Intelligent bad driving behavior identification method and system

Abstract

The invention relates to an intelligent bad driving behavior identification method and system. A data acquisition device acquires vehicle state information during a vehicle operation period; the artificial intelligence voice conversation device sends a voice early warning signal to a driver; playing a preset question to the driver and receiving an answer; the image acquisition device acquires actions made by the driver aiming at the problem; the behavior feedback receiving device receives answers and actions; the information processing device performs preliminary identification on bad driving behaviors according to the vehicle state information acquired by the data acquisition device; obtaining an answer of the driver for a preset question collected by the behavior feedback receiving device, and comparing the answer with a standard answer to confirm that the preliminary identification result of the bad driving behavior is valid; the action information, collected by the behavior feedback receiving device, of the driver is obtained, and the driver is reminded to correct bad driving behaviors by comparing the action information with historical action data and vehicle state information.

Description

Technical Field

[0001] The present invention relates to the field of intelligent driving technology, and in particular to a method and system for intelligently identifying bad driving behavior. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] When a driver exhibits bad driving behavior, his ability to control the vehicle decreases, and it is difficult for him to make timely judgments and responses to road driving risks, which can easily lead to traffic accidents. Some existing technologies use various sensors to identify the driver's status and then issue reminders for bad driving behavior. However, different drivers have different behaviors when driving vehicles. The current identification method is prone to misjudgment during the identification of bad behavior in different scenarios and different drivers, which greatly limits the scope of application of the system. Summary of the invention

[0004] In order to solve the technical problems existing in the above-mentioned background technology, the present invention provides an intelligent identification method and system for bad driving behavior. After using the vehicle status information to determine the preliminary identification result of the bad driving behavior, the driver is asked questions using a question bank to determine whether the driver has reliable driving ability and judgment ability at the moment, and to determine whether the driver at the moment has problems such as "fatigue driving" and "distracted driving" that are not vehicle device changes, thereby eliminating possible misjudgments in the preliminary identification results made by the vehicle status information, so that the system can cope with the different habits of different drivers.

[0005] In order to achieve the above object, the present invention adopts the following technical solution:

[0006] A first aspect of the present invention provides an intelligent identification system for bad driving behavior, comprising:

[0007] A data acquisition device, used to obtain vehicle status information during vehicle operation and send it to an information processing device;

[0008] The artificial intelligence voice dialogue device is used to send a voice warning signal to the driver; play preset questions to the driver based on the question bank and receive the driver's answers to the questions; send the played questions and the received answers to the behavior feedback receiving device;

[0009] An image acquisition device, used to obtain the driver's actions in response to the preset questions and send them to the behavior feedback receiving device;

[0010] A behavior feedback receiving device receives the driver's answer collected by the artificial intelligence voice dialogue device and the driver's action information collected by the image acquisition device, and sends it to the information processing device;

[0011] The information processing device is configured to: preliminarily identify bad driving behavior based on the vehicle status information acquired by the data acquisition device;

[0012] The information processing device is further configured to: obtain the driver's answers to preset questions collected by the behavior feedback receiving device according to the preliminary identification results, and confirm that the preliminary identification results of bad driving behavior are valid by comparing with the standard answers;

[0013] The information processing device is also configured to: obtain the driver's action information collected from the behavior feedback receiving device, compare it with the historical action data and vehicle status information, and use the artificial intelligence voice dialogue device to remind the driver to correct bad driving behavior.

[0014] Furthermore, the vehicle status information includes position, speed, acceleration, steering wheel angle, lane information and turn signal status.

[0015] Furthermore, the data acquisition device includes a differential GPS, an inertial sensor, a vehicle speed sensor, an acceleration rate sensor, a steering wheel angle sensor, a lane detector, and a turn signal activation state detection device.

[0016] Furthermore, the data acquisition device acquires the position information of the vehicle during driving based on differential GPS and inertial sensors.

[0017] Furthermore, the data acquisition device determines the road grade type through the map data, and determines the speed limit information of the road where the vehicle is currently located based on the lane information obtained by the lane detector and the position information.

[0018] Furthermore, preliminary identification of bad driving behavior is performed based on the vehicle status information acquired by the data acquisition device, including at least:

[0019] If the vehicle speed exceeds the speed limit threshold of the corresponding road within the set duration, the bad driving behavior is "speeding";

[0020] If the acceleration change rate exceeds the acceleration change rate threshold, the bad driving behavior is "sudden braking or sudden acceleration";

[0021] When the steering wheel angle exceeds the set threshold, if the turn signal status is off or the steering direction is inconsistent with the direction indicated by the turn signal, the bad driving behavior is "unreasonable use of turn signal";

[0022] When a vehicle is driving in the same lane, if the lateral distance between the vehicle and the lane line is less than a set threshold, the bad driving behavior is determined to be "lane deviation".

[0023] Furthermore, after obtaining the preliminary identification results of bad driving behavior, an artificial intelligence voice dialogue device is used to send a voice warning signal to the driver. After a set time or a set number of times, preset questions are selected from the question bank and played to the driver, and the driver's answers to the questions are received and sent to the behavior feedback receiving device.

[0024] Furthermore, the answers of the driver to the preset questions collected by the behavior feedback receiving device are obtained according to the preliminary identification results, and the preliminary identification results of the bad driving behavior are confirmed to be valid by comparing with the standard answers, specifically:

[0025] Obtain and recognize the driver's answer. If no answer is recognized within the set time, it is judged as "no response"; when the similarity between the answer content and the matching result of the positive or negative standard answer is greater than the set value, it is judged as a positive or negative answer;

[0026] The answer includes voice answer and action answer. If the driver gives an affirmative answer, affirmative action or no response, the preliminary identification result of bad driving behavior is valid and the driver has bad driving behavior. If the driver makes a negative action and voice answer, it is a misjudgment, that is, the driver does not have bad driving behavior.

[0027] Furthermore, the action answer is specifically as follows: the actions taken by the driver in response to the preset questions obtained by the image acquisition device are converted into motion behavior trajectories, and the recognized actions are converted into instructions or feedback by performing similarity comparison with the action trajectory database in the information processing device. At the same time, the action data are grouped into different action modes and associated with instructions or feedback, so as to convert the driver's actions into a single recognizable instruction.

[0028] A second aspect of the present invention provides an intelligent identification method for bad driving behavior, comprising the following steps:

[0029] Obtain vehicle status information during vehicle operation and make preliminary identification of bad driving behavior based on the vehicle status information;

[0030] The artificial intelligence voice dialogue device is activated based on the preliminary recognition results, and questions are played to the driver based on the question bank;

[0031] The artificial intelligence voice dialogue device is used to obtain the driver's voice answer to the question, and the image acquisition device is used to obtain the driver's action answer to the question;

[0032] By comparing with the standard answer, the preliminary identification results of bad driving behavior are confirmed to be valid;

[0033] Based on the comparison between the driver's action response, historical action data and vehicle status information, an artificial intelligence voice dialogue device is used to remind the driver to correct bad driving behavior.

[0034] Compared with the prior art, one or more of the above technical solutions have the following beneficial effects:

[0035] 1. Considering that different drivers have their own driving habits, when the vehicle exhibits certain conditions, it is not necessarily caused by the driver's bad driving behavior. Therefore, after using the vehicle status information to determine the preliminary identification results of bad driving behavior, the driver is asked questions using the question bank to determine whether the driver has reliable driving ability and judgment ability at the moment, and to determine whether the driver at the moment has problems such as "fatigue driving" and "distracted driving" that are not caused by vehicle device changes, thereby eliminating possible misjudgments in the preliminary identification results made by the vehicle status information, so that the system can cope with the different habits of different drivers.

[0036] 2. By continuously collecting the action data of the driver when answering questions, a historical action database can be formed. When a question is asked, the action made by the driver in response to the question is compared with the historical action data and the current state of the vehicle to further determine whether the driver has corrected the bad driving behavior. If not, the information processing device triggers the artificial intelligence voice dialogue device again to issue an in-car voice alarm to remind the driver to correct the bad driving behavior. Since it is difficult to directly obtain the driving habits of different drivers, the use of historical action data can obtain different response actions made by different drivers when responding to the same question. By accumulating historical action data in this way, the answers to the questions can be updated regularly or irregularly, thereby improving the accuracy of system recognition and avoiding misjudgment when the system recognizes actions. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0038] Figure 1 A schematic diagram of the architecture of an intelligent identification system for bad driving behavior provided by one or more embodiments of the present invention;

[0039] Figure 2 A schematic diagram of the intelligent identification process of bad driving behavior provided by one or more embodiments of the present invention. DETAILED DESCRIPTION

[0040] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0041] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0042] Embodiment 1:

[0043] like Figure 1 As shown, the intelligent identification system for bad driving behavior provided in this embodiment includes a data acquisition device, an artificial intelligence voice dialogue device, an image acquisition device, a behavior feedback receiving device, and an information processing device.

[0044] A data acquisition device, used to obtain vehicle status information during vehicle operation and send it to an information processing device;

[0045] The artificial intelligence voice dialogue device is used to send a voice warning signal to the driver; play preset questions to the driver based on the question bank and receive the driver's answers to the questions; send the played questions and the received answers to the behavior feedback receiving device;

[0046] An image acquisition device, used to obtain the driver's actions in response to the preset questions and send them to the behavior feedback receiving device;

[0047] A behavior feedback receiving device receives the driver's answer collected by the artificial intelligence voice dialogue device and the driver's action information collected by the image acquisition device, and sends it to the information processing device;

[0048] The information processing device is configured to: preliminarily identify bad driving behavior based on the vehicle status information acquired by the data acquisition device;

[0049] The information processing device is further configured to: obtain the driver's answers to preset questions collected by the behavior feedback receiving device according to the preliminary identification results, and confirm that the preliminary identification results of bad driving behavior are valid by comparing with the standard answers;

[0050] The information processing device is also configured to: obtain the driver's action information collected from the behavior feedback receiving device, compare it with the historical action data and vehicle status information, and use the artificial intelligence voice dialogue device to remind the driver to correct bad driving behavior.

[0051] The data acquisition device includes a differential GPS and an inertial sensor (IMU), a vehicle speed sensor, an acceleration rate sensor, a steering wheel angle sensor, a lane detector, and a turn signal activation status detection device.

[0052] The artificial intelligence voice dialogue device includes a language recognition processing module, a speaker and a microphone. The language recognition processing module is responsible for voice recognition processing and semantic analysis, the speaker is responsible for outputting voice signals, and the microphone is responsible for capturing the user's voice input.

[0053] The image acquisition device includes multiple cameras to capture the driver's movements from different angles.

[0054] The behavior feedback receiving device receives data transmitted by the driving behavior data acquisition device, the artificial intelligence voice dialogue device and the image acquisition device.

[0055] The data acquisition device acquires the position information of the vehicle during driving based on differential GPS and IMU; and the turn signal activation state detection device detects the vehicle turn signal state signal and the direction indicated by the turn signal.

[0056] like Figure 2 As shown, the working process of the identification system is as follows:

[0057] Based on the vehicle speed sensor, acceleration rate sensor, steering wheel angle sensor, lane detector and turn signal activation status signal, the vehicle's speed, acceleration rate, steering wheel angle, lane position, and turn signal activation status data are respectively acquired, and these data are integrated and transmitted to the information processing device.

[0058] The information processing device receives data from the data acquisition device, accurately locates the real-time position of the vehicle through map data, obtains road speed limit information based on the vehicle position and road grade type, and preliminarily identifies bad driving behavior and its type through the information processing device.

[0059] In this embodiment, preliminary identification of bad driving behaviors and their types is performed, specifically:

[0060] During driving, if the vehicle speed is detected to be greater than the road speed limit threshold for 3 seconds, the bad driving behavior of the vehicle is determined to be "speeding";

[0061] During driving, if it is detected that the acceleration change rate of the vehicle is greater than the acceleration change rate threshold, it is determined that the bad driving behavior of the vehicle is "sudden braking or sudden acceleration";

[0062] During driving, if the steering wheel angle of the vehicle exceeds the specified threshold, and the turn signal status is detected to be off or the actual turn is inconsistent with the turn indicated by the turn signal, the bad driving behavior of the vehicle is determined to be "unreasonable use of turn signals";

[0063] When a vehicle is driving in the same lane, if it is detected that the lateral distance between the vehicle and the lane line is less than a threshold, it is determined that the bad driving behavior of the vehicle is "lane deviation".

[0064] This embodiment takes common bad driving behaviors such as speeding, sudden acceleration and deceleration as examples, and more types of bad driving behaviors can be determined according to actual vehicle types and road conditions.

[0065] After the information processing device preliminarily identifies that the driver may be in a bad driving state, the speaker of the artificial intelligence voice dialogue device is activated to remind the driver of possible bad driving behavior types through voice alarm. For example, if the driver is speeding, the speaker broadcasts a voice alarm of "Please note that you are speeding."

[0066] After the voice broadcast continues for three times or for a set period of time, the information processing device randomly selects questions from the corresponding question bank based on the type of bad driving behavior, and asks the driver about the driving situation through voice dialogue to improve the accuracy of the system's judgment.

[0067] The behavior feedback receiving device receives the data transmitted by the data acquisition device, the artificial intelligence voice dialogue device and the image acquisition device during the inquiry period, and transmits it to the information processing device.

[0068] The information processing device receives the data transmitted by the behavior feedback receiving device, and further determines the correctness of the bad driving behavior initially identified by the system based on the data.

[0069] If the data results transmitted by the data acquisition device show that the driver still has bad driving behavior, it will be compared with the historical driver behavior and status judgment result data to further determine whether the driver has corrected the bad driving behavior. If not, the information processing device will trigger the speaker of the artificial intelligence voice dialogue device again to issue an in-car voice alarm to remind the driver to correct the bad driving behavior.

[0070] The information processing device is pre-set with questions and corresponding answers for the types of bad driving behaviors, and forms a question and answer question bank. The questions corresponding to each type of bad driving behavior include two types of questions: "dialogue answer" and "action recognition". Specifically, the dialogue answer presets the content of the question for non-vehicle device changes such as "fatigue driving" and "distracted driving", and the answer is set as a simple affirmative or negative word (such as "yes" or "no"); the action recognition will select questions based on the type of bad driving behavior that has been preliminarily judged, and the answer is set as a simple shake of the head, nod, and other simple actions. When asking questions, the artificial intelligence voice device will guide the driver on how to answer or act.

[0071] Based on the preliminary identification type of bad driving behavior, the information processing device randomly selects 2-3 questions from the preset question bank, including two types of questions, triggers the artificial intelligence voice dialogue device, and inquires the driver. In the "dialogue answer" question type, the language recognition processing module in the artificial intelligence voice dialogue device converts the driver's answer content into a text record through text-to-speech conversion technology, and uses the text model established by the BERT (Bidirectional Encoder Representations from Transformers) algorithm to recognize the driver's answer content. If the driver fails to make a system-recognizable voice answer within a fixed time, it is judged as "no response". When the similarity between the driver's answer content and the positive or negative answer match result is greater than 70%, it is considered a positive or negative answer.

[0072] In terms of driver action collection, the image acquisition device uses one or more cameras to capture the driver's actions from different angles. Through image recognition and computer vision analysis technology, the driver's action trajectory is identified, and the similarity is compared with the action trajectory database in the information processing device. The identified actions are converted into instructions or feedback, and the action data is grouped into different action modes through clustering algorithms (such as Density-Based Spatial Clustering of Applications with Noise, DBSCAN), and the driver's actions are converted into a single recognizable instruction, such as simple actions such as "nodding" and "shaking head", which are transmitted to the information processing device. If the driver fails to make a system-recognizable action response within a fixed time, it is judged as "no response".

[0073] The information processing device receives the data transmitted by the data acquisition device, the artificial intelligence voice dialogue device and the image acquisition device transmitted by the behavior feedback receiving device, and makes a comprehensive judgment. In all the question answering links, if the driver makes a positive answer or action or no answer, it is judged that the driver does have bad driving behavior. If the driver makes a negative action and voice answer, it means that the driver does not have bad driving behavior, which is a system misjudgment.

[0074] In particular, the information processing device will form a piece of bad driving behavior data based on the preliminary identification data of bad driving behavior collected by the data acquisition device, the driver's answer to the system inquiry, including voice answer and body movement feedback, and the final judgment result of bad driving behavior, and store it in the database of the information processing device. Establish an automated regular update mechanism, based on the historical bad driving behavior judgment results, and regularly update the preset fixed answer library and bad driving behavior database for individual driving habits, and update the bad driving behavior judgment threshold and BERT text model through machine learning algorithms to improve the model's understanding and recognition capabilities of individual driving behaviors.

[0075] The system integrates multiple data sources to obtain multi-dimensional information about the driver, including voice, body movements, and vehicle operating status, and comprehensively evaluates the driver's vehicle operation status at multiple levels to accurately identify bad driving behaviors. At the same time, it warns the driver of bad driving behaviors, assists the driver to correct bad driving conditions in a timely manner, and further improves the level of safe driving.

[0076] Taking into account that different drivers have their own driving habits, when the vehicle exhibits certain conditions, it is not necessarily caused by the driver's bad driving behavior. Therefore, after using the vehicle status information to determine the preliminary identification results of bad driving behavior, the driver is asked questions using the question bank to determine whether the driver has reliable driving ability and judgment ability at the moment, and to determine whether the driver at the moment has problems such as "fatigue driving" and "distracted driving" that are not vehicle device changes, thereby eliminating possible misjudgments in the preliminary identification results made by the vehicle status information, so that the system can cope with the different habits of different drivers.

[0077] By continuously collecting the action data of the driver when answering questions, a historical action database is formed. When a question is asked, the action made by the driver in response to the question is compared with the historical action data and the current state of the vehicle to further determine whether the driver has corrected the bad driving behavior. If not, the information processing device triggers the artificial intelligence voice dialogue device again to issue an in-car voice alarm to remind the driver to correct the bad driving behavior. Since it is difficult to directly obtain the driving habits of different drivers, the use of historical action data can obtain different response actions made by different drivers when responding to the same question. By accumulating historical action data in this way, the answers to the questions can be updated regularly or irregularly, thereby improving the accuracy of system recognition and avoiding misjudgment when the system recognizes actions.

[0078] Embodiment 2:

[0079] The method for intelligently identifying bad driving behavior provided in this embodiment includes the following steps:

[0080] Obtain vehicle status information during vehicle operation and make preliminary identification of bad driving behavior based on the vehicle status information;

[0081] The artificial intelligence voice dialogue device is activated based on the preliminary recognition results, and questions are played to the driver based on the question bank;

[0082] The artificial intelligence voice dialogue device is used to obtain the driver's voice answer to the question, and the image acquisition device is used to obtain the driver's action answer to the question;

[0083] By comparing with the standard answer, the preliminary identification results of bad driving behavior are confirmed to be valid;

[0084] Based on the comparison between the driver's action response, historical action data and vehicle status information, an artificial intelligence voice dialogue device is used to remind the driver to correct bad driving behavior.

[0085] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An intelligent identification system for bad driving behavior, characterized in that: include: A data acquisition device, used to obtain vehicle status information during vehicle operation and send it to an information processing device; Artificial intelligence voice dialogue device, used to send voice warning signals to the driver; Playing preset questions to the driver based on the question bank and receiving the driver's answers to the questions; sending the played questions and received answers to a behavior feedback receiving device; An image acquisition device, used to obtain the driver's actions in response to the preset questions and send them to the behavior feedback receiving device; A behavior feedback receiving device receives the driver's answer collected by the artificial intelligence voice dialogue device and the driver's action information collected by the image acquisition device, and sends it to the information processing device; The information processing device is configured to: preliminarily identify bad driving behavior based on the vehicle status information acquired by the data acquisition device; The information processing device is further configured to: obtain the driver's answers to preset questions collected by the behavior feedback receiving device according to the preliminary identification results, and confirm that the preliminary identification results of bad driving behavior are valid by comparing with the standard answers; The information processing device is also configured to: obtain the driver's action information collected from the behavior feedback receiving device, compare it with the historical action data and vehicle status information, and use the artificial intelligence voice dialogue device to remind the driver to correct bad driving behavior.

2. The intelligent identification system for bad driving behavior according to claim 1, characterized in that: The vehicle status information includes position, speed, acceleration, steering wheel angle, lane information and turn signal status.

3. The intelligent identification system for bad driving behavior according to claim 1, characterized in that: The data acquisition device includes a differential GPS, an inertial sensor, a vehicle speed sensor, an acceleration rate sensor, a steering wheel angle sensor, a lane detector, and a turn signal activation state detection device.

4. The intelligent identification system for bad driving behavior according to claim 1, characterized in that: The data acquisition device acquires the position information of the vehicle during driving based on differential GPS and inertial sensors.

5. The intelligent identification system for bad driving behavior according to claim 1, characterized in that: The data acquisition device determines the road grade type through map data, and determines the speed limit information of the road where the vehicle is currently located based on the lane information obtained by the lane detector in combination with the position information.

6. The intelligent identification system for bad driving behavior according to claim 1, characterized in that: Based on the vehicle status information obtained by the data acquisition device, preliminary identification of bad driving behavior is carried out, including at least: If the vehicle speed exceeds the speed limit threshold of the corresponding road within the set duration, the bad driving behavior is "speeding"; If the acceleration change rate exceeds the acceleration change rate threshold, the bad driving behavior is "sudden braking or sudden acceleration"; When the steering wheel angle exceeds the set threshold, if the turn signal status is off or the steering direction is inconsistent with the direction indicated by the turn signal, the bad driving behavior is "unreasonable use of turn signal"; When a vehicle is driving in the same lane, if the lateral distance between the vehicle and the lane line is less than a set threshold, the bad driving behavior is determined to be "lane deviation".

7. The intelligent identification system for bad driving behavior according to claim 1, characterized in that: After obtaining the preliminary identification results of bad driving behavior, the artificial intelligence voice dialogue device will send a voice warning signal to the driver. After a set time or a set number of times, preset questions will be selected from the question bank and played to the driver. The driver's answers to the questions will be received and sent to the behavior feedback receiving device.

8. The intelligent identification system for bad driving behavior according to claim 1, characterized in that: The driver's answers to the preset questions collected by the behavior feedback receiving device are obtained according to the preliminary identification results. By comparing with the standard answers, it is confirmed that the preliminary identification results of bad driving behavior are valid, specifically: Obtain and recognize the driver's answer. If no answer is recognized within the set time, it is judged as "no response"; when the similarity between the answer content and the matching result of the positive or negative standard answer is greater than the set value, it is judged as a positive or negative answer; The answer includes voice answer and action answer. If the driver gives an affirmative answer, an affirmative action or does not answer, the preliminary identification result of bad driving behavior is valid and the driver has bad driving behavior. If the driver makes a negative action and voice response, it is a misjudgment, which means that the driver does not have bad driving behavior.

9. The intelligent identification system for bad driving behavior as claimed in claim 8, characterized in that: The action answer is specifically as follows: the actions taken by the driver in response to preset questions obtained by the image acquisition device are converted into motion behavior trajectories, and the recognized actions are converted into instructions or feedback by performing similarity comparison with the action trajectory database in the information processing device. At the same time, the action data are grouped into different action modes and associated with instructions or feedback, so that the driver's actions are converted into a single recognizable instruction.

10. A method for intelligently identifying bad driving behavior based on the system according to any one of claims 1 to 9, characterized in that: The following steps are involved: Obtain vehicle status information during vehicle operation and make preliminary identification of bad driving behavior based on the vehicle status information; The artificial intelligence voice dialogue device is activated based on the preliminary recognition results, and questions are played to the driver based on the question bank; The artificial intelligence voice dialogue device is used to obtain the driver's voice answer to the question, and the image acquisition device is used to obtain the driver's action answer to the question; By comparing with the standard answer, the preliminary identification results of bad driving behavior are confirmed to be valid; Based on the comparison between the driver's action response, historical action data and vehicle status information, an artificial intelligence voice dialogue device is used to remind the driver to correct bad driving behavior.

Images