Illegal driving behavior prediction method, device and equipment

By constructing a causal relationship matrix for historical illegal driving behaviors and using association rules mining algorithms, predicting and reminding illegal driving behaviors in advance, the problem of insufficient prediction accuracy in the existing technology is solved and driving safety is improved.

CN119066585APending Publication Date: 2024-12-03CHONGQING CHANGAN TECH CO LTD
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411185705.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-27
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

The prior art relies too much on the accuracy and real-timeness of physical devices when predicting and reminding of illegal driving, making it difficult for users to avoid violations in a timely manner and insufficient prediction accuracy.

Method used

By analyzing the target user's historical illegal driving behavior, building a causal relationship matrix, using association rule mining algorithms and entropy weighting methods to screen out high-value violations, and predict and remind likely violations in advance.

Benefits of technology

It improves the accuracy of predicting illegal driving behaviors, avoids the continuous occurrence of irregular behaviors, enhances driving safety, and reduces the dependence on the accuracy of physical devices.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119066585A_ABST
    Figure CN119066585A_ABST
Patent Text Reader

Abstract

The invention discloses an illegal driving behavior prediction method, device and equipment, and the method is applied to vehicle equipment, and comprises the steps: detecting a first illegal driving behavior of a target user in the vehicle equipment under the condition that the first illegal driving behavior of the target user in the vehicle equipment is detected; determining R target causal relationships taking the first illegal driving behavior as a reason item in at least one reference causal relationship; the at least one reference causal relationship is determined based on a causal relationship among P historical illegal driving behaviors of the target user; r and P are positive integers; predicting a target illegal driving behavior of the target user at the next moment based on R result items in the R target causal relationships; and outputting target prompt information, wherein the target prompt information is used for prompting the target user to avoid the target illegal driving behavior. According to the scheme, the accuracy of illegal driving behavior prediction can be improved, continuous occurrence of illegal driving behaviors is avoided, and the driving safety is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of vehicle technology, and in particular to, but not limited to, a method, device and equipment for predicting illegal driving behavior. Background Art

[0002] With the rapid development of vehicle technology, cars have become an indispensable means of travel for people in their daily travel. At present, illegal driving behaviors such as speeding, drunk driving, and fatigue driving are the main causes of frequent traffic accidents. Therefore, it is very important to predict and remind users to avoid illegal driving behaviors.

[0003] In the related art, the user's current driving data is usually collected in real time through a physical device, and the user's upcoming illegal driving behavior is predicted based on the user's current driving data, and the user is reminded. However, the above solution relies too much on the accuracy and real-time performance of the physical device. Even if the user is found to have an illegal driving behavior and is notified, it is difficult for the user to react in time to avoid the illegal driving behavior. Summary of the invention

[0004] One of the purposes of the present invention is to provide a method, device and equipment for predicting illegal driving behavior, which can improve the accuracy of predicting illegal driving behavior, avoid the continuous occurrence of illegal driving behavior, and improve driving safety.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is as follows:

[0006] In a first aspect, the present invention provides a method for predicting illegal driving behavior, which is applied to vehicle equipment and includes:

[0007] In the case where a first illegal driving behavior of a target user in a vehicle device is detected, R target causal relationships with the first illegal driving behavior as a cause item are determined in at least one reference causal relationship; at least one reference causal relationship is determined based on the causal relationship between P historical illegal driving behaviors of the target user; R and P are positive integers; based on R result items in the R target causal relationships, the target illegal driving behavior of the target user at the next moment is predicted; and target prompt information is output, where the target prompt information is used to prompt the target user to avoid the target illegal driving behavior.

[0008] According to the above technical means, by determining at least one reference causal relationship based on the causal relationship between multiple historical illegal driving behaviors in advance, it is possible to predict and remind possible illegal driving behaviors in advance based on at least one reference causal relationship. Since historical illegal driving behaviors can reflect the user's bad driving habits, by mining the influence and occurrence rules between multiple historical illegal driving behaviors, it is possible not only to predict possible illegal driving behaviors based on the user's bad driving habits, effectively improve the accuracy of predicting illegal driving behaviors, but also to avoid the continuous occurrence of illegal driving behaviors and improve driving safety.

[0009] In a possible implementation, before determining R target causal relationships with the first illegal driving behavior as the cause item in at least one reference causal relationship, the illegal driving behavior prediction method further includes: combining P historical illegal driving behaviors in pairs to obtain Q frequent item sets; Q is a positive integer; for each frequent item set in the Q frequent item sets, performing the following processing to obtain at least one reference causal relationship: based on the first frequent item and the second frequent item in the frequent item set, determining the support and confidence of the frequent item set; wherein the occurrence timestamp of the first frequent item is earlier than the occurrence timestamp of the second frequent item; the support is used to characterize the probability of the first frequent item and the second frequent item occurring at the same time; the confidence is used to characterize the proportion of the probability of the first frequent item and the second frequent item occurring at the same time to the probability of the first frequent item occurring; when the support is greater than or equal to the support threshold and the confidence is greater than or equal to the confidence threshold, establishing a causal relationship with the first frequent item as the cause item and the second frequent item as the result item, and obtaining a reference causal relationship.

[0010] According to the above technical means, by combining multiple historical illegal driving behaviors in pairs to obtain multiple frequent item sets, and determining the support and confidence of each frequent item set, it is possible to determine the association rules between multiple historical illegal driving behaviors based on the association rule mining algorithm, and obtain at least one reference causal relationship. Since the association rule mining algorithm has the advantages of simple implementation, convenient generalization and strong screening ability, by using the association rule mining algorithm to determine at least one reference causal relationship, it is possible to effectively screen out association rules with practical significance and high reliability based on multiple historical illegal driving behaviors, avoid the interference of a large amount of worthless information, and achieve accurate evaluation of the causal relationship between multiple historical illegal driving behaviors.

[0011] In a possible implementation, the illegal driving behavior prediction method further includes: constructing a behavior matrix based on the historical illegal driving data of the target user; the historical illegal driving data is used to characterize the behavior data corresponding to N illegal driving behaviors of the target user in M ​​driving trips; M and N are both positive integers; based on the behavior matrix, determining N weight values ​​corresponding to the N illegal driving behaviors; and determining the illegal driving behaviors whose weight values ​​among the N illegal driving behaviors are greater than a first threshold as P historical illegal driving behaviors.

[0012] According to the above technical means, since some illegal driving behaviors may only occur occasionally and cannot truly reflect the user's bad driving habits, by determining the N weight values ​​corresponding to N illegal driving behaviors, all illegal driving behaviors can be screened based on the weight values ​​to obtain P historical illegal driving behaviors with high value. In this way, the interference caused by accidental behaviors can be effectively eliminated, and illegal driving behaviors with high value that can reflect the user's normalized driving tendency can be determined more comprehensively and accurately, so as to more effectively evaluate the user's bad driving habits and provide more valuable reference for the prediction of illegal driving behaviors that are about to occur and assisted driving.

[0013] In a possible implementation, a behavior matrix is ​​constructed based on the historical illegal driving data of the target user, including: determining the number of behaviors and the behavior penalty value of the j-th illegal driving behavior of the target user in the i-th driving trip based on the historical illegal driving data; multiplying the number of behaviors and the behavior penalty value to obtain the i×j-th behavior matrix element of the behavior matrix; wherein the i×j-th behavior matrix element is any matrix element of the behavior matrix; i is a positive integer less than or equal to M; and j is a positive integer less than or equal to N.

[0014] According to the above technical means, considering the complexity and diversity of different illegal driving behaviors, by determining the number of illegal driving behaviors and the deduction points of each illegal driving behavior of the target user in each driving trip, it is possible to construct a behavior matrix based on the type of illegal driving behavior and the number of illegal driving behaviors and the deduction points of illegal driving behaviors. In this way, by evaluating the user's bad driving habits through a comprehensive consideration method, it is possible to more accurately distinguish the degree and nature of bad driving habits, avoid the bias and one-sidedness brought about by a single consideration method, and more comprehensively, accurately and scientifically characterize the user's bad driving habits.

[0015] In one possible implementation, based on the behavior matrix, N weight values ​​corresponding to N illegal driving behaviors are determined, including: based on the first reference expression of the entropy weight method, the behavior matrix is ​​standardized to obtain a reference matrix; based on the second reference expression of the entropy weight method, the reference matrix is ​​normalized to obtain a normalized matrix; based on the third reference expression of the entropy weight method and the normalized matrix, N information entropies corresponding to the N illegal driving behaviors are determined; based on the fourth reference expression of the entropy weight method and the N information entropies, N weight values ​​corresponding to the N illegal driving behaviors are determined.

[0016] According to the above technical means, by adopting the entropy weight method to determine the N weight values ​​corresponding to N illegal driving behaviors, the interference of human subjective factors can be effectively avoided, and each of the N illegal driving behaviors can be objectively analyzed. Moreover, higher weights can be given to illegal driving behaviors with larger amounts of information and higher degrees of variation, making the determination of the weight value corresponding to each of the N illegal driving behaviors more objective, fair and scientific.

[0017] In a possible implementation, based on R result items in R target causal relationships, predicting the target illegal driving behavior of the target user at the next moment includes: when R is equal to 1, determining the result item in the target causal relationship as the target illegal driving behavior; when R is greater than 1, determining the R result items in the R target causal relationships as R candidate illegal driving behaviors; and screening the R candidate illegal driving behaviors to obtain the target illegal driving behavior.

[0018] According to the above technical means, by determining whether to screen the R result items in the R target causal relationships to obtain the target illegal driving behavior according to different values ​​of the number of result items R, a more refined prediction and management of the target illegal driving behavior can be achieved. In this way, the predicted low-value illegal driving behaviors can be removed, and the focus can be placed on the truly critical illegal driving behaviors, making the prediction results more targeted and reliable, thereby realizing the reminder of high-value illegal driving behaviors.

[0019] In a possible implementation, R candidate illegal driving behaviors are screened to obtain a target illegal driving behavior, including: determining the probability of occurrence of each candidate illegal driving behavior among the R candidate illegal driving behaviors to obtain R occurrence probability values; and determining the candidate illegal driving behavior with the highest occurrence probability value among the R candidate illegal driving behaviors as the target illegal driving behavior.

[0020] According to the above technical means, by determining the probability of occurrence of each candidate illegal driving behavior among the R candidate illegal driving behaviors, it is possible to determine the candidate illegal driving behavior with the highest probability of occurrence among the R candidate illegal driving behaviors as the target illegal driving behavior. Since too much reminder information may make it difficult for users to focus on key illegal driving behaviors, by strictly distinguishing the probability of occurrence of different illegal driving behaviors, it is possible to accurately focus on the illegal driving behaviors that are most likely to occur, so that the predicted target illegal driving behaviors are more focused, thereby achieving targeted reminders for the target illegal driving behaviors.

[0021] In a possible implementation, R candidate illegal driving behaviors are screened to obtain a target illegal driving behavior, including: sorting the R candidate illegal driving behaviors based on a reference priority rule to obtain a sorting result; the reference priority rule is determined based on a front-and-back logic or a harm degree logic between the R candidate illegal driving behaviors; and the first X candidate illegal driving behaviors in the sorting result are determined as the target illegal driving behaviors.

[0022] According to the above technical means, considering the inherent correlation between illegal driving behaviors and the severity of the consequences, by determining the pre- and post-logic or harm degree logic between the R candidate illegal driving behaviors, it is possible to screen the R candidate illegal driving behaviors based on the pre- and post-logic or harm degree logic between the R candidate illegal driving behaviors to obtain the target illegal driving behavior. In this way, the screening mechanism of the candidate illegal driving behaviors can be made more scientific, reasonable, systematic, comprehensive, orderly and balanced, thereby improving the orderliness and emphasis of the predicted target driving behavior.

[0023] In a second aspect, the present invention provides a device for predicting illegal driving behavior, which is deployed in a vehicle device and includes:

[0024] A determination unit is used to determine, in the case where a first illegal driving behavior of a target user in the vehicle device is detected, R target causal relationships with the first illegal driving behavior as a cause item in at least one reference causal relationship; at least one reference causal relationship is determined based on the causal relationship between P historical illegal driving behaviors of the target user; R and P are positive integers;

[0025] A prediction unit, used for predicting a target illegal driving behavior of a target user at a next moment based on R result items in the R target causal relationships;

[0026] The output unit is used to output target prompt information, and the target prompt information is used to prompt the target user to avoid the target illegal driving behavior.

[0027] In a third aspect, the present invention further provides an electronic device, comprising: a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and the processor implements the above-mentioned illegal driving behavior prediction method when executing the program.

[0028] In a fourth aspect, the present invention further provides a storage medium having a computer program stored thereon. When the computer program on the storage medium is executed, the above-mentioned illegal driving behavior prediction method is implemented.

[0029] In a fifth aspect, the present invention further provides a computer program product, including a computer program or instructions, which, when executed by a processor, implements the above-mentioned illegal driving behavior prediction method.

[0030] It should be noted that the technical effects of the second to fifth aspects can refer to the detailed description of the first aspect above, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 A first optional flow chart of a method for predicting illegal driving behavior provided by an embodiment of the present invention;

[0032] Figure 2 A second optional flow chart of the method for predicting illegal driving behavior provided by an embodiment of the present invention;

[0033] Figure 3 A third optional flow chart of the method for predicting illegal driving behavior provided in an embodiment of the present invention;

[0034] Figure 4 A fourth optional flow chart of the method for predicting illegal driving behavior provided by an embodiment of the present invention;

[0035] Figure 5 A fifth optional flow chart of the method for predicting illegal driving behavior provided by an embodiment of the present invention;

[0036] Figure 6 A sixth optional flow chart of the method for predicting illegal driving behavior provided by an embodiment of the present invention;

[0037] Figure 7 A seventh optional flow chart of the method for predicting illegal driving behavior provided by an embodiment of the present invention;

[0038] Figure 8 An eighth optional flow chart of the method for predicting illegal driving behavior provided by an embodiment of the present invention;

[0039] Fig. 9 A ninth optional flow chart of the method for predicting illegal driving behavior provided by an embodiment of the present invention;

[0040] Fig.10 An optional structural diagram of a device for predicting illegal driving behavior provided by an embodiment of the present invention;

[0041] Fig.11 An optional structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0042] To make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the specific technical solution of the application will be further described in detail below in conjunction with the drawings in the embodiments of the present invention. The following embodiments are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0043] In the following description, reference is made to “some embodiments”, which describe a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0044] In the following description, the terms "first\second\third" are used only as examples to distinguish different objects, and do not represent a specific order for the objects, nor do they have a limitation on the order of precedence. It is understood that "first\second\third" can be interchanged with a specific order or order of precedence where permitted, so that the embodiments of the present invention described herein can be implemented in an order other than that illustrated or described herein.

[0045] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. The terms used herein are only for the purpose of describing the embodiments of the present invention and are not intended to limit the present invention.

[0046] The embodiments of the present invention provide a method, device, equipment, storage medium and computer program product for predicting illegal driving behavior. In practical applications, the method for predicting illegal driving behavior can be implemented by a device for predicting illegal driving behavior, and each functional entity in the device for predicting illegal driving behavior can be collaboratively implemented by hardware resources of electronic equipment, such as computing resources such as processors, and communication resources (such as for supporting various communication methods such as optical cables and cellular).

[0047] The following describes various embodiments of the method, device, equipment, storage medium, and computer program product for predicting illegal driving behavior provided by the embodiments of the present invention.

[0048] In the first aspect, an embodiment of the present invention provides a method for predicting illegal driving behavior. The function implemented by the method can be implemented by calling a program code by a processor in an electronic device (such as a vehicle device or a controller), and of course the program code can be stored in a computer storage medium. It can be seen that the electronic device at least includes a processor and a storage medium.

[0049] In a possible implementation, the electronic device may be a vehicle device or a controller in a vehicle device.

[0050] Next, the method for predicting illegal driving behavior provided by an embodiment of the present invention is described by taking a vehicle device as an execution subject as an example.

[0051] The illegal driving behavior prediction method is applied to vehicle equipment. The embodiment of the present invention does not limit the specific type of vehicle equipment, and it can be set according to actual conditions.

[0052] refer to Figure 1 As shown in the content, the process may include but is not limited to the following S101 to S103.

[0053] S101. When a first illegal driving behavior of a target user in the vehicle device is detected, the vehicle device determines R target causal relationships with the first illegal driving behavior as a cause item in at least one reference causal relationship.

[0054] The at least one reference causal relationship is determined based on the causal relationship between P historical illegal driving behaviors of the target user; and R and P are positive integers.

[0055] The target user may be any user who uses the vehicle. The embodiment of the present invention does not limit the specific type and number of the target users, which can be set according to actual conditions. For example, the target user may be a driver or a passenger.

[0056] Illegal driving behavior refers to driving operations or behaviors that violate traffic laws and relevant regulations formulated by the country or region.

[0057] The first driving violation behavior may be any driving violation behavior. The embodiment of the present invention does not limit the specific type and number of the first driving violation behavior, which can be set according to the actual situation. For example, the first driving violation behavior may include: speeding, failure to maintain a safe distance, driving under the influence of alcohol or drunk driving, fatigue driving, running a red light, not driving according to traffic markings, driving without a license, using a mobile phone while driving, not wearing a seat belt while driving, grabbing the steering wheel, deliberately blocking the driver's line of sight, fierce quarrels or physical conflicts with the driver, etc.

[0058] The reference causal relationship is used to characterize the causal relationship between different illegal driving behaviors. The embodiment of the present invention does not limit the specific expression form and quantity of the reference causal relationship, which can be set according to the actual situation. Exemplarily, the expression form of the reference causal relationship can be: cause item --> result item, indicating that the illegal driving behavior represented by the cause item leads to the illegal driving behavior represented by the result item, and the cause item and the result item are different illegal driving behaviors.

[0059] For example, reference causal relationship A: fatigue driving --> not driving according to traffic markings, means that the illegal driving behavior of fatigue driving leads to the illegal driving behavior of not driving according to traffic markings, where fatigue driving is the cause item of reference causal relationship A, and not driving according to traffic markings is the result item of reference causal relationship A.

[0060] It should be noted that the relationship between cause item and result item can be one-to-one, which means that one cause item leads to one result item; or one-to-many, which means that one cause item leads to multiple result items. Moreover, the cause item of a reference causal relationship may also be the result item of another reference causal relationship.

[0061] It can be understood that in a reference causal relationship, the cause item and the result item are in a one-to-one relationship; in multiple reference causal relationships, multiple reference causal relationships can be combined to be represented as a tree structure.

[0062] The target causal relationship refers to the causal relationship with the first illegal driving behavior as the cause item. The embodiment of the present invention does not limit the specific content and number of the target causal relationship, which can be set according to the actual situation. For example, in the case where the first illegal driving behavior is fatigue driving, the target causal relationship may include: fatigue driving--> not driving according to traffic markings, fatigue driving--> not maintaining a safe distance, fatigue driving--> running a red light, etc.

[0063] Historical driving violations refer to driving violations committed in the past.

[0064] It is understandable that a user's historical illegal driving behavior reflects the user's driving habits to a certain extent.

[0065] In practical applications, the causal relationship between the target user's P historical illegal driving behaviors can be determined in advance through human experience or a specific algorithm or a specific model to obtain at least one reference causal relationship. In this way, the target user's illegal driving behavior that is about to occur at the next moment can be predicted and reminded directly by finding at least one reference causal relationship.

[0066] Exemplarily, the vehicle device predetermines and stores at least one reference causal relationship. Correspondingly, S101 can be implemented as follows: the vehicle device detects in real time whether the target user has illegal driving behavior through a sensor or a camera; when any illegal driving behavior is detected, the any illegal driving behavior is determined as the first illegal driving behavior, and a reference causal relationship with the first illegal driving behavior as the cause item is searched in at least one reference causal relationship; after the search is completed, R target causal relationships are obtained.

[0067] S102. The vehicle device predicts the target illegal driving behavior of the target user at the next moment based on the R result items in the R target causal relationships.

[0068] The target illegal driving behavior refers to the illegal driving behavior that the target user is about to perform at the next moment. The embodiment of the present invention does not limit the specific content and number of the target illegal driving behaviors, which can be set according to actual conditions. For example, when the first illegal driving behavior is fatigue driving, the target illegal driving behavior may include: not driving according to traffic markings, not maintaining a safe distance, running a red light, etc.

[0069] In a possible implementation, S102 may be implemented as: determining the R result items in the R target causal relationships as the target illegal driving behaviors of the target user at the next moment.

[0070] In another possible implementation, S102 may be implemented as follows: when R is equal to 1, the result item in the target causal relationship is determined as the target illegal driving behavior of the target user at the next moment; when R is greater than 1, the R result items in the R target causal relationships are screened based on preset rules to obtain the target illegal driving behavior of the target user at the next moment.

[0071] The embodiment of the present invention does not limit the specific method of screening the R result items in the R target causal relationships according to the preset rules, and can be set according to actual conditions.

[0072] It is understandable that since the first illegal driving behavior may lead to multiple other illegal driving behaviors, all of the multiple illegal driving behaviors can be determined as the target illegal driving behaviors of the target user at the next moment, or the multiple other illegal driving behaviors can be screened and some of them can be determined as the target illegal driving behaviors of the target user at the next moment.

[0073] S103. The vehicle device outputs target prompt information, where the target prompt information is used to prompt the target user to avoid the target illegal driving behavior.

[0074] The embodiment of the present invention does not limit the specific type, content and output mode of the target prompt information, which can be set according to actual conditions. For example, the target prompt information can be in text type or voice type.

[0075] In actual applications, when there are multiple target illegal driving behaviors, a total target prompt information can be generated based on the multiple target illegal driving behaviors and then output; or a target prompt information can be generated for each target illegal driving behavior to obtain multiple target prompt information, and the multiple target prompt information can be sorted and output in order of the degree of harm of the illegal driving behaviors from large to small and / or in the order of the front and back of the illegal driving behaviors.

[0076] For example, when the first illegal driving behavior is fatigue driving and the target illegal driving behaviors include not driving according to traffic markings, not keeping a safe distance, and running a red light, outputting multiple target prompt information in order of the degree of harm of the illegal driving behaviors from large to small may be: running a red light > not driving according to traffic markings > not keeping a safe distance; outputting multiple target prompt information in the front-to-back order of the illegal driving behaviors may be: not keeping a safe distance > not driving according to traffic markings > running a red light.

[0077] Exemplarily, S103 may be implemented as follows: the vehicle device generates target prompt information for prompting the target user to avoid the target illegal driving behavior based on the target illegal driving behavior; and outputs the target prompt information using a display screen or a speaker.

[0078] An embodiment of the present invention provides a method for predicting illegal driving behavior, which is applied to vehicle equipment, including: when a first illegal driving behavior of a target user in the vehicle equipment is detected, R target causal relationships with the first illegal driving behavior as a cause item are determined in at least one reference causal relationship; the at least one reference causal relationship is determined based on the causal relationship between P historical illegal driving behaviors of the target user; R and P are positive integers; based on R result items in the R target causal relationships, the target illegal driving behavior of the target user at the next moment is predicted; and target prompt information is output, where the target prompt information is used to prompt the target user to avoid the target illegal driving behavior.

[0079] According to the above technical means, by determining at least one reference causal relationship based on the causal relationship between multiple historical illegal driving behaviors in advance, it is possible to predict and remind possible illegal driving behaviors in advance based on at least one reference causal relationship. Since historical illegal driving behaviors can reflect the user's bad driving habits, by mining the influence and occurrence rules between multiple historical illegal driving behaviors, it is possible not only to predict possible illegal driving behaviors based on the user's bad driving habits, effectively improve the accuracy of predicting illegal driving behaviors, but also to avoid the continuous occurrence of illegal driving behaviors and improve driving safety.

[0080] The embodiment of the present invention may further include at least one reference causal relationship determination process.

[0081] refer to Figure 2 As shown in the content, the process may include but is not limited to the following S201 to S203.

[0082] S201. The vehicle device combines the P historical illegal driving behaviors in pairs to obtain Q frequent item sets.

[0083] The Q is a positive integer.

[0084] Frequent items are items that appear more frequently than a threshold in a given set of items.

[0085] It can be understood that a driving violation behavior is a frequent item, which means that the driving violation behavior must occur frequently during the travel process of the target user.

[0086] Frequent itemsets refer to the collection of frequent items.

[0087] Here, since the causal relationship between multiple historical driving violations is unknown at the beginning, it is necessary to combine the multiple historical driving violations in pairs and evaluate and analyze whether there is a causal relationship between any two of the multiple historical driving violations.

[0088] Exemplarily, S201 may be implemented as follows: the vehicle device traverses P historical illegal driving behaviors; for each historical illegal driving behavior, the historical illegal driving behavior is sequentially combined with other subsequent historical illegal driving behaviors in pairs; after the traversal is completed, Q frequent item sets are obtained.

[0089] It should be noted that the Q frequent item sets are not repeated, and the frequent items included in each frequent item set are not repeated. For example, first, the first historical illegal driving behavior can be selected from the P historical illegal driving behaviors and combined with the subsequent P-1 historical illegal driving behaviors in sequence; secondly, the second historical illegal driving behavior can be selected from the P historical illegal driving behaviors and combined with the subsequent P-2 historical illegal driving behaviors in sequence, and so on, until the P-1th element is selected and combined with the Pth element.

[0090] For example, assuming that P historical driving violations include A, B, C, and D, first combine A with B, C, and D to obtain AB, AC, and AD; then combine B with C and D to obtain BC and BD; and then combine C with D to obtain CD. In this way, the P historical driving violations can be combined in pairs.

[0091] For each of the Q frequent item sets, the following processing is performed to obtain the at least one reference causal relationship:

[0092] S202: The vehicle equipment determines the support and confidence of the frequent item set based on the first frequent item and the second frequent item in the frequent item set.

[0093] The occurrence timestamp of the first frequent item is earlier than the occurrence timestamp of the second frequent item; the support is used to represent the probability of the first frequent item and the second frequent item occurring at the same time; and the confidence is used to represent the ratio of the probability of the first frequent item and the second frequent item occurring at the same time to the probability of the first frequent item occurring.

[0094] The first frequent item refers to the frequent item with an earlier occurrence timestamp among the two frequent items included in the frequent item set; the second frequent item refers to the frequent item with a later occurrence timestamp among the two frequent items included in the frequent item set.

[0095] Here, in the reference causal relationship, since the cause item leads to the result item, the cause item must occur before the result item. Therefore, when evaluating and analyzing whether there is a causal relationship between the two frequent items included in the frequent item set, it is necessary to first assume that the frequent item with an earlier occurrence timestamp is the cause item, and the frequent item with a later occurrence timestamp is the result item, and then evaluate and analyze whether there is a causal relationship.

[0096] Support refers to the frequency with which frequent itemsets appear in a data set and is used to measure the prevalence of frequent itemsets.

[0097] In practical applications, support is equal to the number of transactions containing frequent itemsets divided by the total number of transactions. Assuming that A and B form a frequent itemset, the support of the frequent itemset can be found in the following formula (1):

[0098]

[0099] in, is the support of the frequent item set, A is the first frequent item, B is the second frequent item, and P(A∪B) is the probability of the first frequent item and the second frequent item occurring at the same time.

[0100] For example, in a data set containing 10 driving violations in 20 driving trips, assuming that 5 driving trips contain both driving violations A and B, then the support of the frequent item set "driving violations A and driving violations B" is 5 / 20=25%.

[0101] Confidence refers to the probability that a transaction containing a premise item set also contains a result item set, which is used to measure the reliability and strength of the obtained association rules.

[0102] In practical applications, the confidence is equal to the number of transactions that contain both the premise and the result / the number of transactions that contain the premise. Assuming that A and B form a frequent itemset, the confidence of the frequent itemset is shown in the following formula (2):

[0103]

[0104] in, is the confidence of the frequent item set, A is the first frequent item, B is the second frequent item, P(A∪B) is the probability that the first frequent item and the second frequent item occur at the same time, and P(A) is the probability that the first frequent item occurs.

[0105] For example, in a data set containing 10 driving violations in 20 driving trips, suppose there are 5 driving trips that contain both driving violations A and B, and 10 driving trips that contain driving violations A. Then, the confidence of the frequent item set "driving violations A and driving violations B" is 5 / 10=50%.

[0106] Exemplarily, S202 can be implemented as follows: the vehicle device determines the frequent item with an earlier occurrence timestamp among the two frequent items included in the frequent item set as the first frequent item, and determines the frequent item with a later occurrence timestamp as the second frequent item; calculates the probability of the first frequent item and the second frequent item occurring at the same time to obtain the support of the frequent item set; calculates the quotient of the probability of the first frequent item and the second frequent item occurring at the same time divided by the probability of the first frequent item occurring, to obtain the confidence of the frequent item set.

[0107] S203: When the support is greater than or equal to a support threshold and the confidence is greater than or equal to a confidence threshold, the vehicle equipment establishes a causal relationship with the first frequent item as a cause item and the second frequent item as a result item to obtain a reference causal relationship.

[0108] The support threshold and the confidence threshold are used to assist in determining the reference causal relationship. The embodiment of the present invention does not limit the specific values ​​of the support threshold and the confidence threshold, and they can be set according to actual conditions.

[0109] Here, if the support of the frequent item set is greater than or equal to the support threshold and the confidence of the frequent item set is greater than or equal to the confidence threshold, it means that there is an association relationship between the two frequent items included in the frequent item set, and further, it means that there is an association rule from the first frequent item to the second frequent item.

[0110] Exemplarily, S203 can be implemented as follows: the vehicle equipment determines the relationship between the support and the support threshold, and the relationship between the confidence and the confidence threshold; when both are greater than, a causal relationship is established with the first frequent item as the cause item and the second frequent item as the result item to obtain a reference causal relationship.

[0111] Continuing with the above example, if and Then a causal relationship with A as the cause and B as the result is established, and a reference causal relationship is obtained, that is, The association rules are as follows. Where min_supp represents the support threshold and min_conf represents the confidence threshold.

[0112] According to the above technical means, by combining multiple historical illegal driving behaviors in pairs to obtain multiple frequent item sets, and determining the support and confidence of each frequent item set, it is possible to determine the association rules between multiple historical illegal driving behaviors based on the association rule mining algorithm, and obtain at least one reference causal relationship. Since the association rule mining algorithm has the advantages of simple implementation, convenient generalization and strong screening ability, by using the association rule mining algorithm to determine at least one reference causal relationship, it is possible to effectively screen out association rules with practical significance and high reliability based on multiple historical illegal driving behaviors, avoid the interference of a large amount of worthless information, and achieve accurate evaluation of the causal relationship between multiple historical illegal driving behaviors.

[0113] The embodiment of the present invention may also include a process for determining P historical illegal driving behaviors.

[0114] refer to Figure 3 As shown in the content, the process may include but is not limited to the following S301 to S303.

[0115] S301. The vehicle device constructs a behavior matrix based on the historical illegal driving data of the target user.

[0116] The historical illegal driving data is used to represent the behavior data corresponding to N illegal driving behaviors of the target user within M driving trips; both M and N are positive integers.

[0117] The behavior matrix refers to a matrix constructed based on historical illegal driving data. The embodiment of the present invention does not limit the specific matrix elements, data format, and size of the behavior matrix, which can be set according to actual conditions. For example, the behavior matrix is ​​an M×N matrix, which can be expressed as the following formula (3):

[0118]

[0119] Among them, b ij is the i×jth element of the behavior matrix B, determined based on the behavior data corresponding to the jth illegal driving behavior of the target user in the i-th driving trip. i is a positive integer less than or equal to M, and j is a positive integer less than or equal to N.

[0120] Exemplarily, S301 can be implemented as follows: Sub-step A, the vehicle equipment obtains the behavior data corresponding to N illegal driving behaviors of the target user in M ​​driving trips, and obtains M×N historical illegal driving data; Sub-step 2, the vehicle equipment determines the i×jth element of the behavior matrix based on the behavior data corresponding to the jth illegal driving behavior of the target user in the i-th driving trip, and the initial values ​​of i and j are both 1; Sub-step 3, the vehicle equipment increases j by 1, and re-executes sub-step 2 until j is greater than N; Sub-step 3, the vehicle equipment increases i by 1, and re-executes sub-step 2 and sub-step 3 until i is greater than M, and obtains the behavior matrix.

[0121] For example, vehicle equipment can collect the target user's historical driving data in the past period of time through sensors or cameras; analyze these historical driving data through human experience or specific algorithms or specific models to screen out historical illegal driving data; determine the multiple complete and continuous driving trips included in the historical illegal driving data as M driving trips, and determine the multiple illegal driving behaviors involved in the historical illegal driving data as N illegal driving behaviors; analyze and integrate the historical illegal driving data in the driving trip dimension and the driving behavior dimension to obtain the behavior data corresponding to the jth illegal driving behavior of the target user in the i-th driving trip; determine each element of the initial behavior matrix based on the behavior data corresponding to each illegal driving behavior of the target user in each driving trip. Among them, the historical driving data can include driving information, location information, external environment information, in-vehicle environment information, etc. The driving information can include vehicle speed, acceleration, deceleration, throttle opening and closing, steering or pedal force angle, etc. The external environment information can include traffic light information, traffic markings or vehicle surrounding information, and the in-vehicle environment information can include personnel image information and voice information.

[0122] S302. The vehicle device determines N weight values ​​corresponding to the N illegal driving behaviors based on the behavior matrix.

[0123] The embodiment of the present invention does not limit the specific method for the vehicle device to determine the N weight values ​​corresponding to N days based on the target behavior matrix, and can be set according to actual conditions. For example, the entropy weight method can be used to process the target behavior matrix to obtain N weight values ​​corresponding to N days.

[0124] The entropy weight method is an objective weighting method used to determine the weights of multiple indicators in comprehensive evaluation. The entropy weight method is based on the concept of information entropy. It determines the weight value corresponding to each indicator by calculating the information entropy of each indicator. The smaller the entropy value, the greater the degree of variation of the indicator, the more information provided, and the greater the weight; conversely, the larger the entropy value, the smaller the weight. The entropy weight method can effectively avoid the interference of subjective factors and make the evaluation results more objective and scientific. It is often used in the field of multi-indicator comprehensive evaluation, such as economy, management, engineering, etc.

[0125] Exemplarily, S302 can be implemented as follows: the vehicle equipment standardizes and normalizes the behavior matrix to obtain a normalized matrix; based on the normalized matrix, N information entropies corresponding to N illegal driving behaviors are calculated; based on the N information entropies, N weight values ​​corresponding to the N illegal driving behaviors are calculated.

[0126] It should be noted that for specific methods of standardization, normalization, calculation of indicator information entropy, and calculation of indicator weight values, reference may be made to the detailed description of the entropy weight method in the relevant technology.

[0127] S303: The vehicle device determines the illegal driving behaviors whose weight values ​​are greater than a first threshold among the N illegal driving behaviors as the P historical illegal driving behaviors.

[0128] The first threshold is used to assist in determining the P historical illegal driving behaviors. The embodiment of the present invention does not limit the specific value of the first threshold, and it can be set according to actual conditions.

[0129] Here, since some illegal driving behaviors may occur accidentally and cannot truly reflect the user's bad driving habits, in order to reduce the interference of these occasional illegal driving behaviors in judging bad driving habits, it is necessary to screen all illegal driving behaviors based on weight values ​​and determine illegal driving behaviors with high value.

[0130] Exemplarily, S303 can be implemented as follows: the vehicle equipment traverses N illegal driving behaviors; for each illegal driving behavior, it is determined whether the weight value corresponding to the illegal driving behavior is greater than a first threshold, and if so, the illegal driving behavior is determined as a historical illegal driving behavior; after the traversal is completed, P historical illegal driving behaviors are obtained.

[0131] According to the above technical means, since some illegal driving behaviors may only occur occasionally and cannot truly reflect the user's bad driving habits, by determining the N weight values ​​corresponding to N illegal driving behaviors, all illegal driving behaviors can be screened based on the weight values ​​to obtain P historical illegal driving behaviors with high value. In this way, the interference caused by accidental behaviors can be effectively eliminated, and illegal driving behaviors with high value that can reflect the user's normalized driving tendency can be determined more comprehensively and accurately, so as to more effectively evaluate the user's bad driving habits and provide more valuable reference for the prediction of illegal driving behaviors that are about to occur and assisted driving.

[0132] Next, the process of constructing the behavior matrix based on the historical illegal driving data of the target user by the vehicle device in S301 is described.

[0133] refer to Figure 4 As shown in the content, the process may include but is not limited to the following S401 and S402.

[0134] S401. The vehicle device determines the number of behaviors and behavior penalty points of the j-th illegal driving behavior of the target user in the i-th driving trip based on the historical illegal driving data.

[0135] The deduction points for illegal driving behavior refer to the deduction points for illegal driving behavior according to the traffic laws and regulations formulated by the country or region. For example, if a driver does not overtake or give way according to regulations, 3 points will be deducted at one time; if a driver violates the prohibition signs and prohibition markings, 3 points will be deducted at one time, and so on.

[0136] It can be understood that the number of illegal driving behaviors can reflect the frequency of illegal driving behaviors, and the points deducted for illegal driving behaviors can reflect the degree of harm caused by illegal driving behaviors.

[0137] Exemplarily, the vehicle device collects, analyzes, and integrates the target user's historical illegal driving data in advance to obtain the number of behaviors and behavior penalty points of each illegal driving behavior of the target user in each driving trip, and stores them locally or in the cloud. Correspondingly, S401 can be implemented as follows: the vehicle device obtains the number of behaviors and behavior penalty points of the j-th illegal driving behavior of the target user in the i-th driving trip from the local or cloud.

[0138] S402: The vehicle equipment multiplies the number of behaviors and the behavior deduction value to obtain the i×jth behavior matrix element of the behavior matrix.

[0139] Among them, the i×j-th behavior matrix element is any matrix element of the behavior matrix; the i is a positive integer less than or equal to the M; and the j is a positive integer less than or equal to the N.

[0140] Exemplarily, S402 may be implemented as follows: the vehicle device multiplies the number of behaviors and the behavior penalty value of the j-th illegal driving behavior of the target user in the i-th driving trip to obtain the i×j-th behavior matrix element of the behavior matrix.

[0141] For example, assuming that the target user has 5 violations of driving behavior in the jth driving trip, and the penalty value is 3 points, then the i×jth behavior matrix element of the behavior matrix may be 15.

[0142] According to the above technical means, considering the complexity and diversity of different illegal driving behaviors, by determining the number of illegal driving behaviors and the deduction points of each illegal driving behavior of the target user in each driving trip, it is possible to construct a behavior matrix based on the type of illegal driving behavior and the number of illegal driving behaviors and the deduction points of illegal driving behaviors. In this way, by evaluating the user's bad driving habits through a comprehensive consideration method, it is possible to more accurately distinguish the degree and nature of bad driving habits, avoid the bias and one-sidedness brought about by a single consideration method, and more comprehensively, accurately and scientifically characterize the user's bad driving habits.

[0143] Next, the process of determining the N weight values ​​corresponding to the N illegal driving behaviors by the vehicle device based on the behavior matrix in S302 is described.

[0144] refer to Figure 5 As shown in the content, the process may include but is not limited to the following S501 to S504.

[0145] S501. The vehicle equipment normalizes the behavior matrix based on a first reference expression of an entropy weight method to obtain a reference matrix.

[0146] The reference matrix refers to the matrix obtained by standardizing the behavior matrix.

[0147] The first reference expression is used to standardize the matrix or element. The embodiment of the present invention does not limit the specific formula of the first reference expression, and it can be set according to the actual situation. Exemplarily, the first reference expression can be an indicator standardization formula of the entropy weight method, see the following formula (4):

[0148]

[0149] Among them, b i ' j is the i×jth element of the reference matrix B', b ij is the i×jth element of the behavior matrix B, and M is the number of rows of the behavior matrix B.

[0150] It should be noted that 0.998 and 0.002 are correction coefficients used to avoid calculation deviations, make the standardized matrix elements more reasonable, and avoid the phenomenon of too many 0s in the standardized matrix elements.

[0151] Exemplarily, S501 can be implemented as follows: the vehicle equipment traverses the M×N behavior matrix elements of the behavior matrix; for each behavior matrix element, the indicator standardization formula of the entropy weight method is used to standardize the behavior matrix element to obtain a corresponding reference matrix element; after the traversal is completed, a reference matrix is ​​obtained.

[0152] S502. The vehicle equipment normalizes the reference matrix based on the second reference expression of the entropy weight method to obtain a normalized matrix.

[0153] The normalized matrix refers to the matrix obtained by normalizing the reference matrix.

[0154] The second reference expression is used to normalize the matrix or element. The embodiment of the present invention does not limit the specific formula of the second reference expression, and it can be set according to the actual situation. Exemplarily, the second reference expression can be the index weight formula of the entropy weight method, see the following formula (5):

[0155]

[0156] Among them, b i ' j ' is the i×jth element of the normalized matrix B", b i ' j is the i×jth element of the reference matrix B', and M is the number of rows of the reference matrix B'.

[0157] Exemplarily, S502 can be implemented as follows: the vehicle equipment traverses the M×N reference matrix elements of the reference matrix; for each reference matrix element, the index weight formula of the entropy weight method is used to normalize the reference matrix element to obtain a corresponding normalized matrix element; after the traversal is completed, a normalized matrix is ​​obtained.

[0158] S503. The vehicle equipment determines N information entropies corresponding to the N illegal driving behaviors based on the third reference expression of the entropy weight method and the normalized matrix.

[0159] The third reference expression is used to calculate the information entropy of a matrix or element. The embodiment of the present invention does not limit the specific formula of the third reference expression, and it can be set according to actual conditions. Exemplarily, the third reference expression can be the indicator information entropy formula of the entropy weight method, see the following formula (6):

[0160]

[0161] Among them, H j is the information entropy corresponding to the j-th illegal driving behavior, b″ ij is the i×jth element of the normalized matrix B”, and M is the number of rows of the reference matrix B’.

[0162] Exemplarily, S503 can be implemented as follows: the vehicle equipment traverses the N columns of the normalized matrix; for each column, the information entropy corresponding to the column is calculated using the indicator information entropy formula of the entropy weight method to obtain the information entropy corresponding to the current illegal driving behavior among N illegal driving behaviors; after the traversal is completed, N information entropies corresponding to the N illegal driving behaviors are obtained.

[0163] S504. The vehicle equipment determines N weight values ​​corresponding to the N illegal driving behaviors based on the fourth reference expression of the entropy weight method and the N information entropies.

[0164] The fourth reference expression is used to calculate the weight value of the matrix or element. The embodiment of the present invention does not limit the specific formula of the fourth reference expression, and it can be set according to actual conditions. Exemplarily, the fourth reference expression can be the indicator weight formula of the entropy weight method, see the following formula (7):

[0165]

[0166] Among them, ω j is the weight value corresponding to the j-th illegal driving behavior, H j is the information entropy corresponding to the j-th illegal driving behavior, and N is the total number of illegal driving behaviors.

[0167] Exemplarily, S504 can be implemented as follows: the vehicle equipment traverses the most recent N illegal driving behaviors; for each illegal driving behavior, the index weight formula of the entropy weight method is used to calculate the weight value corresponding to the illegal driving behavior; after the traversal is completed, N weight values ​​corresponding to the N illegal driving behaviors are obtained.

[0168] By adopting the entropy weight method to determine the N weight values ​​corresponding to N illegal driving behaviors, the interference of human subjective factors can be effectively avoided, and each of the N illegal driving behaviors can be objectively analyzed. Moreover, higher weights can be given to illegal driving behaviors with larger amounts of information and higher degrees of variation, making the determination of the weight value corresponding to each of the N illegal driving behaviors more objective, fair and scientific.

[0169] Next, the process of predicting the target illegal driving behavior of the target user at the next moment by the vehicle device based on the R result items in the R target causal relationships in S102 is described.

[0170] refer to Figure 6 The content shown in the above Figure 1 On the basis of, step S102 is replaced by S601 or S602. The process may include but is not limited to the following S601 and S602.

[0171] S601. When R is equal to 1, the vehicle equipment determines the result item in the target causal relationship as the target illegal driving behavior.

[0172] Exemplarily, S601 may be implemented as follows: the vehicle device determines whether R is greater than 1; if not, the result item in the target causal relationship is determined as the target illegal driving behavior.

[0173] It can be understood that when R is equal to 1, it means that there is only one illegal driving behavior predicted based on at least one reference causal relationship, then the illegal driving behavior is determined as the target illegal driving behavior of the target user at the next moment.

[0174] S602: When R is greater than 1, the vehicle equipment determines the R result items in the R target causal relationships as R candidate illegal driving behaviors; and screens the R candidate illegal driving behaviors to obtain the target illegal driving behavior.

[0175] Candidate illegal driving behaviors refer to illegal driving behaviors that are initially predicted based on causal relationships.

[0176] Exemplarily, S602 may be implemented as follows: the vehicle device determines whether R is greater than 1; if so, determines the R result items in the R target causal relationships as R candidate illegal driving behaviors; and screens the R candidate illegal driving behaviors based on preset rules to obtain the target illegal driving behavior.

[0177] The embodiment of the present invention does not limit the specific method of screening R candidate illegal driving behaviors based on preset rules to obtain the target illegal driving behavior, which can be set according to actual conditions. For example, the R candidate illegal driving behaviors can be screened based on the highest probability rule or priority information between multiple candidate illegal driving behaviors to obtain the target illegal driving behavior.

[0178] According to the above technical means, by determining whether to screen the R result items in the R target causal relationships to obtain the target illegal driving behavior according to different values ​​of the number of result items R, a more refined prediction and management of the target illegal driving behavior can be achieved. In this way, the predicted low-value illegal driving behaviors can be removed, and the focus can be placed on the truly critical illegal driving behaviors, making the prediction results more targeted and reliable, thereby realizing the reminder of high-value illegal driving behaviors.

[0179] Next, the process in which the vehicle device screens the R candidate illegal driving behaviors to obtain the target illegal driving behavior in S602 is described.

[0180] refer to Figure 7 As shown in the content, the process may include but is not limited to the following S701 and S702.

[0181] S701. The vehicle device determines the occurrence probability of each of the R candidate illegal driving behaviors to obtain R occurrence probability values.

[0182] Here, the probability of occurrence of a driving violation refers to the ratio of the number of occurrences of a certain driving violation to the number of occurrences of all driving violations.

[0183] In actual application, assuming that the total number of N driving violations of the target user in M ​​driving trips is 100, and the number of speeding violations is 5, then the probability of speeding violations is 5 / 100=5%.

[0184] Exemplarily, the vehicle device analyzes and integrates the historical illegal driving data of the target user in advance to obtain the number of occurrences of each illegal driving behavior and the total number of occurrences of all illegal driving behaviors. S701 can be implemented as follows: the vehicle device traverses R candidate illegal driving behaviors; for each candidate illegal driving behavior, the quotient of the number of occurrences of the candidate illegal driving behavior divided by the total number of occurrences of all illegal driving behaviors is determined as the probability value of the candidate illegal driving behavior; after the traversal is completed, R probability values ​​of occurrence are obtained.

[0185] S702: The vehicle device determines the candidate illegal driving behavior with the highest probability of occurrence among the R candidate illegal driving behaviors as the target illegal driving behavior.

[0186] Exemplarily, S702 may be implemented as follows: the vehicle device sorts the R candidate illegal driving behaviors in descending order of occurrence probability values; and determines the first candidate illegal driving behavior in the sorting result as the target illegal driving behavior.

[0187] According to the above technical means, by determining the probability of occurrence of each candidate illegal driving behavior among the R candidate illegal driving behaviors, it is possible to determine the candidate illegal driving behavior with the highest probability of occurrence among the R candidate illegal driving behaviors as the target illegal driving behavior. Since too much reminder information may make it difficult for users to focus on key illegal driving behaviors, by strictly distinguishing the probability of occurrence of different illegal driving behaviors, it is possible to accurately focus on the illegal driving behaviors that are most likely to occur, so that the predicted target illegal driving behaviors are more focused, thereby achieving targeted reminders for the target illegal driving behaviors.

[0188] Next, the process in which the vehicle device screens the R candidate illegal driving behaviors to obtain the target illegal driving behavior in S602 is described.

[0189] refer to Figure 8 As shown in the content, the process may include but is not limited to the following S801 and S802.

[0190] S801. The vehicle device sorts the R candidate illegal driving behaviors based on a reference priority rule to obtain a sorting result.

[0191] The reference priority rule is determined based on the preceding and following logic or the degree of harm logic among the R candidate illegal driving behaviors.

[0192] The preceding and following logic between driving violations refers to the order in which different driving violations occur. For example, when multiple driving violations include failure to maintain a safe distance and failure to avoid illegal lane changes, the preceding and following logic between the driving violations may include: the driving violation of failure to maintain a safe distance occurs first, and then the driving violation of failure to avoid illegal lane changes occurs.

[0193] It is understandable that there is an essential difference between the causal relationship between illegal driving behaviors and the before-and-after logic between illegal driving behaviors. They are completely different concepts.

[0194] The logic of the degree of harm between illegal driving behaviors refers to the degree of harm caused by different illegal driving behaviors. For example, when multiple illegal driving behaviors include not driving according to traffic markings, not keeping a safe distance, and running a red light, the logic of the degree of harm between illegal driving behaviors may include: running a red light has the greatest degree of harm, not driving according to traffic markings has the second greatest degree of harm, and not keeping a safe distance has the least degree of harm.

[0195] In a possible implementation, S801 may be implemented as follows: the vehicle device sorts the R candidate illegal driving behaviors according to the preceding and following logic among the R candidate illegal driving behaviors to obtain a first sorting result.

[0196] In another possible implementation, S801 can be implemented as follows: the vehicle equipment evaluates and quantifies the degree of harm of R candidate illegal driving behaviors through human experience or a specific algorithm or a specific model to obtain R degree of harm scores; the R candidate illegal driving behaviors are sorted in descending order according to the degree of harm scores to obtain a second sorting result.

[0197] S802: The vehicle device determines the first X candidate illegal driving behaviors in the sorting result as the target illegal driving behaviors.

[0198] In a possible implementation, S802 may be implemented as follows: when a first sorting result is obtained by sorting the R candidate illegal driving behaviors based on the front and back logic between the R candidate illegal driving behaviors, the vehicle equipment determines the first X candidate illegal driving behaviors in the first sorting result as the target illegal driving behaviors.

[0199] In another possible implementation, S802 may be implemented as follows: when a second sorting result is obtained by sorting the R candidate illegal driving behaviors based on the logic of their degree of harm, the vehicle equipment determines the first X candidate illegal driving behaviors in the second sorting result as target illegal driving behaviors.

[0200] According to the above technical means, considering the inherent correlation between illegal driving behaviors and the severity of the consequences, by determining the pre- and post-logic or harm degree logic between the R candidate illegal driving behaviors, it is possible to screen the R candidate illegal driving behaviors based on the pre- and post-logic or harm degree logic between the R candidate illegal driving behaviors to obtain the target illegal driving behavior. In this way, the screening mechanism of the candidate illegal driving behaviors can be made more scientific, reasonable, systematic, comprehensive, orderly and balanced, thereby improving the orderliness and emphasis of the predicted target driving behavior.

[0201] The illegal driving behavior prediction method provided by the present invention is described below through a detailed embodiment.

[0202] Nowadays, driving has become an indispensable means of transportation in today's society. Therefore, it is very important to ensure that drivers drive their vehicles safely. Bad driving behavior of drivers can not only easily cause traffic accidents, but also easily cause damage to vehicles. Analyzing illegal driving behavior is an effective way to discover the causes of accidents and improve vehicle driving safety. Related technology discloses a method for intelligent detection of driving behavior habits and safe driving scoring. The method obtains real-time data of users through devices and core modules, and then promptly reminds and feedbacks users' driving behavior habits. This method obtains real-time data through physical devices to predict possible violations that may occur. It relies too much on the accuracy and real-time performance of physical devices. Even if it is discovered that the user is about to violate the rules and the user is notified, it is difficult for the user to react in time to avoid the occurrence of violations.

[0203] Association rule mining is an effective data mining method that can mine the influence and rules between different violations. Therefore, in view of the problems in the prior art, the present invention provides a driving behavior analysis method that is highly objective, efficient, and applicable to different drivers and different scenarios. It can predict possible violations in advance according to different user driving habits, and promptly remind users to drive in a standardized manner, help users improve their driving habits, and improve driving safety.

[0204] Based on this, the present invention provides a method for analyzing driving violation behavior, which relates to the technical field of driving behavior data processing, and in particular to a method for predicting driving violation behavior based on an Apriori algorithm. Fig. 9 The method mainly includes steps S901 to S905:

[0205] S901. Collect vehicle driving data of the driver during the statistical period.

[0206] The vehicle driving data of the user during the statistical period is obtained by using monitoring modules such as GPS module, vehicle panoramic view system, in-vehicle camera, call information or voice information and various sensors. Driving information includes vehicle speed, acceleration, deceleration, throttle opening and closing degree, steering or pedal force angle, etc., among which vehicle speed, acceleration and deceleration can be obtained through vehicle speed sensor, throttle opening and closing degree can be obtained through throttle sensor, steering can be obtained through steering sensor, and pedal force angle can be obtained through pedal sensor; real-time location information of target vehicle can be obtained through GPS module; vehicle external environment information includes traffic light information, traffic markings or vehicle surrounding information, and external environment information can be obtained through vehicle panoramic view system; in-vehicle environment information includes call information or voice information, and call information is obtained through call module in vehicle computer, and voice information is obtained through voice recognition module.

[0207] A complete statistical cycle refers to the driving scoring cycle time customized by the target driver, such as one day, one week, or one month, during which all driving behaviors of the target vehicle are counted.

[0208] S902: Determine each illegal driving behavior information of the vehicle according to the vehicle driving data.

[0209] The target driver's violation of traffic rules or dangerous driving behavior during the driving trip is used as the evaluation index, the number of complete driving processes is used as the evaluation index, and the target driver is used as the evaluation object. By analyzing the vehicle driving data within the statistical period, the illegal driving behaviors within the statistical period are screened out. Among them, the illegal driving behavior information mainly includes the type of illegal driving behavior and the number of illegal driving behaviors.

[0210] S903. Calculate the scoring weight of each illegal driving behavior using the entropy weight method.

[0211] The entropy weight method is an objective weighting method. Using the entropy weight method to assign weights can avoid subjectivity as much as possible. The calculation method and solution steps are as follows:

[0212] 1) Construct a judgment matrix based on the n illegal driving behavior data of m complete driving records within the statistical period. The judgment matrix is ​​shown in the following formula (8):

[0213]

[0214] In the formula, x mn The behavior data of the nth violation corresponding to the mth driving record.

[0215] 2) Standardize the judgment matrix. Use the following formula (9) for each x ij Do the standardization and get the standardized matrix Z = (r ij )m×n :

[0216]

[0217] 3) Normalize the standardized matrix. Use the following formula (10) for each r ij Do normalization processing to get the normalized matrix A = (y ij ) m×n :

[0218]

[0219] The normalized matrix A is shown in the following formula (11), which is easy to verify: That is to say, it ensures that the sum of the probabilities corresponding to each indicator is 1.

[0220]

[0221] 4) Determine the information entropy of the evaluation index in the evaluated driving behavior object, and determine the entropy weight of the evaluation index according to the defined entropy of the evaluation index. In the evaluation problem of m evaluated objects and n evaluation indicators, the entropy of the i-th evaluation indicator is determined according to the following formula (12):

[0222]

[0223] According to the following formula (13), the entropy weight of the i-th evaluation index is determined as:

[0224]

[0225] in,

[0226] In this way, the evaluation index weight vector based on entropy weight can be obtained ω=(ω 1 ,ω 2 ,…,ω m ).

[0227] According to the above analysis, the judgment matrix of the violation behavior index in the driving process is constructed. It is assumed that the n data items in the evaluation index model are measured m times, thus forming the judgment matrix X; the judgment matrix is ​​standardized and normalized to obtain the normalized matrix A; the information entropy of the evaluation index is determined from the evaluation driving behavior object, and the entropy weight of the evaluation index is determined according to the entropy of the defined evaluation index. In the evaluation problem of m evaluated objects and n evaluation indicators, the entropy of the i-th evaluation indicator is determined, and the evaluation index weight vector based on the entropy weight is obtained. 1 ,ω 2 ,…,ω m ).

[0228] S904: Calculate the potential relationship between each illegal driving behavior using the Apriori association rule mining algorithm.

[0229] Combined with the Apriori association rule algorithm, the weight is used as a measure of the frequency of violations, the support threshold is set, the frequent item sets are calculated, the confidence threshold is set, and finally the association rules are generated. The Apriori association rule mining algorithm is used to form frequent item sets for all violations, and then the algorithm is used to generate association rules for violations. Specifically, the following steps are included:

[0230] a) Perform initialization operations.

[0231] It mainly includes traversing the driving violation database and setting the minimum weight threshold min_sup. Find all items in the evaluation index weight vector whose weight is greater than the minimum weight threshold. These items constitute a set I, I = {I 1 ,I 2 ,…,I m}.

[0232] b) Construct frequent itemsets.

[0233] Combine two items in set I to obtain multiple frequent item sets.

[0234] c) Calculate the support of frequent itemsets.

[0235] For each frequent itemset, let A and B form a frequent itemset, but is the desired association rule. The support of frequent itemsets is calculated according to the following formula (14):

[0236]

[0237] d) If the support is greater than or equal to the support threshold, calculate the confidence of the frequent itemset.

[0238] For each frequent item set, the confidence of the frequent item set is calculated according to the following formula (15):

[0239]

[0240] e) If the confidence is greater than or equal to the confidence threshold, an association rule is established.

[0241] like B≠Φ, and Association rules Indicates that A causes B, where min_conf is the confidence threshold.

[0242] S905. Send information in a timely manner to remind the user to avoid illegal driving behavior.

[0243] Obtain the potential correlation between each violation, predict the violation that the user may trigger, and remind the user in time to avoid the violation. By using the potential correlation between violations, when the user triggers the illegal driving behavior A, send a reminder and feedback of the illegal driving behavior B to the user in time, so as to avoid the continuous occurrence of violations.

[0244] By sorting and analyzing the violation-related data, we can obtain user-specific violation predictions. During the user's driving process, we can use the message module and alarm module to notify the user in real time that they should avoid certain dangerous operations and avoid the occurrence of violations in advance.

[0245] The advantages of the present invention include: providing a driving behavior scoring method based on the entropy weight method, calculating the weight of the violation behavior index by analyzing the driving behavior within the statistical period, using the Apriori association rule mining algorithm to dynamically combine real-time driving data, and using the final calculation result as the predicted violation behavior, reminding users to avoid violations in time and reduce the incidence of traffic accidents.

[0246] In a second aspect, an embodiment of the present invention provides a device for predicting illegal driving behavior, which is deployed in a vehicle device, such as Fig.10 As shown, the illegal driving behavior prediction device 100 includes: a determination unit 1001, a prediction unit 1002, and an output unit 1003. Among them:

[0247] The determining unit 1001 is configured to, when a first illegal driving behavior of a target user in the vehicle device is detected, determine R target causal relationships with the first illegal driving behavior as a cause item in at least one reference causal relationship; the at least one reference causal relationship is determined based on the causal relationship between P historical illegal driving behaviors of the target user; R and P are positive integers;

[0248] A prediction unit 1002 is used to predict the target illegal driving behavior of the target user at the next moment based on the R result items in the R target causal relationships;

[0249] The output unit 1003 is used to output target prompt information, where the target prompt information is used to prompt the target user to avoid the target illegal driving behavior.

[0250] In some embodiments, before determining R target causal relationships with the first illegal driving behavior as the cause item in at least one reference causal relationship, the determining unit 1001 is further used to: combine the P historical illegal driving behaviors in pairs to obtain Q frequent item sets; Q is a positive integer; for each of the Q frequent item sets, perform the following processing to obtain the at least one reference causal relationship: based on the first frequent item and the second frequent item in the frequent item set, determine the support and confidence of the frequent item set; wherein the occurrence timestamp of the first frequent item is earlier than the occurrence timestamp of the second frequent item; the support is used to characterize the probability of the first frequent item and the second frequent item occurring at the same time; the confidence is used to characterize the proportion of the probability of the first frequent item and the second frequent item occurring at the same time to the probability of the first frequent item occurring; when the support is greater than or equal to a support threshold and the confidence is greater than or equal to a confidence threshold, establish a causal relationship with the first frequent item as the cause item and the second frequent item as the result item, and obtain one reference causal relationship.

[0251] In some embodiments, the determination unit 1001 is also used to: construct a behavior matrix based on the historical illegal driving data of the target user; the historical illegal driving data is used to characterize the behavior data corresponding to N illegal driving behaviors of the target user in M ​​driving trips; both M and N are positive integers; based on the behavior matrix, determine the N weight values ​​corresponding to the N illegal driving behaviors; and determine the illegal driving behaviors whose weight values ​​among the N illegal driving behaviors are greater than a first threshold as the P historical illegal driving behaviors.

[0252] In some embodiments, the determination unit 1001 is also used to: determine the number of behaviors and the behavior penalty value of the j-th illegal driving behavior of the target user in the i-th driving trip based on the historical illegal driving data; multiply the number of behaviors and the behavior penalty value to obtain the i×j-th behavior matrix element of the behavior matrix; wherein the i×j-th behavior matrix element is any matrix element of the behavior matrix; i is a positive integer less than or equal to M; and j is a positive integer less than or equal to N.

[0253] In some embodiments, the determination unit 1001 is also used to: based on the first reference expression of the entropy weight method, standardize the behavior matrix to obtain a reference matrix; based on the second reference expression of the entropy weight method, normalize the reference matrix to obtain a normalized matrix; based on the third reference expression of the entropy weight method and the normalized matrix, determine the N information entropies corresponding to the N illegal driving behaviors; based on the fourth reference expression of the entropy weight method and the N information entropies, determine the N weight values ​​corresponding to the N illegal driving behaviors.

[0254] In some embodiments, the prediction unit 1002 is specifically used to: when R is equal to 1, determine the result item in the target causal relationship as the target illegal driving behavior; when R is greater than 1, determine the R result items in the R target causal relationships as R candidate illegal driving behaviors; and screen the R candidate illegal driving behaviors to obtain the target illegal driving behavior.

[0255] In some embodiments, the prediction unit 1002 is also used to: determine the occurrence probability of each of the R candidate illegal driving behaviors to obtain R occurrence probability values; and determine the candidate illegal driving behavior with the highest occurrence probability value among the R candidate illegal driving behaviors as the target illegal driving behavior.

[0256] In some embodiments, the prediction unit 1002 is also used to: sort the R candidate illegal driving behaviors based on a reference priority rule to obtain a sorting result; the reference priority rule is determined based on the front and back logic or the degree of harm logic between the R candidate illegal driving behaviors; and the first X candidate illegal driving behaviors in the sorting result are determined as the target illegal driving behaviors.

[0257] It should be noted that the various units included in the illegal driving behavior prediction device provided in the embodiment of the present invention can be implemented by a processor in an electronic device; of course, it can also be implemented by a specific logic circuit; in the implementation process, the processor can be a central processing unit (CPU, Central Processing Unit), a microprocessor (MPU, MicroProcessor Unit), a digital signal processor (DSP, Digital Signal Processor) or a field programmable gate array (FPGA, Field-Programmable Gate Array), etc.

[0258] The description of the above device embodiment is similar to the description of the above method embodiment, and has similar beneficial effects as the method embodiment. For technical details not disclosed in the device embodiment of the present invention, please refer to the description of the method embodiment of the present invention for understanding.

[0259] It should be noted that, in the embodiment of the present invention, if the above-mentioned illegal driving behavior prediction method is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the relevant technology. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the illegal driving behavior prediction method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), disk or optical disk, etc. Various media that can store program codes. In this way, the embodiment of the present invention is not limited to any specific combination of hardware and software.

[0260] In a third aspect, an embodiment of the present invention provides an electronic device comprising at least a memory and a processor, wherein the memory stores a computer program executable on the processor, and when the processor executes the program, the steps in the method for predicting illegal driving behavior provided in the above embodiment are implemented.

[0261] Combine the following Fig.11 The electronic device 110 shown is used to illustrate a structural diagram of the electronic device.

[0262] In one example, if Fig.11 As shown, the electronic device 110 includes: a processor 1101, at least one communication bus 1102, at least one external communication interface 1103 and a memory 1104. The communication bus 1102 is configured to realize connection and communication between these components. The external communication interface 1103 may include a standard wired interface and a wireless interface.

[0263] The memory 1104 is configured to store instructions and applications executable by the processor 1101, and can also cache data to be processed or processed by the processor 1101 and various modules in the electronic device (for example, image data, audio data, voice communication data, and video communication data), which can be implemented through flash memory (FLASH) or random access memory (Random Access Memory, RAM).

[0264] In another example, the electronic device may be a vehicle device or a controller. The vehicle device or the controller is used to execute the steps in the illegal driving behavior prediction method provided in the above embodiment.

[0265] In a fourth aspect, the present invention further provides a storage medium having a computer program stored thereon. When the computer program on the storage medium is executed, the steps in the method for predicting illegal driving behavior provided in an embodiment of the present invention are implemented.

[0266] In a fifth aspect, the present invention further provides a computer program product, including a computer program or instructions, which, when executed by a processor, implement the steps in the above-mentioned illegal driving behavior prediction method.

[0267] It should be noted that the description of the above storage medium and device embodiments is similar to the description of the above method embodiments, and has similar beneficial effects as the method embodiments. For technical details not disclosed in the storage medium and device embodiments of the present invention, please refer to the description of the method embodiments of the present invention for understanding.

[0268] It should be understood that "one embodiment" or "an embodiment" mentioned throughout the specification means that specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present invention. Therefore, "in one embodiment" or "in some embodiments" appearing throughout the specification does not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. It should be understood that in various embodiments of the present invention, the size of the serial number of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention. The serial numbers of the above-mentioned embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments.

[0269] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0270] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0271] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0272] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0273] A person skilled in the art can understand that all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM), magnetic disks or optical disks, etc., various media that can store program codes.

[0274] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the relevant technology. The computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the illegal driving behavior prediction method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROMs, magnetic disks, or optical disks.

[0275] The above is only an embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for predicting illegal driving behavior, characterized in that: The method is applied to vehicle equipment, and the method comprises: In the case where a first illegal driving behavior of a target user in the vehicle device is detected, R target causal relationships with the first illegal driving behavior as a cause item are determined in at least one reference causal relationship; the at least one reference causal relationship is determined based on the causal relationship between P historical illegal driving behaviors of the target user; R and P are positive integers; Based on the R result items in the R target causal relationships, predict the target illegal driving behavior of the target user at the next moment; Output target prompt information, where the target prompt information is used to prompt the target user to avoid the target illegal driving behavior.

2. The method for predicting illegal driving behavior according to claim 1, characterized in that: Before determining R target causal relationships with the first illegal driving behavior as a cause item in at least one reference causal relationship, the method further includes: The P historical illegal driving behaviors are combined in pairs to obtain Q frequent item sets; Q is a positive integer; For each of the Q frequent item sets, the following processing is performed to obtain the at least one reference causal relationship: Based on the first frequent item and the second frequent item in the frequent item set, determining the support and confidence of the frequent item set; wherein the occurrence timestamp of the first frequent item is earlier than the occurrence timestamp of the second frequent item; the support is used to represent the probability of the first frequent item and the second frequent item occurring at the same time; the confidence is used to represent the ratio of the probability of the first frequent item and the second frequent item occurring at the same time to the probability of the first frequent item occurring; When the support is greater than or equal to a support threshold and the confidence is greater than or equal to a confidence threshold, a causal relationship is established with the first frequent item as a cause item and the second frequent item as a result item to obtain a reference causal relationship.

3. The method for predicting illegal driving behavior according to claim 1, characterized in that: The method further comprises: Based on the historical illegal driving data of the target user, a behavior matrix is ​​constructed; the historical illegal driving data is used to represent the behavior data corresponding to N illegal driving behaviors of the target user in M ​​driving trips; M and N are both positive integers; Based on the behavior matrix, determining N weight values ​​corresponding to the N illegal driving behaviors; The illegal driving behaviors whose weight values ​​are greater than a first threshold among the N illegal driving behaviors are determined as the P historical illegal driving behaviors.

4. The method for predicting illegal driving behavior according to claim 3, characterized in that: The constructing of a behavior matrix based on the target user's historical illegal driving data includes: Based on the historical illegal driving data, determine the number of behaviors and behavior penalty points of the j-th illegal driving behavior of the target user in the i-th driving trip; Multiply the number of behaviors and the behavior deduction value to obtain the i×jth behavior matrix element of the behavior matrix; Among them, the i×j-th behavior matrix element is any matrix element of the behavior matrix; the i is a positive integer less than or equal to the M; and the j is a positive integer less than or equal to the N.

5. The method for predicting illegal driving behavior according to claim 3, characterized in that: The determining, based on the behavior matrix, N weight values ​​corresponding to the N illegal driving behaviors includes: Based on a first reference expression of an entropy weight method, the behavior matrix is ​​normalized to obtain a reference matrix; Based on the second reference expression of the entropy weight method, normalizing the reference matrix to obtain a normalized matrix; Determining N information entropies corresponding to the N illegal driving behaviors based on the third reference expression of the entropy weight method and the normalized matrix; Based on the fourth reference expression of the entropy weight method and the N information entropies, N weight values ​​corresponding to the N illegal driving behaviors are determined.

6. The method for predicting illegal driving behavior according to claim 1, characterized in that: The predicting the target illegal driving behavior of the target user at the next moment based on the R result items in the R target causal relationships includes: When R is equal to 1, the result item in the target causal relationship is determined as the target illegal driving behavior; When R is greater than 1, the R result items in the R target causal relationships are determined as R candidate illegal driving behaviors; the R candidate illegal driving behaviors are screened to obtain the target illegal driving behavior.

7. The method for predicting illegal driving behavior according to claim 6, characterized in that: The screening of the R candidate illegal driving behaviors to obtain the target illegal driving behavior includes: Determine the occurrence probability of each of the R candidate illegal driving behaviors to obtain R occurrence probability values; The candidate driving violation behavior with the highest occurrence probability value among the R candidate driving violation behaviors is determined as the target driving violation behavior.

8. The method for predicting illegal driving behavior according to claim 6, characterized in that: The screening of the R candidate illegal driving behaviors to obtain the target illegal driving behavior includes: Sorting the R candidate illegal driving behaviors based on a reference priority rule to obtain a sorting result; the reference priority rule is determined based on a preceding and following logic or a harm degree logic between the R candidate illegal driving behaviors; The first X candidate illegal driving behaviors in the sorting result are determined as the target illegal driving behaviors.

9. A device for predicting illegal driving behavior, characterized in that: The device is deployed on vehicle equipment, and the device includes: a determining unit, configured to, when a first illegal driving behavior of a target user in the vehicle device is detected, determine R target causal relationships with the first illegal driving behavior as a cause item in at least one reference causal relationship; the at least one reference causal relationship is determined based on the causal relationship between P historical illegal driving behaviors of the target user; R and P are positive integers; A prediction unit, configured to predict a target illegal driving behavior of the target user at a next moment based on the R result items in the R target causal relationships; The output unit is used to output target prompt information, where the target prompt information is used to prompt the target user to avoid the target illegal driving behavior.

10. An electronic device comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, and the processor implements the method according to any one of claims 1 to 8 when executing the program.

Citation Information

Cited By

  • Urban traffic violation recognition and voice notification method, system and device and medium

    CN120014845A

  • Vehicle violation supervision method, system and device based on mac address and storage medium

    CN122715414A