Automobile data platform management method and system based on Internet of Things

Abstract

The invention discloses an Internet of Things-based automobile data platform management method and system, and relates to the technical field of automobile data processing, and the method comprises the steps: intercepting a driving path containing a classic driving situation from driving path data as a classic driving situation path, matching the driving data of a user with the classic driving situation path, and carrying out the matching of the driving data of the user and the classic driving situation path; and a basis is provided for subsequent model training and optimization. According to the method, the attribute information of the driving habit features of the user is analyzed, the fitting proportion is determined through the attribute information of the driving habit features of the user, the feature set is divided into the training set and the test set, the problem of under-fitting or over-fitting of the model is avoided, and the reliability of model training is improved. Automatic driving optimization, auxiliary making of a vehicle maintenance plan and pushing of a customized navigation route are realized through the driving habit model of the user, personalized vehicle service management is realized through the driving habit model, the management adaptability and the management effect are improved, and the user is helped to have better service experience and driving experience.

Description

Technical Field

[0001] The present invention relates to the technical field of automobile data processing, and in particular to an automobile data platform management method and system based on the Internet of Things. Background Art

[0002] Relying on the rapid development and popularization of Internet of Things technology. The Internet of Things realizes real-time monitoring, data collection and interaction of vehicles through various information sensing devices such as radio frequency identification, global positioning system, sensors, and integrates with the Internet to form a huge network. This technology not only improves the efficiency and accuracy of vehicle management, but also promotes the development of intelligent transportation systems. Internet of Things technology provides strong support for automobile data platform management solutions. By collecting vehicle operation data in real time and analyzing vehicle status, it realizes intelligent scheduling, fault diagnosis and predictive maintenance, thereby optimizing traffic management, improving road transportation efficiency, reducing energy consumption and emissions, and is of great significance to environmental protection.

[0003] In the prior art, assistance in formulating vehicle maintenance plans and pushing navigation routes is often achieved only based on user wishes or map information, without considering the large amount of car-related data on the car data platform, which reflects the user's driving habits. As a result, the current vehicle maintenance and route navigation management effects are poor, the management adaptability is low, and it is impossible to provide users with a personalized and good service experience.

[0004] Therefore, how to improve the management effect and adaptability of related automobile services is a technical problem that needs to be solved. Summary of the invention

[0005] The purpose of the present invention is to solve the problems of poor management effect and low management adaptability in the prior art, and the proposed automobile data platform management method based on the Internet of Things includes:

[0006] Acquire the user's driving data, driving events, driving records and driving path data, analyze the driving events and driving records to build a classic driving scenario, intercept the driving path containing the classic driving scenario from the driving path data as the classic driving scenario path, and match the user's driving data with the classic driving scenario path;

[0007] Preprocess the user's driving data under the classic driving scenario path, extract the user's driving habit characteristics, analyze the attribute information of the user's driving habit characteristics, quantify the attribute information of the user's driving habit characteristics, determine the fitting ratio through the attribute information of the user's driving habit characteristics, and divide the user's driving habit characteristics into a training set and a test set based on the fitting ratio;

[0008] The user's driving habit model is trained based on the training set and the test set. The user's driving habit model is used to optimize autonomous driving, assist in formulating vehicle maintenance plans, and push customized navigation routes. For pushing customized navigation routes, the user's navigation request is received. The navigation request includes the current location and destination, and multiple optional paths are planned based on the current location and destination.

[0009] Each optional path is evaluated based on the user's driving habit model and path information to select a path that suits the user and push that path to the user as the preferred path. Based on the user's vehicle information, other information and driving habit model, the system completes autonomous driving optimization and assists in formulating vehicle maintenance plans.

[0010] In some embodiments of the present application, driving events and driving records are analyzed to build classic driving scenarios, including:

[0011] Define driving events that reflect driving habits during driving, use the user's driving record as a timeline, and mark driving events that reflect driving habits on the timeline. Associate driving events based on traffic violation records, and divide driving events that reflect driving habits into normal driving events and abnormal driving events.

[0012] Divide the time axis into multiple complete driving process time axes according to the start and stop time nodes of the automobile, set thresholds for normal driving events and abnormal driving events respectively, count the number of repetitions of normal driving events and abnormal driving events in a single complete driving process time axis, and count the number of repetitions of normal driving events and abnormal driving events in multiple complete driving process time axes, and record them as the first number and the second number respectively;

[0013] According to the thresholds of normal driving events and abnormal driving events and their respective first and second times, driving events reflecting driving habits are regarded as classic driving situations, and the habitual degree of the classic driving situations is described.

[0014] In some embodiments of the present application, a driving path including a classic driving scenario is intercepted from the driving path data as a classic driving scenario path, and the user's driving data is matched with the classic driving scenario path, including:

[0015] Determine the starting point and the end point of the classic driving scenario on the driving path to obtain the original path, extend both ends of the original path according to the degree of habit corresponding to the classic driving scenario to obtain the classic driving scenario path, and retain part of the user's driving data corresponding to the classic driving scenario path.

[0016] In some embodiments of the present application, the attribute information of the user's driving habit characteristics is analyzed and the attribute information of the user's driving habit characteristics is quantified, including:

[0017] Determine multiple driving safety dimensions of the user, calculate the correlation between each driving habit feature and each driving safety dimension, integrate the correlation between the driving habit feature and all driving safety dimensions to obtain the overall correlation, and convert the overall correlation into a safety index of the driving habit feature, thereby determining the safety index of all driving habit features, and using the safety index as one of the attribute information of the driving habit feature;

[0018] Construct a box plot of each driving habit feature of the user, determine the abnormal data, interquartile range, upper bound and lower bound in the box plot, divide the abnormal data into upper abnormal data and lower abnormal data according to the positional relationship between the abnormal data and the upper bound and the lower bound, determine the relative discreteness of the driving habit feature through the abnormal data, interquartile range, upper bound and lower bound, and convert the relative discreteness into a stability index of the driving habit feature, thereby determining the stability index of all driving habit features, and use the stability index as one of the attribute information of the driving habit feature;

[0019]

[0020] Among them, W is the stability index of driving habit characteristics, τ is the stability conversion coefficient, IQR is the interquartile range, exp is the exponential function, n1 is the number of abnormal data, is the value of the i1th upper abnormal data, E1 is the upper bound, n2 is the number of lower abnormal data, is the value of the i2th lower abnormal data, E2 is the lower bound size, and k is a preset constant.

[0021] In some embodiments of the present application, the fitting ratio is determined by attribute information of the user's driving habit characteristics, including:

[0022] Divide the driving habit characteristics into multiple levels of feature sets according to the safety index and stability index of the user's driving habit characteristics, determine the feature quantity and feature type under each level of the feature set, and determine the mapping value by combining the average value of the safety index and the stability index of each level of the feature set, and obtain the fitting ratio through mapping the mapping values;

[0023] R j =α1m1 j +α2m2 j +α3(β1T1 j +β2T2 j );

[0024] Among them, R j is the mapping value of the feature set at the jth level, α1 is the conversion coefficient of the feature quantity, m1 jis the feature quantity of the feature set at the jth level, α2 is the conversion coefficient of the number of feature types, m2 j is the number of feature types in the feature set of the jth level, α3 is the conversion coefficient between the security index and the stability index, β1 is the weight of the security index, T1 j is the average security index of the feature set at the jth level, β2 is the weight of the stability index, T2 j is the average stability index of the feature set at the jth level.

[0025] In some embodiments of the present application, the user's driving habit characteristics are divided into a training set and a test set by means of a fitting ratio, including:

[0026] The feature sets at each level are stratified and sampled according to the fitting ratio of the feature sets at each level. The training set and the test set are randomly divided out from the feature sets at each level according to the fitting ratio. All the training sets and the test sets are summarized to obtain the training set and the test set.

[0027] In some embodiments of the present application, each optional path is evaluated according to the user's driving habit model and path information, so as to select a path suitable for the user, including:

[0028] The first evaluation index of each optional path is evaluated according to the path information, the user's driving conditions on each optional path are simulated through the user's driving habit model, and the second evaluation index of each optional path is obtained. The first evaluation index and the second evaluation index of each optional path are weightedly summed to obtain a comprehensive evaluation index, and the path suitable for the user is selected based on the comprehensive evaluation index.

[0029] In some embodiments of the present application, after the path is pushed to the user as the preferred path, the method further includes:

[0030] Each optional path is sorted according to the comprehensive evaluation index. The first-ranked optional path is pushed to the user first, and the remaining rankings are visually displayed to the user as options.

[0031] Correspondingly, the present application also provides an automobile data platform management system based on the Internet of Things, including:

[0032] A matching module, used to obtain the user's driving data, driving events, driving records and driving path data, analyze the driving events and driving records to build a classic driving scenario, intercept the driving path containing the classic driving scenario from the driving path data as the classic driving scenario path, and match the user's driving data with the classic driving scenario path;

[0033] A partitioning module is used to pre-process the driving data of the user under the classic driving scenario path, extract the user's driving habit characteristics, analyze the attribute information of the user's driving habit characteristics, quantify the attribute information of the user's driving habit characteristics, determine the fitting ratio through the attribute information of the user's driving habit characteristics, and divide the user's driving habit characteristics into a training set and a test set based on the fitting ratio;

[0034] The screening module is used to train the user's driving habit model based on the training set and the test set, and realize automatic driving optimization, assist in making vehicle maintenance plans and push customized navigation routes through the user's driving habit model. For pushing customized navigation routes, the module receives the user's navigation request, which includes the current location and destination, and plans multiple optional paths based on the current location and destination;

[0035] The management module is used to evaluate each optional path based on the user's driving habit model and path information, so as to select the path that is suitable for the user and push the path to the user as the preferred path. It also completes the automatic driving optimization and assists in formulating the vehicle maintenance plan based on the user's vehicle information, other information and driving habit model.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] 1. In the driving path data, the driving paths containing classic driving scenarios are intercepted as classic driving scenario paths, the user's driving data is matched with the classic driving scenario paths, and some data related to the classic driving scenarios that can reflect the user's driving habits are screened out to provide a basis for subsequent model training and optimization.

[0038] 2. Analyze the attribute information of the user's driving habit characteristics, determine the fitting ratio based on the attribute information of the user's driving habit characteristics, and divide the feature set into a training set and a test set to avoid under-fitting or over-fitting of the model and improve the reliability of model training.

[0039] 3. Through the user's driving habit model, we can achieve automatic driving optimization, assist in vehicle maintenance plan formulation and push customized navigation routes. Through the driving habit model, we can achieve personalized car service management, improve the adaptability and management effect of management, and help users have a better service experience and driving experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 This is a flow chart of the automobile data platform management method based on the Internet of Things proposed by the present invention;

[0041] Figure 2 This is a structural diagram of the automobile data platform management system based on the Internet of Things proposed by the present invention. DETAILED DESCRIPTION

[0042] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0043] Reference Figure 1 , an automobile data platform management method and system based on the Internet of Things, comprising the following steps,

[0044] Step S101, obtaining the user's driving data, driving events, driving records and driving path data, analyzing the driving events and driving records to build a classic driving scenario, intercepting the driving path containing the classic driving scenario on the driving path data as the classic driving scenario path, and matching the user's driving data with the classic driving scenario path.

[0045] In this embodiment, driving data includes vehicle data, driving status data, vehicle parts data, etc., and driving events include driving behaviors (starting, accelerating, turning, changing lanes, waiting for traffic lights, etc.) and driving behaviors in special situations (driving behaviors in bad weather conditions such as rainy days, foggy days, snowy days, and driving behaviors in congested situations). The user's driving events and driving records are analyzed to determine which events can better reflect the user's driving habits for the user, and these are used as classic driving scenarios. The driving data in the classic driving scenarios are retained.

[0046] In some embodiments of the present application, driving events and driving records are analyzed to build classic driving scenarios, including:

[0047] Define driving events that reflect driving habits during driving, use the user's driving record as a timeline, and mark driving events that reflect driving habits on the timeline. Associate driving events based on traffic violation records, and divide driving events that reflect driving habits into normal driving events and abnormal driving events.

[0048] Divide the time axis into multiple complete driving process time axes according to the start and stop time nodes of the automobile, set thresholds for normal driving events and abnormal driving events respectively, count the number of repetitions of normal driving events and abnormal driving events in a single complete driving process time axis, and count the number of repetitions of normal driving events and abnormal driving events in multiple complete driving process time axes, and record them as the first number and the second number respectively;

[0049] According to the thresholds of normal driving events and abnormal driving events and their respective first and second times, driving events reflecting driving habits are regarded as classic driving situations, and the habitual degree of the classic driving situations is described.

[0050] In this embodiment, the driving record records the information of each start and stop of the car. Each driving process (multiple complete driving process time axes) is determined according to the start and stop information. The driving events are associated according to the traffic violation records to determine which driving events (driving events) are prone to violations, such as excessive acceleration, speeding, etc. The first number and the second number are both the number of times a certain driving event occurs, but the two have different meanings. The first number is the frequent occurrence in one driving process, and the second number is the frequent occurrence in multiple driving processes. The two have different meanings. The two numbers are used to determine which driving events that reflect driving habits belong to the classic driving scenario, and the habit degree of the classic driving scenario is defined according to the situation where these two numbers exceed the threshold (describing the degree of habit of the user).

[0051] In some embodiments of the present application, a driving path including a classic driving scenario is intercepted from the driving path data as a classic driving scenario path, and the user's driving data is matched with the classic driving scenario path, including:

[0052] Determine the starting point and the end point of the classic driving scenario on the driving path to obtain the original path, extend both ends of the original path according to the degree of habit corresponding to the classic driving scenario to obtain the classic driving scenario path, and retain part of the user's driving data corresponding to the classic driving scenario path.

[0053] In this embodiment, different degrees of habit correspond to different path extension distances. The two ends of the original path are extended by the path extension distance. The greater the degree of habit, the longer the extension, which means that the event that better reflects the user's habit is more likely to have other situations that reflect the user's habit before and after.

[0054] Step S102, preprocessing the driving data of the user under the classic driving scenario path, extracting the user's driving habit characteristics, analyzing the attribute information of the user's driving habit characteristics, quantifying the attribute information of the user's driving habit characteristics, determining the fitting ratio through the attribute information of the user's driving habit characteristics, and dividing the user's driving habit characteristics into a training set and a test set based on the fitting ratio.

[0055] In this embodiment, the driving habit features include average vehicle speed, number of sudden acceleration / deceleration, lane change frequency, etc. The attribute information of the user's driving habit features includes safety and stability. Safety indicates that the feature corresponds to a safety risk, and stability indicates the fluctuation of the feature. The fitting ratio is determined in combination with safety and stability. The fitting ratio is the ratio that is neither overfitting nor underfitting, and the ratio is the ratio (data volume) of the training set to the test set.

[0056] In some embodiments of the present application, the attribute information of the user's driving habit characteristics is analyzed and the attribute information of the user's driving habit characteristics is quantified, including:

[0057] Determine multiple driving safety dimensions of the user, calculate the correlation between each driving habit feature and each driving safety dimension, integrate the correlation between the driving habit feature and all driving safety dimensions to obtain the overall correlation, and convert the overall correlation into a safety index of the driving habit feature, thereby determining the safety index of all driving habit features, and using the safety index as one of the attribute information of the driving habit feature;

[0058] Construct a box plot of each driving habit feature of the user, determine the abnormal data, interquartile range, upper bound and lower bound in the box plot, divide the abnormal data into upper abnormal data and lower abnormal data according to the positional relationship between the abnormal data and the upper bound and the lower bound, determine the relative discreteness of the driving habit feature through the abnormal data, interquartile range, upper bound and lower bound, and convert the relative discreteness into a stability index of the driving habit feature, thereby determining the stability index of all driving habit features, and use the stability index as one of the attribute information of the driving habit feature;

[0059]

[0060] Among them, W is the stability index of driving habit characteristics, τ is the stability conversion coefficient, IQR is the interquartile range, exp is the exponential function, n1 is the number of abnormal data, is the value of the i1th upper abnormal data, E1 is the upper bound, n2 is the number of lower abnormal data, is the value of the i2th lower abnormal data, E2 is the lower bound size, and k is a preset constant.

[0061] In this embodiment, multiple driving safety dimensions include compliance with traffic regulations, distance maintenance, speeding, etc.

[0062] Comply with traffic rules: observe whether the user strictly abides by traffic rules, such as stopping at red lights, going at green lights, and whether the user gives way according to regulations. This can be obtained by analyzing the user's driving trajectory, number of stops, and interaction with other vehicles.

[0063] Keeping a safe distance from the vehicle in front is the key to preventing rear-end collisions. We can analyze the average distance between the user and the vehicle in front during driving, as well as the change in distance at different speeds, to evaluate the user's habit of keeping a safe distance.

[0064] Do not exceed the speed limit: Speeding is one of the main causes of traffic accidents. We can count the average speed of users on different roads and at different time periods, and compare it with the road speed limit to evaluate the user's speeding behavior.

[0065] In this embodiment, the box plot is used to describe the fluctuation of the driving habit characteristics. In the past, only the interquartile range was used to describe the degree of dispersion, and the influence of outliers (dangerous values) was not considered. Here, outliers are considered to define the stability index. Outliers are values ​​outside the upper and lower bounds in the box plot. According to the positional relationship between the abnormal data and the upper and lower bounds, the abnormal data are divided into upper abnormal data and lower abnormal data. The abnormal data above the upper bound is the upper abnormal data, and the abnormal data below the lower bound is the lower abnormal data.

[0066] In this embodiment, The stability index is determined by the correction of the abnormal degree of abnormal data after the interquartile range transformation.

[0067] In some embodiments of the present application, the fitting ratio is determined by attribute information of the user's driving habit characteristics, including:

[0068] Divide the driving habit characteristics into multiple levels of feature sets according to the safety index and stability index of the user's driving habit characteristics, determine the feature quantity and feature type under each level of the feature set, and determine the mapping value by combining the average value of the safety index and the stability index of each level of the feature set, and obtain the fitting ratio through mapping the mapping values;

[0069] R j =α1m1 j +α2m2 j +α3(β1T1 j +β2T2 j );

[0070] Among them, R j is the mapping value of the feature set at the jth level, α1 is the conversion coefficient of the feature quantity, m1 j is the feature quantity of the feature set at the jth level, α2 is the conversion coefficient of the number of feature types, m2 j is the number of feature types in the feature set of the jth level, α3 is the conversion coefficient between the security index and the stability index, β1 is the weight of the security index, T1 j is the average security index of the feature set at the jth level, β2 is the weight of the stability index, T2 j is the average stability index of the feature set at the jth level.

[0071] In this embodiment, the driving habit characteristics are divided into multiple levels of feature sets according to the safety index and stability index of the user's driving habit characteristics, and the multiple levels include high safety and high stability, high safety and low stability, low safety and high stability, low safety and low stability, etc., and more levels can be divided according to actual needs. Considering the feature quantity (data volume), feature type (number of types), safety index and stability index of each level feature set to comprehensively determine a mapping value, and the fitting ratio (for example, 7:3, 8:2, etc.) is obtained by mapping the mapping value, and the feature sets in different situations have different division ratios.

[0072] In some embodiments of the present application, the user's driving habit characteristics are divided into a training set and a test set by means of a fitting ratio, including:

[0073] The feature sets at each level are stratified and sampled according to the fitting ratio of the feature sets at each level. The training set and the test set are randomly divided out from the feature sets at each level according to the fitting ratio. All the training sets and the test sets are summarized to obtain the training set and the test set.

[0074] Step S103, training the user's driving habit model based on the training set and the test set, and realizing automatic driving optimization, assisting in vehicle maintenance plan formulation and pushing customized navigation routes through the user's driving habit model. For pushing customized navigation routes, receiving the user's navigation request, the navigation request includes the current location and destination, and planning multiple optional paths based on the current location and destination.

[0075] In this embodiment, a user's driving habit model is trained. This driving habit model can help implement multiple different car service items and provide specific implementation steps for customized navigation route selection.

[0076] Step S104, evaluate each optional path according to the user's driving habit model and path information, so as to select a path suitable for the user, and push the path to the user as the preferred path, and complete the automatic driving optimization and assist in formulating the vehicle maintenance plan based on the user's vehicle information, other information and driving habit model.

[0077] In this embodiment, the user's driving habit model is used to simulate the user's driving conditions on each optional path, and the path information (distance, congestion level, number of traffic lights, road flatness, etc.) is combined to obtain a comprehensive evaluation index to select a path suitable for the user.

[0078] In some embodiments of the present application, each optional path is evaluated according to the user's driving habit model and path information, so as to select a path suitable for the user, including:

[0079] The first evaluation index of each optional path is evaluated according to the path information, the user's driving conditions on each optional path are simulated through the user's driving habit model, and the second evaluation index of each optional path is obtained. The first evaluation index and the second evaluation index of each optional path are weightedly summed to obtain a comprehensive evaluation index, and the path suitable for the user is selected based on the comprehensive evaluation index.

[0080] In this embodiment, all path information is comprehensively considered to determine the first evaluation index, and the user's driving conditions (driving safety, stability, violations, speed, time, etc.) on each optional path are simulated through the user's driving habit model to determine the second evaluation index, and the two indicators are combined to determine the comprehensive evaluation index.

[0081] In some embodiments of the present application, after the path is pushed to the user as the preferred path, the method further includes:

[0082] Each optional path is sorted according to the comprehensive evaluation index. The first-ranked optional path is pushed to the user first, and the remaining rankings are visually displayed to the user as options.

[0083] In this embodiment, automatic driving optimization and auxiliary vehicle maintenance plan are completed based on the user's vehicle information, other information and driving habit model. The specific implementation process is as follows:

[0084] Autonomous driving optimization

[0085] Path planning and navigation:

[0086] Based on the user's departure and destination, combined with traffic information and driving habit models, the optimal route is planned for the user.

[0087] During navigation, traffic information is updated in real time and the route is adjusted to avoid congested areas.

[0088] Driving behavior adjustment:

[0089] Based on the user's driving habit model, the driving behavior of the autonomous driving system is adjusted to match the user's driving style.

[0090] In specific situations (such as rainy and snowy weather, driving on mountain roads, etc.), adjust driving strategies to improve safety and comfort.

[0091] Energy saving and efficiency optimization:

[0092] Based on vehicle information and driving habit models, the energy consumption management of the autonomous driving system is optimized to reduce fuel or electricity consumption.

[0093] Improve driving efficiency and reduce unnecessary parking and waiting time.

[0094] Assist in formulating vehicle maintenance plans

[0095] Predictive Maintenance:

[0096] Leverage machine learning algorithms and vehicle status data to predict vehicle maintenance needs and failure risks.

[0097] Based on the prediction results, maintenance plans can be made in advance to avoid unexpected failures and unnecessary repair costs.

[0098] Maintenance recommendations:

[0099] Provide users with personalized maintenance recommendations based on their driving habits, vehicle information and road conditions.

[0100] Recommendations include regularly replacing parts, checking battery health, tire wear, etc.

[0101] Maintenance reminder:

[0102] When the vehicle reaches the scheduled maintenance mileage or time, the user is automatically reminded to perform maintenance.

[0103] Provide maintenance appointment service to facilitate users to arrange maintenance time.

[0104] Compared with the prior art, the present invention has the following beneficial effects:

[0105] 1. In the driving path data, the driving paths containing classic driving scenarios are intercepted as classic driving scenario paths, the user's driving data is matched with the classic driving scenario paths, and some data related to the classic driving scenarios that can reflect the user's driving habits are screened out to provide a basis for subsequent model training and optimization.

[0106] 2. Analyze the attribute information of the user's driving habit characteristics, determine the fitting ratio based on the attribute information of the user's driving habit characteristics, and divide the feature set into a training set and a test set to avoid under-fitting or over-fitting of the model and improve the reliability of model training.

[0107] 3. Through the user's driving habit model, we can achieve automatic driving optimization, assist in vehicle maintenance plan formulation and push customized navigation routes. Through the driving habit model, we can achieve personalized car service management, improve the adaptability and management effect of management, and help users have a better service experience and driving experience.

[0108] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present invention can be implemented by hardware, or by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present invention.

[0109] Correspondingly, this application also provides an automobile data platform management system based on the Internet of Things, such as Figure 2 As shown, including,

[0110] A matching module, used to obtain the user's driving data, driving events, driving records and driving path data, analyze the driving events and driving records to build a classic driving scenario, intercept the driving path containing the classic driving scenario from the driving path data as the classic driving scenario path, and match the user's driving data with the classic driving scenario path;

[0111] A partitioning module is used to pre-process the driving data of the user under the classic driving scenario path, extract the user's driving habit characteristics, analyze the attribute information of the user's driving habit characteristics, quantify the attribute information of the user's driving habit characteristics, determine the fitting ratio through the attribute information of the user's driving habit characteristics, and divide the user's driving habit characteristics into a training set and a test set based on the fitting ratio;

[0112] The screening module is used to train the user's driving habit model based on the training set and the test set, and realize automatic driving optimization, assist in making vehicle maintenance plans and push customized navigation routes through the user's driving habit model. For pushing customized navigation routes, the module receives the user's navigation request, which includes the current location and destination, and plans multiple optional paths based on the current location and destination;

[0113] The management module is used to evaluate each optional path based on the user's driving habit model and path information, so as to select the path that is suitable for the user and push the path to the user as the preferred path. It also completes the automatic driving optimization and assists in formulating the vehicle maintenance plan based on the user's vehicle information, other information and driving habit model.

[0114] Those skilled in the art will appreciate that the accompanying drawings are merely schematic diagrams of a preferred implementation scenario, and the modules or processes in the accompanying drawings are not necessarily required for implementing the present invention.

[0115] Those skilled in the art will appreciate that the modules in the system in the implementation scenario can be distributed in the system of the implementation scenario according to the implementation scenario description, or can be changed accordingly and located in one or more systems different from the implementation scenario. The modules in the above implementation scenario can be combined into one module, or can be further split into multiple submodules.

[0116] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. The automobile data platform management method based on the Internet of Things is characterized by: include, Acquire the user's driving data, driving events, driving records and driving path data, analyze the driving events and driving records to build a classic driving scenario, intercept the driving path containing the classic driving scenario from the driving path data as the classic driving scenario path, and match the user's driving data with the classic driving scenario path; Preprocess the user's driving data under the classic driving scenario path, extract the user's driving habit characteristics, analyze the attribute information of the user's driving habit characteristics, quantify the attribute information of the user's driving habit characteristics, determine the fitting ratio through the attribute information of the user's driving habit characteristics, and divide the user's driving habit characteristics into a training set and a test set based on the fitting ratio; The user's driving habit model is trained based on the training set and the test set. The user's driving habit model is used to optimize autonomous driving, assist in formulating vehicle maintenance plans, and push customized navigation routes. For pushing customized navigation routes, the user's navigation request is received. The navigation request includes the current location and destination, and multiple optional paths are planned based on the current location and destination. Each optional path is evaluated based on the user's driving habit model and path information to select a path that suits the user and push that path to the user as the preferred path. Based on the user's vehicle information, other information and driving habit model, the system completes autonomous driving optimization and assists in formulating vehicle maintenance plans.

2. The method for managing an automobile data platform based on the Internet of Things according to claim 1, characterized in that: Analyze driving events and driving records to build classic driving scenarios, including: Define driving events that reflect driving habits during driving, use the user's driving record as a timeline, and mark driving events that reflect driving habits on the timeline. Associate driving events based on traffic violation records, and divide driving events that reflect driving habits into normal driving events and abnormal driving events. Divide the time axis into multiple complete driving process time axes according to the start and stop time nodes of the automobile, set thresholds for normal driving events and abnormal driving events respectively, count the number of repetitions of normal driving events and abnormal driving events in a single complete driving process time axis, and count the number of repetitions of normal driving events and abnormal driving events in multiple complete driving process time axes, and record them as the first number and the second number respectively; According to the thresholds of normal driving events and abnormal driving events and their respective first and second times, driving events reflecting driving habits are regarded as classic driving situations, and the habitual degree of the classic driving situations is described.

3. The method for managing an automobile data platform based on the Internet of Things according to claim 2, characterized in that: The driving path containing the classic driving scenario is intercepted from the driving path data as the classic driving scenario path, and the user's driving data is matched with the classic driving scenario path, including: Determine the starting point and the end point of the classic driving scenario on the driving path to obtain the original path, extend both ends of the original path according to the degree of habit corresponding to the classic driving scenario to obtain the classic driving scenario path, and retain part of the user's driving data corresponding to the classic driving scenario path.

4. The method for managing an automobile data platform based on the Internet of Things according to claim 1, characterized in that: Analyze the attribute information of the user's driving habit characteristics, and quantify the attribute information of the user's driving habit characteristics, including: Determine multiple driving safety dimensions of the user, calculate the correlation between each driving habit feature and each driving safety dimension, integrate the correlation between the driving habit feature and all driving safety dimensions to obtain the overall correlation, and convert the overall correlation into a safety index of the driving habit feature, thereby determining the safety index of all driving habit features, and using the safety index as one of the attribute information of the driving habit feature; Construct a box plot of each driving habit feature of the user, determine the abnormal data, interquartile range, upper bound and lower bound in the box plot, divide the abnormal data into upper abnormal data and lower abnormal data according to the positional relationship between the abnormal data and the upper bound and the lower bound, determine the relative discreteness of the driving habit feature through the abnormal data, interquartile range, upper bound and lower bound, and convert the relative discreteness into a stability index of the driving habit feature, thereby determining the stability index of all driving habit features, and use the stability index as one of the attribute information of the driving habit feature; Among them, W is the stability index of driving habit characteristics, τ is the stability conversion coefficient, IQR is the interquartile range, exp is the exponential function, n1 is the number of abnormal data, is the value of the i1th upper abnormal data, E1 is the upper bound, n2 is the number of lower abnormal data, is the value of the i2th lower abnormal data, E2 is the lower bound size, and k is a preset constant.

5. The method for managing an automobile data platform based on the Internet of Things according to claim 4, characterized in that: The fitting ratio is determined by the attribute information of the user's driving habit characteristics, including: Divide the driving habit characteristics into multiple levels of feature sets according to the safety index and stability index of the user's driving habit characteristics, determine the feature quantity and feature type under each level of the feature set, and determine the mapping value by combining the average value of the safety index and the stability index of each level of the feature set, and obtain the fitting ratio through mapping the mapping values; R j =α1m1 j +α2m2 j +α3(β1T1 j +β2T2 j ); Among them, R j is the mapping value of the feature set at the jth level, α1 is the conversion coefficient of the feature quantity, m1 j is the feature quantity of the feature set at the jth level, α2 is the conversion coefficient of the number of feature types, m2 j is the number of feature types in the feature set of the jth level, α3 is the conversion coefficient between the security index and the stability index, β1 is the weight of the security index, T1 j is the average security index of the feature set at the jth level, β2 is the weight of the stability index, T2 j is the average stability index of the feature set at the jth level.

6. The method for managing an automobile data platform based on the Internet of Things according to claim 5, characterized in that: The user's driving habit characteristics are divided into training set and test set based on the fitting ratio, including: The feature sets at each level are stratified and sampled according to the fitting ratio of the feature sets at each level. The training set and the test set are randomly divided out from the feature sets at each level according to the fitting ratio. All the training sets and the test sets are summarized to obtain the training set and the test set.

7. The method for managing an automobile data platform based on the Internet of Things according to claim 1, characterized in that: Each optional path is evaluated based on the user's driving habit model and path information to select the path that is suitable for the user, including: The first evaluation index of each optional path is evaluated according to the path information, the user's driving conditions on each optional path are simulated through the user's driving habit model, and the second evaluation index of each optional path is obtained. The first evaluation index and the second evaluation index of each optional path are weightedly summed to obtain a comprehensive evaluation index, and the path suitable for the user is selected based on the comprehensive evaluation index.

8. The method for managing an automobile data platform based on the Internet of Things according to claim 1, characterized in that: After the path is pushed to the user as the preferred path, the method further includes: Each optional path is sorted according to the comprehensive evaluation index. The first-ranked optional path is pushed to the user first, and the remaining rankings are visually displayed to the user as options.

9. The automobile data platform management system based on the Internet of Things is characterized by: include, A matching module, used to obtain the user's driving data, driving events, driving records and driving path data, analyze the driving events and driving records to build a classic driving scenario, intercept the driving path containing the classic driving scenario from the driving path data as the classic driving scenario path, and match the user's driving data with the classic driving scenario path; A partitioning module is used to pre-process the driving data of the user under the classic driving scenario path, extract the user's driving habit characteristics, analyze the attribute information of the user's driving habit characteristics, quantify the attribute information of the user's driving habit characteristics, determine the fitting ratio through the attribute information of the user's driving habit characteristics, and divide the user's driving habit characteristics into a training set and a test set based on the fitting ratio; The screening module is used to train the user's driving habit model based on the training set and the test set, and realize automatic driving optimization, assist in making vehicle maintenance plans and push customized navigation routes through the user's driving habit model. For pushing customized navigation routes, the module receives the user's navigation request, which includes the current location and destination, and plans multiple optional paths based on the current location and destination; The management module is used to evaluate each optional path based on the user's driving habit model and path information, so as to select the path that is suitable for the user and push the path to the user as the preferred path. It also completes the automatic driving optimization and assists in formulating the vehicle maintenance plan based on the user's vehicle information, other information and driving habit model.

Images