A driver emotional risk early warning intervention system

Abstract

The application discloses a driver emotion risk early warning intervention system and relates to the technical field of driving emotion intervention.The application analyzes historical emotion monitoring data of different drivers in different interactive scenes, constructs emotion parameter sets of the drivers in the interactive scenes, including emotion risk factors, a characteristic parameter-emotion type mapping table and a risk intervention reference table, detects the drivers and the interactive scenes that are driving, identifies and intervenes in the emotions of the drivers, analyzes emotion danger by combining different interactive scenes and driver individual differences, realizes dynamic evaluation of the risk of the emotions of the drivers, can effectively distinguish the actual danger of the emotions in specific scenes from the emotions themselves, reduces the mechanization of driver emotion identification and the probability of misjudgment, simultaneously targets the intervention of the risk in different interactive scenes, and improves the accuracy, effectiveness and applicability of early warning intervention.

Description

Technical Field

[0001] This invention relates to the field of driving emotion intervention technology, specifically to a driver emotion risk early warning and intervention system. Background Technology

[0002] Negative emotions are one of the core contributing factors to traffic accidents. A driver's emotional state directly affects driving behavior. By recognizing, issuing warnings, and intervening in driver emotions, the incidence of emotion-related accidents can be reduced, ensuring the safety of drivers' lives and property and minimizing social losses.

[0003] Existing technologies, such as the AI-based intelligent neural network-based voice user emotion recognition method and system disclosed in application CN119028379A, relate to the field of voice recognition technology. The method includes steps such as data acquisition, first modeling, first training, and emotion recognition. The system includes modules such as a data acquisition module, a first modeling module, a first training module, and an emotion recognition module. This application trains the emotion recognition model based on historical voice data and historical driving data, which can improve the accuracy of recognizing driver emotions.

[0004] Existing technology, such as the driving control method, device, equipment, and storage medium based on emotion recognition disclosed in application CN115626167A, includes: in response to detecting the start of a target vehicle, performing emotion recognition on the driver in the target vehicle based on a preset frequency to obtain the driver's emotion type; if the driver's emotion type is a target emotion type, calculating the target emotion intensity level of the driver; obtaining a safe driving strategy corresponding to the target emotion intensity level; and performing driving control on the target vehicle according to the safe driving strategy. This invention can proactively sense the driver's emotions and actively switch the vehicle's driving control mode when the driver is in an unstable emotional state (i.e., an overreaction state, such as excessive joy, excessive anger, excessive sadness, etc.). Therefore, it can solve the safety hazard problem of existing driving control methods when the driver is in an unstable emotional state.

[0005] The above solutions have at least the following shortcomings: 1. While the solutions disclose the driver's emotion recognition and model training process, different in-vehicle interaction scenarios have varying impacts on the driver's emotions. The solutions only train the driver's emotion recognition model without considering different interaction scenarios and individual driver differences to analyze the danger of emotions. Emotions themselves are not inherently dangerous; their danger depends on the interaction scenario's requirements for driver focus and the actual degree of interference the emotion causes with driving behavior. Without analyzing the interaction scenario, safe emotions may be misjudged as risks, or high-risk emotions may be ignored. Single emotion detection essentially identifies emotions in isolation, rather than dynamically assessing risk, failing to distinguish between the emotion itself and its actual danger in a specific scenario, leading to a significant decrease in the accuracy, effectiveness, and applicability of subsequent warning interventions.

[0006] 2. The facial and vocal features of drivers at high risk vary in different interaction scenarios, resulting in differences in the intensity of physiological reactions and the degree of constraint on emotional expression. Ultimately, this leads to different manifestations of facial and vocal features during high-risk driving. However, the above-mentioned solution sets specific thresholds for facial and vocal features for different drivers in different interaction scenarios, which makes driver emotion recognition mechanical, increases the probability of misjudgment, and reduces the accuracy of emotion recognition and intervention. Summary of the Invention

[0007] To address the aforementioned technical shortcomings, the present invention aims to provide a driver emotional risk early warning and intervention system.

[0008] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The present invention provides a driver emotion risk early warning and intervention system, including: a scenario setting module, used to acquire vehicle historical emotion monitoring data, classify the vehicle historical emotion monitoring data, acquire the historical emotion monitoring data of each driver in each interaction scenario, and set the emotion parameter set of each driver in each interaction scenario.

[0009] The emotion recognition module is used to identify the driver and interaction scenario of the current vehicle, and to denote the driver and interaction scenario as the target driver and target interaction scenario, respectively. It performs emotion recognition monitoring on the target driver, obtains the emotion recognition monitoring dataset of the target driver, and analyzes the emotion type and risk level of the target driver based on the emotion parameter set of the target driver in the target interaction scenario.

[0010] The risk intervention module is used to determine whether the target driver needs emotional risk intervention. If so, it analyzes the driver's risk intervention plan based on the target driver's emotional parameter set in the target interaction scenario and executes it. At the same time, it performs emotion recognition monitoring on the target driver after risk intervention, obtains the risk intervention effect of the target driver, and provides corresponding feedback.

[0011] The beneficial effects of this invention are as follows: This invention provides a driver emotion risk early warning and intervention system. By analyzing historical emotion monitoring data of different drivers in different interaction scenarios, it constructs an emotion parameter set for each driver in each interaction scenario, including emotion risk factors, feature parameters, emotion type mapping tables, and risk intervention reference tables. Then, it detects the driver and the interaction scenario while driving, and then performs targeted identification and risk intervention for the driver's emotions. By combining different interaction scenarios and individual driver differences to analyze the emotional danger, it realizes dynamic risk assessment of driver emotions, effectively distinguishes between the emotion itself and the actual danger of the emotion in a specific scenario, reduces the mechanization of driver emotion recognition, and lowers the probability of misjudgment. At the same time, it provides targeted intervention for risks in different interaction scenarios, improving the accuracy, effectiveness, and applicability of early warning and intervention. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a schematic diagram of the system structure connection of the present invention. Detailed Implementation

[0014] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0015] See Figure 1 As shown, a driver emotion risk early warning and intervention system includes the following modules: a scenario setting module, an emotion recognition module, a risk intervention module, and a database.

[0016] The scenario setting module is used to acquire historical vehicle emotion monitoring data, classify the historical vehicle emotion monitoring data, acquire historical emotion monitoring data of each driver in each interaction scenario, and set the emotion parameter set of each driver in each interaction scenario.

[0017] In one specific embodiment, the scene setting module includes a data classification unit and a data analysis unit.

[0018] The data classification unit is used to obtain historical vehicle emotion monitoring data from the database, classify the historical vehicle emotion monitoring data based on the driver and interaction status, and obtain historical emotion monitoring data of each driver in each interaction scenario. The historical emotion monitoring data includes the emotion type, facial feature parameters, voice feature parameters, risk intervention plan and feature persistence parameters of each emotion risk intervention.

[0019] It should be noted that the emotion types include anger, panic, and excitement.

[0020] Facial feature parameters include eyelid opening and closing degree and mouth corner curvature, etc.

[0021] Voice characteristic parameters include fundamental frequency, volume, loudness, and speech rate.

[0022] The risk intervention plan includes multiple intervention measures, such as voice reminders to the driver, slight vibrations in the steering wheel, signal interruption, and voice reminders to passengers.

[0023] The persistent feature parameters are the facial feature parameters and voice feature parameters after the risk intervention plan is implemented.

[0024] Preferably, the process of classifying vehicle historical emotion monitoring data based on driver and interaction status is as follows: The driver, interaction status, emotion type, facial feature parameters, voice feature parameters, risk intervention plan, and feature persistence parameters corresponding to each emotion risk intervention are obtained from the vehicle historical emotion monitoring data. First, the data is classified according to the driver. Then, each driver's emotion risk intervention is classified according to the interaction status, resulting in the emotion type, facial feature parameters, voice feature parameters, risk intervention plan, and feature persistence parameters for each driver's emotion risk intervention in each interaction status.

[0025] It should be noted that the interaction status includes communication with passengers, communication with the outside world, and communication with electronic devices. The interaction status can be obtained through visual detection using cameras inside the vehicle.

[0026] The data analysis unit is used to analyze the emotional risk factors, characteristic parameters and emotional type mapping table and risk intervention reference table of each driver in each interactive scenario based on the historical emotional monitoring data of each driver in each interactive scenario, and to serve as the emotional parameter set of each driver in each interactive scenario.

[0027] Preferably, the specific analysis process of the data analysis unit is as follows: S101, analyze the emotion type, facial feature parameters, voice feature parameters and feature persistence parameters of each driver in each interaction state for each emotional risk intervention, and obtain the emotional risk factors of each driver in each interaction scenario.

[0028] In the above, the emotion types of the same emotional risk intervention are integrated to obtain the facial feature parameters, voice feature parameters and feature persistence parameters of each driver for each emotion type in each interaction state. At the same time, the facial feature parameters and voice feature parameters of each driver for each interaction state without emotional risk intervention are obtained from the database. The median is selected as the reference value for the facial feature parameters and the voice feature parameters, and is denoted as M and S, respectively.

[0029] The facial feature parameters of each driver for each emotion type in each interaction state are denoted as follows: Where j represents the driver's ID, y represents the interaction state ID, x represents the emotion type ID, and f represents the facial feature parameter ID. j, y, x, and f are all positive integers. The calculation formula is as follows: The first emotional risk factor of the j-th driver in the y-th interaction scenario is obtained. In the formula, F and X represent the number of facial feature parameters and the number of emotion types, respectively. This represents the number of risk interventions for the j-th driver in the y-th interaction scenario for the x-th emotion type. This represents the total number of emotional risk interventions for the j-th driver in the y-th interaction scenario.

[0030] The voice feature parameters and feature duration parameters for each driver's emotional type under each interaction state are arranged according to... The calculation method is to calculate the second and third emotional risk factors of each driver in each interaction scenario, and then calculate the average of the first, second and third emotional risk factors of each driver in each interaction scenario to obtain the emotional risk factor of each driver in each interaction scenario.

[0031] S102. Analyze the emotion type, facial feature parameters, and voice feature parameters of each driver in each interaction state for each emotional risk intervention, and obtain a mapping table of feature parameters and emotion type for each driver in each interaction scenario.

[0032] Preferably, the facial feature parameters and voice feature parameters of each driver for each emotion type in each interaction state are obtained, and the facial feature parameters of each driver for each emotion type in each interaction state are displayed through a histogram. The distribution characteristics of the facial feature parameters of each driver for each emotion type in each interaction state are obtained from the histogram; wherein, the distribution characteristics include normal distribution, skewed distribution and discrete distribution.

[0033] When the distribution is normal, the range of the mean ± k × standard deviation of each facial feature parameter is taken as the range of facial feature parameters for intermediate risk. The range greater than the upper limit of this range is the range of facial feature parameters for high risk, and the range less than the lower limit of this range is the range of facial feature parameters for low risk.

[0034] Where k represents the standard deviation multiple of the deviation from the mean, the value of which is determined by the engineer based on actual needs, and no specific numerical limit is imposed here.

[0035] When the distribution is skewed, the values ​​at 50% and 90% are obtained from the histogram and used as the lower and upper limits of the facial feature parameter range for intermediate risk, respectively. The range greater than the upper limit is the range of facial feature parameters for high risk, and the range less than the lower limit is the range of facial feature parameters for low risk.

[0036] When the distribution is discrete, the facial feature parameters are clustered into three ranges. These three ranges are then sorted from largest to smallest, representing the facial feature parameter ranges for high, medium, and low risk levels, respectively. This method is used to obtain the facial feature parameter ranges corresponding to each driver's emotional type and risk level in various interaction scenarios.

[0037] Clustering technology is an existing technology and will not be elaborated upon here.

[0038] Based on the analysis method of facial feature parameter ranges corresponding to risk levels for each driver's emotional type in each interaction scenario, the voice feature parameter ranges corresponding to risk levels for each driver's emotional type in each interaction scenario are obtained. The facial feature parameter ranges and voice feature parameter ranges corresponding to risk levels for each driver's emotional type in each interaction scenario constitute a feature parameter and emotion type mapping table for each driver in each interaction scenario.

[0039] S103. Analyze the emotion type, facial feature parameters, voice feature parameters, risk intervention plan and feature persistence parameters of each driver in each interaction state for each emotional risk intervention, obtain the effect level of each risk intervention plan in each emotion type of each driver in each interaction state, and set a risk intervention reference table for each driver in each interaction scenario.

[0040] In the above, the emotion type, facial feature parameters, and voice feature parameters of each driver in each interactive state for each emotional risk intervention are compared with the feature parameters and emotion type mapping table of each driver in each interactive scenario to obtain the risk level of the emotion type of each driver in each interactive state for each emotional risk intervention. The feature persistence parameters of each driver in each interactive state for each emotional risk intervention are compared with the feature parameters and emotion type mapping table of each driver in each interactive scenario to obtain the post-intervention risk level of the emotion type of each driver in each interactive state for each emotional risk intervention.

[0041] If the risk level is medium or high, and the risk level after intervention is low, the risk intervention plan's effectiveness level is 3; if the risk level is high, and the risk level after intervention is medium, the risk intervention plan's effectiveness level is 2; if the risk level and the risk level after intervention are the same, the risk intervention plan's effectiveness level is 1; and for all other cases, the risk intervention plan's effectiveness level is -1. This method is used to obtain the effectiveness level of each driver's emotional risk intervention plan in each interaction state. The mode is then selected as the effectiveness level of each risk intervention plan. This allows us to obtain the effectiveness level of each driver's use of each risk intervention plan for each emotional type in each interaction scenario. Finally, the effectiveness levels of each driver's use of each risk intervention plan for each emotional type in each interaction scenario constitute a risk intervention reference table.

[0042] The emotion recognition module is used to identify the driver and interaction scenario of the current vehicle, and to denote the driver and interaction scenario as the target driver and target interaction scenario, respectively. It performs emotion recognition monitoring on the target driver, obtains the emotion recognition monitoring dataset of the target driver, and analyzes the emotion type and risk level of the target driver based on the emotion parameter set of the target driver in the target interaction scenario.

[0043] In one specific embodiment, the emotion recognition module includes an emotion monitoring unit and an emotion analysis unit.

[0044] The emotion monitoring unit is used to acquire the current driver and interaction scene of the vehicle using the monitoring equipment in the vehicle. The current driver and interaction scene are respectively recorded as the target driver and the target interaction scene. While the vehicle is in motion, the facial image and voice data of the target driver are acquired using the monitoring equipment in the vehicle, and image processing and voice processing are performed to acquire the facial feature parameters and voice feature parameters of the target driver as an emotion recognition monitoring dataset.

[0045] It should be noted that the driver's facial image is captured by a camera and compared with facial feature images of other drivers in a database to identify the current driver. Simultaneously, the interaction scenario of the current vehicle is visually detected using the vehicle's in-vehicle camera to obtain the interaction status. The specific detection process is existing technology and will not be elaborated upon here.

[0046] The emotion analysis unit is used to extract the set of emotion parameters of the target driver in the target interaction scenario, and to analyze the emotion type and risk level of the target driver using the emotion recognition monitoring dataset of the target driver.

[0047] Preferably, the analysis process of the target driver's emotion type is as follows: extract feature parameters and emotion type mapping table from the target driver's emotion parameter set in the target interaction scenario, compare the target driver's facial feature parameters and voice feature parameters with the target driver's emotion type mapping table in the target interaction scenario, and obtain the target driver's emotion type.

[0048] Preferably, the risk level analysis process for the target driver is as follows: the emotional risk factor of the target driver in the target interaction scenario is multiplied by the facial feature parameters and the voice feature parameters respectively to obtain the predicted changes in the facial feature parameters and the predicted changes in the voice feature parameters of the target driver in the target interaction scenario. Then, the facial feature parameters and the voice feature parameters are added to obtain the predicted facial feature parameters and the predicted voice feature parameters of the target driver in the target interaction scenario. Finally, the predicted facial feature parameters and the predicted voice feature parameters of the target driver are compared with the emotional type mapping table of the target driver in the target interaction scenario to obtain the risk level of the target driver.

[0049] The risk intervention module is used to determine whether the target driver needs emotional risk intervention. If so, it analyzes the driver's risk intervention plan based on the target driver's emotional parameter set in the target interaction scenario and executes it. At the same time, it performs emotion recognition monitoring on the target driver after risk intervention, obtains the risk intervention effect of the target driver, and provides corresponding feedback.

[0050] In one specific embodiment, the risk intervention module includes a risk assessment unit and an emotion intervention unit.

[0051] The risk assessment unit is used to determine whether the target driver needs emotional risk intervention based on the target driver's emotional type and risk level, and if so, to execute the emotional intervention unit.

[0052] Preferably, the process for determining whether the target driver needs emotional risk intervention is as follows: when the target driver's emotional type is a risk type or the risk level is medium or high risk level, it is determined that the target driver needs emotional risk intervention; if the target driver's emotional type is a safe type and the risk level is low risk level, it is determined that the target driver does not need emotional risk intervention.

[0053] It should be noted that the emotion type in the risk type and safety type is set by the engineer according to the warning requirements, and no specific restrictions are imposed here.

[0054] The emotion intervention unit is used to analyze the target driver's risk intervention plan based on the target driver's emotion parameter set, emotion type and risk level in the target interaction scenario, and execute it. After execution, it performs emotion recognition monitoring on the target driver after the risk intervention, obtains the emotion recognition monitoring dataset after intervention, analyzes the risk intervention effect on the target driver, and provides intervention feedback.

[0055] Preferably, a risk intervention reference table is extracted from the target driver's emotional parameters set in the target interaction scenario. Then, the target driver's emotional type and risk level are compared with the risk intervention reference table to obtain the optimal set of risk intervention measures as the target driver's risk intervention plan.

[0056] In the above, the target driver's emotional type is compared with the risk intervention reference table to obtain the effect level of each risk intervention plan for the target driver's corresponding emotional type in the target interaction scenario; if the target driver's risk level is high risk level, the union of each intervention measure in each risk intervention plan with a level 3 effect level is selected as the optimal risk intervention measure set; if the target driver's risk level is medium risk level, the intersection of each intervention measure in each risk intervention plan with a level 3 effect level is selected as the optimal risk intervention measure set.

[0057] It should be noted that the effectiveness level of the risk intervention plan for the target driver is obtained by analyzing the effectiveness level of each driver's use of each risk intervention plan under each emotional type in each interaction scenario. When the effectiveness level is 3, it indicates that the risk intervention is effective. When the effectiveness level is 2, it indicates that the risk intervention is moderate. Conversely, it indicates that the risk intervention is ineffective.

[0058] When the risk intervention effect is mediocre, the intersection of each intervention measure in each risk intervention plan of level 3 and level 2 effect is selected as the set of secondary risk intervention measures, which is then used as the secondary intervention plan for intervention and intervention feedback.

[0059] The database is used to store historical emotion monitoring data of vehicles, facial feature parameters and voice feature parameters of each driver in each interaction state without emotional risk intervention, and facial feature images of each driver.

[0060] The above description is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they should all fall within the protection scope of the present invention.

Claims

1. A driver emotional risk early warning and intervention system, characterized in that, Includes the following modules: The scenario setting module is used to acquire historical vehicle emotion monitoring data, classify the historical vehicle emotion monitoring data, acquire historical emotion monitoring data of each driver in each interaction scenario, and set the emotion parameter set of each driver in each interaction scenario. The scene setting module includes a data classification unit and a data analysis unit; The data classification unit is used to obtain historical vehicle emotion monitoring data from the database, classify the historical vehicle emotion monitoring data based on the driver and interaction status, and obtain historical emotion monitoring data of each driver in each interaction scenario. The historical emotion monitoring data includes the emotion type, facial feature parameters, voice feature parameters, risk intervention plan and feature persistence parameters of each emotion risk intervention. The specific process for classifying vehicle historical emotion monitoring data based on driver and interaction status is as follows: The driver, interaction state, emotion type, facial feature parameters, voice feature parameters, risk intervention plan, and feature persistence parameters corresponding to each emotion risk intervention are obtained from the vehicle's historical emotion monitoring data. First, the data is classified according to the driver. Then, each driver's emotion risk intervention is classified according to the interaction state to obtain the emotion type, facial feature parameters, voice feature parameters, risk intervention plan, and feature persistence parameters for each driver's emotion risk intervention in each interaction state. The data analysis unit is used to analyze the emotional risk factors, characteristic parameters and emotional type mapping table and risk intervention reference table of each driver in each interactive scenario based on the historical emotional monitoring data of each driver in each interactive scenario, and to serve as the emotional parameter set of each driver in each interactive scenario. The specific analysis process of the data analysis unit is as follows: S101, Analyze the emotion type, facial feature parameters, voice feature parameters and feature persistence parameters of each driver in each interaction state for each emotional risk intervention, and obtain the emotional risk factors of each driver in each interaction scenario. The same emotional risk intervention is integrated to obtain the facial feature parameters, voice feature parameters and feature persistence parameters of each driver in each emotional type under each interaction state. At the same time, the facial feature parameters and voice feature parameters of each driver in each interaction state without emotional risk intervention are obtained from the database. The median is selected as the reference value of the facial feature parameters and the voice feature parameters, and is denoted as M and S respectively. The facial feature parameters of each driver for each emotion type in each interaction state are denoted as follows: Where j represents the driver's ID, y represents the interaction state ID, x represents the emotion type ID, and f represents the facial feature parameter ID. j, y, x, and f are all positive integers. The calculation formula is as follows: The first emotional risk factor of the j-th driver in the y-th interaction scenario is obtained. In the formula, F and X represent the number of facial feature parameters and the number of emotion types, respectively. This represents the number of risk interventions for the j-th driver in the y-th interaction scenario for the x-th emotion type. This represents the total number of emotional risk interventions for the j-th driver in the y-th interaction scenario; The voice feature parameters and feature duration parameters for each driver's emotional type under each interaction state are arranged according to... The calculation method is to calculate the second and third emotional risk factors of each driver in each interaction scenario, and then calculate the average of the first, second and third emotional risk factors of each driver in each interaction scenario to obtain the emotional risk factor of each driver in each interaction scenario. S102. Analyze the emotion type, facial feature parameters and voice feature parameters of each driver in each interaction state for each emotional risk intervention, and obtain the feature parameter and emotion type mapping table of each driver in each interaction scenario. The facial feature parameters and voice feature parameters of each driver for each emotion type under each interaction state are obtained. The facial feature parameters of each driver for each emotion type under each interaction state are displayed through histograms. The distribution characteristics of the facial feature parameters of each driver for each emotion type under each interaction state are obtained from the histograms. The distribution characteristics include normal distribution, skewed distribution and discrete distribution. When the distribution is normal, the range of the mean ± k × standard deviation of each facial feature parameter is taken as the range of facial feature parameters for intermediate risk. The range greater than the upper limit of this range is the range of facial feature parameters for high risk, and the range less than the lower limit of this range is the range of facial feature parameters for low risk. When the distribution is skewed, the values ​​at 50% and 90% are obtained from the histogram and are used as the lower and upper limits of the range of facial feature parameters for intermediate risk, respectively. The range greater than the upper limit of the range is the range of facial feature parameters for high risk, and the range less than the lower limit of the range is the range of facial feature parameters for low risk. When the distribution is discrete, the facial feature parameters are clustered into three ranges. The three ranges are then sorted from largest to smallest, and then identified as the ranges of facial feature parameters for high, medium, and low risk levels. In this way, the ranges of facial feature parameters corresponding to each driver's emotional type and risk level in each interaction scenario are obtained. Based on the analysis method of the range of facial feature parameters of each driver in each interaction scenario corresponding to each risk level of each emotion type, the range of voice feature parameters of each driver in each interaction scenario corresponding to each risk level of each emotion type is obtained. The range of facial feature parameters and the range of voice feature parameters of each driver in each interaction scenario corresponding to each risk level of each emotion type constitute a mapping table of feature parameters and emotion type of each driver in each interaction scenario. S103. Analyze the emotion type, facial feature parameters, voice feature parameters, risk intervention plan and feature persistence parameters of each driver in each interaction state for each emotional risk intervention, obtain the effect level of each risk intervention plan in each emotion type of each driver in each interaction state, and set a risk intervention reference table for each driver in each interaction scenario. The emotion type, facial feature parameters, and voice feature parameters of each driver in each interactive state for each emotional risk intervention are compared with the feature parameters and emotion type mapping table of each driver in each interactive scenario to obtain the risk level of the emotion type of each driver in each interactive state for each emotional risk intervention. The feature persistence parameters of each driver in each interactive state for each emotional risk intervention are compared with the feature parameters and emotion type mapping table of each driver in each interactive scenario to obtain the post-intervention risk level of the emotion type of each driver in each interactive state for each emotional risk intervention. If the risk level is medium or high, and the risk level after intervention is low, the risk intervention plan's effectiveness level is 3; if the risk level is high, and the risk level after intervention is medium, the risk intervention plan's effectiveness level is 2; if the risk level is the same as the risk level after intervention, the risk intervention plan's effectiveness level is 1; and for situations other than the above, the risk intervention plan's effectiveness level is -1. This is used to obtain the effectiveness level of each driver's emotional risk intervention plan in each interactive state. Then, the mode is selected as the effectiveness level of each risk intervention plan. This is used to obtain the effectiveness level of each driver's use of each risk intervention plan for each emotional type in each interactive scenario. The effectiveness levels of each driver's use of each risk intervention plan for each emotional type in each interactive scenario constitute a risk intervention reference table. The emotion recognition module is used to identify the driver and interaction scenario of the current vehicle, and to record the driver and interaction scenario of the current vehicle as the target driver and the target interaction scenario, respectively. It performs emotion recognition monitoring on the target driver, obtains the emotion recognition monitoring dataset of the target driver, and analyzes the emotion type and risk level of the target driver based on the emotion parameter set of the target driver in the target interaction scenario. The risk intervention module is used to determine whether the target driver needs emotional risk intervention. If so, it analyzes the target driver's risk intervention plan based on the target driver's emotional parameter set in the target interaction scenario, and executes it. At the same time, it performs emotion recognition monitoring on the target driver after risk intervention, obtains the risk intervention effect of the target driver, and provides corresponding feedback.

2. The driver emotional risk early warning and intervention system according to claim 1, characterized in that, The emotion recognition module includes an emotion monitoring unit and an emotion analysis unit; The emotion monitoring unit is used to acquire the current driver and interaction scenario of the vehicle using the monitoring equipment in the vehicle. The current driver and interaction scenario are recorded as the target driver and target interaction scenario, respectively. While the vehicle is in motion, the facial image and voice data of the target driver are acquired using the monitoring equipment in the vehicle, and image processing and voice processing are performed to acquire the facial feature parameters and voice feature parameters of the target driver as an emotion recognition monitoring dataset. The emotion analysis unit is used to extract the set of emotion parameters of the target driver in the target interaction scenario, and to analyze the emotion type and risk level of the target driver using the emotion recognition monitoring dataset of the target driver.

3. The driver emotional risk early warning and intervention system according to claim 2, characterized in that, The analysis process for the target driver's emotion type is as follows: The target driver's emotional parameters are extracted from the set of emotional parameters in the target interaction scenario and mapped to an emotional type. The target driver's facial feature parameters and voice feature parameters are compared with the target driver's emotional type mapping table in the target interaction scenario to obtain the target driver's emotional type.

4. A driver emotional risk early warning and intervention system according to claim 2, characterized in that, The process of analyzing the risk level of the target driver is as follows: The target driver's emotional risk factor in the target interaction scenario is multiplied by the facial feature parameters and voice feature parameters respectively to obtain the predicted changes in the facial feature parameters and voice feature parameters in the target interaction scenario. Then, the facial feature parameters and voice feature parameters are added to obtain the predicted facial feature parameters and predicted voice feature parameters of the target driver in the target interaction scenario. Finally, these are compared with the target driver's emotional type mapping table in the target interaction scenario to obtain the target driver's risk level.

5. A driver emotional risk early warning and intervention system according to claim 1, characterized in that, The risk intervention module includes a risk assessment unit and an emotion intervention unit; The risk assessment unit is used to determine whether the target driver needs emotional risk intervention based on the target driver's emotional type and risk level. If so, the emotional intervention unit is executed. The emotion intervention unit is used to analyze the target driver's risk intervention plan based on the target driver's emotion parameter set, emotion type and risk level in the target interaction scenario, and execute it. After execution, it performs emotion recognition monitoring on the target driver after the risk intervention, obtains the emotion recognition monitoring dataset after intervention, analyzes the risk intervention effect on the target driver, and provides intervention feedback.

6. A driver emotional risk early warning and intervention system according to claim 1, characterized in that, The process for determining whether the target driver needs emotional risk intervention is as follows: When the target driver's emotional type is classified as risky or the risk level is medium or high, it is determined that the target driver needs emotional risk intervention. If the target driver's emotional type is classified as safe and the risk level is low, it is determined that the target driver does not need emotional risk intervention.

7. A driver emotional risk early warning and intervention system according to claim 6, characterized in that, The analysis process for the driver risk intervention plan is as follows: A risk intervention reference table is extracted from the target driver's emotional parameters in the target interaction scenario. Then, the target driver's emotional type and risk level are compared with the risk intervention reference table to obtain the optimal set of risk intervention measures as the target driver's risk intervention plan.

Images