Driver guidance compliance feature real-time analysis method and apparatus

Abstract

The application discloses a driver guidance compliance feature real-time analysis method and device, the method comprises the following steps: collecting the multi-source data of the guidance instruction, the target vehicle state and the surrounding traffic environment element state of the vehicle during the driving process of the vehicle; clearly expecting the future driving state of the target vehicle, and predicting the future driving state of the target vehicle; constructing a weight coefficient matrix fusing the actual driving state and the predicted driving state of the vehicle; fusing the actual driving state and the predicted driving state of the vehicle, and the weight coefficient matrix, and calculating the comprehensive compliance feature of the driver to the current guidance instruction. The application realizes closed-loop optimization of the guidance instruction, and improves the passing efficiency.

Description

Technical Field

[0001] This application relates to real-time analysis technology of driver guidance compliance characteristics, and in particular to a real-time analysis method of driver guidance compliance characteristics based on fusion state prediction, so as to realize early quantitative calculation of driver compliance characteristics with connected guidance instructions. Background Technology

[0002] In a vehicle-road cooperative environment, intelligent connected vehicles can receive guidance commands from cloud platforms or roadside units, enabling collaborative operation of "people-vehicle-road" within a fixed area, significantly improving traffic efficiency and operational safety. When unexpected traffic incidents occur or temporary traffic control is required, guidance commands can also assist roadside police in more efficiently directing traffic. Furthermore, intelligent connected vehicles can be categorized into different levels based on their degree of automation. The national standard "Classification of Driving Automation for Automobiles" classifies them into six levels based on their automation level: emergency assistance, partial driving assistance, combined driving assistance, conditional automated driving, highly automated driving, and fully automated driving. However, due to technological and market factors, intelligent connected vehicles cannot directly leap to highly automated and fully automated driving levels in the short term. Currently, most intelligent connected vehicles on the market still belong to the combined driving assistance level, where the vehicle is still primarily controlled by the driver, and the effectiveness of guidance commands depends on the driver's compliance with those commands.

[0003] Scholars have studied changes in driver compliance characteristics under different traffic scenarios and guidance instructions, exploring modeling methods to quantify driver compliance with guidance instructions. Driver compliance characteristics typically include two aspects: compliance rate and compliance accuracy. Compliance rate qualitatively describes whether a driver is subjectively willing to comply with guidance instructions, while compliance accuracy quantitatively analyzes whether the driver effectively follows the guidance instructions. When both compliance rate and accuracy are low, guidance instructions need to be adjusted to allow for a greater behavioral range for the driver, avoiding traffic chaos caused by excessive deviations in guidance outcomes. Modeling methods for quantifying driver compliance characteristics typically involve constructing models based on both driver and vehicle driving states. Some studies use physiological instruments and eye trackers to collect data such as skin conductance, electrocardiogram (ECG), electroencephalogram (EEG), and eye movement patterns in drivers, analyzing this data to understand driver compliance characteristics with different guidance instructions in different scenarios. This method primarily analyzes the driver's subjective compliance with guidance instructions, mainly focusing on the compliance rate. It analyzes the impact of factors such as driver attributes, workload, and in-vehicle human-machine interaction design on the driver's compliance rate. Furthermore, this method requires additional specialized equipment and has high requirements for data noise processing, necessitating adjustments for practical application. Another research attempt to extract vehicle driving status through in-vehicle positioning systems or external video recordings, objectively quantifying the driver's compliance with guidance instructions through data analysis to simultaneously analyze compliance rate and accuracy. However, limited by traditional data acquisition methods, it typically requires collecting the complete vehicle driving status after the driver receives the guidance instructions. It cannot identify abnormalities in the driver's compliance with guidance instructions in advance, limiting its use to post-event feature analysis and preventing real-time correction of guidance instructions during the guidance process.

[0004] In recent years, with the continuous development of intelligent connected vehicles, corresponding regulatory technologies and platforms have been gradually advanced. Various regions have progressively implemented intelligent connected vehicle operation safety monitoring platforms, enabling real-time collection of vehicle driving status data. This allows for advance analysis of driver compliance characteristics during the response to guidance instructions, providing a basis for real-time correction of guidance commands. However, due to the need for real-time analysis of compliance characteristics, and considering that the driver may not have fully responded to guidance instructions and the vehicle's driving status is incomplete, how to predict and complete the vehicle's driving status and integrate this completed data with existing data remains a key technical challenge that urgently needs to be overcome. This is fundamental to ensuring the rationality and effectiveness of real-time analysis results. Summary of the Invention

[0005] In view of this, this application proposes a method and device for real-time analysis of driver guidance compliance characteristics, which realizes early calculation and real-time update of response accuracy and response rate by introducing a vehicle state prediction model and a dynamic data fusion mechanism.

[0006] According to a first aspect of this application, a method for real-time analysis of driver guidance compliance characteristics is provided, comprising: During vehicle operation, multi-source data are collected, including vehicle guidance instructions, target vehicle status, and status of surrounding traffic environment elements. The desired future driving state of the target vehicle is determined according to the guidance instructions, and the future driving state of the target vehicle is predicted based on the current state of the target vehicle, the historical state of the target vehicle, and the state of the surrounding traffic environment elements. Based on the historical performance of the guidance instructions and prediction model, a weighted coefficient matrix that integrates the actual driving state and the predicted driving state of the vehicle is constructed. By integrating the actual driving status and predicted driving status of the vehicle with a weighted coefficient matrix, the driver's comprehensive compliance characteristics with the current guidance instructions are calculated in real time. The comprehensive compliance characteristics include at least response accuracy and response rate.

[0007] In some exemplary embodiments, the multi-source data collected, including vehicle guidance instructions, target vehicle status, and surrounding traffic environment element status, includes: The system acquires guidance instructions and timestamps received by the target vehicle, the target vehicle's status and timestamp, and the status of surrounding traffic environmental elements through vehicle-mounted data acquisition devices, roadside data acquisition devices, and cloud-based devices. The guidance instructions include the instruction type, instruction content, and the desired future driving state of the target vehicle. The guidance instruction types include pure longitudinal control instructions, pure lateral control instructions, and composite control instructions. When the guidance instruction only contains the recommended vehicle control strategy and not the recommended driving state, the desired future driving state of the target vehicle is calculated based on the vehicle kinematics formula. The target vehicle's status information includes its lateral and longitudinal positions, speed, acceleration, and heading angle; The status of surrounding traffic environment elements includes high-precision map information and status information of other motor vehicles, non-motor vehicles, and pedestrians; The acquired multi-source data is aligned by resampling and interpolation based on a set time series or the time series of one of the data.

[0008] In some exemplary embodiments, predicting the future driving state of the target vehicle based on its current state, historical state, and the state of surrounding traffic environment elements includes: The shortest of the following factors is determined as the effective time window for the target vehicle to respond to the guidance command: the guidance duration involved in the guidance command, the guidance area targeted by the guidance command, and the maximum prediction duration of the prediction model. Based on the target vehicle's current driving state, historical driving state, and surrounding traffic environment conditions, a vehicle driving state prediction model or a vehicle driving behavior model is used to predict the target vehicle's future driving state. The prediction time window is from the current moment to the end of the effective time window of the guidance command, expressed as: ,in, This indicates the predicted future driving status of the target vehicle. Indicates the current driving status of the target vehicle. This indicates the historical driving status of the target vehicle. x , y , v , a and ψ These represent the target vehicle's longitudinal position, lateral position, velocity, acceleration, and azimuth angle, respectively. This indicates the status of traffic environment elements surrounding the target vehicle. This refers to a vehicle trajectory prediction model or a vehicle driving behavior model.

[0009] Vehicle trajectory prediction models include Long Short-Term Memory Networks and Transformer models, while vehicle driving behavior models include Expected Safety Margin Models and Intelligent Driver Models.

[0010] In some exemplary embodiments, constructing the weighted coefficient matrix that integrates the actual driving state and the predicted driving state of the vehicle includes: The weight coefficients of the vehicle state dimension in the weight matrix are calculated based on the guidance instructions, as follows: For pure longitudinal control commands, the vehicle's longitudinal position, velocity, and acceleration have a significant impact on the calculated result of the conformation feature, while the vehicle's lateral position and heading angle have a smaller impact. The expression for the weighting coefficient matrix is ​​as follows: In the formula, This represents the weighting coefficient matrix for the vehicle state dimension in purely longitudinal control commands, with a dimension of 5×1. i x , j x , k x , m x and n x The parameters are in the range [0,1], representing the weights of the vehicle's longitudinal position, lateral position, velocity, acceleration, and heading angle, respectively. For pure lateral control commands, the vehicle's lateral position and heading angle have a significant impact on the calculated result of the conformation feature, while the vehicle's longitudinal position, velocity, and acceleration have a smaller impact. The expression for the weighting coefficient matrix is ​​as follows: , This represents the weighting coefficient matrix for the vehicle state dimension in purely lateral control commands, with a dimension of 5×1. i y , j y , k y , m y and n y The parameters are in the range [0,1], representing the weights of the vehicle's longitudinal position, lateral position, velocity, acceleration, and heading angle, respectively. For composite control commands, the vehicle's lateral and longitudinal position, velocity, acceleration, and heading angle have varying degrees of influence on the result of the feature calculation. The expression for the weighting coefficient matrix is ​​as follows: , This is a weighting coefficient matrix for the vehicle state dimension in composite control commands, with a dimension of 5×1. i xy , j xy , k xy , m xy and n xy The parameters are in the range [0,1], representing the weights of the vehicle's longitudinal position, lateral position, velocity, acceleration, and heading angle, respectively. The weight coefficients of the prediction time dimension in the weight matrix are calculated based on the historical performance of the prediction model. The expression is as follows: Where E is a matrix of all 1s with dimension 1. The time elapsed from the moment the instruction was issued to the current moment. This is the weight coefficient matrix for the prediction time dimension, with dimension 1. The value ranges from [0,1], and decreases as the prediction time increases, showing a negative correlation with the accuracy of the prediction model. The time length from the current moment to the last point in the effective prediction time window. To predict the weighting coefficients for the time dimension, the dimension is... This refers to the time from the issuance of the guiding instruction to the last time point of the predicted effective time window. By combining the weight coefficients of the vehicle state dimension and the prediction time dimension in the weighting function, a weight matrix that can be used for weighted fusion of vehicle trajectories is obtained, as expressed below: , where QF The resulting weight matrix has dimensions n×5, and Q... I These are the weighting coefficients for the vehicle state dimension, derived from Q based on the corresponding control command type. Ix Q Iy and Q Ixy Choose from.

[0011] In some exemplary embodiments, the method further includes: The driving states that the target vehicle has completed after receiving the guidance command and the predicted driving states in the steps are spliced ​​and fused together, and the expression is: S o ( x , y , v , a , ψ This indicates the driving state that the target vehicle has completed, with the dimension being... S F ( x , y , v , a , ψ The n×5 dimension represents the driving state after the target vehicle has completed its driving state and the predicted driving state are spliced ​​and fused together. The weighted root mean square error (RMSE) method is used to calculate the error between the stitched and fused vehicle driving state and the expected future driving state of the target vehicle as specified in the guidance instructions. This error serves as a quantification of the driver's response accuracy to the guidance instructions at that moment, and its expression is as follows: , d ( t S represents the accuracy of the driver's response to the guidance instructions at the current moment. d ( x , y , v , a , ψ The dimension represents the expected future driving state of the target vehicle as specified in the guidance instructions. n ×5, mean (·) represents the average of the matrices in the function. sum (·) represents summing the matrices in the function; By comparing the driver's response accuracy to guidance instructions at the current moment with a set threshold, a percentage response rate is calculated. This rate is used to adjust guidance instructions to the target vehicle or surrounding vehicles in a timely manner. The calculation expression is as follows: , This indicates the driver's compliance rate with guidance instructions at the current moment. This indicates the set threshold for determining compliance rate.

[0012] According to a second aspect of this application, a real-time driver guidance compliance characteristic analysis device is provided, comprising: The data acquisition unit is used to collect multi-source data on vehicle guidance instructions, target vehicle status, and surrounding traffic environment status during vehicle operation. The prediction unit is used to determine the desired future driving state of the target vehicle according to the guidance instructions, and to predict the future driving state of the target vehicle based on the current state of the target vehicle, the historical state of the target vehicle and the state of the surrounding traffic environment elements. The construction unit is used to construct a weight coefficient matrix that integrates the actual driving state and the predicted driving state of the vehicle based on the historical performance of the guidance instructions and the prediction model. The computing unit is used to integrate the actual driving state and predicted driving state of the vehicle, as well as the weight coefficient matrix, to calculate the driver's comprehensive compliance characteristics with the current guidance instructions in real time. The comprehensive compliance characteristics include at least response accuracy and response rate.

[0013] In some exemplary embodiments, the acquisition unit is further configured to: The system acquires guidance instructions and timestamps received by the target vehicle, the target vehicle's status and timestamp, and the status of surrounding traffic environmental elements through vehicle-mounted data acquisition devices, roadside data acquisition devices, and cloud-based devices. The guidance instructions include the instruction type, instruction content, and the desired future driving state of the target vehicle. The guidance instruction types include pure longitudinal control instructions, pure lateral control instructions, and composite control instructions. When the guidance instruction only contains the recommended vehicle control strategy and not the recommended driving state, the desired future driving state of the target vehicle is calculated based on the vehicle kinematics formula. The target vehicle state information includes the vehicle's lateral and longitudinal position, speed, acceleration, and heading angle. The state of the surrounding traffic environment elements includes high-precision map information and the state information of other motor vehicles, non-motor vehicles, and pedestrians. The acquired multi-source data is aligned by resampling and interpolation based on a set time series or the time series of one of the data.

[0014] In some exemplary embodiments, the building unit is further configured to: The shortest of the following factors is determined as the effective time window for the target vehicle to respond to the guidance command: the guidance duration involved in the guidance command, the guidance area targeted by the guidance command, and the maximum prediction duration of the prediction model. Based on the target vehicle's current driving state, historical driving state, and surrounding traffic environment conditions, a vehicle driving state prediction model or a vehicle driving behavior model is used to predict the target vehicle's future driving state. The prediction time window is from the current moment to the end of the effective time window of the guidance command, expressed as: ,in, This indicates the predicted future driving status of the target vehicle. Indicates the current driving status of the target vehicle. This represents the historical driving state of the target vehicle, where x, y, v, a, and ψ represent the target vehicle's longitudinal position, lateral position, velocity, acceleration, and azimuth angle, respectively. This indicates the status of traffic environment elements surrounding the target vehicle. This refers to a vehicle trajectory prediction model or a vehicle driving behavior model.

[0015] Vehicle trajectory prediction models include Long Short-Term Memory Networks and Transformer models, while vehicle driving behavior models include Expected Safety Margin Models and Intelligent Driver Models.

[0016] In some exemplary embodiments, the building unit is further configured to: The weight coefficients of the vehicle state dimension in the weight matrix are calculated based on the guidance instructions, as follows: For pure longitudinal control commands, the vehicle's longitudinal position, velocity, and acceleration have a significant impact on the calculated result of the conformation feature, while the vehicle's lateral position and heading angle have a smaller impact. The expression for the weighting coefficient matrix is ​​as follows: In the formula, This represents the weighting coefficient matrix for the vehicle state dimension in purely longitudinal control commands, with a dimension of 5×1. i x , j x , k x , m x and n y The parameters are in the range [0,1], representing the weights of the vehicle's longitudinal position, lateral position, velocity, acceleration, and heading angle, respectively. For pure lateral control commands, the vehicle's lateral position and heading angle have a significant impact on the calculated result of the conformation feature, while the vehicle's longitudinal position, velocity, and acceleration have a smaller impact. The expression for the weighting coefficient matrix is ​​as follows: , This represents the weighting coefficient matrix for the vehicle state dimension in purely lateral control commands, with a dimension of 5×1. i y , jy , k y , m y and n y The parameters are in the range [0,1], representing the weights of the vehicle's longitudinal position, lateral position, velocity, acceleration, and heading angle, respectively. For composite control commands, the vehicle's lateral and longitudinal position, velocity, acceleration, and heading angle have varying degrees of influence on the result of the feature calculation. The expression for the weighting coefficient matrix is ​​as follows: , This is a weighting coefficient matrix for the vehicle state dimension in composite control commands, with a dimension of 5×1. i xy , j xy , k xy , m xy and n xy The parameters are in the range [0,1], representing the weights of the vehicle's longitudinal position, lateral position, velocity, acceleration, and heading angle, respectively. The weight coefficients of the prediction time dimension in the weight matrix are calculated based on the historical performance of the prediction model. The expression is as follows: Where E is a matrix of all 1s with dimension 1. The time elapsed from the moment the instruction was issued to the current moment. This is the weight coefficient matrix for the prediction time dimension, with dimension 1. The value ranges from [0,1], and decreases as the prediction time increases, showing a negative correlation with the accuracy of the prediction model. The time length from the current moment to the last point in the effective prediction time window. To predict the weighting coefficients for the time dimension, the dimension is... This refers to the time from the issuance of the guiding instruction to the last time point of the predicted effective time window. By combining the weight coefficients of the vehicle state dimension and the prediction time dimension in the weighting function, a weight matrix that can be used for weighted fusion of vehicle trajectories is obtained, as expressed below: , where Q F The resulting weight matrix has dimensions n×5, and Q... I These are the weighting coefficients for the vehicle state dimension, derived from Q based on the corresponding control command type. Ix Q Iy and Q Ixy Choose from.

[0017] In some exemplary embodiments, the computing unit is further configured to: The driving states that the target vehicle has completed after receiving the guidance command and the predicted driving states in the steps are spliced ​​and fused together, and the expression is: S o ( x , y , v , a , ψ This indicates the driving state that the target vehicle has completed, with the dimension being... S F ( x , y , v , a , ψ The n×5 dimension represents the driving state after the target vehicle has completed its driving state and the predicted driving state are spliced ​​and fused together. The weighted root mean square error (RMSE) method is used to calculate the error between the stitched and fused vehicle driving state and the expected future driving state of the target vehicle as specified in the guidance instructions. This error serves as a quantification of the driver's response accuracy to the guidance instructions at that moment, and its expression is as follows: , d ( t S represents the accuracy of the driver's response to the guidance instructions at the current moment. d ( x , y , v , a , ψ The dimension represents the expected future driving state of the target vehicle as specified in the guidance instructions. n ×5, mean (·) represents the average of the matrices in the function. sum (·) represents summing the matrices in the function; By comparing the driver's response accuracy to guidance instructions at the current moment with a set threshold, a percentage response rate is calculated. This rate is used to adjust guidance instructions to the target vehicle or surrounding vehicles in a timely manner. The calculation expression is as follows: , This indicates the driver's compliance rate with guidance instructions at the current moment. This indicates the set threshold for determining compliance rate.

[0018] This application presents a real-time analysis method and apparatus for driver guidance compliance characteristics. It employs vehicle driving state data analysis to quantify driver compliance characteristics with guidance instructions, avoiding the need for additional driver physiological data acquisition equipment and effectively reducing analysis costs. By predicting vehicle driving state to complete the future driving state of the target vehicle within the effective range of the guidance instruction, it can determine the driver's compliance characteristics before the target vehicle leaves the effective range, avoiding analysis lag and providing a quantitative basis for timely adjustments to the guidance instruction. The weighting coefficient matrix is ​​designed to consider prediction model errors, dynamically balancing the actual driving state of the target vehicle after the guidance instruction is issued with its expected future driving state, effectively improving the reliability and stability of the analysis results. Based on the prediction of the vehicle's current and historical states and future states of environmental factors, real-time calculation of response accuracy and response rate can be achieved after the guidance instruction is issued but before the driver has fully executed the instruction, thereby predicting compliance trends in advance and reducing traffic risks caused by delayed compliance. For three types of commands—pure longitudinal (e.g., acceleration / deceleration), pure lateral (e.g., lane change / lane keeping), and composite control (e.g., curve deceleration + lane centering)—differentiated vehicle state dimension weights are configured for each. This ensures that key parameters, such as speed in longitudinal commands and heading angle in lateral commands, receive higher weights. This makes error calculations more closely aligned with the core requirements of the commands, avoids interference from long-term prediction errors on the overall analysis results, and reduces errors in response accuracy calculations. The real-time output compliance feature data in this embodiment can be directly fed back to the vehicle-road cooperative system, achieving closed-loop optimization of guidance commands and improving traffic efficiency. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the real-time analysis method for driver guidance compliance features according to an embodiment of this application. Figure 2 This is a schematic diagram showing the expected longitudinal position and speed of the target vehicle in the embodiments of this application; Figure 3 The target vehicle driving state predicted by the model in this application embodiment; Figure 4 This refers to the error range of the model predicting the driving state of the target vehicle in the embodiments of this application; Figure 5 This is a schematic diagram of key time points in the embodiments of this application; Figure 6 This is a schematic diagram of the driver guidance compliance feature real-time analysis device according to an embodiment of this application. Figure 7 This is a schematic diagram of the composition structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0020] Various embodiments and features of this application are described herein with reference to the accompanying drawings.

[0021] These and other features of this application will become apparent from the following description of preferred forms of embodiments given as non-limiting examples, with reference to the accompanying drawings. Although this application has been described with reference to specific examples, those skilled in the art will be able to implement many other equivalent forms of this application.

[0022] The above and other aspects, features and advantages of this application will become more apparent when taken in conjunction with the accompanying drawings and in view of the following detailed description.

[0023] Specific embodiments of this application are described thereafter with reference to the accompanying drawings; however, it should be understood that the claimed embodiments are merely examples of this application, which can be implemented in various ways. Well-known and / or repeated functions and structures are not described in detail to avoid unnecessary or redundant details that could obscure the application. Therefore, the specific structural and functional details claimed herein are not intended to be limiting, but merely serve as the basis and representative basis for the claims to teach those skilled in the art to use this application in a variety of substantially any suitable detailed structures.

[0024] Figure 1 This is a flowchart illustrating the real-time analysis method for driver guidance compliance features according to an embodiment of this application. Figure 1 As shown in the figure, the real-time analysis method for driver guidance compliance features according to an embodiment of this application includes the following processing steps: Step 101: During the vehicle's operation, collect multi-source data on vehicle guidance instructions, target vehicle status, and the status of surrounding traffic environment elements.

[0025] In this embodiment, multi-source data on guidance instructions, target vehicle status, and surrounding traffic environment status are simultaneously collected through vehicle-mounted, roadside, and cloud-based data collection units. Vehicle-mounted collection units can be vehicle-mounted cameras, vehicle-mounted speech acquisition devices, etc., while roadside collection units can be traffic cameras from transportation departments, etc.

[0026] The guidance instructions include the instruction type, instruction content, and the desired future driving state of the target vehicle. Guidance instruction types include pure longitudinal control instructions, pure lateral control instructions, and composite control instructions. When the guidance instruction only includes a recommended vehicle control strategy but not a recommended driving state, the desired future driving state of the target vehicle is calculated based on vehicle kinematics formulas. The desired future driving state of the target vehicle refers to the driving state that the target vehicle, as recommended in the guidance instruction, should follow over a subsequent period. When the guidance instruction only provides a recommended vehicle control strategy without a specific recommended driving state, it can be calculated using vehicle kinematics formulas.

[0027] The target vehicle status information includes the vehicle's lateral and longitudinal position, speed, acceleration, and heading angle; the status of surrounding traffic environment elements includes high-precision map information and status information of other motor vehicles, non-motor vehicles, and pedestrians.

[0028] In this embodiment of the application, the data collected in step 101 is further aligned by resampling and interpolation based on a set time series or the time series of one of the data, so as to establish a correlation between data of different dimensions.

[0029] As an example, the guidance instructions and timestamps received by the target vehicle are obtained through onboard, roadside, and cloud units. Assuming the received timestamp is 1758677175000 milliseconds, the received guidance instruction type is a pure longitudinal control instruction, and the content of the received guidance instruction is "reduce the vehicle speed to 10 m / s in 10 seconds and maintain a constant speed for 1 minute," if no specific recommended driving state exists, it can be calculated using vehicle kinematics formulas. Assuming the vehicle's current speed is 20 m / s and it reduces its speed to 10 m / s exactly in 10 seconds, the vehicle acceleration is -1 m / s². 2 From this, the desired future driving state of the target vehicle can be calculated, including the desired longitudinal position and speed of the target vehicle, as shown in the figure. Figure 2 As shown, the desired driving state expression is as follows: .

[0030] The target vehicle's status and timestamp are obtained through onboard, roadside, and cloud units. Assuming the received timestamp is 1758677178500 milliseconds, and the received target vehicle status information includes longitudinal position, lateral position, velocity, acceleration, and heading angle of 50.0m, 3.5m, 20m / s, and 0.0m / s², respectively. 2 and 0°.

[0031] The surrounding traffic environment status is acquired through vehicle-mounted, roadside, and cloud-based units. Assuming the main relevant traffic elements are vehicles moving in front of the target vehicle, and the received timestamp is 1758677178600 milliseconds, the received information on the vehicle ahead includes longitudinal position, lateral position, velocity, acceleration, and heading angle of 150.0 m, 3.5 m, 10 m / s, and 0.0 m / s, respectively. 2 and 0°.

[0032] Assuming the receiving frequency of the above data is 1Hz, and using the receiving time of the guidance command as a reference, linear interpolation is used to obtain the data of the target vehicle at timestamp 1758677178000 milliseconds. Assuming the target vehicle's status information at timestamp 1758677177500 milliseconds is 30.0m, 3.5m, 20m / s, and 0.0m / s respectively. 2 If 0° is used, then by linear interpolation, the state information of the target vehicle at timestamp 1758677178000 milliseconds can be obtained as 40.0m, 3.5m, 20m / s, and 0.0m / s, respectively. 2 and 0° The above is merely an illustrative example and is not intended to limit the technical means of this application.

[0033] Step 102: Determine the desired future driving state of the target vehicle according to the guidance instructions, and predict the future driving state of the target vehicle based on the current state of the target vehicle, the historical state of the target vehicle, and the state of the surrounding traffic environment elements.

[0034] In this embodiment of the application, the future driving state of the target vehicle is predicted based on the current state of the target vehicle, the historical state of the target vehicle, and the state of the surrounding traffic environment elements. Specifically, this includes: Based on the guidance duration involved in the guidance command, the guidance area targeted by the guidance command, and the maximum prediction duration of the prediction model, the shortest of these factors is determined as the effective time window for the target vehicle to respond to the guidance command. T p ;Specifically, ;in, T p This indicates the effective time window for the target vehicle to respond to guidance instructions. T th This indicates the maximum duration set to consider the performance of the prediction model. T t Indicates the boot duration involved in the boot commands. T d This indicates the travel time for the area covered by the guidance instruction. T dIt can be calculated based on the control area involved in the guidance instructions and the average traffic speed in the area. min(·) means taking the minimum value.

[0035] Based on the target vehicle's current driving state, historical driving state, and surrounding traffic environment conditions, a vehicle driving state prediction model or a vehicle driving behavior model is used to predict the target vehicle's future driving state. The prediction time window is from the current moment to the end of the effective time window of the guidance command, expressed as: ,in, This indicates the predicted future driving status of the target vehicle. Indicates the current driving status of the target vehicle. This indicates the historical driving status of the target vehicle. x , y , v , a and ψ These represent the target vehicle's longitudinal position, lateral position, velocity, acceleration, and azimuth angle, respectively. This indicates the status of traffic environment elements surrounding the target vehicle. This refers to a vehicle trajectory prediction model or a vehicle driving behavior model.

[0036] In this embodiment of the application, the vehicle trajectory prediction model may include a long short-term memory network or a Transformer model, and the vehicle driving behavior model may include a desired safety margin model or an intelligent driver model.

[0037] As an example, the effective time window for the target vehicle to respond to the guidance instruction is determined based on the guidance instruction information. Taking into account the time window set by the guidance instruction, the guidance area targeted by the instruction, and the maximum prediction duration of the prediction model, based on the aforementioned guidance instruction content, the guidance duration involved in this instruction is 70 seconds. Assuming the length of the guidance area targeted by the instruction is 1 km and the average traffic speed in the area is 20 m / s, then the travel time in the area involved by the instruction is 50 seconds. Assuming the maximum duration set by the prediction model is 120 seconds, the effective time window for the target vehicle to respond to the guidance instruction is 70 seconds. The calculation expression is as follows: ; Step S22: Based on the target vehicle's current driving state, historical driving state, and surrounding traffic environment conditions, predict the target vehicle's future driving state. Considering that the guidance instruction in step S11 mainly involves following the target vehicle's preceding vehicle, the driving behavior model used for prediction can be the expected safety margin model and intelligent driver model from existing research. The vehicle driving state prediction model used can be the long short-term memory network and Transformer model from existing research. Considering that the current time is 8s, the guidance instruction is issued at 5s, and the effective time window for the target vehicle to respond to the guidance instruction is 70s, the required prediction time is 67s. Assuming the predicted target vehicle driving state is as follows... Figure 3 As shown, the expression is as follows: .

[0038] Step 103: Based on the historical performance of the guidance instructions and prediction model, construct a weighted coefficient matrix that integrates the actual driving state and the predicted driving state of the vehicle.

[0039] In this embodiment of the application, the weight coefficients of the vehicle state dimension in the weight matrix are calculated according to the guidance instructions, as follows: For pure longitudinal control commands, the vehicle's longitudinal position, velocity, and acceleration have a significant impact on the calculated result of the conformation feature, while the vehicle's lateral position and heading angle have a smaller impact. The expression for the weighting coefficient matrix is ​​as follows: In the formula, This represents the weighting coefficient matrix for the vehicle state dimension in purely longitudinal control commands, with a dimension of 5×1. i x , j x , k x , m x and n x The parameters are in the range [0,1], representing the weights of the vehicle's longitudinal position, lateral position, velocity, acceleration, and heading angle, respectively.

[0040] For pure lateral control commands, the vehicle's lateral position and heading angle have a significant impact on the calculated result of the conformation feature, while the vehicle's longitudinal position, velocity, and acceleration have a smaller impact. The expression for the weighting coefficient matrix is ​​as follows: , This represents the weighting coefficient matrix for the vehicle state dimension in purely lateral control commands, with a dimension of 5×1. i y , j y , ky , m y and n y The parameters are in the range [0,1], representing the weights of the vehicle's longitudinal position, lateral position, velocity, acceleration, and heading angle, respectively.

[0041] For composite control commands, the vehicle's lateral and longitudinal position, velocity, acceleration, and heading angle have varying degrees of influence on the result of the feature calculation. The expression for the weighting coefficient matrix is ​​as follows: , This is a weighting coefficient matrix for the vehicle state dimension in composite control commands, with a dimension of 5×1. i xy , j xy , k xy , m xy and n xy The parameters are in the range [0,1], representing the weights of the vehicle's longitudinal position, lateral position, velocity, acceleration, and heading angle, respectively.

[0042] The weight coefficients of the prediction time dimension in the weight matrix are calculated based on the historical performance of the prediction model. The expression is as follows: Where E is a matrix of all 1s with dimension 1. The time elapsed from the moment the instruction was issued to the current moment. This is the weight coefficient matrix for the prediction time dimension, with dimension 1. The value ranges from [0,1], and decreases as the prediction time increases, showing a negative correlation with the accuracy of the prediction model. The time length from the current moment to the last point in the effective prediction time window. To predict the weighting coefficients for the time dimension, the dimension is... , is the length of time from the moment the guiding instruction is issued to the last time point of the predicted effective time window.

[0043] By combining the weight coefficients of the vehicle state dimension and the prediction time dimension in the weighting function, a weight matrix that can be used for weighted fusion of vehicle trajectories is obtained, as expressed below: , where Q F The weight matrix after fusion has dimensions n×5, Q I These are the weighting coefficients for the vehicle state dimension, derived from Q based on the corresponding control command type. Ix Q Iy and Q Ixy Choose from.

[0044] As an example, specifically, the weight coefficients of the vehicle state dimension in the weight matrix are calculated based on the guidance instructions. Considering that the guidance instructions include purely longitudinal control instructions, the vehicle's longitudinal position, velocity, and acceleration have a significant impact on the result of the conformation feature calculation, while the vehicle's lateral position and heading angle have a smaller impact. Assume that the weight coefficient matrix of the vehicle state dimension in the weight matrix is ​​as follows: ; The weight coefficients of the prediction time dimension in the weight matrix are calculated based on the historical performance of the prediction model. Assume the relationship between the prediction error and prediction time is as follows: ; In the formula, e ( t The ) represents the prediction error of the prediction model. t p This indicates the prediction time length and the error range of the predicted target vehicle's driving state, such as... Figure 4 As shown.

[0045] Suppose the relationship between the weight coefficients of the prediction time dimension in the weight matrix and the historical performance of the prediction model is expressed as follows: ; In the formula, q p ( t () represents the weight coefficient function of the prediction time dimension in the weight matrix. t p This indicates the predicted time length, with relevant time points such as... Figure 5 As shown, if one data point is taken per second, the weight coefficient matrix can be calculated using this function as follows: ; To supplement the impact of the actual driving state of the vehicle after the guidance command is issued, the weight coefficient expression for the prediction time dimension in the weight matrix is ​​as follows: ; By combining the weight coefficients of the vehicle state dimension and the prediction time dimension in the weighting function, a weight matrix that can be used for weighted fusion of vehicle trajectories is obtained, as expressed below: ,

[0046] Here, in the calculation of the fused weight matrix, Q I Selected from Q Ix .

[0047] Step 104: Integrate the actual driving state and predicted driving state of the vehicle, as well as the weight coefficient matrix, to calculate the driver's comprehensive compliance characteristics with the current guidance instructions in real time. The comprehensive compliance characteristics include at least response accuracy and response rate.

[0048] Specifically, the driving states that the target vehicle has already completed after receiving the guidance command and the predicted driving states in the steps are spliced ​​and fused together, as expressed by: S o ( x , y , v , a , ψ This indicates the driving state that the target vehicle has completed, with the dimension being... S F ( x , y , v , a , ψ The n×5 dimension represents the driving state after the target vehicle has completed its driving state and the predicted driving state are spliced ​​and fused together. The weighted root mean square error (RMSE) method is used to calculate the error between the stitched and fused vehicle driving state and the expected future driving state of the target vehicle as specified in the guidance instructions. This error serves as a quantification of the driver's response accuracy to the guidance instructions at that moment, and its expression is as follows: , d ( t S represents the accuracy of the driver's response to the guidance instructions at the current moment. d ( x , y , v , a , ψ The dimension represents the desired future driving state of the target vehicle as indicated by the guidance command. n ×5, mean (·) represents the average of the matrices in the function. sum (·) represents summing the matrices in the function; By comparing the driver's response accuracy to guidance instructions at the current moment with a set threshold, a percentage response rate is calculated. This rate is used to adjust guidance instructions to the target vehicle or surrounding vehicles in a timely manner. The calculation expression is as follows: , This indicates the driver's compliance rate with guidance instructions at the current moment. This indicates the set threshold for determining compliance rate.

[0049] As an example, the following expression is used to concatenate and fuse the driving state that the target vehicle has already completed after receiving the guidance command with the predicted driving state: ; The weighted root mean square error (RMSE) method is used to calculate the error between the stitched and fused vehicle driving state and the expected future driving state of the target vehicle as specified in the guidance instructions. This error serves as a quantification of the driver's response accuracy to the guidance instructions at that moment, and its expression is as follows: ; By comparing the driver's response accuracy to guidance instructions at the current moment with a set threshold, a percentage response rate can be calculated. This rate is used to adjust guidance instructions to the target vehicle or surrounding vehicles in a timely manner. Assuming the set threshold is 0.5, the response rate calculation expression is as follows: .

[0050] This application's embodiments utilize vehicle driving state data analysis to quantify the driver's compliance characteristics with guidance instructions. This avoids the need for additional driver physiological data collection equipment, effectively reducing analysis and quantification costs. By predicting the vehicle's future driving state within the effective range of the guidance instructions, it can determine the driver's compliance characteristics before the target vehicle leaves the effective range, avoiding analysis lag and providing a quantitative basis for timely adjustments to the guidance instructions. The weighting coefficient matrix is ​​designed to consider prediction model errors, dynamically balancing the actual driving state of the target vehicle after the guidance instructions are issued with the expected future driving state, effectively improving the reliability and stability of the analysis results. Based on the prediction of the vehicle's current and historical states and environmental factors, real-time calculation of response accuracy and response rate can be achieved after the guidance instructions are issued but before the driver has fully executed them, thereby predicting compliance trends in advance and reducing traffic risks caused by delayed compliance. For three types of commands—pure longitudinal (e.g., acceleration / deceleration), pure lateral (e.g., lane change / lane keeping), and composite control (e.g., curve deceleration + lane centering)—differentiated vehicle state dimension weights are configured for each. This ensures that key parameters, such as speed in longitudinal commands and heading angle in lateral commands, receive higher weights. This makes error calculations more closely aligned with the core requirements of the commands, avoids interference from long-term prediction errors on the overall analysis results, and reduces errors in response accuracy calculations. The real-time output compliance feature data in this embodiment can be directly fed back to the vehicle-road cooperative system, achieving closed-loop optimization of guidance commands and improving traffic efficiency.

[0051] Figure 6 This is a schematic diagram of the driver guidance compliance feature real-time analysis device according to an embodiment of this application, as shown below. Figure 6As shown, the real-time driver guidance compliance feature analysis device of this application embodiment includes: The data acquisition unit 60 is used to collect multi-source data on vehicle guidance instructions, target vehicle status, and surrounding traffic environment status during vehicle operation. The prediction unit 61 is used to determine the desired future driving state of the target vehicle according to the guidance instructions, and to predict the future driving state of the target vehicle based on the current state of the target vehicle, the historical state of the target vehicle and the state of the surrounding traffic environment elements. Construction unit 62 is used to construct a weight coefficient matrix that integrates the actual driving state and the predicted driving state of the vehicle based on the historical performance of the guidance instructions and the prediction model. The calculation unit 63 is used to fuse the actual driving state and predicted driving state of the vehicle, as well as the weight coefficient matrix, to calculate the driver's comprehensive compliance characteristics with the current guidance instructions in real time. The comprehensive compliance characteristics include at least response accuracy and response rate.

[0052] In some exemplary embodiments, the acquisition unit 60 is further configured to: The system acquires guidance instructions and timestamps received by the target vehicle, the target vehicle's status and timestamp, and the status of surrounding traffic environmental elements through vehicle-mounted data acquisition devices, roadside data acquisition devices, and cloud-based devices. The guidance instructions include the instruction type, instruction content, and the desired future driving state of the target vehicle. The guidance instruction types include pure longitudinal control instructions, pure lateral control instructions, and composite control instructions. When the guidance instruction only contains the recommended vehicle control strategy and not the recommended driving state, the desired future driving state of the target vehicle is calculated based on the vehicle kinematics formula. The target vehicle state information includes the vehicle's lateral and longitudinal position, speed, acceleration, and heading angle. The state of the surrounding traffic environment elements includes high-precision map information and the state information of other motor vehicles, non-motor vehicles, and pedestrians. The acquired multi-source data is aligned by resampling and interpolation based on a set time series or the time series of one of the data.

[0053] In some exemplary embodiments, the building unit 62 is further configured to: The shortest of the following factors is determined as the effective time window for the target vehicle to respond to the guidance command: the guidance duration involved in the guidance command, the guidance area targeted by the guidance command, and the maximum prediction duration of the prediction model. Based on the target vehicle's current driving state, historical driving state, and surrounding traffic environment conditions, a vehicle driving state prediction model or a vehicle driving behavior model is used to predict the target vehicle's future driving state. The prediction time window is from the current moment to the end of the effective time window of the guidance command, expressed as: ,in, This indicates the predicted future driving status of the target vehicle. Indicates the current driving status of the target vehicle. This indicates the historical driving status of the target vehicle. x , y , v , a and ψ These represent the target vehicle's longitudinal position, lateral position, velocity, acceleration, and azimuth angle, respectively. This indicates the status of traffic environment elements surrounding the target vehicle. This refers to a vehicle trajectory prediction model or a vehicle driving behavior model.

[0054] In this embodiment, the vehicle trajectory prediction model includes a long short-term memory network and a Transformer model, and the vehicle driving behavior model includes a desired safety margin model and an intelligent driver model.

[0055] In some exemplary embodiments, the building unit 62 is further configured to: The weight coefficients of the vehicle state dimension in the weight matrix are calculated based on the guidance instructions, as follows: For pure longitudinal control commands, the vehicle's longitudinal position, velocity, and acceleration have a significant impact on the calculated result of the conformation feature, while the vehicle's lateral position and heading angle have a smaller impact. The expression for the weighting coefficient matrix is ​​as follows: In the formula, This represents the weighting coefficient matrix for the vehicle state dimension in purely longitudinal control commands, with a dimension of 5×1. i x , j x , k x , m x and n x The parameters are in the range [0,1], representing the weights of the vehicle's longitudinal position, lateral position, velocity, acceleration, and heading angle, respectively. For pure lateral control commands, the vehicle's lateral position and heading angle have a significant impact on the calculated result of the conformation feature, while the vehicle's longitudinal position, velocity, and acceleration have a smaller impact. The expression for the weighting coefficient matrix is ​​as follows: , This represents the weighting coefficient matrix for the vehicle state dimension in purely lateral control commands, with a dimension of 5×1. i y , j y , k y , my and n y The parameters are in the range [0,1], representing the weights of the vehicle's longitudinal position, lateral position, velocity, acceleration, and heading angle, respectively. For composite control commands, the vehicle's lateral and longitudinal position, velocity, acceleration, and heading angle have varying degrees of influence on the result of the feature calculation. The expression for the weighting coefficient matrix is ​​as follows: , This is a weighting coefficient matrix for the vehicle state dimension in composite control commands, with a dimension of 5×1. i xy , j xy , k xy , m xy and n xy The parameters are in the range [0,1], representing the weights of the vehicle's longitudinal position, lateral position, velocity, acceleration, and heading angle, respectively. The weight coefficients of the prediction time dimension in the weight matrix are calculated based on the historical performance of the prediction model. The expression is as follows: Where E is a matrix of all 1s with dimension 1. The time elapsed from the moment the instruction was issued to the current moment. This is the weight coefficient matrix for the prediction time dimension, with dimension 1. The value ranges from [0,1], and decreases as the prediction time increases, showing a negative correlation with the accuracy of the prediction model. The time length from the current moment to the last point in the effective prediction time window. To predict the weighting coefficients for the time dimension, the dimension is... This refers to the time from the issuance of the guiding instruction to the last time point of the predicted effective time window. By combining the weight coefficients of the vehicle state dimension and the prediction time dimension in the weighting function, a weight matrix that can be used for weighted fusion of vehicle trajectories is obtained, as expressed below: , where Q F The weight matrix after fusion has dimensions n×5, Q I These are the weighting coefficients for the vehicle state dimension, derived from Q based on the corresponding control command type. Ix Q Iy and Q Ixy Choose from.

[0056] In some exemplary embodiments, the computing unit 64 is further configured to: The driving states that the target vehicle has completed after receiving the guidance command and the predicted driving states in the steps are spliced ​​and fused together, and the expression is: S o ( x , y , v , a , ψ This indicates the driving state that the target vehicle has completed, with the dimension being... S F ( x , y , v , a , ψ The n×5 dimension represents the driving state after the target vehicle has completed its driving state and the predicted driving state are spliced ​​and fused together. The weighted root mean square error (RMSE) method is used to calculate the error between the stitched and fused vehicle driving state and the expected future driving state of the target vehicle as specified in the guidance instructions. This error serves as a quantification of the driver's response accuracy to the guidance instructions at that moment, and its expression is as follows: , d ( t S represents the accuracy of the driver's response to the guidance instructions at the current moment. d ( x , y , v , a , ψ The dimension represents the desired future driving state of the target vehicle as indicated by the guidance command. n ×5, mean (·) represents the average of the matrices in the function. sum (·) represents summing the matrices in the function; By comparing the driver's response accuracy to guidance instructions at the current moment with a set threshold, a percentage response rate is calculated. This rate is used to adjust guidance instructions to the target vehicle or surrounding vehicles in a timely manner. The calculation expression is as follows: , This indicates the driver's compliance rate with guidance instructions at the current moment. This indicates the set threshold for determining compliance rate.

[0057] In an exemplary embodiment, the aforementioned units may be implemented by one or more central processing units (CPUs), graphics processing units (GPUs), application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), general-purpose processors, controllers, micro controller units (MCUs), microprocessors, or other electronic components.

[0058] Regarding the apparatus in the above embodiments, the specific manner in which each module and unit performs its operations has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0059] Figure 7 This is a schematic diagram of the composition structure of the electronic device according to an embodiment of this application, such as... Figure 7 As shown, the electronic device 800 supports multi-screen output and may include one or more of the following components: processing component 802, memory 804, power supply component 806, multimedia component 808, audio component 810, input / output (I / O) interface 812, sensor component 814, and communication component 816.

[0060] Processing component 802 typically controls the overall operation of electronic device 800, such as operations associated with display, telephone calls, data communication, camera operation, and recording operations. Processing component 802 may include one or more processors 820 to execute instructions to complete all or part of the steps of the methods described above. Furthermore, processing component 802 may include one or more modules to facilitate interaction between processing component 802 and other components. For example, processing component 802 may include a multimedia module to facilitate interaction between multimedia component 808 and processing component 802.

[0061] Memory 804 is configured to store various types of data to support the operation of electronic device 800. Examples of such data include instructions for any application or method operating on electronic device 800, contact data, phonebook data, messages, pictures, videos, etc. Memory 804 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0062] Power supply component 806 provides power to various components of electronic device 800. Power supply component 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to electronic device 800.

[0063] Multimedia component 808 includes a screen that provides an output interface between the electronic device 800 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touchscreen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensors may sense not only the boundaries of the touch or swipe action but also the duration and pressure associated with the touch or swipe operation. In some embodiments, multimedia component 808 includes a front-facing camera and / or a rear-facing camera. When the electronic device 800 is in an operating mode, such as a shooting mode or a video mode, the front-facing camera and / or the rear-facing camera may receive external multimedia data. Each front-facing camera and rear-facing camera may be a fixed optical lens system or have focal length and optical zoom capabilities.

[0064] Audio component 810 is configured to output and / or input audio signals. For example, audio component 810 includes a microphone (MIC) configured to receive external audio signals when electronic device 800 is in an operating mode, such as call mode, recording mode, and voice recognition mode. The received audio signals may be further stored in memory 804 or transmitted via communication component 816. In some embodiments, audio component 810 also includes a speaker for outputting audio signals.

[0065] I / O interface 812 provides an interface between processing component 802 and peripheral interface modules, such as keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to, home buttons, volume buttons, power buttons, and lock buttons.

[0066] Sensor assembly 814 includes one or more sensors for providing state assessments of various aspects of electronic device 800. For example, sensor assembly 814 can detect the on / off state of electronic device 800, the relative positioning of components such as the display and keypad of electronic device 800, changes in position of electronic device 800 or a component of electronic device 800, the presence or absence of user contact with electronic device 800, orientation or acceleration / deceleration of electronic device 800, and temperature changes of electronic device 800. Sensor assembly 814 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. Sensor assembly 814 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, sensor assembly 814 may also include an accelerometer, gyroscope, magnetometer, pressure sensor, or temperature sensor.

[0067] Communication component 816 is configured to facilitate wired or wireless communication between electronic device 800 and other devices. Electronic device 800 can access wireless networks based on communication standards, such as Wi-Fi, 2G, or 3G, or combinations thereof. In one exemplary embodiment, communication component 816 receives broadcast signals or broadcast-related information from an external broadcast management system via a broadcast channel. In one exemplary embodiment, communication component 816 also includes a near-field communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on radio frequency identification (RFID) technology, Infrared Data Association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.

[0068] In an exemplary embodiment, the electronic device 800 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the steps of the stand-based driver guidance compliance feature real-time analysis method of the above embodiments.

[0069] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 804 including instructions, which can be executed by a processor 820 of an electronic device 800 to complete the steps of the real-time driver guidance compliance feature analysis method of the above embodiments. For example, the non-transitory computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0070] This application also describes a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the real-time analysis method for driver guidance compliance features described in the embodiment.

[0071] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of the invention. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification do not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. It should be understood that in the various embodiments of the invention, the sequence numbers of the above-described processes do not imply a sequential order of execution; the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application. The sequence numbers of the above-described embodiments are merely descriptive and do not represent the superiority or inferiority of the embodiments.

[0072] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0073] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not present.

[0074] The units described above as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0075] The above description is merely an embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for real-time analysis of driver guidance compliance characteristics, characterized in that, The method includes: During vehicle operation, multi-source data are collected, including vehicle guidance instructions, target vehicle status, and status of surrounding traffic environment elements. The desired future driving state of the target vehicle is determined according to the guidance instructions, and the future driving state of the target vehicle is predicted based on the current state of the target vehicle, the historical state of the target vehicle, and the state of the surrounding traffic environment elements. Based on the historical performance of the guidance instructions and prediction model, a weighted coefficient matrix that integrates the actual driving state and the predicted driving state of the vehicle is constructed. By integrating the actual driving status and predicted driving status of the vehicle with a weighted coefficient matrix, the driver's comprehensive compliance characteristics with the current guidance instructions are calculated in real time. The comprehensive compliance characteristics include at least response accuracy and response rate.

2. The method according to claim 1, characterized in that, The multi-source data collected, including vehicle guidance instructions, target vehicle status, and surrounding traffic environment status, includes: The system acquires guidance instructions and timestamps received by the target vehicle, the target vehicle's status and timestamp, and the status of surrounding traffic environmental elements through vehicle-mounted data acquisition devices, roadside data acquisition devices, and cloud-based devices. The guidance instructions include the instruction type, instruction content, and the desired future driving state of the target vehicle. The guidance instruction types include pure longitudinal control instructions, pure lateral control instructions, and composite control instructions. When the guidance instruction only contains the recommended vehicle control strategy and not the recommended driving state, the desired future driving state of the target vehicle is calculated based on the vehicle kinematics formula. The target vehicle's status information includes its lateral and longitudinal positions, speed, acceleration, and heading angle; The status of surrounding traffic environment elements includes high-precision map information and status information of other motor vehicles, non-motor vehicles, and pedestrians; The acquired multi-source data is aligned by resampling and interpolation based on a set time series or the time series of one of the data.

3. The method according to claim 1, characterized in that, The prediction of the future driving state of the target vehicle based on its current state, historical state, and surrounding traffic environment factors includes: The shortest of the following factors is determined as the effective time window for the target vehicle to respond to the guidance command: the guidance duration involved in the guidance command, the guidance area targeted by the guidance command, and the maximum prediction duration of the prediction model. Based on the target vehicle's current driving state, historical driving state, and surrounding traffic environment conditions, a vehicle driving state prediction model or a vehicle driving behavior model is used to predict the target vehicle's future driving state. The prediction time window is from the current moment to the end of the effective time window of the guidance command, expressed as: ,in, This indicates the predicted future driving status of the target vehicle. Indicates the current driving status of the target vehicle. This represents the historical driving state of the target vehicle, where x, y, v, a, and ψ represent the target vehicle's longitudinal position, lateral position, velocity, acceleration, and azimuth angle, respectively. This indicates the status of traffic environment elements surrounding the target vehicle. This refers to a vehicle trajectory prediction model or a vehicle driving behavior model.

4. The method according to claim 1, characterized in that, The construction of the weighted coefficient matrix that integrates the actual driving state and the predicted driving state of the vehicle includes: The weight coefficients of the vehicle state dimension in the weight matrix are calculated based on the guidance instructions, as follows: For pure longitudinal control commands, the vehicle's longitudinal position, velocity, and acceleration have a significant impact on the calculated result of the conformation feature, while the vehicle's lateral position and heading angle have a smaller impact. The expression for the weighting coefficient matrix is ​​as follows: In the formula, This represents the weighting coefficient matrix for the vehicle state dimension in purely longitudinal control commands, with a dimension of 5×1. i x , j x , k x , m x and n x The parameters are in the range [0,1], representing the weights of the vehicle's longitudinal position, lateral position, velocity, acceleration, and heading angle, respectively. For purely lateral control commands, the vehicle's lateral position and heading angle have a significant impact on the calculated result of the conformation feature, while the vehicle's longitudinal position, velocity, and acceleration have a smaller impact. The expression for the weighting coefficient matrix is ​​as follows: , This represents the weighting coefficient matrix for the vehicle state dimension in purely lateral control commands, with a dimension of 5×1. i y , j y , k y , m y and n y The parameters are in the range [0,1], representing the weights of the vehicle's longitudinal position, lateral position, velocity, acceleration, and heading angle, respectively. For composite control commands, the vehicle's lateral and longitudinal position, velocity, acceleration, and heading angle have varying degrees of influence on the result of the feature calculation. The expression for the weighting coefficient matrix is ​​as follows: , This is a weighting coefficient matrix for the vehicle state dimension in composite control commands, with a dimension of 5×1. i xy , j xy , k xy , m xy and n xy The parameters are in the range [0,1], representing the weights of the vehicle's longitudinal position, lateral position, velocity, acceleration, and heading angle, respectively. The weight coefficients of the prediction time dimension in the weight matrix are calculated based on the historical performance of the prediction model. The expression is as follows: Where E is a matrix of all 1s with dimension 1. The time elapsed from the moment the instruction was issued to the current moment. This is the weight coefficient matrix for the prediction time dimension, with dimension 1. The value ranges from [0,1], and decreases as the prediction time increases, showing a negative correlation with the accuracy of the prediction model. The time length from the current moment to the last point in the effective prediction time window. To predict the weighting coefficients for the time dimension, the dimension is... This refers to the time from the issuance of the guiding instruction to the last time point of the predicted effective time window. By combining the weight coefficients of the vehicle state dimension and the prediction time dimension in the weighting function, a weight matrix that can be used for weighted fusion of vehicle trajectories is obtained, as expressed below: , where Q F The weight matrix after fusion has dimensions n×5, Q I These are the weighting coefficients for the vehicle state dimension, derived from Q based on the corresponding control command type. Ix Q Iy and Q Ixy Choose from.

5. The method according to claim 1, characterized in that, The method further includes: The driving states that the target vehicle has completed after receiving the guidance command and the predicted driving states in the steps are spliced ​​and fused together, and the expression is: S o ( x , y , v , a , ψ This indicates the driving state that the target vehicle has completed, with the dimension being... S F ( x , y , v , a , ψ The n×5 dimension represents the driving state after the target vehicle has completed its driving state and the predicted driving state are spliced ​​and fused together. The weighted root mean square error (RMSE) method is used to calculate the error between the stitched and fused vehicle driving state and the expected future driving state of the target vehicle as specified in the guidance instructions. This error serves as a quantification of the driver's response accuracy to the guidance instructions at that moment, and its expression is as follows: ， δ ( t S represents the accuracy of the driver's response to the guidance instructions at the current moment. d ( x , y , v , a , ψ The dimension represents the desired future driving state of the target vehicle as indicated by the guidance command. n ×5, mean (·) represents the average of the matrices in the function. sum (·) represents summing the matrices in the function; By comparing the driver's response accuracy to guidance instructions at the current moment with a set threshold, a percentage response rate is calculated. This rate is used to adjust guidance instructions to the target vehicle or surrounding vehicles in a timely manner. The calculation expression is as follows: , This indicates the driver's compliance rate with guidance instructions at the current moment. This indicates the set threshold for determining compliance rate.

6. A real-time analysis device for driver guidance compliance characteristics, characterized in that, The device includes: The data acquisition unit is used to collect multi-source data on vehicle guidance instructions, target vehicle status, and surrounding traffic environment status during vehicle operation. The prediction unit is used to determine the desired future driving state of the target vehicle according to the guidance instructions, and to predict the future driving state of the target vehicle based on the current state of the target vehicle, the historical state of the target vehicle and the state of the surrounding traffic environment elements. The construction unit is used to construct a weight coefficient matrix that integrates the actual driving state and the predicted driving state of the vehicle based on the historical performance of the guidance instructions and the prediction model. The computing unit is used to integrate the actual driving state and predicted driving state of the vehicle, as well as the weight coefficient matrix, to calculate the driver's comprehensive compliance characteristics with the current guidance instructions in real time. The comprehensive compliance characteristics include at least response accuracy and response rate.

7. The apparatus according to claim 6, characterized in that, The acquisition unit is also used for: The system acquires guidance instructions and timestamps received by the target vehicle, the target vehicle's status and timestamp, and the status of surrounding traffic environmental elements through vehicle-mounted data acquisition devices, roadside data acquisition devices, and cloud-based devices. The guidance instructions include the instruction type, instruction content, and the desired future driving state of the target vehicle. The guidance instruction types include pure longitudinal control instructions, pure lateral control instructions, and composite control instructions. When the guidance instruction only contains the recommended vehicle control strategy and not the recommended driving state, the desired future driving state of the target vehicle is calculated based on the vehicle kinematics formula. The target vehicle state information includes the vehicle's lateral and longitudinal position, speed, acceleration, and heading angle. The state of the surrounding traffic environment elements includes high-precision map information and the state information of other motor vehicles, non-motor vehicles, and pedestrians. The acquired multi-source data is aligned by resampling and interpolation based on a set time series or the time series of one of the data.

8. The apparatus according to claim 6, characterized in that, The building unit is also used for: The shortest of the following factors is determined as the effective time window for the target vehicle to respond to the guidance command: the guidance duration involved in the guidance command, the guidance area targeted by the guidance command, and the maximum prediction duration of the prediction model. Based on the target vehicle's current driving state, historical driving state, and surrounding traffic environment conditions, a vehicle driving state prediction model or a vehicle driving behavior model is used to predict the target vehicle's future driving state. The prediction time window is from the current moment to the end of the effective time window of the guidance command, expressed as: ,in, This indicates the predicted future driving status of the target vehicle. Indicates the current driving status of the target vehicle. This represents the historical driving state of the target vehicle, where x, y, v, a, and ψ represent the target vehicle's longitudinal position, lateral position, velocity, acceleration, and azimuth angle, respectively. This indicates the status of traffic environment elements surrounding the target vehicle. This refers to a vehicle trajectory prediction model or a vehicle driving behavior model. Vehicle trajectory prediction models include Long Short-Term Memory Networks and Transformer models, while vehicle driving behavior models include Expected Safety Margin Models and Intelligent Driver Models.

9. The apparatus according to claim 6, characterized in that, The building unit is also used for: The weight coefficients of the vehicle state dimension in the weight matrix are calculated based on the guidance instructions, as follows: For pure longitudinal control commands, the vehicle's longitudinal position, velocity, and acceleration have a significant impact on the calculated result of the conformation feature, while the vehicle's lateral position and heading angle have a smaller impact. The expression for the weighting coefficient matrix is ​​as follows: In the formula, This represents the weighting coefficient matrix for the vehicle state dimension in purely longitudinal control commands, with a dimension of 5×1. i x , j x , k x , m x and n x The parameters are in the range [0,1], representing the weights of the vehicle's longitudinal position, lateral position, velocity, acceleration, and heading angle, respectively. For purely lateral control commands, the vehicle's lateral position and heading angle have a significant impact on the calculated result of the conformation feature, while the vehicle's longitudinal position, velocity, and acceleration have a smaller impact. The expression for the weighting coefficient matrix is ​​as follows: , This represents the weighting coefficient matrix for the vehicle state dimension in purely lateral control commands, with a dimension of 5×1. i y , j y , k y , m y and n y The parameters are in the range [0,1], representing the weights of the vehicle's longitudinal position, lateral position, velocity, acceleration, and heading angle, respectively. For composite control commands, the vehicle's lateral and longitudinal position, velocity, acceleration, and heading angle have varying degrees of influence on the result of the feature calculation. The expression for the weighting coefficient matrix is ​​as follows: , This is a weighting coefficient matrix for the vehicle state dimension in composite control commands, with a dimension of 5×1. i xy , j xy , k xy , m xy and n xy The parameters are in the range [0,1], representing the weights of the vehicle's longitudinal position, lateral position, velocity, acceleration, and heading angle, respectively. The weight coefficients of the prediction time dimension in the weight matrix are calculated based on the historical performance of the prediction model. The expression is as follows: Where E is a matrix of all 1s with dimension 1. The time elapsed from the moment the instruction was issued to the current moment. This is the weight coefficient matrix for the prediction time dimension, with dimension 1. The value ranges from [0,1], and decreases as the prediction time increases, showing a negative correlation with the accuracy of the prediction model. The time length from the current moment to the last point in the effective prediction time window. To predict the weighting coefficients for the time dimension, the dimension is... This refers to the time from the issuance of the guiding instruction to the last time point of the predicted effective time window. By combining the weight coefficients of the vehicle state dimension and the prediction time dimension in the weighting function, a weight matrix that can be used for weighted fusion of vehicle trajectories is obtained, as expressed below: , where Q F The weight matrix after fusion has dimensions n×5, Q I These are the weighting coefficients for the vehicle state dimension, derived from Q based on the corresponding control command type. Ix Q Iy and Q Ixy Choose from.

10. The apparatus according to claim 6, characterized in that, The computing unit is also used for: The driving states that the target vehicle has completed after receiving the guidance command and the predicted driving states in the steps are spliced ​​and fused together, and the expression is: S o ( x , y , v , a , ψ This indicates the driving state that the target vehicle has completed, with the dimension being... S F ( x , y , v , a , ψ The n×5 dimension represents the driving state after the target vehicle has completed its driving state and the predicted driving state are spliced ​​and fused together. The weighted root mean square error (RMSE) method is used to calculate the error between the stitched and fused vehicle driving state and the expected future driving state of the target vehicle as specified in the guidance instructions. This error serves as a quantification of the driver's response accuracy to the guidance instructions at that moment, and its expression is as follows: ， δ ( t S represents the accuracy of the driver's response to the guidance instructions at the current moment. d ( x , y , v , a , ψ The dimension represents the desired future driving state of the target vehicle as indicated by the guidance command. n ×5, mean (·) represents the average of the matrices in the function. sum (·) represents summing the matrices in the function; By comparing the driver's response accuracy to guidance instructions at the current moment with a set threshold, a percentage response rate is calculated. This rate is used to adjust guidance instructions to the target vehicle or surrounding vehicles in a timely manner. The calculation expression is as follows: , This indicates the driver's compliance rate with guidance instructions at the current moment. This indicates the set threshold for determining compliance rate.

Images