Drunk driving and poisoning prevention method and system

Abstract

The invention relates to the technical field of vehicle safety, and discloses a drunk driving and poisoning prevention method and system.The method comprises the following steps that in-vehicle environment data and driver data are collected in real time; performing multi-sensor data fusion on the in-vehicle environment data and the driver data to obtain a comprehensive state value; based on the comprehensive state value, generating a dynamic estimation value by using an extended Kalman filtering algorithm; performing nonlinear dynamic modeling on the dynamic estimation value, and optimizing a drunk driving risk assessment result in combination with an optimal control theory; according to the drunk driving risk assessment result, the drunk driving risk of the driver is assessed in real time and fed back in real time. According to the method, the in-vehicle environment data and the driver pupil reflex data are fused, and then the dynamic estimation value is generated by using the extended Kalman filtering algorithm, so that the state of the driver can be estimated in real time by accurately combining the alcohol concentration and based on the pupil reflex and driving behavior data, and the detection precision is greatly improved; and the problems of false alarm and missing alarm caused by external environment interference are avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle safety, and in particular to a method and system for preventing drunk driving and poisoning. Background Art

[0002] As society pays more and more attention to traffic safety, especially in preventing accidents caused by drunk driving and air pollution in the car, intelligent vehicle monitoring technology has developed rapidly and played an important role in ensuring driving safety. At present, some preliminary vehicle alcohol detection systems have appeared on the market, which can monitor whether the driver is drunk driving to a certain extent.

[0003] However, existing vehicle monitoring systems still have some significant problems: first, the detection accuracy is low and is easily affected by the external environment, resulting in false alarms and missed alarms; second, these systems have a slow response speed and fail to take effective measures in an emergency; in addition, the functions are relatively simple, and most systems are limited to monitoring alcohol content, ignoring the real-time monitoring of toxic gases such as carbon monoxide in the car, resulting in an insufficient safety protection system and an inability to fully guarantee the safety of drivers and passengers. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a method and system for preventing drunk driving and poisoning, which solves the problems of low detection accuracy, susceptibility to external interference, slow response speed, and limited to alcohol concentration monitoring while ignoring real-time monitoring of other harmful gases in the car.

[0005] To achieve the above objectives, the present invention is implemented through the following technical scheme: a method for preventing drunk driving and poisoning, comprising the following steps: Collect in-vehicle environment data and driver data in real time; Perform multi-sensor data fusion on the in-vehicle environment data and driver data to obtain a comprehensive state value; Based on the comprehensive state value, the dynamic estimation value is generated using the extended Kalman filter algorithm; Conduct nonlinear dynamic modeling on dynamic estimation values ​​and optimize the risk assessment results of drunk driving by combining with optimal control theory; Based on the drunk driving risk assessment results, the driver's drunk driving risk is assessed in real time and real-time feedback is provided.

[0006] Preferably, the in-vehicle environment data includes alcohol concentration and carbon monoxide concentration, the driver data includes driver's facial physiological characteristics and driving behavior data, and the driver's facial physiological characteristics include pupil reflex data.

[0007] Preferably, the pupil reflex data is used to establish an oculomotor model, and the risk of drunk driving is determined based on the oculomotor model and driving behavior data.

[0008] Preferably, the state equation of the extended Kalman filter algorithm is: ; in, Indicates time The estimated value of the state; Indicates time based on The predicted value of the state at the moment; is the Kalman gain, which is used to balance the weight between the predicted value and the observed value; represents the actual observed data; A function that represents the nonlinear relationship between the predicted value and the observed data.

[0009] Preferably, the objective function of the optimal control theory is: ; in, is the objective function, which represents the cost or loss to be minimized in the control process; is the system state variable, indicating that at time The status of the system; is the control input, indicating that at time the controls exerted on the system; is the cost function, which represents the system state variable and control input At the moment Variable loss; is the end time of the time interval and is the total duration of the control process; To represent the cost in the entire time interval during the control process, the integral sum is performed to obtain the overall cost.

[0010] Preferably, the eye movement model detects the risk of drunk driving through nonlinear changes in pupil reflexes, and the pupil reflex change equation includes: ; in, is the pupil diameter; is the initial pupil diameter; is the amplitude of change; the reflex decay rate constant, which indicates the rate at which the pupillary response returns to its normal state; is the exponential decay function of the reflection recovery, which represents the change of pupil diameter over time.

[0011] Preferably, the real-time feedback includes warning information displayed on the vehicle display screen, voice prompts and vehicle speed limits.

[0012] Preferably, the vehicle speed limitation includes activating an emergency brake system, limiting the vehicle speed and automatically locking the vehicle.

[0013] Anti-drunk driving and anti-intoxication system, including: Multi-sensor module for collecting in-vehicle environment data and driver data in real time; The data fusion module is used to fuse the in-vehicle environment data and the driver data and output a comprehensive state estimate; The risk assessment module uses nonlinear dynamics modeling and optimal control algorithms to assess the risk of drunk driving based on comprehensive state estimation; The feedback module provides real-time feedback based on the results of the drunk driving risk assessment, including warning messages on the vehicle display, voice prompts and vehicle speed limits.

[0014] Preferably, the multi-sensor module comprises: Alcohol concentration sensor: used to monitor the alcohol concentration in the air inside the car in real time; Carbon monoxide concentration sensor: used to monitor the carbon monoxide concentration in the air inside the car in real time; Camera: used to collect driver's pupil reflection data.

[0015] The present invention provides a method and system for preventing drunk driving and poisoning. It has the following beneficial effects: 1. The present invention fuses the in-vehicle environment data and the driver's pupil reflex data, and then uses the extended Kalman filter algorithm to generate a dynamic estimate, which can accurately combine the alcohol concentration and estimate the driver's state in real time based on the pupil reflex and driving behavior data. Compared with the existing drunk driving detection system that relies on only a single sensor, the present invention greatly improves the detection accuracy and avoids the problem of false alarms and missed alarms caused by external environmental interference.

[0016] 2. The present invention uses an oculomotor model to combine pupil reflex data with driving behavior to accurately analyze whether the driver is driving under the influence of alcohol. Compared with the prior art method of only detecting alcohol concentration, the present invention solves the problem of ignoring other influencing factors by relying only on alcohol concentration, and improves the comprehensiveness and accuracy of risk determination.

[0017] 3. The present invention dynamically optimizes the drunk driving risk assessment strategy through nonlinear dynamic modeling and optimal control theory to ensure that the system always makes the best response under different environments and driving behaviors. Compared with the traditional drunk driving monitoring system, the existing system has a slow response speed and lacks flexibility. The present invention can respond quickly in an emergency and reduce the risk of potential accidents.

[0018] 4. The present invention uses voice prompts, speed limits and automatic braking functions to remind the driver and perform emergency braking when it is determined that the driver is at risk of drunk driving. This solves the problem of overly single intervention methods in the prior art, ensures timely and effective responses in different drunk driving risk situations, and improves driving safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a schematic diagram of the method flow of the present invention; Figure 2 Schematic diagram of the system architecture of the present invention. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the specification of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] In order to better understand the present invention, the above contents are described in detail below in conjunction with specific embodiments.

[0022] Please refer to the attached Figure 1 The embodiment of the present invention provides a method for preventing drunk driving and poisoning, comprising the following steps: Collect in-vehicle environment data and driver data in real time; In this embodiment, in order to achieve real-time monitoring of whether the driver has a risk of drunk driving and the risk of intoxication in the car, the present invention first collects in-car environment data and driver data through on-board sensors to ensure that the system can make accurate assessments and timely responses based on real-time data. The collection of real-time data provides the necessary basis for subsequent data fusion, dynamic estimation and risk assessment, and lays the foundation for the implementation of more precise control measures.

[0023] Specifically, the in-car environment data mainly includes alcohol concentration and carbon monoxide concentration. The alcohol concentration sensor can monitor the alcohol concentration in the air in the car in real time and transmit the data to the data processing unit. The working principle of the alcohol concentration sensor is usually based on infrared spectroscopy or semiconductor gas sensor technology, which can accurately measure the trace alcohol content in the air in the car, thereby providing a direct basis for the risk assessment of drunk driving.

[0024] In addition, the carbon monoxide concentration sensor is used to detect the concentration of carbon monoxide in the car to determine whether the driver is at risk of carbon monoxide poisoning. As a possible implementation method, the carbon monoxide concentration sensor can use an electrochemical sensor or a metal oxide semiconductor sensor, which has good sensitivity and response speed and can quickly detect the potential concentration of harmful gases in the car. Specifically, the collection of in-car environmental data is automatically monitored through these sensors to ensure the safety of the in-car environment.

[0025] In addition, driver data includes driver facial physiological features and driving behavior data. The driver's facial physiological features, especially pupil reflex data, can be collected in real time through the on-board camera. The on-board camera detects the driver's eye area through image recognition technology and then analyzes changes in the pupil. Changes in the driver's pupil reflex data can reflect whether the driver is under the influence of alcohol or other drugs. For example, under the influence of alcohol, the driver's pupils usually show dilation or change characteristics, which can be captured by the image recognition algorithm and converted into pupil reflex values ​​for further analysis.

[0026] Driving behavior data mainly involves acceleration, braking, steering wheel rotation and other behavior data. Such data is monitored in real time through sensors installed in the car (such as accelerometers, gyroscopes, steering wheel rotation sensors, etc.) to provide feedback on the driver's control status. Driving behavior data reflects the driver's driving behavior, such as frequent emergency braking and acceleration, which are potential dangerous behaviors of drunk driving or drug driving.

[0027] During the data collection phase, real-time data from all sensors will be collected and transmitted to the onboard central control unit. The control unit will respond in a timely manner based on the collected data, combined with the vehicle's status and environmental conditions. Through real-time monitoring and data collection, the system can continuously update the driver's physiological characteristics and behavior patterns to provide accurate and real-time input data for subsequent drunk driving risk assessment.

[0028] Therefore, real-time collection of in-vehicle environment and driver data can provide a reliable basis for subsequent drunk driving risk assessment and safety intervention.

[0029] Perform multi-sensor data fusion on the in-vehicle environment data and driver data to obtain a comprehensive state value; In this embodiment, multi-sensor data fusion of in-vehicle environment data and driver data is a key link to improve system accuracy and response speed. Through the collaborative work of multiple sensors, the system can obtain data from different sources and fuse them to obtain an accurate comprehensive state estimate. This comprehensive state estimate provides an accurate basis for subsequent drunk driving risk assessment and safety intervention.

[0030] Specifically, the in-vehicle environment data includes alcohol concentration and carbon monoxide concentration, which are collected in real time through corresponding gas sensors. Driver data includes the driver's facial physiological characteristics, especially pupil reflex data, and driving behavior data (such as acceleration, braking, vehicle speed, etc.). These data are monitored in real time through on-board sensors and transmitted to the control system for processing.

[0031] In order to accurately fuse the data from different sensors, the present invention uses a particle filter algorithm, which can effectively process highly nonlinear systems and is particularly suitable for processing dynamic and complex real-time data. This algorithm discretizes the state space, uses multiple particles to represent the system state, and updates the state estimate by weighted average.

[0032] Specifically, the particle filter algorithm performs data fusion and state estimation through the following steps: Initialization: At the initial moment, a set of particles is generated as a discrete representation of the state space, each particle represents a possible state. Each particle is assigned a weight, and the weights of all particles are equal at the beginning.

[0033] Prediction step: According to the dynamic model of the system and the control input (driver's operation), each particle is predicted and updated. Assume that the state transition model of the system is: ; in, Indicates Particles at time Status; Indicates Particles at time Status; is the state transfer function; Represents process noise, which is used to model the uncertainty in the system; represents the control variable.

[0034] Resampling step: In order to avoid particle degradation (i.e. only a few particles play a decisive role in the final estimation), the particle filter algorithm ensures the diversity of particles through resampling operations. The resampling process reselects particles according to their weights and generates new particles. .

[0035] State estimation: Finally, the state estimation of all particles is combined to obtain the comprehensive state of the system , which is usually the weighted average of the particles: ; in, is the number of particles, It is a particle The weight of .

[0036] Through the particle filter algorithm, the system can comprehensively consider the in-car environment (alcohol concentration, carbon monoxide concentration) and the driver's physiological characteristics and behavior (pupillary reflex, driving behavior), and calculate the comprehensive status in real time. , that is, the comprehensive state estimation value, which provides an important basis for subsequent drunk driving risk assessment, dynamic decision-making and safety intervention.

[0037] In some embodiments, particle filtering can provide stronger robustness than traditional Kalman filtering algorithms, especially in the face of system nonlinearity and highly complex environments. Since particle filtering does not rely on linear assumptions and can handle non-Gaussian distributions and nonlinear dynamic systems, particle filtering can provide more accurate results in the process of multi-sensor data fusion and dynamic estimation.

[0038] Specifically, particle filtering can effectively fuse data from different sensors, such as data from alcohol concentration sensors, data from carbon monoxide concentration sensors, pupil reflex data, and driving behavior data, and update and correct state estimates at each moment, thereby reducing errors and ensuring the accuracy and real-time nature of driver drunk driving risk assessment.

[0039] This data fusion method not only improves the accuracy of the system, but also ensures that it can respond quickly in complex and dynamic driving environments. Particle filtering provides strong support in dealing with sensor noise, missing data or time synchronization issues between sensors.

[0040] Therefore, through the multi-sensor data fusion of the particle filter algorithm, the present invention can accurately process the in-vehicle environment data and driver data to obtain a comprehensive state estimation value, providing reliable data support for drunk driving risk assessment, dynamic decision-making and safety intervention.

[0041] Based on the comprehensive state value, the dynamic estimation value is generated using the extended Kalman filter algorithm; In this embodiment, the calculation of the comprehensive state value provides an accurate data basis for subsequent dynamic estimation. By performing multi-sensor fusion on data from different sensors, the system can obtain a comprehensive state estimation value, which is then used for subsequent dynamic estimation. In order to further improve the estimation accuracy, the system adopts the extended Kalman filter algorithm (EKF), which can handle state estimation in nonlinear systems and ensure the accuracy and reliability of dynamic estimation values.

[0042] Specifically, the extended Kalman filter algorithm uses the following steps to dynamically estimate the comprehensive state value. First, the system predicts the current state and uses the state transition equation to infer the current state. Then, the system updates the predicted state and corrects the state estimate through the Kalman gain in combination with the observed data. The update formula is as follows: ; in, Indicates time The estimated value of the state; Indicates time based on The predicted value of the state at the moment; is the Kalman gain, which is used to balance the weight between the predicted value and the observed value; represents the actual observed data; A function that represents the nonlinear relationship between the predicted value and the observed data.

[0043] As an option, during implementation, the system can also adjust the accuracy of state estimation based on real-time data. For example, if the alcohol concentration or pupil reflex changes abnormally, the system will pay more attention to the weight of these data and quickly adjust the current state estimation value, thereby improving the accuracy of the drunk driving risk assessment.

[0044] In one possible implementation, in order to further improve the accuracy and reliability of the system, the extended Kalman filter algorithm can be used in conjunction with other optimization algorithms (such as unscented Kalman filter), so that under different sensor and system conditions, the system can select the most suitable algorithm for dynamic estimation based on actual conditions.

[0045] Therefore, by using the extended Kalman filter algorithm, the present invention can perform accurate state estimation in complex and dynamic environments, timely reflect changes in the in-vehicle environment and the driver's physiological state, ensure that the system can respond quickly, and timely judge the risks of drunk driving and poisoning, thereby improving traffic safety.

[0046] Conduct nonlinear dynamic modeling on dynamic estimation values ​​and optimize the risk assessment results of drunk driving by combining with optimal control theory; In this embodiment, the dynamic estimation value processed by the extended Kalman filter algorithm provides accurate input for the subsequent drunk driving risk assessment. In order to further optimize the drunk driving risk assessment results, the system accurately assesses the drunk driving risk through nonlinear dynamic modeling and optimal control theory. This process ensures that the system can respond quickly in complex driving environments and make reasonable safety decisions based on real-time data.

[0047] Specifically, the comprehensive state estimation value obtained through the aforementioned multi-sensor data fusion and extended Kalman filtering process reflects the in-vehicle environment and the physiological characteristics and behavior patterns of the driver. Then, the system dynamically models these estimation values ​​through nonlinear dynamic modeling to further improve the accuracy and robustness of the evaluation results.

[0048] Specifically, the state transition of the system is described by a nonlinear dynamic equation, which is based on the behavior of the system and the interaction of various data sources. Assume that the state transition equation of the system is: ; in, The system is at time System state variables, including information such as alcohol concentration, pupil reflex, and driver behavior; is the system state variable at time The rate of change of is a nonlinear function describing the state of the system. is the control input, which represents the driver's operation (such as acceleration, braking, etc.), is the process noise, representing the uncertainty in the system.

[0049] Through nonlinear dynamic models, the system can simulate the changes in the driver's state and behavior during drunk driving, and analyze whether the driver is in a drunk driving state, thereby providing dynamic data support for subsequent drunk driving risk assessment.

[0050] In one possible implementation, the system further combines optimal control theory to optimize the drunk driving risk assessment results. Optimal control theory minimizes the risk of drunk driving by defining an objective function and optimizing the control input. For example, in the present invention, the optimal control objective function can be expressed as: ; in, is the objective function, which represents the cost or loss to be minimized in the control process; is the system state variable, indicating that at time The status of the system; is the control input, indicating that at time the controls exerted on the system; is the cost function, which represents the system state variable and control input At the moment Variable loss; is the end time of the time interval and is the total duration of the control process; To represent the cost in the entire time interval during the control process, the integral sum is performed to obtain the overall cost.

[0051] As an option, each term in the cost function can be weighted based on real-time observations and historical data to ensure that the system can perform risk assessment more efficiently in emergency situations. For example, abnormal changes in pupil reflexes may be given higher weights to prioritize the impact of physiological state on the risk of drunk driving. In this way, the system can optimize the results of risk assessment based on the comprehensive state estimate and initiate relevant safety intervention measures in a timely manner.

[0052] Therefore, by combining nonlinear dynamic modeling with optimal control theory, the present invention not only achieves real-time assessment of drunk driving risk, but also ensures that the system can adaptively adjust the assessment strategy in complex and dynamic driving environments, thereby providing more accurate risk prediction and control response. Through this optimized assessment mechanism, the system can greatly improve the accuracy and real-time performance of drunk driving prevention, and maintain efficient and stable performance in different driving scenarios.

[0053] Based on the drunk driving risk assessment results, the driver's drunk driving risk is assessed in real time and real-time feedback is provided.

[0054] In this embodiment, in order to ensure real-time detection of drunk driving risk during driving, the system generates a comprehensive state value through multi-sensor data fusion and dynamic estimation, and then conducts real-time assessment of the driver's drunk driving risk. This process not only improves the accuracy of risk assessment, but also enables the system to respond promptly according to the driver's state changes during driving, thereby providing effective protection for drunk driving prevention.

[0055] After obtaining the comprehensive state estimation value through methods such as extended Kalman filtering, the system will evaluate the driver's risk of drunk driving in real time based on the estimation value. Specifically, the system , pupil reflection data , Carbon monoxide concentration and driving behavior data The information is comprehensively analyzed to calculate the risk assessment value of drunk driving. The risk assessment value is usually obtained by the following calculation model: ; in, It is a comprehensive assessment value that indicates the driver’s risk of driving under the influence of alcohol; is a function that performs a weighted assessment of the risk of drunk driving based on various data sources; It is the driver's pupil reflection data.

[0056] Therefore, the weights of different data sources are adjusted according to the actual situation. For example, abnormal changes in alcohol concentration and pupil reflex may have a greater impact on the risk assessment of drunk driving, so the system will increase the weights of these data sources accordingly.

[0057] Specifically, the alcohol concentration When the set safety threshold is exceeded, the system will increase the risk assessment value by increasing the weight of this parameter in the assessment. Unusual changes (e.g., abnormal pupil dilation or constriction) can also increase the risk assessment and indicate that the driver may be under the influence of alcohol or other substances. If a higher concentration is detected, it may mean that the driver is at risk of carbon monoxide poisoning, and the system will adjust the comprehensive assessment value of the driver's risk of drunk driving. , thereby enhancing the response to poisoning risks.

[0058] In one possible implementation, when the system calculates driving behavior data After that, if the value exceeds a certain preset threshold, the system will start the real-time feedback mechanism and send a warning message to the driver. The feedback content may include warning information displayed on the vehicle display, voice prompts or other visual prompts to remind the driver of the risk of drunk driving. Depending on the actual situation, the system can also initiate emergency intervention measures based on the results of the drunk driving risk assessment, such as automatically limiting the speed and initiating emergency braking.

[0059] Generally, the system will set multiple risk levels and assess the driver's risk of drunk driving based on the comprehensive assessment value. The size of the feedback intensity is adjusted. For example, if the comprehensive assessment value of the driver's risk of drunk driving is If the risk of drunk driving is high, the system may issue a more urgent warning, such as a voice prompt "The risk of drunk driving is high, please stop and check immediately", while limiting the vehicle speed to ensure the driver's safety.

[0060] Through this real-time evaluation and feedback mechanism, the system can not only provide real-time drunk driving risk judgment, but also automatically take safety intervention measures when necessary, thereby effectively preventing traffic accidents caused by drunk driving or intoxication.

[0061] As an option, in order to ensure the accuracy of the drunk driving risk assessment, the system may be dynamically adjusted according to the driver's driving history or personalized behavior pattern. For example, the system can adjust the assessment strategy according to the driver's driving style (such as frequent sudden acceleration or braking) to adapt to the driver's behavior pattern. This personalized assessment strategy enables the system to more accurately determine the risk of drunk driving.

[0062] Therefore, through real-time drunk driving risk assessment and feedback mechanism, the system can monitor and respond to the driver's drunk driving risk in a timely and effective manner, reduce the probability of accidents, and greatly improve driving safety.

[0063] Please see attached Figure 2 The embodiment of the present invention provides an anti-drunk driving and anti-poisoning system, including: Multi-sensor module for collecting in-vehicle environment data and driver data in real time; In this embodiment, the multi-sensor module in the present invention provides accurate information for subsequent drunk driving risk assessment and safety intervention by collecting in-vehicle environment data and driver data in real time. The module includes an alcohol concentration sensor, a carbon monoxide concentration sensor, a pupil reflex sensor, and a driving behavior sensor.

[0064] In this embodiment, the alcohol concentration sensor monitors the alcohol concentration in the air in the car to determine the risk of drunk driving; the carbon monoxide concentration sensor monitors the carbon monoxide concentration in the car to determine whether there is a risk of poisoning; the pupil reflex sensor monitors the driver's pupil changes through a camera to analyze whether the driver is affected by alcohol or drugs; and the speed, acceleration, braking, steering wheel rotation, etc. These data are collected through the accelerometer, gyroscope and speed sensor in the car to reflect the driver's control method. Abnormal driving behavior (such as sudden braking, excessive acceleration, etc.) may mean that the driver's judgment is affected, and these data will serve as an important reference factor for the risk assessment of drunk driving.

[0065] The data fusion module is used to fuse the in-vehicle environment data and the driver data and output a comprehensive state estimate; In this embodiment, the data fusion module is responsible for fusing the in-vehicle environment data and the driver data to generate a comprehensive state estimation value for subsequent drunk driving risk assessment and safety decision-making.

[0066] The risk assessment module uses nonlinear dynamics modeling and optimal control algorithms to assess the risk of drunk driving based on comprehensive state estimation; In this embodiment, the risk assessment module further assesses the driver's risk of drunk driving based on the comprehensive state estimation value. Through the combination of nonlinear dynamics modeling and optimal control algorithm, the system can accurately assess whether the driver has the risk of drunk driving in real time and take corresponding safety intervention measures based on the assessment results.

[0067] The feedback module provides real-time feedback based on the results of the drunk driving risk assessment, including warning messages on the vehicle display, voice prompts and vehicle speed limits.

[0068] In this embodiment, the feedback module is responsible for providing real-time feedback to the driver based on the results of the drunk driving risk assessment. The feedback module can help the driver to understand the current drunk driving risk in a timely manner and take necessary countermeasures through various forms of prompts, such as warning information displayed on the vehicle display, voice prompts and vehicle speed limits.

[0069] Specifically, the feedback module provides the following feedback methods: The car display shows a warning message When the risk assessment result of drunk driving reaches a certain threshold, the system will display a warning message in real time on the in-vehicle display screen to remind the driver of the risk of drunk driving. For example, the display screen may show warning messages such as "High risk of drunk driving, please stop and check" or "Drunk driving risk detected, please slow down". These messages ensure that the driver can obtain relevant risk information in a timely manner through clear visual prompts.

[0070] Voice prompts In addition to the on-board display, the feedback module also provides voice prompts through the car's audio system to help drivers pay attention to the risk of drunk driving while driving. Voice prompts are usually enabled when the risk of drunk driving is high, and the prompts include "Drunk driving risk detected, please slow down and stop" or "Please stop driving and check your alcohol concentration". The voice prompts are designed to ensure that the driver can quickly understand the risk level and respond.

[0071] Vehicle speed limit When the risk of drunk driving is high, the system will ensure the safety of the driver and passengers by limiting the speed. Based on the real-time risk assessment results, the system adjusts the vehicle's power system, limits the vehicle's maximum speed or forcibly initiates emergency braking. Specifically, when the risk of drunk driving is high, the system will limit the vehicle's maximum speed to prevent the driver from making dangerous driving behaviors due to drunk driving. In addition, the system can also activate the automatic braking function to quickly reduce the speed in an emergency and reduce the possibility of an accident.

[0072] As an option, when the risk of drunk driving is assessed to be high, the feedback module can also start multiple feedback modes at the same time. For example, the system can display warning information, provide voice prompts and limit the speed at the same time to ensure that the driver receives a full range of warnings. Therefore, the system of the present invention can timely deliver clear and intuitive information to the driver when the risk of drunk driving is high, and take necessary control measures, greatly improving driving safety.

[0073] In order to improve the detection accuracy of the vehicle monitoring system and reduce the false alarm rate, this technical solution can also introduce an environmental adaptive adjustment mechanism and user personalized setting options to ensure that the system can still maintain efficient and stable monitoring performance in different driving environments, and provide a customized monitoring experience based on the individual needs of the driver.

[0074] The environmental adaptive adjustment mechanism is designed to cope with the impact of changes in the in-vehicle environment on the sensor's detection accuracy. Factors such as temperature and humidity in the car may interfere with the sensor's response, especially when the temperature changes greatly, the sensor may produce false positives or miss reports. This solution monitors the in-vehicle environment in real time by equipping it with temperature and humidity sensing components, and automatically adjusts the sensor's detection threshold according to the preset rule base. For example, in a high temperature environment, the alcohol concentration sensor may have errors due to the thermal expansion of the air. The system will automatically lower the alcohol concentration alarm threshold to avoid false positives caused by excessive temperature; in a low temperature environment, the system will appropriately increase the threshold to avoid unnecessary warnings due to excessive sensitivity of the sensor.

[0075] The user personalization setting option further enhances the adaptability and flexibility of the system. Each driver has different behavioral habits and sensitivity to safety warnings, so drivers are allowed to adjust the parameters of the monitoring system according to their personal needs to meet the needs of different groups of people. For example, drivers can set different drunk driving warning thresholds through the in-vehicle touch screen or the supporting mobile application. For example, in some specific cases, users can adjust the trigger value of the alcohol concentration warning according to their personal preferences. For some drivers who are particularly sensitive to environmental changes, users can also set different safety upper limits for carbon monoxide concentrations to ensure that the system can provide the most appropriate warning strategy based on the individual differences of the driver.

[0076] To further enhance the intelligence level of the system, this technical solution also combines cloud computing platforms with remote operation and maintenance support. Using cloud big data analysis, the system can regularly upgrade the detection algorithm of the vehicle monitoring system and continuously optimize monitoring efficiency and accuracy. For example, the system will automatically adjust the detection model and rules based on the analysis of a large amount of accumulated historical data, thereby improving detection accuracy and system response speed.

[0077] On the other hand, the system also integrates a cloud service platform, allowing vehicle administrators or users to remotely view the operating status of the vehicle within the authorized scope. This includes historical driving records, vehicle health reports, and fault diagnosis suggestions. Through cloud services, administrators can promptly discover potential safety hazards, conduct remote diagnosis and maintenance, and ensure that the vehicle always maintains a good operating condition. This function not only improves the efficiency of vehicle maintenance, but also prevents safety hazards caused by equipment failure or sensor failure.

[0078] In summary, through the combination of environmental adaptive adjustment mechanism, personalized settings and cloud computing platform, this technical solution not only improves detection accuracy and response speed, but also fully considers different driving environments and driver needs, providing a more intelligent, flexible and personalized solution for the vehicle-mounted drunk driving prevention system. At the same time, the introduction of the cloud platform further improves the maintainability of the system and its reliability in long-term use, ensuring that the monitoring system can always provide stable and secure services.

[0079] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for preventing drunk driving and poisoning, characterized in that: The following steps are involved: Collect in-vehicle environment data and driver data in real time; Perform multi-sensor data fusion on the in-vehicle environment data and driver data to obtain a comprehensive state value; Based on the comprehensive state value, the dynamic estimation value is generated using the extended Kalman filter algorithm; Conduct nonlinear dynamic modeling on dynamic estimation values ​​and optimize the risk assessment results of drunk driving by combining with optimal control theory; Based on the drunk driving risk assessment results, the driver's drunk driving risk is assessed in real time and real-time feedback is provided.

2. The method for preventing drunk driving and poisoning according to claim 1, characterized in that: The in-vehicle environment data includes alcohol concentration and carbon monoxide concentration, the driver data includes driver's facial physiological characteristics and driving behavior data, and the driver's facial physiological characteristics include pupil reflex data.

3. The method for preventing drunk driving and poisoning according to claim 2, characterized in that: The pupil reflex data is used to establish an oculomotor model, and the risk of drunk driving is determined based on the oculomotor model and driving behavior data.

4. The method for preventing drunk driving and poisoning according to claim 1, characterized in that: The state equation of the extended Kalman filter algorithm is: ； in, Indicates time The estimated value of the state; Indicates time based on The predicted value of the state at the moment; is the Kalman gain, which is used to balance the weight between the predicted value and the observed value; represents the actual observed data; A function that represents the nonlinear relationship between the predicted value and the observed data.

5. The method for preventing drunk driving and poisoning according to claim 1, characterized in that: The objective function of the optimal control theory is: ； in, is the objective function, which represents the cost or loss to be minimized in the control process; is the system state variable, indicating that at time The status of the system; is the control input, indicating that at time the controls exerted on the system; is the cost function, which represents the system state variable and control input At the moment Variable loss; is the end time of the time interval and is the total duration of the control process; To represent the cost in the entire time interval during the control process, the integral sum is performed to obtain the overall cost.

6. The method for preventing drunk driving and poisoning according to claim 3, characterized in that: The eye movement model detects the risk of drunk driving through the nonlinear change of pupil reflex, and the pupil reflex change equation includes: ； in, is the pupil diameter; is the initial pupil diameter; is the amplitude of change; the reflex decay rate constant, which indicates the rate at which the pupillary response returns to its normal state; is the exponential decay function of the reflection recovery, which represents the change of pupil diameter over time.

7. The method for preventing drunk driving and poisoning according to claim 1, characterized in that: The real-time feedback includes warning information displayed on the vehicle display screen, voice prompts and vehicle speed limits.

8. The method for preventing drunk driving and poisoning according to claim 7, characterized in that: The vehicle speed limit includes activating the emergency brake system, limiting the vehicle speed and automatically locking the vehicle.

9. A system for preventing drunk driving and poisoning, based on the method for preventing drunk driving and poisoning according to any one of claims 1 to 8, characterized in that: include: Multi-sensor module for collecting in-vehicle environment data and driver data in real time; The data fusion module is used to fuse the in-vehicle environment data and the driver data and output a comprehensive state estimate; The risk assessment module uses nonlinear dynamics modeling and optimal control algorithms to assess the risk of drunk driving based on comprehensive state estimation; The feedback module provides real-time feedback based on the results of the drunk driving risk assessment, including warning messages on the vehicle display, voice prompts and vehicle speed limits.

10. The anti-drunk driving and anti-poisoning system according to claim 9, characterized in that: The multi-sensor module comprises: Alcohol concentration sensor: used to monitor the alcohol concentration in the air inside the car in real time; Carbon monoxide concentration sensor: used to monitor the carbon monoxide concentration in the air inside the car in real time; Camera: used to collect driver's pupil reflection data.

Images