Heavy truck driver behavior evaluation and economical driving guide method

Abstract

The invention discloses a heavy truck driver behavior evaluation and economical driving guidance method, which comprises the following steps of: extracting vehicle operation data required by a model from a multi-source data set by using a big data extraction technology so as to finish data abnormal frame identification, data cleaning and vehicle feature expansion, and constructing a driving behavior data set based on operation interruption moment segmentation; constructing a clustering analysis model of the driving behavior data set by adopting a K-means algorithm, and generating four types of driving behavior clusters of high-speed driving, relieving, moderating and aggressive driving behavior clusters; for aggressive driving behavior segments, a Transform instantaneous fuel consumption rate prediction model is adopted to evaluate the energy consumption level, then a dynamic planning algorithm model with the optimal energy consumption as the target is established, and energy consumption level comparison and optimal fuel economy strategy planning of the driving behavior segments are completed. Driving behavior type identification is realized based on vehicle networking data, an economical operation scheme is provided for a high-energy-consumption driving segment, on-road driving operation of a driver can be standardized, and a reliable strategy is provided for optimization of fuel consumption.

Description

Technical Field

[0001] The present invention belongs to the field of vehicle economical driving, and in particular relates to a heavy truck driver behavior assessment and economical driving guidance method. Background Art

[0002] Social and economic development has driven the improvement of consumption levels, and the improvement of transportation facilities has corresponded to the progress of transportation tools. The number of motor vehicles in my country has continued to increase. Fuel-based cars developed in the industrial era still dominate motor vehicles and have become a means of transportation for more and more people. By 2023, the number of cars in use has reached 336 million. The fuel consumption and environmental pollution caused by car driving have become more severe, and the demand for improving vehicle fuel economy and energy conservation and emission reduction is also increasing. The current transportation industry is mainly based on shipping, air transportation, rail transportation, road transportation, and pipeline transportation. Road transportation ranks first among various modes of transportation with 73.8% of the total comprehensive transportation freight volume. Heavy-duty trucks dominate highway freight with their superior transportation capacity, transportation efficiency, and a wide range of transportation. The high operating cost of heavy-duty trucks restricts the development of the heavy-duty truck market, and fuel consumption cost is an important reason for the high operating cost of heavy-duty trucks. Therefore, the present invention is studied with the purpose of reducing the fuel consumption of heavy-duty trucks.

[0003] There are many factors that affect the fuel consumption of heavy-duty trucks, and the relationships between these factors are complex and mutually influential. The vehicle structure has become more complete, and it is difficult to effectively improve the external environmental factors in a short period of time. Driving behavior, as a relatively controllable factor, has attracted the attention of many scholars and engineers at a time when automation research is a hot topic. Scientific driving behavior can effectively reduce unnecessary fuel consumption, body wear and exhaust emissions, improve transportation efficiency and vehicle service life, and due to technological development, the on-board terminal OBD (On Board Diagnostics) equipment can collect real-time vehicle operation status data, which contains parameters related to driver behavior, such as gear information, engine start and braking time, throttle opening, etc., making complete driver behavior analysis possible.

[0004] The vehicle fuel consumption prediction model can map driving behavior through fuel consumption and driving characteristics, thereby predicting fuel consumption by simulating driving behavior and comparing fuel-saving strategies. Summarizing existing driving data and corresponding driving behavior fuel consumption can provide a more accurate basis for fuel-saving driving suggestions or driving behavior corrections, and using dynamic programming to prompt drivers before driving behaviors occur, so as to achieve effective application of fuel-saving strategies. Therefore, the study of driving behavior can focus on the vehicle fuel consumption prediction model, and on this basis, use dynamic programming to optimize fuel-saving strategies. Summary of the invention

[0005] Purpose of the invention: The present invention provides a heavy truck driver behavior assessment and economical driving guidance method, which can realize the identification of heavy truck driver behavior type and economical speed planning of high energy consumption driving segments.

[0006] Technical solution: A heavy truck driver behavior assessment and economical driving guidance method according to the present invention comprises the following steps:

[0007] (1) Multi-source data acquisition: Data acquired through the GPS module includes timestamp, elevation, and vehicle speed; data acquired through the CAN bus includes engine speed, engine required torque, and gear information;

[0008] (2) Data cleaning and feature expansion: Implement corresponding cleaning strategies for missing, noisy and other erroneous frame data in the operating data, and take pre-processing measures such as interpolation and filtering for the operating data that does not meet the continuity conditions; then obtain the road slope characteristics, vehicle acceleration and jerk characteristics of the entire road section at all times through the pre-processed data;

[0009] (3) Construction of driver behavior dataset: The trip segments are divided according to the parking and shutdown time, and the extreme value, mean, standard deviation and other characteristic indicators of the vehicle operation data in each trip segment are obtained. The strongly related features are extracted based on the principal component analysis algorithm to form a driver behavior feature dataset;

[0010] (4) Clustering evaluation of driver behavior characteristics: Based on the K-means algorithm, cluster analysis is performed on the segmented driving behavior data set, and four cluster centers are constructed: high-speed driving, slow, moderate, and aggressive. The clustering results of each driving behavior segment are then output;

[0011] (5) Driving segment fuel consumption prediction: A Transformer-based instantaneous fuel consumption rate prediction model is constructed. The model inputs the preprocessed vehicle operation characteristics and extended road structure characteristics, and outputs the instantaneous fuel consumption rate level of each driving behavior segment;

[0012] (6) Economical driving speed planning for driving segments: A dynamic programming model is constructed with the goal of minimizing energy consumption. The cost function selects the instantaneous fuel consumption rate model constructed above to solve and output the economical driving speed plan with the optimal energy consumption.

[0013] Furthermore, the implementation process of step (2) is as follows:

[0014] Considering the differences in the operating characteristics of non-homologous data, artificial features are extracted and marked for the error frames in the GPS source data and CAN bus source data respectively, and the error frame data samples are separated based on the continuity criterion. The error frame features and cleaning strategies of multi-source data are formulated as follows.

[0015] GPS source data: Error frames mainly come from data interruptions. Combined with the data feedback timestamp, they can be divided into short-term data loss and long-term signal interruption. The time threshold δ is set to 30s. For short-term data loss less than this threshold, linear interpolation is performed using the nearest non-error data value before and after the error frame. For long-term signal interruptions exceeding this threshold, the data during the period is set to zero and discarded.

[0016] CAN bus source data: Error frame data is mainly characterized by short-term data loss, high-frequency measurement noise and discontinuous mutations.

[0017] Short-term data missing is mainly aimed at vehicle gear data detection. Based on the principle of continuous change of vehicle gear, short-term single-point missing data is filled with non-abnormal data at the previous moment.

[0018] High-frequency measurement noise is mainly used for engine speed and engine torque data detection, and a Savitzky-Golay filter is designed to process the noise data.

[0019] Discontinuous mutations are mainly aimed at engine torque data detection, and the characteristic detection standards are set as follows:

[0020]

[0021] Among them, R is the engine torque data at a certain moment, T is the timestamp at the corresponding moment, t is the calculation moment, z is the time step, and δ is the set detection threshold; for the data points obtained according to the above detection standards, the linear interpolation method is used to cover the data values ​​at the corresponding moment.

[0022] After the vehicle operation data is cleaned, the vehicle operation speed characteristics are obtained based on the vehicle parameter attributes and operation characteristics:

[0023]

[0024] Then the vehicle running acceleration and jerk characteristics are obtained:

[0025]

[0026] Where n is the engine speed, r is the wheel rolling radius, i g is the transmission ratio, i0 is the rear axle speed ratio, Δt is the time step, V t is the vehicle speed, a t is the acceleration value, j is the jerk;

[0027] Use the preprocessed speed data and elevation data to obtain the road slope:

[0028]

[0029] Where l is the time step of backward calculation; h t is the elevation value at time t, h t+n is the elevation value at time t+n, in meters; v i is the vehicle speed value, in units of

[0030] The point slope calculation is performed according to the distance span of 100m, and the elevation data is guaranteed to change monotonically within a single calculation step, and the time step k satisfies:

[0031]

[0032] (h i -h i-1 )·(h i+1 -h i )>0,(i=t+1,t+2,…,t+n-1)

[0033] Based on the above calculation strategy, the road slope characteristics, vehicle acceleration and jerk characteristics of the entire road section and at all times are obtained.

[0034] Furthermore, the implementation process of step (3) is as follows:

[0035] A parking and ignition-off duration threshold Δt is set, the time point when the parking and ignition-off duration exceeds the threshold is marked, and the vehicle operation data between the two parking time points is recorded as a driving journey segment, which is the driving behavior segment division.

[0036] Based on the pre-processed vehicle operation data, the extreme value, mean, standard deviation and other characteristic indicators of speed, acceleration, throttle and other data (throttle data obtained from the CAN bus) are expanded. The complete 20 driving behavior characteristic parameter indicators are shown in the following table.

[0037] The Pearson correlation coefficient is used to evaluate the correlation between the 20 characteristic parameters and the fuel consumption rate, and then the principal component analysis algorithm is used to evaluate the cumulative variance contribution rate of the complete characteristic parameters. According to the driving trip segments divided above, the dimension reduction feature group of the vehicle operation data in each segment is obtained to form the driver behavior feature data set.

[0038] Furthermore, the implementation process of step (4) is as follows:

[0039] Based on K-means, a clustering model is constructed to divide data objects with high similarity into the same cluster and data with high dissimilarity into different clusters, so as to effectively distinguish different types of vehicle operating conditions. The specific implementation steps are as follows:

[0040] (1) The number of clusters is set to K = 4, corresponding to four types (driver behavior characteristic types): high-speed driving, slow, moderate, and aggressive, and m characteristic parameters are used.

[0041] (2) Randomly select K initial cluster centers μ.

[0042] (3) For each driving behavior segment in the driver behavior feature dataset, the Euclidean distance is used to calculate the distance to each cluster center. The calculation formula is:

[0043]

[0044] In the formula, x i is the sample, μ j is the center cluster, x ik is the kth parameter of the sample coordinates.

[0045] (4) Assign each driving behavior segment to the cluster to which the nearest cluster center belongs, that is, the corresponding cluster with the smallest Euclidean distance, and update the cluster center with the mean of all samples in the cluster. The minimum loss function that converges the function is:

[0046]

[0047] (5) Repeat step 4 until the change in cluster allocation is less than the preset threshold.

[0048] Through the above steps, K-means cluster analysis based on the driver behavior feature data set is completed, and four types of cluster centers are generated: high-speed driving, slow, moderate, and aggressive. The aggressive driving behavior cluster is characterized by high-energy driving behavior in vehicle operation. In actual vehicle operation conditions, the driving behavior segment data is input into the pre-trained clustering model in a unified format to obtain the driver behavior type corresponding to the driving behavior segment; the driving behavior data set under the aggressive cluster is specifically extracted as the driving behavior segment to be optimized for subsequent economical vehicle speed planning.

[0049] Furthermore, the implementation process of step (5) is as follows:

[0050] An instantaneous fuel consumption rate prediction model is constructed based on the Transformer encoder to evaluate the vehicle energy consumption level under different driving speeds and operating conditions.

[0051] The input parameters of the Transformer encoder are the vehicle operation data after the above preprocessing. The driving behavior segmentation criteria consistent with the K-means clustering model are used, and the input sequence length is consistent with the clustering model length. A decoder with a multi-head attention mechanism is used to obtain the sequence prediction results, and the output result is the instantaneous fuel consumption rate prediction value corresponding to each moment of the driving behavior segment.

[0052] Furthermore, the implementation process of step (6) is as follows:

[0053] A dynamic programming solution model with the goal of minimizing energy consumption is constructed, and the optimal vehicle speed is derived and solved at each stage based on the Bellman optimality principle, thereby solving the global problem. Iterative dynamic programming is a multi-stage decision-making problem. The steps of the dynamic programming algorithm proposed in this patent are as follows:

[0054] (1) Stage design and parameter selection: The global energy consumption minimization problem is decomposed into sub-problems. The speed planning scheme that minimizes energy consumption is solved in each time step. The slope is taken into account as an influencing factor of the driver's behavior. The road slope extension feature is used as the system input variable, the vehicle running speed is used as the system state variable, and the engine output torque is used as the system decision variable.

[0055] (2) Establishment of state transfer equation: The state transfer equation is established based on the vehicle longitudinal dynamics. The vehicle longitudinal dynamics equation used is:

[0056]

[0057] (3) Cost function design: The instantaneous fuel consumption rate prediction model based on Transformer constructed above is used. The input parameters are adjusted to the speed variables and engine demand torque variables of each stage of the dynamic programming model. The output of the prediction model is the instantaneous fuel consumption rate corresponding to each stage of the dynamic programming. The predicted value of fuel consumption per 100 kilometers is calculated based on the predicted value of the instantaneous fuel consumption rate as the decision cost function.

[0058] (4) Recursive solution: Based on the state transfer equation, the optimal solution for the vehicle speed in each stage is derived sequentially to obtain the economical driving speed plan with the optimal energy consumption.

[0059] Based on the above dynamic planning solution steps, the global economic speed planning of the driving behavior segment is completed, and the economic driving plan, the global instantaneous fuel consumption rate prediction value and the fuel consumption prediction value per 100 kilometers are output to the driver. The original driving behavior speed spectrum and fuel consumption rate comparison are provided to provide the driver with a reference for driving behavior and vehicle operation plans.

[0060] Beneficial effects: Compared with the prior art, the beneficial effects of the present invention are as follows: 1. Taking into account the technical advantages and technical level of existing heavy-duty freight vehicles, the present invention combines the content and characteristics of Internet of Vehicles data, and adopts mature data preprocessing methods for typical Internet of Vehicles data types. Based on the acquired vehicle and road status information, the economical speed planning of high-energy driving segments is completed on the basis of accurate identification of driver behavior characteristics; 2. The present invention adopts a K-means clustering model to construct a driving behavior data set by preprocessing Internet of Vehicles data, which can accurately and efficiently realize high-precision identification of driver behavior types, and form a classification of four types of driving behavior types: high-speed driving, soothing, moderate, and aggressive, which can basically cover the classification needs of various traffic scenes; 3. Based on the identification of high-energy driving behavior segments, the present invention adopts a dynamic programming method to construct a minimum energy consumption problem solving model, and combines the Transformer instantaneous fuel consumption rate prediction model to ensure the accuracy and optimality of the driving speed plan at each time step in the driving behavior segment. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0062] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0063] The present invention provides a heavy truck driver behavior assessment and economical driving guidance method, which can realize the identification of heavy truck driver behavior type and economical vehicle speed planning for high energy consumption driving segments.

[0064] Technical solution: A heavy truck driver behavior assessment and economical driving guidance method according to the present invention comprises the following steps:

[0065] (1) Multi-source data acquisition: Data acquired through the GPS module includes timestamp, elevation, and vehicle speed; data acquired through the CAN bus includes engine speed, engine required torque, and gear information;

[0066] (2) Data cleaning and feature expansion: Implement corresponding cleaning strategies for missing, noisy and other erroneous frame data in the operating data, and take pre-processing measures such as interpolation and filtering for the operating data that does not meet the continuity conditions; then obtain the road slope characteristics, vehicle acceleration and jerk characteristics of the entire road section at all times through the pre-processed data;

[0067] (3) Construction of driver behavior dataset: The trip segments are divided according to the parking and shutdown time, and the extreme value, mean, standard deviation and other characteristic indicators of the vehicle operation data in each trip segment are obtained. The strongly related features are extracted based on the principal component analysis algorithm to form a driver behavior feature dataset;

[0068] (4) Clustering evaluation of driver behavior characteristics: Based on the K-means algorithm, cluster analysis is performed on the segmented driving behavior data set, and four cluster centers are constructed: high-speed driving, slow, moderate, and aggressive. The clustering results of each driving behavior segment are then output;

[0069] (5) Driving segment fuel consumption prediction: A Transformer-based instantaneous fuel consumption rate prediction model is constructed. The model inputs the preprocessed vehicle operation characteristics and extended road structure characteristics, and outputs the instantaneous fuel consumption rate level of each driving behavior segment;

[0070] (6) Economical driving speed planning for driving segments: A dynamic programming model is constructed with the goal of minimizing energy consumption. The cost function selects the instantaneous fuel consumption rate model constructed above to solve and output the economical driving speed plan with the optimal energy consumption.

[0071] Furthermore, the implementation process of step (2) is as follows:

[0072] Considering the differences in the operating characteristics of non-homologous data, artificial features are extracted and marked for the error frames in the GPS source data and CAN bus source data respectively, and the error frame data samples are separated based on the continuity criterion. The error frame features and cleaning strategies of multi-source data are formulated as follows.

[0073] GPS source data: Error frames mainly come from data interruptions. Combined with the data feedback timestamp, they can be divided into short-term data loss and long-term signal interruption. The time threshold δ is set to 30s. For short-term data loss less than this threshold, linear interpolation is performed using the nearest non-error data value before and after the error frame. For long-term signal interruptions exceeding this threshold, the data during the period is set to zero and discarded.

[0074] CAN bus source data: Error frame data is mainly characterized by short-term data loss, high-frequency measurement noise and discontinuous mutations.

[0075] Short-term data missing is mainly aimed at vehicle gear data detection. Based on the principle of continuous change of vehicle gear, short-term single-point missing data is filled with non-abnormal data at the previous moment.

[0076] The high-frequency measurement noise is mainly used for engine speed and engine torque data detection. A Savitzky-Golay filter is designed to process the noise data. The filter is constructed using a quadratic polynomial, and the filter window length is set to 19.

[0077] Discontinuous mutations are mainly aimed at engine torque data detection, and the characteristic detection standards are set as follows:

[0078]

[0079] Among them, R is the engine torque data at a certain moment, T is the timestamp at the corresponding moment, t is the calculation moment, z is the time step, and δ is the set detection threshold; for the data points obtained according to the above detection standards, the linear interpolation method is used to cover the data values ​​at the corresponding moment.

[0080] After the vehicle operation data is cleaned, the vehicle operation speed characteristics are obtained based on the vehicle parameter attributes and operation characteristics:

[0081]

[0082] Then the vehicle running acceleration and jerk characteristics are obtained:

[0083]

[0084] Where n is the engine speed, r is the wheel rolling radius, i g is the transmission ratio, i0 is the rear axle speed ratio, Δt is the time step, V t is the vehicle speed, a t is the acceleration value, j is the jerk;

[0085] Use the preprocessed speed data and elevation data to obtain the road slope:

[0086]

[0087] Where l is the time step of backward calculation; h t is the elevation value at time t, h t+n is the elevation value at time t+n, in meters; v i is the vehicle speed value, in units of

[0088] The point slope calculation is performed according to the distance span of 100m, and the elevation data is guaranteed to change monotonically within a single calculation step, and the time step k satisfies:

[0089]

[0090] (h i -h i-1 )·(h i+1 -h i )>0,(i=t+1,t+2,…,t+n-1)

[0091] Based on the above calculation strategy, the road slope characteristics, vehicle acceleration and jerk characteristics of the entire road section and at all times are obtained.

[0092] Furthermore, the implementation process of step (3) is as follows:

[0093] A parking and ignition-off duration threshold Δt is set, the time point when the parking and ignition-off duration exceeds the threshold is marked, and the vehicle operation data between the two parking time points is recorded as a driving journey segment, which is the driving behavior segment division.

[0094] Based on the pre-processed vehicle operation data, the extreme value, mean, standard deviation and other characteristic indicators of speed, acceleration, throttle and other data (throttle data obtained from the CAN bus) are expanded. The complete 20 driving behavior characteristic parameter indicators are shown in the following table.

[0095]

[0096] The Pearson correlation coefficient was used to evaluate the correlation between the 20 feature parameters and the fuel consumption rate, and then the principal component analysis algorithm was used to evaluate the cumulative variance contribution rate of the complete feature parameters. The top 10 general features with a cumulative variance contribution rate of 95% were used to form a dimensionality reduction feature group. According to the driving trip segments divided above, the dimensionality reduction feature group of the vehicle operation data in each segment was obtained to form a driver behavior feature data set.

[0097] Furthermore, the implementation process of step (4) is as follows:

[0098] Based on K-means, a clustering model is constructed to divide data objects with high similarity into the same cluster and data with high dissimilarity into different clusters, so as to effectively distinguish different types of vehicle operating conditions. The specific implementation steps are as follows:

[0099] (1) The number of clusters is set to K = 4, corresponding to four types (driver behavior characteristic types): high-speed driving, slow, moderate, and aggressive, and m = 10 characteristic parameters are used.

[0100] (2) Randomly select K initial cluster centers μ.

[0101] (3) For each driving behavior segment in the driver behavior feature dataset, the Euclidean distance is used to calculate the distance to each cluster center. The calculation formula is:

[0102]

[0103] In the formula, x i is the sample, μ j is the center cluster, x ik is the kth parameter of the sample coordinates.

[0104] (4) Assign each driving behavior segment to the cluster to which the nearest cluster center belongs, that is, the corresponding cluster with the smallest Euclidean distance, and update the cluster center with the mean of all samples in the cluster. The minimum loss function that converges the function is:

[0105]

[0106] (5) Repeat step 4 until the change in cluster allocation is less than the preset threshold.

[0107] Through the above steps, K-means cluster analysis based on the driver behavior feature data set is completed, and four types of cluster centers are generated: high-speed driving, slow, moderate, and aggressive. The aggressive driving behavior cluster is characterized by high-energy driving behavior in vehicle operation. In actual vehicle operation conditions, the driving behavior segment data is input into the pre-trained clustering model in a unified format to obtain the driver behavior type corresponding to the driving behavior segment; the driving behavior data set under the aggressive cluster is specifically extracted as the driving behavior segment to be optimized for subsequent economical vehicle speed planning.

[0108] Furthermore, the implementation process of step (5) is as follows:

[0109] An instantaneous fuel consumption rate prediction model is constructed based on the Transformer encoder to evaluate the vehicle energy consumption level under different driving speeds and operating conditions.

[0110] The input parameters of the Transformer encoder are the vehicle operation data after the above preprocessing. The driving behavior segmentation criteria consistent with the K-means clustering model are used, and the input sequence length is consistent with the clustering model length. A decoder with a multi-head attention mechanism is used to obtain the sequence prediction results, and the output result is the instantaneous fuel consumption rate prediction value corresponding to each moment of the driving behavior segment.

[0111] Furthermore, the implementation process of step (6) is as follows:

[0112] A dynamic programming solution model with the goal of minimizing energy consumption is constructed, and the optimal vehicle speed is derived and solved at each stage based on the Bellman optimality principle, thereby solving the global problem. Iterative dynamic programming is a multi-stage decision-making problem. The steps of the dynamic programming algorithm proposed in this patent are as follows:

[0113] (1) Stage design and parameter selection: The global energy consumption minimization problem is decomposed into sub-problems. The speed planning scheme that minimizes energy consumption is solved in each time step. The slope is taken into account as an influencing factor of the driver's behavior. The road slope extension feature is used as the system input variable, the vehicle running speed is used as the system state variable, and the engine output torque is used as the system decision variable.

[0114] (2) Establishment of state transfer equation: The state transfer equation is established based on the vehicle longitudinal dynamics. The vehicle longitudinal dynamics equation used is:

[0115]

[0116] Use the vehicle acceleration as an intermediate variable:

[0117]

[0118] Construct the state transfer equation based on the kinematic relationship between speed and distance:

[0119] u k+1 =u k +a k t k

[0120]

[0121] In the formula, a k is the vehicle acceleration calculated for the optimal fuel economy in the kth stage, u k is the vehicle running speed in the kth stage, d k is the mileage of the vehicle that migrates from the kth stage to the k+1th stage.

[0122] (3) Cost function design: The instantaneous fuel consumption rate prediction model based on Transformer constructed above is used. The input parameters are adjusted to the speed variables and engine required torque variables of each stage of the dynamic programming model. The output of the prediction model is the instantaneous fuel consumption rate corresponding to each stage of the dynamic programming. The predicted value of fuel consumption per 100 kilometers is calculated based on the predicted value of instantaneous fuel consumption rate as the decision cost function. The formula for calculating fuel consumption per 100 kilometers is:

[0123]

[0124] (4) Recursive solution: Based on the state transfer equation, the optimal solution for the vehicle speed in each stage is derived sequentially to obtain the economical driving speed plan with the optimal energy consumption.

[0125] Based on the above dynamic planning solution steps, the global economic speed planning of the driving behavior segment is completed, and the economic driving plan, the global instantaneous fuel consumption rate prediction value and the fuel consumption prediction value per 100 kilometers are output to the driver. The original driving behavior speed spectrum and fuel consumption rate comparison are provided to provide the driver with a reference for driving behavior and vehicle operation plans.

[0126] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A heavy truck driver behavior assessment and economical driving guidance method, characterized in that: The following steps are involved: (1) Multi-source data acquisition: Data acquired through the GPS module includes timestamp, elevation, and vehicle speed; data acquired through the CAN bus includes engine speed, engine required torque, and gear information; (2) Data cleaning and feature expansion: Implement corresponding cleaning strategies for missing, noisy and other erroneous frame data in the operating data, and take pre-processing measures such as interpolation and filtering for the operating data that does not meet the continuity conditions; then obtain the road slope characteristics, vehicle acceleration and jerk characteristics of the entire road section at all times through the pre-processed data; (3) Construction of driver behavior dataset: The trip segments are divided according to the parking and shutdown time, and the extreme value, mean, standard deviation and other characteristic indicators of the vehicle operation data in each trip segment are obtained. The strongly related features are extracted based on the principal component analysis algorithm to form a driver behavior feature dataset; (4) Clustering evaluation of driver behavior characteristics: Based on the K-means algorithm, cluster analysis is performed on the segmented driving behavior data set, and four cluster centers are constructed: high-speed driving, slow, moderate, and aggressive. The clustering results of each driving behavior segment are then output; (5) Driving segment fuel consumption prediction: A Transformer-based instantaneous fuel consumption rate prediction model is constructed. The model inputs the preprocessed vehicle operation characteristics and extended road structure characteristics, and outputs the instantaneous fuel consumption rate level of each driving behavior segment; (6) Economical driving speed planning for driving segments: A dynamic programming model is constructed with the goal of minimizing energy consumption. The cost function selects the instantaneous fuel consumption rate model constructed above to solve and output the economical driving speed plan with the optimal energy consumption.

2. A heavy truck driver behavior assessment and economical driving guidance method according to claim 1, characterized in that: The implementation process of step (2) is as follows: Considering the differences in the operating characteristics of non-homologous data, artificial feature extraction and marking are performed on the error frames in the GPS source data and CAN bus source data respectively, and the error frame data samples are separated based on the continuity criterion. The error frame features and cleaning strategies of multi-source data are formulated as follows; GPS source data: Error frames mainly come from data interruption. Combined with the data feedback timestamp, they can be divided into short-term data loss and long-term signal interruption. The time threshold is set to δ = 30s; For short-term data loss less than the threshold, linear interpolation is performed using the nearest non-erroneous data value before and after the error frame; For long signal interruptions exceeding this threshold, the data during the period is set to zero and discarded; CAN bus source data: Error frame data is mainly characterized by short-term data loss, high-frequency measurement noise and discontinuous mutation; Short-term data missing is mainly for vehicle gear data detection. Based on the principle of continuous change of vehicle gear, short-term single-point missing data is filled with non-abnormal data at the previous moment; High-frequency measurement noise is mainly used for engine speed and engine torque data detection. Savitzky-Golay filter is designed to process the noise data. Discontinuous mutations are mainly aimed at engine torque data detection, and the characteristic detection standards are set as follows: Wherein, R is the engine torque data at a certain moment, T is the timestamp at the corresponding moment, t is the calculation moment, z is the time step, and δ is the set detection threshold; for the data points obtained according to the above detection standards, the linear interpolation method is used to cover the data values ​​at the corresponding moment; After the vehicle operation data is cleaned, the vehicle operation speed characteristics are obtained based on the vehicle structure and operation characteristics: Then the vehicle running acceleration and jerk characteristics are obtained: Where n is the engine speed, r is the wheel rolling radius, i g is the transmission ratio, i0 is the rear axle speed ratio, Δt is the time step, V t is the vehicle speed, a t is the acceleration value, j is the jerk; Use the preprocessed speed data and elevation data to obtain the road slope: Where l is the time step of backward calculation; h t is the elevation value at time t, h t+n is the elevation value at time t+n, in meters; v i is the vehicle speed value, in units of The point slope calculation is performed according to the distance span of 100m, and the elevation data is guaranteed to change monotonically within a single calculation step, and the time step k satisfies: (h i -h i-1 )·(h i+1 -h i )>0,(i=t+1,t+2,…,t+n-1) Based on the above calculation strategy and formula, all running time period data are traversed to obtain the road slope characteristics, vehicle acceleration and jerk characteristics of the entire journey.

3. A heavy truck driver behavior assessment and economical driving guidance method according to claim 2, characterized in that: The implementation process of step (3) is as follows: A parking and ignition-off duration threshold Δt is set, and the time point when the parking and ignition-off duration exceeds the threshold is marked, and the vehicle operation data between the two parking time points is recorded as a driving trip segment, that is, the driving behavior segment division; Based on the pre-processed vehicle operation data, the extreme value, mean, and standard deviation characteristic indicators of speed, acceleration, and throttle data are expanded, and the complete 20 driving behavior characteristic parameter indicators are statistically analyzed; The Pearson correlation coefficient was used to evaluate the correlation between the 20 characteristic parameters and the fuel consumption rate, and then the principal component analysis algorithm was used to evaluate the cumulative variance contribution rate of the complete characteristic parameters; according to the driving trip segments divided as mentioned above, the dimensionality reduction feature group of the vehicle operation data in each segment was obtained to form a driver behavior characteristic data set.

4. A heavy truck driver behavior assessment and economical driving guidance method according to claim 3, characterized in that: The implementation process of step (4) is as follows: Based on K-means, a clustering model is constructed to divide data objects with high similarity into the same cluster and data with high dissimilarity into different clusters, so as to effectively distinguish different types of vehicle operating conditions. The specific implementation steps are as follows: (1) Set the number of clusters K = 4, corresponding to the four types of high-speed driving, soothing, moderate, and aggressive; (2) Randomly select K initial cluster centers μ; (3) For each driving behavior segment in the driver behavior feature dataset, the Euclidean distance is used to calculate the distance to each cluster center. The calculation formula is: In the formula, x i is the sample, μ j is the center cluster, x ik is the kth parameter of the sample coordinate; (4) Assign each driving behavior segment to the cluster to which the nearest cluster center belongs, that is, the corresponding cluster with the smallest Euclidean distance, and update the cluster center with the mean of all samples in the cluster. The minimum loss function that converges the function is: (5) Repeat step 4 until the change in cluster allocation is less than a preset threshold; Through the above steps, K-means cluster analysis based on the driver behavior feature data set is completed, and four types of cluster centers, namely high-speed driving, slow, moderate, and aggressive, are generated. The aggressive driving behavior cluster is characterized by high-energy-consuming driving behavior of vehicle operation. In actual vehicle operation conditions, the driving behavior segment data is input into the pre-trained clustering model in a unified format, and the driver behavior type corresponding to the driving behavior segment is obtained; The driving behavior data set under the radical cluster is extracted specifically as the driving behavior fragment to be optimized for subsequent economical speed planning.

5. A heavy truck driver behavior assessment and economical driving guidance method according to claim 4, characterized in that: The implementation process of step (5) is as follows: An instantaneous fuel consumption rate prediction model is built based on the Transformer encoder to evaluate the vehicle energy consumption level at different driving speeds and operating conditions; The input parameters of the Transformer encoder are the vehicle operation data after the above preprocessing. The driving behavior segmentation criteria consistent with the K-means clustering model are used, and the input sequence length is consistent with the clustering model length. A decoder with a multi-head attention mechanism is used to obtain the sequence prediction results, and the output result is the instantaneous fuel consumption rate prediction value corresponding to each moment of the driving behavior segment.

6. A heavy truck driver behavior assessment and economical driving guidance method according to claim 1, characterized in that: The implementation process of step (6) is as follows: A dynamic programming solution model with the goal of minimizing energy consumption is constructed. Based on the Bellman optimality principle, the optimal vehicle speed is derived and solved at each stage, thereby solving the global problem. Iterative dynamic programming is a multi-stage decision-making problem. The steps of the dynamic programming algorithm proposed in this patent are as follows: (1) Stage design and parameter selection: The global energy consumption minimization problem is decomposed into sub-problems, and the speed planning scheme that minimizes energy consumption is solved in each time step. The slope is taken into account as the influencing factor of the driver's behavior, the road slope extension feature is used as the system input variable, the vehicle running speed is used as the system state variable, and the engine output torque is used as the system decision variable; (2) Establishment of state transfer equation: The state transfer equation is established based on the vehicle longitudinal dynamics. The vehicle longitudinal dynamics equation used is: (3) Cost function design: The instantaneous fuel consumption rate prediction model based on Transformer constructed above is used. The input parameters are adjusted to the speed variables and engine demand torque variables of each stage of the dynamic programming model. The output of the prediction model is the instantaneous fuel consumption rate corresponding to each stage of the dynamic programming. The predicted value of fuel consumption per 100 kilometers is calculated based on the predicted value of the instantaneous fuel consumption rate as the decision cost function. (4) Recursive solution: Based on the state transition equation, the optimal speed solution for each stage is derived sequentially to obtain the economical driving speed plan with the optimal energy consumption; Based on the above dynamic planning solution steps, the global economic speed planning of the driving behavior segment is completed, and the economic driving plan, the global instantaneous fuel consumption rate prediction value and the fuel consumption prediction value per 100 kilometers are output to the driver. The original driving behavior speed spectrum and fuel consumption rate comparison are provided to provide the driver with a reference for driving behavior and vehicle operation plans.

Images