Commercial vehicle driver state monitoring and health management intelligent system and method

Abstract

The invention discloses a commercial vehicle driver state monitoring and health management intelligent system and method, relates to the technical field of automobiles, and solves the problems that an existing driver state monitoring method adopts single data detection, the real dangerous state is difficult to effectively judge, the judgment standard is fixed, and the safety is poor. The technical problem that the driver state detection management is low in comprehensive efficiency due to the fact that the judgment accuracy of the driver state detection management is low when the driver state detection management is applied to different conditions is solved. Comprising the following steps: acquiring electrocardiosignal data, performing identity recognition on a driver based on the electrocardiosignal data to obtain a driver ID, and binding the driver ID with a current travel; generating a hand holding state of the driver based on the hand leaving detection data; acquiring electrocardiosignal data in real time, and generating an electrocardiogram state of the driver based on the electrocardiosignal data; acquiring driving data, and performing graded early warning based on the driving data, the holding state and the electrocardio state; and the comprehensive efficiency of driver detection state detection management is improved.

Description

Technical Field

[0001] This application belongs to the field of automotive technology, specifically an intelligent system and method for monitoring the condition and health management of commercial vehicle drivers. Background Technology

[0002] Commercial vehicles are the backbone of my country's economic transportation system, and their safe operation is of paramount importance. Authoritative data shows that over 40% of major commercial vehicle traffic accidents are directly related to driver fatigue, distraction, and sudden health problems.

[0003] Existing technology (invention patent CN 113183971 B) discloses a driver state monitoring method, device, vehicle terminal, and readable storage medium, belonging to the field of autonomous driving technology. It determines the corresponding alarm threshold duration based on the driver's hand state information; different hand state information corresponds to different alarm threshold durations. This allows for monitoring of the first driver state information, ensuring that an alarm is triggered only when the first driver state information meets the alarm conditions within an alarm threshold duration that better matches the driver's hand state. By introducing a fusion processing mechanism for hand state information and the first driver state information, the alarm results are more accurate. This effectively reduces false alarms and improves the reliability of driver state monitoring results while achieving driver state monitoring.

[0004] The aforementioned driver status detection method detects the driver's hands-off state and overall driver status, and analyzes the results using a fixed three-level judgment standard. However, the high degree of independence between hands-off detection and driver status detection leads to fragmented monitoring dimensions and insufficient accuracy in early warnings. It also makes it difficult to effectively judge true dangerous situations. Furthermore, the existing method for determining whether a driver is abnormal relies on a fixed three-level judgment standard, triggering an alarm only when the standard is met. This lack of adaptability results in an excessively high alarm trigger frequency and numerous false alarms, leading to low overall efficiency in driver status monitoring and management. Therefore, an intelligent system and method for commercial vehicle driver status monitoring and health management are needed. Summary of the Invention

[0005] This application provides an intelligent system and method for monitoring the condition and health management of commercial vehicle drivers. It solves the technical problems of existing driver condition monitoring methods, which rely on single data detection, making it difficult to effectively judge real dangerous conditions, and fixed judgment criteria, resulting in low judgment accuracy in different situations and ultimately low overall efficiency of driver condition detection and management.

[0006] To achieve the above objectives, this application adopts the following technical solution: Firstly, a smart method for monitoring the condition and health management of commercial vehicle drivers is provided, including: Acquire electrocardiogram (ECG) signal data, identify the driver based on the ECG signal data to obtain the driver ID, and bind the driver ID to the current trip; Acquire hands-off detection data, and generate the driver's hand grip state based on the hands-off detection data; Real-time acquisition of electrocardiogram (ECG) signal data, and generation of the driver's ECG status based on the ECG signal data; Acquire driving data and provide tiered warnings based on driving data, hand grip status, and ECG status.

[0007] Based on the above technical solutions, in the intelligent system and method for monitoring and managing the health of commercial vehicle drivers provided in this application, the system acquires electrocardiogram (ECG) signal data, identifies the driver based on the ECG signal data to obtain a driver ID, and binds the driver ID to the current trip; acquires hands-off detection data, and generates the driver's hand-holding state based on the hands-off detection data; acquires ECG signal data in real time, and generates the driver's ECG state based on the ECG signal data; acquires driving data, and performs graded early warnings based on driving data, hand-holding state, and ECG state; it uses multi-source data, including hands-off detection and ECG signal detection, to analyze the driver's state, ensuring the non-singularity of the analysis results, thereby significantly improving the accuracy of the analysis results; at the same time, it uses the driver's own relevant data and real-time driving status for dynamic graded early warnings, breaking through the traditional fixed discrimination criteria and further increasing the accuracy of the discrimination results; thus improving the overall efficiency of driver status detection and management.

[0008] In conjunction with the first aspect above, in one possible implementation, the step of identifying the driver based on electrocardiogram signal data to obtain a driver ID includes: After feature extraction of the electrocardiogram signal, it is converted into a binary hash code. The binary hash code is then matched with the hash codes in a pre-stored binary hash code library for drivers to obtain the driver ID corresponding to the binary hash code. The binary hash code is 256 bits. The driver binary hash code library is determined by several pairs of driver IDs and the binary hash codes corresponding to the driver IDs.

[0009] In conjunction with the first aspect above, in one possible implementation, generating the driver's hand grip state based on the hand-off detection data includes: Extract a capacitance array from the off-hand detection data for several time periods within a set time period, wherein the capacitance array contains a set number of capacitance values; Obtain the set reference capacitance value, and correct the reference capacitance value based on the capacitance array to obtain the corrected reference capacitance value; The capacitance array of the current time period in the off-hand detection data is acquired in real time. Each capacitance value in the capacitance array and the reference capacitance value are input into the hand grip state recognition model to obtain the corresponding hand grip state. The hand grip state includes off-hand state, one-finger touch, two-finger touch, three-finger touch, four-finger touch, palm touch, three-finger grip, single-hand grip and two-hand grip, etc.

[0010] In conjunction with the first aspect above, in one possible implementation, the step of correcting the reference capacitance value based on the capacitance array to obtain the corrected reference capacitance value includes: Obtain each capacitance value in each capacitance array and calculate the difference ratio between each capacitance value and its corresponding comparison capacitance value. When the difference ratio of more than a set number of capacitor arrays is less than the set offset ratio threshold, the difference ratio is used as the offset ratio, and the reference capacitor value is corrected using the difference ratio to obtain the corrected reference capacitor value; otherwise, the reference capacitor value is not corrected.

[0011] In conjunction with the first aspect above, in one possible implementation, a training method for the hand-holding state recognition model includes: Acquire several hand-grip states and several capacitance values ​​and reference capacitance values ​​corresponding to the hand-grip states; integrate the several capacitance values ​​and reference capacitance values ​​of the same hand-grip state into several feature data belonging to the hand-grip state; The feature data of each hand grip state are integrated into training data and test data. The training data is used to train the artificial intelligence model, and the test data is used to test the trained artificial intelligence model. Finally, a hand grip state recognition model is obtained with the capacitance value and the reference capacitance value as inputs and the hand grip state as output. The artificial intelligence model includes at least a BP neural network model and an RBF neural network model.

[0012] In conjunction with the first aspect above, in one possible implementation, generating the driver's electrocardiogram (ECG) state based on ECG signal data includes: Obtain ECG signal test data corresponding to the driver ID, the ECG signal test data including several ECG signals and their corresponding ECG states; integrate the ECG signals and their corresponding ECG states into correction data; obtain a pre-trained ECG signal recognition model; use the correction data to correct the ECG signal recognition model to obtain an ECG signal recognition model suitable for the driver ID; The electrocardiogram (ECG) signal is extracted from the ECG signal data and input into an ECG signal recognition model to obtain the corresponding ECG state. The ECG state includes a normal state or an abnormal state. The normal state includes a negative state, a calm state, and an active state. Negative states include fatigue and drowsiness, active states include excitement and agitation, and a calm state is the driver's normal state. The input of the ECG signal recognition model is the ECG signal, and the output is the ECG state corresponding to the ECG signal.

[0013] In conjunction with the first aspect above, in one possible implementation, the graded early warning based on driving data, hand grip status, and ECG status includes: If the ECG status is abnormal, a body abnormality warning signal is generated, and the cumulative number of abnormal ECG statuses is obtained. When the cumulative number of abnormal statuses exceeds the set abnormal status threshold, a cloud emergency abnormality signal is generated. The cloud emergency abnormality signal includes the driver ID, vehicle location, driving data, trip details, and real-time ECG signal. This facilitates timely location and rescue of the driver when physical problems occur. No, when the hand-holding state is not a two-hand grip, driving data is acquired and input into the driving state analysis and evaluation model to obtain the driving state risk. The higher the driving state risk value, the greater the probability of danger in the current driving state, requiring the driver to be more vigilant. The driving data includes real-time vehicle speed, turn signal status, accelerator pedal opening, and brake pedal opening. The driving state risk score is substituted into a set off-hand warning duration correction function to obtain the corrected off-hand warning duration. One expression of the off-hand warning duration correction function includes: ; in, This is the revised off-hand warning duration. For driving condition risks, The original hand-off warning duration is set by an expert. In this embodiment, the original hand-off warning duration is set to 10 seconds. When the ECG status is in the calm state within the normal state, the duration conversion ratio and state duration of each hand grip state set in the calm state are obtained; the cumulative time off the hand is calculated based on the duration conversion ratio and state duration of each hand grip state; otherwise, the duration conversion ratio and state duration of each hand grip state set in the non-calm state are obtained; the cumulative time off the hand is calculated based on the duration conversion ratio and state duration of each hand grip state. When the cumulative time spent away from the device exceeds the set time for the warning of being away from the device, an emergency signal for being away from the device will be generated. When the cumulative time spent away from the device is less than or equal to the set time for the warning of being away from the device, and is greater than K times the time for the warning of being away from the device, an alarm signal for being away from the device is generated. When the cumulative time away from the hand is less than or equal to K times the time of the warning for leaving the hand, and is greater than 0, a time away from the hand prompt signal is generated; K is the alarm coefficient, and K∈(0,1). The specific value is set by experience. In this embodiment, the alarm coefficient K is set to 0.6.

[0014] In conjunction with the first aspect above, in one possible implementation, the calculation of the cumulative hand-holding time based on the duration conversion ratio and state duration of each hand-holding state includes: Obtain the duration and transition ratio of each hand-holding state, and mark the transition ratio as... The state duration is marked as ; Through the formula: ; The cumulative time away from the device was calculated. ; Aside from the off-hand and two-handed gripping states, for the other identical gripping states, the duration conversion rate in the calm state was greater than that in the non-calm state, and the duration conversion rate was also greater. ∈[0,1], where the duration conversion ratio for the off-hand state is 1 and the duration conversion ratio for the two-hand grip state is 0; it is worth noting that the duration of the states calculated above are the duration of the corresponding grip state collected after obtaining the corrected off-hand warning duration.

[0015] Understandably, this can also be achieved through changes in capacitance.

[0016] Secondly, this application provides an intelligent device for monitoring and managing the condition and health of commercial vehicle drivers, comprising: a processor and a storage medium; the storage medium includes instructions, and the processor is used to execute the instructions to implement the methods described in the first aspect and any possible implementation thereof. This intelligent device for monitoring and managing the condition and health of commercial vehicle drivers can be an electronic device or a chip within an electronic device.

[0017] Thirdly, this application provides an intelligent system for monitoring the status and health management of commercial vehicle drivers, including: a vehicle terminal system and a cloud management platform; The vehicle terminal system includes a hands-off detection module, an electrocardiogram monitoring module, and a data processing center; The off-hand detection module acquires off-hand monitoring data, which is the capacitance value collected by the hybrid capacitive sensing electrode array set in the steering wheel; and generates a hand grip state based on the off-hand monitoring data; the hand grip state is the state in which the driver holds the steering wheel, including one-finger touch, two-finger touch, three-finger touch, four-finger touch, palm touch, three-finger grip, single-hand grip, and two-hand grip, etc. The electrocardiogram (ECG) monitoring module includes an ECG monitoring unit and an identity recognition unit; The identity recognition unit: identifies the driver based on electrocardiogram signal data, obtains the driver ID, and binds the driver ID to the current trip; The electrocardiogram (ECG) monitoring unit: acquires ECG monitoring data and ECG signal data in real time, and generates the driver's ECG state based on the ECG signal data; The data processing center acquires driving data in real time and provides graded warnings based on driving data, hand grip status, and ECG status.

[0018] In conjunction with the first aspect mentioned above, one possible implementation also includes a cloud management platform, which is used to acquire cloud emergency anomaly signals, issue warnings for cloud emergency anomaly signals on the cloud management platform, and store driving data and related data corresponding to driver IDs.

[0019] Fourthly, this application provides a computer-readable storage medium storing instructions that, when executed on a commercial vehicle driver status monitoring and health management intelligent device, cause the commercial vehicle driver status monitoring and health management intelligent device to perform the methods described in the first aspect and any possible implementation thereof.

[0020] Fifthly, this application provides a computer program product containing instructions that, when the computer program product is run on a commercial vehicle driver status monitoring and health management intelligent device, causes the commercial vehicle driver status monitoring and health management intelligent device to perform the methods described in the first aspect and any possible implementation thereof.

[0021] This application provides an intelligent method and system for monitoring the status and health management of commercial vehicle drivers. It can acquire electrocardiogram (ECG) signal data, identify the driver based on the ECG signal data to obtain a driver ID, and bind the driver ID to the current trip; acquire hands-off detection data, and generate the driver's hand-holding status based on the hands-off detection data; acquire ECG signal data in real time, and generate the driver's ECG status based on the ECG signal data; acquire driving data, and provide graded warnings based on driving data, hand-holding status, and ECG status; and use multi-source data, including hands-off detection and ECG signal detection, to analyze the driver's status, ensuring the non-singularity of the analysis results and thus significantly improving the accuracy of the analysis results. Simultaneously, it uses the driver's own relevant data and real-time driving status for dynamic graded warnings, breaking through traditional fixed discrimination criteria and further increasing the accuracy of the discrimination results.

[0022] It should be understood that the descriptions of technical features, technical solutions, beneficial effects, or similar language in this application do not imply that all features and advantages can be achieved in any single embodiment. Rather, it is understood that the description of a feature or beneficial effect means that a specific technical feature, technical solution, or beneficial effect is included in at least one embodiment. Therefore, the descriptions of technical features, technical solutions, or beneficial effects in this specification do not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions, and beneficial effects described in this embodiment can be combined in any suitable manner. Those skilled in the art will understand that embodiments can be implemented without one or more specific technical features, technical solutions, or beneficial effects of a particular embodiment. In other embodiments, additional technical features and beneficial effects may be identified in specific embodiments that do not embody all embodiments. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of this application 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 this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a schematic diagram illustrating the steps of the intelligent method for monitoring the status and health management of vehicle drivers in this application; Figure 2 This is a schematic diagram of the module connections of the intelligent system for monitoring the status and health management of vehicle drivers in this application. Detailed Implementation

[0025] The technical solutions of this application will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0026] Please see Figure 1 The first aspect of this application provides an intelligent method for monitoring the status and health management of commercial vehicle drivers, including: Acquire electrocardiogram (ECG) signal data, identify the driver based on the ECG signal data to obtain the driver ID, and bind the driver ID to the current trip; The driver's hand grip state is generated based on the off-hand detection data. The off-hand detection data is collected by a sensor array set on the steering wheel. Specifically, a hybrid capacitive sensing electrode array combining self-capacitance and mutual capacitance is optimized and arranged inside the steering wheel rim according to the principles of ergonomics. Real-time acquisition of electrocardiogram (ECG) signal data, and generation of the driver's ECG status based on the ECG signal data; ECG signal data is acquired through a wearable ECG signal detection device. Acquire driving data and provide tiered warnings based on driving data, hand grip status, and ECG status.

[0027] Based on the above technical solutions, in the intelligent system and method for monitoring the status and health management of commercial vehicle drivers provided in this application, the system acquires electrocardiogram (ECG) signal data, identifies the driver based on the ECG signal data to obtain a driver ID, and binds the driver ID to the current trip; acquires hands-off detection data, and generates the driver's hand-holding status based on the hands-off detection data; acquires ECG signal data in real time, and generates the driver's ECG status based on the ECG signal data; acquires driving data, and performs graded warnings based on driving data, hand-holding status, and ECG status; and uses multi-source data, including hands-off detection and ECG signal detection, to analyze the driver's status, ensuring the non-singularity of the analysis results, thereby significantly improving the accuracy of the analysis results; at the same time, it uses the driver's own relevant data and real-time driving status for dynamic graded warnings, breaking through the traditional fixed discrimination criteria and further increasing the accuracy of the discrimination results.

[0028] In one possible implementation, driver identification based on electrocardiogram (ECG) signal data to obtain a driver ID includes: extracting features from the ECG signal, converting it into a binary hash code, and then matching the binary hash code with hash codes in a pre-stored driver binary hash code library to determine the corresponding driver ID.

[0029] In a preferred embodiment, to ensure uniqueness while maintaining matching efficiency, the binary hash code uses 256 bits. The driver binary hash code library is pre-established through registration and includes several pairs of driver IDs and their corresponding binary hash codes.

[0030] This embodiment ensures the accuracy of driver ID recognition through the driver ID recognition method described above, thereby improving the accuracy of subsequent model-specific corrections based on the driver ID.

[0031] In one possible implementation, generating the driver's hand grip state based on the off-hand detection data includes: extracting a capacitance array of several time periods within a set time period from the off-hand detection data, wherein the capacitance array contains a set number of capacitance values; the set time period includes at least 5 time periods, each time period includes a capacitance value group, and the duration of each time period is fixed, wherein the duration of each time period in this embodiment is 3 seconds; Obtain the set reference capacitance value, and correct the reference capacitance value based on the capacitance array to obtain the corrected reference capacitance value; The capacitance array of the current time period in the off-hand detection data is acquired in real time. Each capacitance value in the capacitance array and the reference capacitance value are input into the hand grip state recognition model to obtain the corresponding hand grip state. The hand grip state includes off-hand state, one-finger touch, two-finger touch, three-finger touch, four-finger touch, palm touch, three-finger grip, single-hand grip and two-hand grip, etc.

[0032] In one possible implementation, the corrected reference capacitance value is obtained by correcting the reference capacitance value based on the capacitance array, including: obtaining each capacitance value in each capacitance array and calculating the difference ratio between each capacitance value and its corresponding comparison capacitance value. When the difference ratio of more than a set number of capacitor arrays is less than the set offset ratio threshold, the difference ratio is used as the offset ratio, and the reference capacitor value is corrected using the difference ratio to obtain the corrected reference capacitor value; otherwise, the reference capacitor value is not corrected.

[0033] Since changes in the temperature and humidity of the environment around the steering wheel can cause deviations in the reference capacitance value, and these deviations are generally within a certain range, adjusting the reference capacitance value based on the actual environment to determine the user's hand grip status can better ensure the accuracy of hand grip recognition. This avoids situations where changes in the ambient temperature and humidity inside the driver's cabin, such as when the driver turns on the air conditioning or opens the windows, lead to incorrect hand grip recognition. Furthermore, it increases the accuracy of subsequent operations based on hand grip status.

[0034] In one possible implementation, a training method for the hand-holding state recognition model includes: acquiring several hand-holding states and several capacitance values ​​and reference capacitance values ​​corresponding to the hand-holding states; integrating the several capacitance values ​​and reference capacitance values ​​of the same hand-holding state into several feature data belonging to the hand-holding state; The feature data of each hand grip state are integrated into training data and test data. The training data is used to train the artificial intelligence model, and the test data is used to test the trained artificial intelligence model. Finally, a hand grip state recognition model is obtained with the capacitance value and the reference capacitance value as inputs and the hand grip state as output. The artificial intelligence model includes at least a BP neural network model and an RBF neural network model.

[0035] It is understandable that hand grip states include at least the off-hand state, one-finger touch, two-finger touch, three-finger touch, four-finger touch, palm touch, three-finger grip, single-hand grip, and two-hand grip. Different touch or grip methods result in different contact areas between the driver's hand and the steering wheel. These different contact areas lead to different capacitance differences in the capacitance detection unit set on the steering wheel. By determining the range of capacitance differences corresponding to different hand grip states based on the range of contact areas between different fingers or palms and the steering wheel, the hand grip state can be identified through changes in capacitance.

[0036] In one possible implementation, generating the driver's ECG state based on ECG signal data includes: acquiring ECG signal trial data corresponding to the driver ID, wherein the ECG signal trial data includes several ECG signals and their corresponding ECG states; integrating the ECG signals and their corresponding ECG states into correction data; acquiring a pre-trained ECG signal recognition model; and using the correction data to correct the ECG signal recognition model to obtain an ECG signal recognition model suitable for the driver ID. The electrocardiogram (ECG) signal is extracted from the ECG signal data and input into an ECG signal recognition model to obtain the corresponding ECG state. The ECG state includes a normal state or an abnormal state. The normal state includes a negative state, a calm state, and an active state. Negative states include fatigue and drowsiness, active states include excitement and agitation, and a calm state is the driver's normal state. The input of the ECG signal recognition model is the ECG signal, and the output is the ECG state corresponding to the ECG signal.

[0037] The specific pre-trained ECG signal recognition model is a 1D-CNN hybrid network structure. It uses several ECG signals from several standard personnel and manually labels the ECG signals with corresponding ECG states. The ECG signals and ECG states are then integrated into several training and testing data to train and test the hybrid model. Alternatively, the hybrid model can be pre-trained on public datasets such as MIT-BIH and PTB-XL. The training process is a relatively mature technology and will not be described in detail here.

[0038] Specifically, the model includes at least an input layer: the input is a preprocessed electrocardiogram (ECG) signal, which is a one-dimensional time-series data of the ECG signal. It needs to undergo preprocessing such as denoising and standardization, and the data is in the form of a continuous sampling point sequence; a 1D-CNN feature extraction module: it uses 1D convolution to adapt to the temporal characteristics of the ECG signal and extracts local waveform features, including QRS complex, P wave, and T wave related features; a classification module: after flattening the features through a Flatten layer, it fuses global features through a fully connected layer and finally outputs the ECG status through Softmax.

[0039] Model training consists of two phases: a pre-training phase and a fine-tuning phase. The pre-training phase trains a basic model based on publicly available datasets or standard personnel ECG data, learning general ECG state mapping patterns. The fine-tuning phase uses "ECG signal trial data" from target drivers to adjust model parameters, adapting to individual ECG signal differences, such as differences in heart rate baseline and waveform morphology, resulting in a model with good individual adaptability. This balances general generalization ability with individual adaptability, reducing the misjudgment rate among different drivers.

[0040] This embodiment modifies the ECG signal recognition model using the driver's own relevant data to obtain an ECG signal recognition model adapted to the driver. This avoids the recognition error caused by using a uniform standard ECG signal recognition model due to differences in the physical functions of different drivers. This increases the accuracy of ECG signal recognition.

[0041] In one possible implementation, a tiered early warning system is implemented based on driving data, hand grip status, and electrocardiogram (ECG) status. This includes: determining whether the ECG status is abnormal; if so, generating a physical abnormality warning signal; and acquiring the cumulative number of abnormal ECG statuses. When the cumulative number exceeds a set abnormal status threshold, a cloud emergency abnormality signal is generated. The cloud emergency abnormality signal includes the driver ID, vehicle location, driving data, trip details, and real-time ECG signal. This facilitates timely location and rescue of the driver when health problems occur. No, when the hand-holding state is not a two-hand grip, driving data is acquired and input into the driving state analysis and evaluation model to obtain the driving state risk. The higher the driving state risk value, the greater the probability of danger in the current driving state, requiring the driver to be more vigilant. The driving data includes real-time vehicle speed, turn signal status, accelerator pedal opening, and brake pedal opening. The driving state risk score is substituted into a set off-hand warning duration correction function to obtain the corrected off-hand warning duration. One expression of the off-hand warning duration correction function includes: ; in, This is the revised off-hand warning duration. For driving condition risks, The original hand-off warning duration is set by an expert. In this embodiment, the original hand-off warning duration is set to 10 seconds. When the ECG status is in the calm state within the normal state, the duration conversion ratio and state duration of each hand grip state set in the calm state are obtained; the cumulative time off the hand is calculated based on the duration conversion ratio and state duration of each hand grip state; otherwise, the duration conversion ratio and state duration of each hand grip state set in the non-calm state are obtained; the cumulative time off the hand is calculated based on the duration conversion ratio and state duration of each hand grip state. When the cumulative time spent away from the device exceeds the set time for the warning of being away from the device, an emergency signal for being away from the device will be generated. When the cumulative time spent away from the device is less than or equal to the set time for the warning of being away from the device, and is greater than K times the time for the warning of being away from the device, an alarm signal for being away from the device is generated. When the cumulative time away from the hand is less than or equal to K times the time of the warning for leaving the hand, and is greater than 0, a time away from the hand prompt signal is generated; K is the alarm coefficient, and K∈(0,1). The specific value is set by experience. In this embodiment, the alarm coefficient K is set to 0.6. It can be understood that when the value of K is in the range of 0 to 1, all values ​​are feasible values, ensuring that at least K times the off-hand warning time is less than the original off-hand warning time. Yes, the above can be achieved, and the judgment logic can be performed normally.

[0042] One training method for a driving state analysis and evaluation model includes: acquiring several driving data points and corresponding driving state risks. These driving state risks are scores given by experts based on real-time vehicle speed, steering status, throttle opening, and brake pedal opening from the driving data, assessing the degree of danger of the current vehicle's driving state. The higher the degree of danger, the larger the corresponding driving state risk value. For example, when the vehicle speed is high, the probability of a dangerous situation is greater, thus the degree of danger for the vehicle in that state is higher, and the corresponding driving state risk value is set higher. When the vehicle's turn signal is on, it indicates that the vehicle may be turning, overtaking, or changing lanes, and the vehicle's driving state changes. At this time, the probability of a dangerous situation is also higher, so the corresponding driving state risk value is also set higher. The larger the throttle opening or brake pedal opening, the greater the vehicle's acceleration. The greater the deceleration, the faster the vehicle's speed changes, and the higher the probability of danger. Therefore, the corresponding driving state risk value is set higher. The driving state risk value ranges from 0 to 1, with values ​​closer to 1 indicating a greater driving risk. Specific mapping relationships and scoring criteria will not be elaborated further. Several driving data points and their corresponding driving risks are integrated into training and testing data. The training data is used to train the artificial intelligence model, and the testing data is used to test the trained model. The final result is a driving state analysis and evaluation model that takes driving data as input and outputs the driving state risk corresponding to that data. The artificial intelligence model includes a neural network model. It is understood that the more training data samples used, the closer the accuracy of the trained driving state analysis and evaluation model will be to expert empirical assessments.

[0043] This embodiment uses different levels of alarm signals to provide tiered early warnings, enabling targeted detection and reminders of the driver's status. Simultaneously, this embodiment adjusts the warning criteria based on the detected vehicle driving status, facilitating targeted warnings under different driving conditions, such as issuing a warning 2 seconds after the driver releases their hands at high speeds and 5 seconds after the driver releases their hands at low speeds.

[0044] In one possible implementation, the cumulative time off the hand is calculated based on the duration conversion ratio and state duration of each hand-holding state, including: obtaining the state duration and state conversion ratio of each hand-holding state, and marking the state conversion ratio as... The state duration is marked as ; Through the formula: ; The cumulative time away from the device was calculated. ; Aside from the off-hand and two-handed gripping states, for the other identical gripping states, the duration conversion rate in the calm state was greater than that in the non-calm state, and the duration conversion rate was also greater. ∈[0,1], where the duration conversion ratio for the off-hand state is 1, and the duration conversion ratio for the two-hand grip state is 0. It is worth noting that, understandably, the weaker the control over the vehicle in the grip state, the larger the corresponding duration conversion ratio is set, i.e., the closer it is to 1. For example, the duration conversion ratio for one-finger touch is greater than the duration conversion ratio for two-finger touch; the duration conversion ratio for two-finger touch is greater than the duration conversion ratio for three-finger grip. The state durations calculated above are the durations in the corresponding grip state collected after obtaining the corrected off-hand warning duration.

[0045] Secondly, this application provides an intelligent device for monitoring and managing the condition and health of commercial vehicle drivers, comprising: a processor and a storage medium; the storage medium includes instructions, and the processor is used to execute the instructions to implement the methods described in the first aspect and any possible implementation thereof. This intelligent device for monitoring and managing the condition and health of commercial vehicle drivers can be an electronic device or a chip within an electronic device.

[0046] Please see Figure 2 Thirdly, this application provides an intelligent system for monitoring the status and health management of commercial vehicle drivers, including: a vehicle terminal system and a cloud management platform; The vehicle terminal system includes a hands-off detection module, an electrocardiogram monitoring module, and a data processing center; Hands-off detection module: Acquires hands-off monitoring data, which is the capacitance value collected by a hybrid capacitive sensing electrode array installed in the steering wheel; generates hand grip state based on the hands-off monitoring data; the hand grip state refers to the driver's grip on the steering wheel, including one-finger touch, two-finger touch, three-finger touch, four-finger touch, palm touch, three-finger grip, single-hand grip, and two-hand grip, etc.; the sensor array uses a high-quality clock spring as the connection hub between the rotating mechanism and the fixed body; through this clock spring, a stable and reliable power supply is continuously provided to the sensing system in the steering wheel, and the real-time, lossless transmission of the capacitive sensing signal to the processing unit at the vehicle body end is ensured. The off-hand detection module is used to detect when a hand is removed from its grasp and to generate a hand-holding status. The ECG monitoring module includes an ECG monitoring unit and an identity recognition unit; Identity recognition unit: Based on electrocardiogram signal data, the driver's identity is recognized to obtain the driver ID, and the driver ID is bound to the current trip; ECG monitoring unit: acquires ECG monitoring data and ECG signal data in real time, and generates the driver's ECG state based on the ECG signal data; The ECG monitoring module is used for driver identification and status monitoring, and generates an identification code and ECG status. Data processing center: Acquires driving data in real time and provides tiered warnings based on driving data, hand grip status, and ECG status.

[0047] In conjunction with the first aspect mentioned above, one possible implementation also includes a cloud management platform. The cloud management platform is used to acquire cloud emergency anomaly signals, issue warnings for cloud emergency anomaly signals on the cloud management platform, and store driving data and related data corresponding to driver IDs.

[0048] Fourthly, this application provides a computer-readable storage medium storing instructions that, when executed on a commercial vehicle driver status monitoring and health management intelligent device, cause the commercial vehicle driver status monitoring and health management intelligent device to perform the methods described in the first aspect and any possible implementation thereof.

[0049] Fifthly, this application provides a computer program product containing instructions that, when run on a commercial vehicle driver status monitoring and health management intelligent device, causes the commercial vehicle driver status monitoring and health management intelligent device to execute the methods described in the first aspect and any possible implementation thereof. Some data in the above formulas are calculated by removing dimensions and obtaining numerical values. The formulas are derived from a large amount of collected data through software simulation to obtain a formula that most closely approximates the real situation. The preset parameters and preset thresholds in the formulas are set by those skilled in the art based on actual conditions or obtained through simulation with a large amount of data.

[0050] How this application works: By acquiring electrocardiogram (ECG) signal data, the driver is identified based on the ECG signal data to obtain a driver ID, which is then bound to the current trip. Hand-off detection data is acquired, and the driver's hand-holding status is generated based on this data. Real-time ECG signal data is acquired, and the driver's ECG status is generated based on this data. Driving data is acquired, and graded warnings are issued based on driving data, hand-holding status, and ECG status. The use of multi-source data, including hand-off detection and ECG signal detection, to analyze the driver's status ensures the non-singularity of the analysis results, thus significantly improving the accuracy of the analysis. Simultaneously, dynamic graded warnings are issued using the driver's own relevant data and real-time driving status, breaking through traditional fixed judgment criteria and further increasing the accuracy of the judgment results.

[0051] The above embodiments are only used to illustrate the technical methods of this application and are not intended to limit it. Although this application has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of this application without departing from the spirit and scope of the technical methods of this application.

Claims

1. An intelligent method for monitoring the status and health management of commercial vehicle drivers, characterized in that, include: Acquire electrocardiogram (ECG) signal data, identify the driver based on the ECG signal data to obtain the driver ID, and bind the driver ID to the current trip; Acquire hands-off detection data, and generate the driver's hand grip state based on the hands-off detection data; Real-time acquisition of electrocardiogram (ECG) signal data, and generation of the driver's ECG status based on the ECG signal data; Acquire driving data and provide tiered warnings based on driving data, hand grip status, and ECG status.

2. The intelligent method for monitoring the status and health management of commercial vehicle drivers according to claim 1, characterized in that, The driver identification based on electrocardiogram signal data, to obtain the driver ID, includes: After feature extraction of the electrocardiogram signal, it is converted into a binary hash code. The binary hash code is then matched with the hash codes in a pre-stored binary hash code library for drivers to obtain the driver ID corresponding to the binary hash code. The driver binary hash code library is determined by several pairs of driver IDs and the binary hash codes corresponding to the driver IDs.

3. The intelligent method for monitoring the status and health management of commercial vehicle drivers according to claim 1, characterized in that, The process of generating the driver's hand grip state based on the off-hand detection data includes: Extract a capacitance array from the off-hand detection data for several time periods within a set time period, wherein the capacitance array contains a set number of capacitance values; Obtain the set reference capacitance value, and correct the reference capacitance value based on the capacitance array to obtain the corrected reference capacitance value; The capacitance array for the current time period is acquired in real time from the hand-free detection data. Each capacitance value in the capacitance array and the reference capacitance value are input into the hand-holding state recognition model to obtain the corresponding hand-holding state.

4. The intelligent method for monitoring the status and health management of commercial vehicle drivers according to claim 3, characterized in that, The process of correcting the reference capacitance value based on the capacitance array to obtain the corrected reference capacitance value includes: Obtain each capacitance value in each capacitance array and calculate the difference ratio between each capacitance value and its corresponding comparison capacitance value. When the difference ratio of more than a set number of capacitor arrays is less than the set offset ratio threshold, the difference ratio is used as the offset ratio, and the reference capacitor value is corrected using the difference ratio to obtain the corrected reference capacitor value; otherwise, the reference capacitor value is not corrected.

5. The intelligent method for monitoring the status and health management of commercial vehicle drivers according to claim 3, characterized in that, One training method for the hand-holding state recognition model includes: Acquire several hand-grip states and several capacitance values ​​and reference capacitance values ​​corresponding to the hand-grip states; integrate the several capacitance values ​​and reference capacitance values ​​of the same hand-grip state into several feature data belonging to the hand-grip state; The feature data of each hand grip state are integrated into training data and test data. The training data is used to train the artificial intelligence model, and the test data is used to test the trained artificial intelligence model. Finally, a hand grip state recognition model is obtained with the capacitance value and the reference capacitance value as inputs and the hand grip state as output.

6. The intelligent method for monitoring the status and health management of commercial vehicle drivers according to claim 1, characterized in that, The process of generating the driver's electrocardiogram (ECG) state based on ECG signal data includes: Obtain ECG signal test data corresponding to the driver ID, the ECG signal test data including several ECG signals and their corresponding ECG states; integrate the ECG signals and their corresponding ECG states into correction data; obtain a pre-trained ECG signal recognition model; use the correction data to correct the ECG signal recognition model to obtain an ECG signal recognition model suitable for the driver ID; The electrocardiogram (ECG) signal is extracted from the ECG signal data, and the ECG signal is input into the ECG signal recognition model to obtain the corresponding ECG state; the ECG state includes a normal state or an abnormal state; the normal state includes a negative state, a calm state, and an active state; the input of the ECG signal recognition model is the ECG signal, and the output is the ECG state corresponding to the ECG signal.

7. The intelligent method for monitoring the status and health management of commercial vehicle drivers according to claim 1, characterized in that, The tiered early warning system based on driving data, hand grip status, and ECG status includes: If the ECG status is abnormal, a body abnormality warning signal is generated. The cumulative number of times the ECG status is abnormal is also obtained. When the cumulative number of times is greater than the set abnormal status threshold, a cloud emergency abnormality signal is generated. The cloud emergency abnormality signal includes the driver ID, vehicle location, driving data, trip details and real-time ECG signal. No, when the hand grip is not a two-hand grip, the driving data is acquired and input into the driving status analysis and evaluation model to obtain the driving status risk; the driving data includes real-time vehicle speed, turn signal status, accelerator opening and brake pedal opening; the driving status risk score is substituted into the set off-hand warning duration correction function to obtain the corrected off-hand warning duration; When the ECG status is in the calm state within the normal state, the duration conversion ratio and state duration of each hand grip state set in the calm state are obtained; the cumulative time off the hand is calculated based on the duration conversion ratio and state duration of each hand grip state; otherwise, the duration conversion ratio and state duration of each hand grip state set in the non-calm state are obtained; the cumulative time off the hand is calculated based on the duration conversion ratio and state duration of each hand grip state. When the cumulative time spent away from the device exceeds the set time for the warning of being away from the device, an emergency signal for being away from the device will be generated. When the cumulative time spent away from the device is less than or equal to the set time for the warning of being away from the device, and is greater than K times the time for the warning of being away from the device, an alarm signal for being away from the device is generated. When the cumulative time away from the hand is less than or equal to K times the time of the warning for leaving the hand, and is greater than zero, a time away from the hand prompt signal is generated; K is the alarm coefficient, and K∈(0,1).

8. The intelligent method for monitoring the status and health management of commercial vehicle drivers according to claim 1, characterized in that, The calculation of cumulative hand-free time based on the duration conversion ratio and state duration for each hand-holding state includes: Obtain the duration and transition ratio of each hand-holding state, and mark the transition ratio as... The state duration is marked as ; By formula: ； Calculate the cumulative time away from the device ; Aside from the off-hand and two-handed gripping states, for all other identical gripping states, the duration conversion rate in the calm state was greater than that in the non-calm state, and the duration conversion rate was also greater. ∈[0,1], where the duration conversion ratio of the off-hand state is 1, and the duration conversion ratio of the two-hand grip state is 0.

9. A smart system for monitoring the status and health management of commercial vehicle drivers, based on the application of the smart method for monitoring the status and health management of commercial vehicle drivers according to any one of claims 1 to 8, characterized in that, include: Vehicle terminal system and cloud management platform; The vehicle terminal system includes a hands-off detection module, an electrocardiogram monitoring module, and a data processing center; The off-hand detection module: acquires off-hand monitoring data; generates hand-holding status based on the off-hand monitoring data; The electrocardiogram (ECG) monitoring module includes an ECG monitoring unit and an identity recognition unit; The identity recognition unit: identifies the driver based on electrocardiogram signal data, obtains the driver ID, and binds the driver ID to the current trip; The electrocardiogram (ECG) monitoring unit: acquires ECG monitoring data and ECG signal data in real time, and generates the driver's ECG state based on the ECG signal data; The data processing center acquires driving data in real time and provides tiered warnings based on driving data, hand grip status, and ECG status.

10. The intelligent system for monitoring the status and health management of commercial vehicle drivers according to claim 9, characterized in that, It also includes a cloud management platform, which is used to acquire cloud emergency anomaly signals and issue warnings for the cloud emergency anomaly signals on the cloud management platform.

Images