicle within a preset time period, analyzing stop points in combination with a clustering algorithm and marking time stamps, further calculating the stop dispersion and repeated stop points, and finally analyzing user behavior data in combination with a deep learning model, accurate rental status recognition is achieved. This method improves the automation and intelligence levels of rental management, can efficiently and accurately identify the usage frequency and status of vehicles, and provides strong support for vehicle scheduling and operation decision-making.

[0103] For the specific limitations on the rental status recognition device for a vehicle, reference can be made to the limitations on the rental status recognition method for a vehicle in the above text, which will not be elaborated here. Each module in the above rental status recognition device for a vehicle can be implemented in whole or in part by software, hardware, and their combination. The above modules can be embedded in or independent of the processor in a computer device in the form of hardware, or stored in the memory in a computer device in the form of software, so as to facilitate the processor to call and execute the operations corresponding to the above modules.

[0104] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in Figure 5 . The computer device includes a processor, a memory, a network interface, and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile and / or volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external client through a network connection. When the computer program is executed by the processor, it realizes the functions or steps on the server side of a vehicle-based rental status recognition method.

[0105] In one embodiment, a computer device is provided. The computer device may be a client, and its internal structure diagram may be as shown in Figure 6 . The computer device includes a processor, a memory, a network interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external server through a network connection. When the computer program is executed by the processor, it realizes the functions or steps on the client side of a vehicle-based rental status recognition method.

[0106] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the following steps are implemented:

[0107] Obtain the positioning information of the target vehicle within a preset time period, and determine several stop points of the target vehicle according to the positioning information through a clustering algorithm, and perform timestamp marking on several of the stop points to obtain a time distribution feature;

[0108] Determine the stop dispersion between several of the stop points, and compare the stop dispersion with a preset dispersion threshold;

[0109] If the stop dispersion is greater than the preset dispersion threshold, determine the repeated stop point value between several of the stop points according to the time distribution feature;

[0110] If the repeated stop point data is greater than a preset threshold, analyze the usage frequency information and user behavior data of the target vehicle through a deep learning model, and output the rental status recognition result of the target vehicle.

[0111] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0112] Obtain the positioning information of the target vehicle within a preset time period, and determine a number of stop points of the target vehicle according to the positioning information through a clustering algorithm, and perform timestamp marking on the number of stop points to obtain a time distribution feature;

[0113] Determine the stop dispersion between the number of stop points, and compare the stop dispersion with a preset dispersion threshold;

[0114] If the stop dispersion is greater than the preset dispersion threshold, determine the repeated stop point value between the number of stop points according to the time distribution feature;

[0115] If the repeated stop point data is greater than a preset threshold, analyze the usage frequency information and user behavior data of the target vehicle through a deep learning model, and output the rental status recognition result of the target vehicle.

[0116] It should be noted that for the functions or steps that can be realized by the above computer-readable storage medium or computer device, reference can be made to the relevant descriptions on the server side and the client side in the foregoing method embodiments. To avoid repetition, they will not be described in detail here.

[0117] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the various embodiments provided in the present application can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0118] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above division of each functional unit and module is used as an example. In actual applications, the above functions can be allocated to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0119] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for identifying the rental status of a vehicle, characterized in that: The method comprises: Acquire the positioning information of the target vehicle within a preset time period, determine a number of stop points of the target vehicle according to the positioning information through a clustering algorithm, and timestamp the stop points to obtain a time distribution feature; Determining a stay discreteness between a plurality of the stay points, and comparing the stay discreteness with a preset discreteness threshold; If the stay discreteness is greater than the preset discreteness threshold, determining the repeated stay point values ​​between the stay points according to the time distribution characteristics; If the repeated stop point data is greater than a preset threshold, the usage frequency information and user behavior data of the target vehicle are analyzed through a deep learning model to output a rental status identification result of the target vehicle.

2. The method according to claim 1, characterized in that The positioning information includes a plurality of positioning points, and the determining of a plurality of stop points of the target vehicle according to the positioning information by using a clustering algorithm includes: Analyzing the plurality of positioning points according to a preset distance threshold by using a clustering algorithm to obtain an analysis result; If the analysis result is that the distance between the positioning points is greater than the preset distance threshold, the positioning points are determined as the stay points; If the analysis result is that the distance between the positioning points is less than or equal to the preset distance threshold, the center point between the positioning points is determined as the stay point.

3. The method according to claim 1, characterized in that The determining of the stay discreteness between the plurality of stay points comprises: determining an average center point of the dwell points; Calculating the Euclidean distance between each of the stop points and the average center point; According to the Euclidean distance, the stay dispersion between the stay points is calculated by the following formula: Wherein, D is the stay discreteness between the stay points, n is the number of the stay points; and d is the Euclidean distance.

4. The method according to claim 1, characterized in that: The comparing the dwell discreteness with a preset discreteness threshold comprises: Normalizing the stay discreteness to obtain a normalized stay discreteness; The normalized dwell discreteness is compared with the preset discreteness threshold.

5. The method according to claim 1, characterized in that The determining of repeated stay point values ​​between the plurality of stay points according to the time distribution characteristics comprises: Segmenting the stay points according to the time distribution characteristics to divide them into several time windows; Determine the occurrence frequency information of the stay point in each of the time windows to obtain the number of repeated occurrences of the stay point; A weighted calculation is performed on the number of repeated occurrences corresponding to each of the stay points to obtain the repeated stay point value.

6. The method according to claim 1, characterized in that The method further comprises: Obtaining a training data set and a pre-trained model, wherein the training data set includes historical usage frequency information and historical user behavior data of a number of vehicles; Labeling the training data set to obtain a labeling result, wherein the labeling result includes the historical usage frequency information and the rental status corresponding to the historical user behavior data; The pre-trained model is trained using the training data set and the annotation results to obtain the deep learning model.

7. The method according to claim 6, characterized in that After obtaining the deep learning model, the method further includes: Iteratively train the natural language processing model based on the training data set and the annotation results to extract data features, and calculate a loss function; Iteratively training the loss function using a preset method for the purpose of reducing the value of the loss function until the expected threshold value is met; Based on the loss function after iterative training, an iterative deep learning model is obtained.

8. A vehicle rental status identification device, characterized in that: include: An acquisition module is used to acquire the positioning information of the target vehicle within a preset time period, and determine a number of stop points of the target vehicle according to the positioning information through a clustering algorithm, and timestamp the several stop points to obtain a time distribution feature; A determination module, used for determining a stay discreteness between the plurality of stay points, and comparing the stay discreteness with a preset discreteness threshold; The determination module is further configured to determine the repeated stay point values ​​between the stay points according to the time distribution characteristics if the stay discreteness is greater than the preset discreteness threshold; The analysis module is used to analyze the usage frequency information and user behavior data of the target vehicle through a deep learning model if the repeated stop point data is greater than a preset threshold, and output the rental status identification result of the target vehicle.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the vehicle rental status identification method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method for identifying the rental status of a vehicle as claimed in any one of claims 1 to 7 are implemented.

Images