Single-pedal driving deceleration intention determination method

Abstract

The present application belongs to the technical field of electric vehicles, and is especially a single-pedal driving deceleration intention determination method. The method comprises the following steps: S1: collecting the deceleration and speed data set of the driver; S2: modeling the commonly used maximum deceleration of the driver at each vehicle speed; S3: constructing the deceleration statistical feature space of the driver; S4: performing cluster analysis on the deceleration statistical features of all test drivers; S5: establishing a linear benchmark mapping of the pedal opening and the expected stable vehicle speed under the single-pedal mode; S6: designing the dynamic critical position of the single-pedal driving deceleration mode; and S7: determining the single-pedal driving deceleration intention. The present application scientifically divides the drivers into three typical styles of "aggressive, ordinary, and cautious" by using the weighted hierarchical clustering method, which overcomes the defect that the single dynamic partition in the prior art cannot adapt to individual differences, and realizes the recognition of the deceleration intention that fits the real operation habits of different drivers.

Description

Technical Field

[0001] This invention relates to the field of electric vehicle technology, specifically to a method for determining deceleration intention during single-pedal driving. Background Technology

[0002] The automotive industry is undergoing a massive transformation, with automotive pedals evolving from mechanical to electrified and then to intelligent systems. One-pedal driving evolved from the intelligentization of traditional pedals, allowing drivers to control acceleration and deceleration with a single accelerator pedal. Research shows that 90% of everyday driving situations can be achieved with a single pedal, while the reserved brake pedal is only used for emergency braking. The use of a single pedal significantly reduces the driver's workload. Under the one-pedal control architecture, the system must rely on characteristic parameters such as pedal opening and the rate of change of opening to accurately determine the driver's deceleration intention. The accuracy of this intention recognition directly affects the coordination between the drive system and the regenerative braking system, thus profoundly impacting overall vehicle safety and the driving experience.

[0003] Currently, the mainstream methods for determining the deceleration intention of electric vehicle single-pedal driving can be divided into three categories: The first is the experience-based fixed partitioning method, which directly divides the accelerator pedal travel into three fixed intervals—acceleration, coasting, and deceleration—based on engineering experience and fixed rules to determine the intention. The second is the speed-based dynamic partitioning method, which introduces a vehicle speed variable on top of pedal opening to dynamically adjust the boundaries of each operating interval, such as the coasting zone, and construct a multi-dimensional adaptive PedalMap mapping relationship. The third is the intelligent algorithm-based reasoning classification method, which introduces fuzzy control theory, neural networks, hidden Markov models, and various optimization algorithms, utilizing feature data such as pedal opening and opening change rate to achieve intelligent recognition and intention parsing of driving behavior.

[0004] Considering the significant individual differences in driving habits among drivers, different driving styles often reflect drastically different deceleration intentions under the same pedal opening and rate of change. Existing one-pedal deceleration intention recognition technologies lack systematic mining of the multidimensional statistical characteristics of a large number of drivers' natural deceleration habits, failing to establish a scientific model of driving deceleration habits. Consequently, they cannot quantitatively cluster and analyze groups of drivers with similar characteristics and summarize them into typical styles that highly match real human driving characteristics. Therefore, introducing statistical analysis methods to deeply mine driving sample data from different driving subjects and constructing multiple differentiated deceleration intention recognition models is the core foundation for meeting personalized driving needs and improving the driving experience. There is an urgent need in this field for a method to determine one-pedal driving deceleration intentions. Summary of the Invention

[0005] The purpose of this section is to outline some aspects of the embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of this section, the abstract, and the title, and such simplifications or omissions should not be used to limit the scope of the invention.

[0006] To address the aforementioned technical problems, according to one aspect of the present invention, the present invention provides the following technical solution:

[0007] A method for determining deceleration intention during single-pedal driving includes the following steps:

[0008] S1: Collect driver deceleration and speed data using vehicle sensors;

[0009] S2: A Gaussian structure function is used to model the maximum deceleration commonly used by the driver at various vehicle speeds;

[0010] S3: Construct the statistical feature space of driver deceleration;

[0011] S4: Cluster analysis of deceleration statistical characteristics of all test drivers based on weighted hierarchical clustering method: The feature vectors obtained in S3 are standardized to obtain standardized feature vectors. A hierarchical clustering tree is constructed. The two nearest classes are repeatedly calculated and merged until all samples are clustered into one class to form a dendrogram. The driver group is divided into three classes. The arithmetic mean of each class is calculated to obtain three cluster centroids. The drivers' deceleration habits are divided into three typical styles: aggressive, normal, and cautious.

[0012] S5: Establish a linear reference mapping between pedal opening and desired stable vehicle speed in single-pedal mode;

[0013] S6: The dynamic critical position design of the single-pedal driving deceleration mode divides the pedal opening into a coasting zone and a braking zone. Under high-speed or low-deceleration conditions, the driver can slightly release the pedal to put the vehicle into a coasting state. The coasting zone serves as a buffer area between acceleration control and braking control.

[0014] S7: Determination of deceleration intention for multi-style single-pedal driving based on clustering of driver deceleration statistical features: The three typical deceleration style curves obtained by S4 clustering are directly mapped to the target deceleration characteristics when the pedal is fully released, providing a larger deceleration in the medium speed range, and a relatively mild deceleration in high-speed and low-speed conditions.

[0015] As a preferred embodiment of the single-pedal driving deceleration intention determination method described in this invention, the driver's deceleration and speed dataset in S1 includes the real-time speed of the driver performing frequent acceleration and deceleration operations under three typical driving cycles: urban, suburban, and highway. With deceleration .

[0016] As a preferred embodiment of the single-pedal driving deceleration intention determination method of the present invention, the specific method of S2 is as follows: The 90th percentile of the deceleration value is calculated for the driver's deceleration behavior data in each speed range to characterize the driver's personalized common maximum deceleration at different vehicle speeds. A Gaussian structure function is used to model the driver's deceleration habits, and its expression is:

[0017]

[0018] In the formula, Current vehicle speed; Current speed The maximum deceleration when the pedal is fully released; and It is related to the peak deceleration and the velocity corresponding to the peak deceleration; The parameter used to control the width of the peak deceleration distribution.

[0019] In a preferred embodiment of the single-pedal driving deceleration intention determination method of the present invention, the specific method of S3 is as follows: performing least-squares fitting on the deceleration habit modeling parameters of each driver to obtain the deceleration habit modeling parameters of each driver. At this time Speed ​​corresponds to the driver's peak deceleration Select a triplet array for each driver As a statistical feature vector of driver deceleration.

[0020] In a preferred embodiment of the single-pedal driving deceleration intention determination method of the present invention, the calculation formula for the standardized feature vector in step S4 is as follows:

[0021]

[0022] In the formula, For the first The driver in The original values ​​on each eigenvalue; and For all drivers in the Mean and standard deviation of each feature;

[0023] Weights are introduced to reconstruct the features, and the expression for the weight vector is as follows:

[0024]

[0025] in, These are the weighting coefficients;

[0026] The weighted eigenvector expression is:

[0027]

[0028] Based on weighted Hierarchical clustering is used, and the Ward variance minimization algorithm is employed as the merging criterion to calculate the inter-class distance. The specific expression is as follows:

[0029] .

[0030] In a preferred embodiment of the single-pedal driving deceleration intention determination method of the present invention, the specific mapping relationship in S5 is expressed as follows:

[0031]

[0032] In the formula, To stabilize the pedal opening; In single-pedal mode, the fixed pedal ensures a stable pedal opening. The final stable speed that the vehicle can maintain; is the proportionality coefficient, representing the linear relationship between the two.

[0033] In a preferred embodiment of the single-pedal driving deceleration intention determination method of the present invention, the width of the coasting zone in step S6 dynamically increases with the increase of vehicle speed, and the specific formula for expressing the coasting zone is as follows:

[0034]

[0035] In the formula, This represents the natural coasting deceleration at the current vehicle speed; The relaxation threshold is set; This is the width coefficient of the taxiing area.

[0036] As a preferred embodiment of the single-pedal driving deceleration intention determination method described in this invention, in step S7, to ensure that the single pedal has both sensitivity and comfort, two key objectives must be met: first, the deceleration resolved during the initial release of the pedal by the driver should be sufficiently small; second, the final deceleration intention determination must conform to the driver's individual expectations; the driving deceleration intention is constructed using Stevens' power law.

[0037]

[0038]

[0039] In the formula, Indicates the driver's desired deceleration; This represents the maximum deceleration when the pedal is fully released at the current speed v. This refers to the amount of pedal release; This represents the current pedal opening. The pedal opening required to stabilize the current vehicle speed; n is the perception index, reflecting the driver's perceived intensity of the deceleration response.

[0040] Compared with existing technologies, the beneficial effects of this invention are as follows: By modeling the driver's deceleration behavior data under natural conditions using a Gaussian structure function, a feature space containing multidimensional statistical features is deeply mined and constructed. A weighted hierarchical clustering method is used to scientifically classify drivers into three typical styles: "aggressive," "normal," and "cautious." This overcomes the deficiency of existing technologies where a single dynamic partition cannot adapt to individual differences, and achieves deceleration intention recognition that aligns with the actual operating habits of different drivers. Attached Figure Description

[0041] To more clearly illustrate the technical solutions of the embodiments of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and detailed embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0042] Figure 1 This is a system functional module architecture diagram of an embodiment of a single-pedal driving deceleration intention determination method of the present invention;

[0043] Figure 2 This is a schematic diagram of the core steps of a method for determining a single-pedal driving deceleration intention according to an embodiment of the present invention.

[0044] Figure 3 This is a clustering result diagram of driver deceleration habits in an embodiment of a single-pedal driving deceleration intention determination method of the present invention;

[0045] Figure 4 This is a mapping diagram of pedal opening and desired vehicle stable speed in an embodiment of a single-pedal driving deceleration intention determination method of the present invention.

[0046] Figure 5 This is a diagram showing the deceleration intention determination result when releasing the pedal at 30% pedal opening, according to an embodiment of the single-pedal driving deceleration intention determination method of the present invention. Detailed Implementation

[0047] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0048] Secondly, the present invention is described in detail with reference to the schematic diagrams. When detailing the embodiments of the present invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not according to the usual scale. Furthermore, the schematic diagrams are merely examples and should not limit the scope of protection of the present invention. In addition, actual fabrication should include three-dimensional spatial dimensions of length, width, and depth.

[0049] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0050] A method for determining single-pedal driving deceleration intention based on statistical characteristics of driver deceleration habits includes the following steps:

[0051] Step 1: Collect real-time speed data from the driver's frequent acceleration and deceleration operations under three typical driving cycles: city, suburbs, and highway, using the vehicle's sensors. With deceleration The deceleration and velocity dataset of the driver is obtained and used to analyze the deceleration distribution characteristics of the driver under natural conditions.

[0052] Step 2: Model the driver's deceleration habits, i.e., the driver's commonly used maximum deceleration at various vehicle speeds, using a Gaussian structure function. Calculate the 90th percentile of the driver's deceleration behavior data within each speed range to characterize the driver's personalized commonly used maximum deceleration at different speeds. The expression for modeling the driver's deceleration habits using a Gaussian structure function is as follows:

[0053]

[0054] In the formula, Current vehicle speed; Current speed The maximum deceleration when the pedal is fully released; and It is related to the peak deceleration and the velocity corresponding to the peak deceleration; Parameters for controlling the width of the peak deceleration distribution;

[0055] Step 3: Construct the statistical feature space of driver deceleration. Perform least-squares fitting on the deceleration habit modeling parameters for each driver to obtain the deceleration habit modeling parameters for each driver. At this time Speed ​​corresponds to the driver's peak deceleration Select a triplet array for each driver As a statistical feature vector of driver deceleration;

[0056] Step 4: Perform cluster analysis on the deceleration statistical characteristics of all test drivers using a weighted hierarchical clustering method. The feature vectors obtained in Step 3 are standardized to obtain standardized feature vectors. The specific calculation formula is as follows:

[0057]

[0058] In the formula, For the first The driver in The original values ​​on each eigenvalue; and For all drivers in the Mean and standard deviation over each feature.

[0059] Weights are introduced to reconstruct the features, and the expression for the weight vector is as follows:

[0060]

[0061] in, These are the weighting coefficients.

[0062] The standardized feature vectors are weighted to construct a weighted feature vector, which is obtained by multiplying each of the three components of the standard feature vector by its corresponding weight coefficient. The expression for the weighted feature vector is as follows:

[0063]

[0064] Based on weighted Hierarchical clustering is used, and the Ward variance minimization algorithm is employed as the merging criterion to calculate the inter-class distance. The specific expression is as follows:

[0065]

[0066] Construct a hierarchical clustering tree, repeatedly calculate and merge the two closest classes until all samples are clustered into one class, forming a dendrogram. Finally, based on the structure of the dendrogram, prune at the nodes where the inter-class distance changes abruptly, divide the driver group into three classes, calculate the arithmetic mean in each class, and obtain three cluster centroids, which respectively classify the drivers' deceleration habits into three typical styles: aggressive, normal, and cautious.

[0067] Step 5: Establish a linear baseline mapping between pedal opening and desired stable vehicle speed in one-pedal mode. A linear relationship is established between the vehicle's stable speed and the single-pedal opening. This mapping does not directly determine acceleration but is used to calculate the theoretical equilibrium pedal opening corresponding to the current vehicle speed, serving as the baseline zero point for subsequent judgment of the driver's acceleration / deceleration intentions. The specific expression for the mapping relationship is as follows:

[0068]

[0069] In the formula, To stabilize the pedal opening; In single-pedal mode, the fixed pedal ensures a stable pedal opening. The final stable speed that the vehicle can maintain; This is the proportionality coefficient, representing the linear relationship between the two.

[0070] Step 6: Dynamic Critical Position Design for One-Pedal Driving Deceleration Mode. The pedal opening is divided into a coasting zone and a braking zone. Under high-speed or low-deceleration conditions, the driver slightly releases the pedal to initiate a coasting state. The coasting zone acts as a buffer between acceleration and braking control, preventing acceleration fluctuations caused by frequent controller switching. The width of the coasting zone dynamically increases with vehicle speed. The specific formula for expressing the coasting zone is as follows:

[0071]

[0072] In the formula, This represents the natural coasting deceleration at the current vehicle speed; The relaxation threshold is set; This is the width coefficient of the taxiing area;

[0073] Step 7: Determining the deceleration intent of multi-style one-pedal driving based on driver deceleration statistical characteristics clustering. The three typical deceleration style curves obtained from the clustering in Step 4 are directly mapped to the target deceleration characteristics when the pedal is fully released. This ensures that the braking intensity can both match the driver's individual habits and adaptively adjust with vehicle speed. A larger deceleration is provided in the medium-speed range (typically corresponding to urban driving conditions) to meet frequent deceleration needs, while a relatively gentle deceleration is used in high-speed and low-speed conditions.

[0074] To ensure both responsiveness and comfort with a single pedal, two key objectives must be met: first, the deceleration resolved during the initial pedal release phase should be sufficiently small to avoid overly sensitive pedal response, thus allowing for a smooth transition to the coasting zone; second, the final deceleration intention must align with the driver's individual expectations. This invention employs Stevens' power law to construct the driving deceleration intention:

[0075]

[0076]

[0077] In the formula, Indicates the driver's desired deceleration; This is the maximum deceleration when the pedal is fully released at the current speed v (derived from the Gaussian model in step 2); This refers to the amount of pedal release; This represents the current pedal opening. The pedal opening required to stabilize the current vehicle speed; n is the perception index, reflecting the driver's perceived intensity of the deceleration response.

[0078] Example:

[0079] like Figure 1 and 2 As shown, a method for determining a single-pedal driving deceleration intention according to an embodiment of the present invention includes the following steps:

[0080] Step 1: Twenty test drivers were selected and performed frequent acceleration and deceleration operations under three typical driving cycles: city, suburbs, and highway. The real-time speed of the drivers was collected by the vehicle's sensors. With deceleration The deceleration and velocity datasets of the driver were obtained. No preceding vehicle was included in the test scenario to ensure that the driver's deceleration behavior was entirely based on their own habits. All collected test data were cleaned, and the active braking segments were extracted to analyze the deceleration distribution characteristics of the driver under natural conditions.

[0081] Step 2: Model driver deceleration habits using a Gaussian structure function. To quantify the expected speed-deceleration relationship for each driver in this mode, the 90th percentile of the deceleration values ​​is calculated for each speed range, thus representing the driver's personalized maximum deceleration at each vehicle speed. Since driver deceleration behavior exhibits a normal distribution, a Gaussian structure function is used to model driver deceleration habits to establish a continuous relationship between speed and the current maximum deceleration. Its expression is:

[0082]

[0083] In the formula, Current vehicle speed; Current speed The maximum deceleration when the pedal is fully released; and It is related to the peak deceleration and the velocity corresponding to the peak deceleration; Parameters for controlling the width of the peak deceleration distribution;

[0084] Step 3: Construct the statistical feature space of driver deceleration. Perform least-squares fitting on the deceleration habit modeling parameters for each driver to obtain the deceleration habit modeling parameters for each driver. When the speed is At that time, the corresponding peak deceleration of the driver In this embodiment of the invention, a ternary array is used for each driver. As a statistical feature vector of driver deceleration. It characterizes the peak deceleration that a driver can accept during daily driving, reflecting the driver's deceleration intensity preference; The speed range that indicates its preference for stronger deceleration; Characterizing the consistency of the maximum deceleration as a function of velocity, A larger value indicates a more stable deceleration style;

[0085] Step 4: Perform cluster analysis on the deceleration statistical characteristics of all test drivers using a weighted hierarchical clustering method. To systematically study different deceleration styles and design differentiated strategies accordingly, this embodiment of the invention summarizes a large number of discrete driver characteristics into typical driving styles during the clustering process. Since different feature components have different dimensions and physical meanings, the feature vectors obtained in Step 3 are standardized. The specific calculation formula is as follows:

[0086]

[0087] In the formula, For the first The driver in The original values ​​on each eigenvalue; and For all drivers in the Mean and standard deviation over each feature.

[0088] Considering that in one-pedal mode, the driver's sensitivity to peak deceleration is much higher than their sensitivity to speed range, this embodiment of the invention introduces weights to reconstruct the features, and the expression for the weight vector is as follows:

[0089]

[0090] In the formula, The weighting coefficients are set to 0.6, 0.1, and 0.3 in this embodiment of the invention.

[0091] The standardized feature vectors are weighted to construct a weighted feature vector, which is obtained by multiplying each of the three components of the standard feature vector by its corresponding weight coefficient. The expression for the weighted feature vector is as follows:

[0092]

[0093] Based on weighted Hierarchical clustering is used, and the Ward variance minimization algorithm is employed as the merging criterion to calculate the inter-class distance. The specific expression is as follows:

[0094]

[0095] A hierarchical clustering tree was constructed, repeatedly calculating and merging the two closest classes until all samples were clustered into one class, forming a dendrogram. Finally, based on the structure of the dendrogram, pruning was performed at nodes where inter-class distances abruptly changed, dividing the driver group into three classes. The arithmetic mean of each class was calculated, resulting in three cluster centroids, which categorized drivers' deceleration habits into three typical styles: aggressive, moderate, and cautious. Aggressive drivers preferred to achieve higher decelerations at higher speeds, with maximum deceleration at each speed varying significantly with speed. Conversely, cautious drivers achieved smaller decelerations at lower speeds, and the larger distribution width of the regions indicated that the maximum deceleration of cautious drivers at each speed was less sensitive to speed changes and exhibited higher consistency. The maximum deceleration values ​​of moderate drivers at each speed varied between the two. The cluster centroids of deceleration habits of 20 drivers and their corresponding deceleration habit representation functions are shown in Table 1 and [Table data missing]. Figure 3 ;

[0096] Table 1 Clustering results of driver deceleration statistical characteristics

[0097] Driving style Driver's Number radical 7,10,12,13,18 -2.3053 61.43 70.62 Standard 1,2,4,5,6,8,9,11,14,16,19,20 -1.9190 55.82 90.46 Cautious 3,17,15 -1.2688 47.75 121.14

[0098] Step 5: Establish a linear reference mapping between pedal opening and desired stable vehicle speed in one-pedal mode. To meet the predictable driving experience for most drivers, this embodiment of the invention sets a linear relationship between stable vehicle speed and one-pedal opening, such as... Figure 4 As shown. This mapping relationship does not directly determine acceleration, but is used to calculate the theoretical balance pedal opening corresponding to the current vehicle speed, serving as the reference zero point for subsequent judgment of the driver's acceleration and deceleration intentions. The specific expression of the mapping relationship is as follows:

[0099]

[0100] In the formula, To stabilize the pedal opening; In single-pedal mode, the fixed pedal ensures a stable pedal opening. The final stable speed that the vehicle can maintain; is the proportionality coefficient, representing the linear relationship between the two.

[0101] Step 6: Dynamic Critical Position Design for Single-Pedal Driving Deceleration Mode. To accurately determine the driver's deceleration intention, this embodiment of the invention divides the pedal opening into a coasting zone and a braking zone. Under high-speed or low-deceleration demand conditions, the driver slightly releases the pedal to put the vehicle into a coasting state. The coasting zone serves as a buffer area between acceleration control and braking control, used to avoid acceleration fluctuations caused by frequent controller switching. The width of the coasting zone dynamically increases with vehicle speed, and the specific formula for expressing the coasting zone is as follows:

[0102]

[0103] In the formula, This represents the natural coasting deceleration at the current vehicle speed; For the set relaxation threshold, take ; This is the taxiing area width coefficient, which is empirically set to 1 or 2.

[0104] Step 7: Determining the deceleration intention of multi-style one-pedal driving based on driver deceleration statistical features clustering. In this embodiment of the invention, the three typical deceleration style curves obtained from clustering in Step 4 above are directly mapped to the target deceleration characteristics when the pedal is fully released. This allows the braking intensity to both match the driver's individual habits and adaptively adjust with vehicle speed: providing greater deceleration in the medium-speed range (typically corresponding to urban driving conditions) to meet frequent deceleration needs, while employing relatively gentle deceleration in high-speed and low-speed conditions.

[0105] To ensure both responsiveness and comfort with a single pedal, two key objectives must be met: first, the determined deceleration during the initial pedal release phase should be sufficiently small to avoid overly sensitive pedal response, thus allowing for a smooth transition to the coasting zone; second, the final deceleration intention must align with the driver's individual expectations. This embodiment of the invention uses Stevens' power law to construct the deceleration intention. The driver's deceleration intention determination result at 30% pedal opening is as follows... Figure 5 As shown:

[0106]

[0107]

[0108] In the formula, Indicates the driver's desired deceleration; This is the maximum deceleration when the pedal is fully released at the current speed v (derived from the Gaussian model in step 2); The amount by which the pedal is released; This represents the current pedal opening. The pedal opening required to stabilize the current vehicle speed; n is the perception index, reflecting the driver's perceived intensity of the deceleration response.

[0109] Although the present invention has been described above with reference to embodiments, various modifications can be made and components can be replaced with equivalents without departing from the scope of the invention. In particular, as long as there is no structural conflict, the features in the disclosed embodiments can be combined with each other in any manner. The lack of an exhaustive description of these combinations in this specification is merely for the sake of brevity and resource conservation. Therefore, the present invention is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A method for determining deceleration intention during single-pedal driving, characterized in that, Includes the following steps: S1: Collect driver deceleration and speed data using vehicle sensors; S2: A Gaussian structure function is used to model the maximum deceleration commonly used by the driver at various vehicle speeds; S3: Construct the statistical feature space of driver deceleration; S4: Cluster analysis of deceleration statistical characteristics of all test drivers based on weighted hierarchical clustering method: The feature vectors obtained in S3 are standardized to obtain standardized feature vectors. A hierarchical clustering tree is constructed. The two nearest classes are repeatedly calculated and merged until all samples are clustered into one class to form a dendrogram. The driver group is divided into three classes. The arithmetic mean of each class is calculated to obtain three cluster centroids. The drivers' deceleration habits are divided into three typical styles: aggressive, normal, and cautious. S5: Establish a linear reference mapping between pedal opening and desired stable vehicle speed in single-pedal mode; S6: The dynamic critical position design of the single-pedal driving deceleration mode divides the pedal opening into a coasting zone and a braking zone. Under high-speed or low-deceleration conditions, the driver can slightly release the pedal to put the vehicle into a coasting state. The coasting zone serves as a buffer area between acceleration control and braking control. S7: Determination of deceleration intention for multi-style single-pedal driving based on clustering of driver deceleration statistical features: The three typical deceleration style curves obtained by S4 clustering are directly mapped to the target deceleration characteristics when the pedal is fully released, providing a larger deceleration in the medium speed range, and a relatively mild deceleration in high-speed and low-speed conditions.

2. The method for determining deceleration intention of a single-pedal driving system according to claim 1, characterized in that, The driver's deceleration and speed dataset in S1 includes the real-time speed of the driver performing frequent acceleration and deceleration operations under three typical driving cycles: urban, suburban, and highway. With deceleration .

3. The method for determining deceleration intention during single-pedal driving according to claim 1, characterized in that, The specific method of S2 is as follows: The 90th percentile of the deceleration value is calculated for the driver's deceleration behavior data within each speed range. This represents the driver's personalized, commonly used maximum deceleration at different vehicle speeds. A Gaussian structure function is used to model the driver's deceleration habits, and its expression is: In the formula, Current vehicle speed; Current speed The maximum deceleration when the pedal is fully released; and It is related to the peak deceleration and the velocity corresponding to the peak deceleration; The parameter used to control the width of the peak deceleration distribution.

4. The method for determining deceleration intention during single-pedal driving according to claim 1, characterized in that, The specific method of S3 is to perform least-squares fitting on the deceleration habit modeling parameters of each driver to obtain the deceleration habit modeling parameters for each driver. At this time Speed ​​corresponds to the driver's peak deceleration Select a triplet array for each driver As a statistical feature vector of driver deceleration.

5. The method for determining deceleration intention during single-pedal driving according to claim 1, characterized in that, The formula for calculating the standardized eigenvector in S4 is as follows: In the formula, For the first The driver in The original values ​​on each eigenvalue; and For all drivers in the Mean and standard deviation of each feature; Weights are introduced to reconstruct the features, and the expression for the weight vector is as follows: in, These are the weighting coefficients; The weighted eigenvector expression is: Based on weighted Hierarchical clustering is used, and the Ward variance minimization algorithm is employed as the merging criterion to calculate the inter-class distance. The specific expression is as follows: 。 6. The method for determining deceleration intention during single-pedal driving according to claim 1, characterized in that, The specific mapping relationship in S5 is expressed as follows: In the formula, To stabilize the pedal opening; In single-pedal mode, the fixed pedal ensures a stable pedal opening. The final stable speed that the vehicle can maintain; is the proportionality coefficient, representing the linear relationship between the two.

7. The method for determining deceleration intention of a single-pedal driving system according to claim 1, characterized in that, The width of the coasting zone in S6 increases dynamically with increasing vehicle speed. The specific formula for expressing the coasting zone is as follows: In the formula, This represents the natural coasting deceleration at the current vehicle speed; The relaxation threshold is set; This is the width coefficient of the taxiing area.

8. The method for determining deceleration intention during single-pedal driving according to claim 1, characterized in that, In the S7 described above, to ensure that the single pedal combines sensitivity and comfort, two key objectives must be met: first, the deceleration resolved during the initial release of the pedal by the driver should be sufficiently small; second, the final deceleration intention must conform to the driver's individual expectations; the driving deceleration intention is constructed using Stevens' power law. In the formula, Indicates the driver's desired deceleration; This represents the maximum deceleration when the pedal is fully released at the current speed v. This refers to the amount of pedal release; This represents the current pedal opening. The pedal opening required to stabilize the current vehicle speed; n is the perception index, reflecting the driver's perceived intensity of the deceleration response.

Images