Driving style recognition method and device

Abstract

The application provides a driving style recognition method and device, comprising: acquiring vehicle driving data at a current time; determining whether the vehicle driving data at the current time is valid data affecting driving style; in response to the vehicle driving data at the current time being valid data, determining a driving style index value at the current time based on the vehicle driving data at the current time; determining a driving style quantitative value at the current time based on the driving style index value at the current time and a driving style index value at a historical time; in response to the vehicle driving data at the current time not being valid data, determining the driving style quantitative value at the current time based on the driving style index value at the historical time; and determining the driving style at the current time based on the driving style quantitative value at the current time. The method can realize real-time recognition of the driving style of a driver, and the accuracy of the recognition result and the efficiency of the recognition are relatively high.

Description

Technical Field

[0001] The present application relates to the field of vehicle technology, and in particular to a driving style recognition method and device. Background Art

[0002] Driving style refers to the behavioral characteristics exhibited by a driver during driving activities, determined by stable characteristics such as driving skills and preferences. This characteristic manifests itself as individual or group-specific driving habits. Driving style significantly impacts the safety, comfort, and energy efficiency of the closed-loop system. With the advancement of intelligent in-vehicle electronic control systems, optimizing driving and control system design based on driving style to achieve individual driver adaptation has become a trend. Accurately identifying a driver's driving style is crucial to achieving this optimized design.

[0003] In the related art, a subjective identification method such as a questionnaire survey or an objective statistical method relying on complete driving cycle data is generally used to analyze a driver's driving style.

[0004] However, current driving style recognition methods have difficulty in determining a driver's driving style in real time, and also have difficulty in ensuring the authenticity and globality of the recognition results. Summary of the Invention

[0005] In view of this, the present application provides a driving style recognition method and device, which can realize quasi-real-time recognition of a driver's driving style, and the accuracy of the recognition result and the efficiency of recognition are high.

[0006] Specifically, the following technical solutions are included:

[0007] In one aspect, a driving style identification method is provided, the method comprising:

[0008] Get the current vehicle driving data;

[0009] determining whether the vehicle driving data at the current moment is valid data that affects the driving style;

[0010] In response to the vehicle driving data at the current moment being valid data, determining a driving style index value at the current moment based on the vehicle driving data at the current moment, the driving style index value indicating a tendency of a driving style;

[0011] determining a quantified driving style value at the current moment based on the driving style index value at the current moment and driving style index values ​​at historical moments, wherein the driving style index values ​​at historical moments are driving style index values ​​determined based on historical valid data;

[0012] in response to the vehicle driving data of the current time being invalid data, determining a driving style quantization value of the current time based on the driving style index value of the historical time;

[0013] determining a driving style of the current time based on the driving style quantization value of the current time.

[0014] Optionally, the obtaining of the vehicle driving data of the current time comprises:

[0015] obtaining original vehicle driving data of the current time;

[0016] performing low-pass filtering processing on a driving parameter signal containing the original vehicle driving data of the current time to obtain the vehicle driving data of the current time after filtering processing.

[0017] Optionally, the vehicle driving data comprises longitudinal acceleration and lateral acceleration; and the determining of whether the vehicle driving data of the current time is valid data affecting driving style comprises:

[0018] calculating an acceleration norm based on the longitudinal acceleration and the lateral acceleration;

[0019] in response to the acceleration norm being greater than a preset threshold, determining that the vehicle driving data of the current time is valid data affecting driving style;

[0020] in response to the acceleration norm being less than or equal to the preset threshold, determining that the vehicle driving data of the current time is not valid data affecting driving style.

[0021] Optionally, the vehicle driving data further comprises vehicle speed; and the determining of the driving style index value of the current time based on the vehicle driving data of the current time comprises:

[0022] determining a maximum vehicle speed value in a working condition level corresponding to the vehicle speed, the working condition level indicating a level corresponding to a vehicle speed interval in which the vehicle speed is located;

[0023] correcting the vehicle speed based on the maximum vehicle speed value to obtain a corrected vehicle speed;

[0024] calculating an acceleration phase based on the longitudinal acceleration and the lateral acceleration;

[0025] correcting the acceleration norm based on the corrected vehicle speed and the acceleration phase, and taking the corrected acceleration norm as the driving style index value of the current time.

[0026] Optionally, the correcting of the acceleration norm based on the corrected vehicle speed and the acceleration phase comprises correcting the acceleration norm according to the following formula:

[0027] DS(k)=a xy ×sin(β)×v',

[0028] Among them, DS(k) is the modified acceleration norm, a xy is the acceleration norm, β is the acceleration phase, and v' is the corrected vehicle speed obtained by correcting the vehicle speed v.

[0029] Optionally, determining the current driving style quantified value based on the current driving style index value and the historical driving style index values ​​includes:

[0030] An average processing is performed on the driving style index value at the current moment and the driving style index values ​​at the historical moments, and the calculated average is used as the quantized value of the driving style at the current moment.

[0031] Optionally, the averaging of the driving style index value at the current moment and the driving style index values ​​at historical moments includes:

[0032] An average processing is performed on the driving style index value at the current moment and the most recent S-1 driving style index values ​​in history, where S is an integer greater than or equal to 2.

[0033] Optionally, in response to the vehicle driving data at the current moment being not valid data, determining the quantified driving style value at the current moment based on the driving style index value at the historical moment includes:

[0034] In response to the vehicle driving data at the current moment being invalid, the vehicle driving data at the current moment is discarded, and an average of S most recent driving style index values ​​in history is calculated, and the calculated average is used as the driving style quantization value at the current moment, where S is an integer greater than or equal to 2.

[0035] Optionally, determining the driving style at the current moment based on the quantified value of the driving style at the current moment includes:

[0036] According to the quantized value of the driving style at the current moment, a pre-stored mapping relationship between the quantized value of the driving style and the driving style is searched to obtain the driving style at the current moment.

[0037] On the other hand, an embodiment of the present application provides a driving style recognition device, the device comprising:

[0038] The acquisition module is used to obtain the vehicle driving data at the current moment;

[0039] Identify modules for:

[0040] determining whether the vehicle driving data at the current moment is valid data that affects the driving style;

[0041] In response to the vehicle driving data at the current moment being valid data, determining a driving style index value at the current moment based on the vehicle driving data at the current moment, the driving style index value indicating a tendency of a driving style;

[0042] determining a quantified driving style value at the current moment based on the driving style index value at the current moment and driving style index values ​​at historical moments, wherein the driving style index values ​​at historical moments are driving style index values ​​determined based on historical valid data;

[0043] In response to the vehicle driving data at the current moment being not valid data, determining a driving style quantization value at the current moment based on the driving style index value at the historical moment;

[0044] The driving style at the current moment is determined based on the quantified value of the driving style at the current moment.

[0045] Embodiments of the present application provide a driving style recognition method and device. In this method, a determination is made as to whether the acquired vehicle driving data at the current moment is valid data that affects the driving style. Only if the data is valid is a driving style index value determined based on the current vehicle driving data. The current driving style is then comprehensively determined based on the current driving style index value and historical driving style index values. If the current vehicle driving data is invalid, the current driving style is determined directly based on the historical driving style index values. This prevents the invalid data from negatively impacting the recognition results, while improving the accuracy and efficiency of the recognition results. Furthermore, the current driving style is always determined based on the current driving style quantization value, achieving near-real-time detection of the driving style. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0047] Figure 1 A flowchart of a first driving style identification method provided in an embodiment of the present application;

[0048] Figure 2 A flowchart of a second driving style identification method provided in an embodiment of the present application;

[0049] Figure 3 A flowchart of a third driving style identification method provided in an embodiment of the present application;

[0050] Figure 4 A flowchart of a fourth driving style identification method provided in an embodiment of the present application;

[0051] Figure 5 A schematic diagram of a driving style recognition device provided in an embodiment of the present application.

[0052] The above drawings illustrate specific embodiments of the present application, which will be described in more detail below. These drawings and the textual description are not intended to limit the scope of the present application in any way, but rather to illustrate the concepts of the present application to those skilled in the art by reference to specific embodiments. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0054] Unless otherwise defined, all technical terms used in the embodiments of the present application have the same meanings as commonly understood by those skilled in the art.

[0055] First, to facilitate understanding of the technical solution provided by this application, the inventor's research ideas are briefly introduced here:

[0056] After conducting in-depth research on the application scenarios of driving style, the inventors found that if driving style is to be applied to the vehicle-mounted electronic control system, the driving style recognition method should meet the following three requirements: (1) Universality. The information required for driving style recognition should be derived from the vehicle-mounted CAN (Controller Area Network, abbreviated as CAN) signal or more common vehicle-mounted external sensors, such as IMU (Inertial Measurement Unit, abbreviated as IMU) and GPS (Global Positioning System, (1) quasi-real-time performance: given that the vehicle-mounted electronic control system needs to adapt to the driving styles of different types of drivers and the same driver's evolving driving styles over time, the recognition of driving style should be a quasi-real-time process that can be carried on the vehicle, so that the vehicle-mounted electronic control system can obtain the changes in the driver's style in a timely manner and match the corresponding system configuration within a certain driving cycle; (2) quasi-real-time performance: given that the vehicle-mounted electronic control system needs to adapt to the driving styles of different types of drivers and the same driver's evolving driving styles over time, the recognition of driving style should be a quasi-real-time process that can be carried on the vehicle, so that the vehicle-mounted electronic control system can obtain the changes in the driver's style in a timely manner and match the corresponding system configuration within a certain driving cycle; (3) globality: insufficient information coverage dimensions usually lead to the recognition of local driving styles. In addition, due to the interference of various internal and external factors, the deviation between the recognition results and the actual driving style often becomes larger. Therefore, comprehensive information coverage dimensions can ensure the authenticity and globality of the recognition results.

[0057] Obviously, both the subjective identification methods using questionnaires and the objective statistical methods relying on complete driving cycle data in related technologies cannot simultaneously meet the above requirements. Considering versatility, near-real-time performance, and globality, the inventors have proposed the following driving style identification method and device.

[0058] In order to make the technical solutions and advantages of the present application clearer, the implementation methods of the present application will be described in further detail below with reference to the accompanying drawings.

[0059] In the first aspect, the embodiment of the present application provides a driving style recognition method. The method can be executed by a global controller such as a vehicle controller, or a dedicated controller such as a chassis pre-controller, hereinafter collectively referred to as a controller. Figure 1 , the method comprising:

[0060] Step 101: Obtain vehicle driving data at the current moment.

[0061] In this step, since various sensors of the vehicle are electrically connected to the controller, the controller can receive the driving data of the vehicle at the current moment transmitted by various sensors.

[0062] Step 102: Determine whether the vehicle driving data at the current moment is valid data that affects the driving style.

[0063] In practice, there are various working conditions during the driving of the vehicle, but the determination of the driving style often depends on the operation of the user under some specific working conditions. For example, the vehicle can run in the working condition of uniform forward movement, and the user operation at this time is relatively single and smooth, but this cannot indicate that the user driving style is "smooth". Therefore, in order to improve the effectiveness of driving style recognition, the vehicle driving data can be evaluated and screened, and after a certain pretreatment, the subsequent driving style recognition is carried out. In this step, the controller can evaluate the vehicle driving data at the current time according to the preset algorithm to determine whether the vehicle driving data at the current time is effective data affecting the driving style. Optionally, the vehicle driving data when the acceleration, deceleration or steering operation occurs can be determined as the effective data affecting the driving style.

[0064] Step 103: In response to the vehicle driving data at the current time being effective data, determining a driving style index value at the current time based on the vehicle driving data at the current time, the driving style index value indicating the tendency of the driving style.

[0065] In this step, after determining that the vehicle driving data at the current time is effective, the driving style index value at the current time can be determined according to the vehicle driving data at the current time, which can indicate the tendency of the user driving style corresponding to the vehicle driving data at the current time. Then, the driving style at the current time and the historical time can be used to comprehensively identify the driving style of the user.

[0066] Step 104: Determining a driving style quantitative value at the current time based on the driving style index value at the current time and the driving style index value at the historical time, wherein the driving style index value at the historical time is a driving style index value determined based on historical effective data.

[0067] In this step, the driving style quantitative value at the current time can be determined according to the driving style index value at the current time and the driving style index value at the historical time, so as to realize the quantification of the concept index of the driving style. It should be noted that the driving style value at the historical time is determined based on the historical effective data, and the historical effective data can be the vehicle driving data evaluated as effective data in the history through step 102. The driving style value at the historical time can be temporarily calculated according to the historical effective data when step 104 is executed, or directly read from the vehicle memory. The vehicle memory can store the vehicle driving data evaluated as effective data each time, or directly store the driving style index value at the historical time calculated based on the historical effective data.

[0068] Step 105: In response to the vehicle driving data at the current time not being effective data, determining a driving style quantitative value at the current time based on the driving style index value at the historical time.

[0069] In this step, if the current vehicle driving data is not valid, the current driving style quantification value can be directly determined based on the driving style index values ​​at previous moments. In other words, if the judgment result indicates that the current vehicle driving data is not valid data that affects driving style, it is assumed that the driving style has not changed and the current vehicle driving data is ignored to avoid reducing the accuracy and effectiveness of the driving style recognition results.

[0070] Step 106: Determine the current driving style based on the current driving style quantification value.

[0071] In this step, after determining the current driving style quantification value, the current driving style can be determined based on the quantification value, achieving the conversion from the quantification value to the concept of driving style. The controller can pre-store response logic corresponding to different driving styles. Based on the determined current driving style, the corresponding response logic is determined. The user operation is responded to according to the response logic adapted to the current driving style, achieving vehicle adaptation to the individual driver.

[0072] In summary, the driving style recognition method provided by the embodiments of the present application can achieve quantitative recognition of the virtual concept of driving style. During the recognition process, invalid data is removed from the data used to identify driving style, ensuring the accuracy and effectiveness of driving style recognition. Furthermore, the current driving style can always be determined based on the current driving style quantification value, ensuring near-real-time driving style recognition.

[0073] In some embodiments, reference Figure 2 , step 101 may include:

[0074] Step 1011: Obtain the original vehicle driving data at the current moment.

[0075] The original vehicle driving data in this step can be data directly detected by various sensors in the vehicle and transmitted to the controller.

[0076] Step 1012: Low-pass filtering is performed on the driving parameter signal containing the original vehicle driving data at the current moment to obtain filtered vehicle driving data at the current moment. The driving parameter signal here may be, for example, an acceleration signal or a vehicle speed signal.

[0077] Because the signals output by various vehicle sensors are often mixed with various high-frequency noise sources, such as electromagnetic interference, these high-frequency noises are ineffective for driving style identification and may even interfere with it. Furthermore, the inevitable presence of environmental disturbances during driving, such as continuous speed bumps, sudden headwinds or crosswinds, or unexpected potholes, can cause medium- and high-frequency interference to accumulate in the vehicle parameter signals. Due to the driver's physiological characteristics and driving habits, the vehicle's response during normal driving generally falls within the low-frequency range (this applies only to ordinary drivers, not to specialized groups such as racing drivers). Low-frequency interference caused by long slopes or constant wind is integrated with the low-frequency response from the driver's actions to form the vehicle's final response, which serves as the basis for the driver's next maneuver decision. Therefore, these low-frequency disturbances can be used as part of the driving style characterization.

[0078] Therefore, in this step, the mid- and high-frequency information in the driving parameter signal is removed to prevent interference with driving style recognition. Furthermore, low-pass filtering effectively smooths the driving parameter signal, facilitating accurate identification of valid data later on.

[0079] Optionally, a low-pass filter with a cutoff frequency of 2 Hz is used to filter the driving parameter signal containing the original vehicle driving data at the current moment, and the filtered vehicle driving data at the current moment is used for subsequent driving style recognition, which can effectively avoid interference signals from interfering with the recognition results and improve the accuracy of the recognition results.

[0080] In some embodiments, the vehicle driving data includes longitudinal acceleration and lateral acceleration. Figure 3 , step 102, comprising:

[0081] Step 1021: Calculate the acceleration norm based on the longitudinal acceleration and the lateral acceleration.

[0082] Optionally, the acceleration norm can be calculated using the following formula:

[0083] Among them, a xy is the acceleration norm, a x is the longitudinal acceleration, a y is the lateral acceleration. It can be seen that the acceleration norm can comprehensively reflect the magnitude of the longitudinal acceleration and lateral acceleration of the vehicle.

[0084] Step 1022: Determine whether the acceleration norm is greater than a preset threshold.

[0085] In this step, the preset threshold is set based on experience or obtained through statistical analysis of a large amount of acceleration data. The preset threshold represents the minimum value of the total norm of the longitudinal acceleration and lateral acceleration of the vehicle when the driver's driving operation causes the vehicle to accelerate, decelerate, or turn. The preset threshold here is generally not 0, because when the vehicle is driving at a constant speed in a straight line, there is also a certain degree of acceleration fluctuation. Optionally, the preset threshold here can be 0.03g, where g is the acceleration due to gravity. That is, in this step, a xy By executing this step, it can be determined whether the vehicle is currently experiencing acceleration, deceleration, or steering, which can be used to identify driving style.

[0086] Step 1023: In response to the acceleration norm being greater than a preset threshold, determining that the vehicle driving data at the current moment is valid data that affects the driving style.

[0087] In this step, if the acceleration norm is greater than the preset threshold, it can be determined that the driver's operation caused the vehicle to accelerate, decelerate, or turn, etc. The vehicle driving data at this time can be used to extract the index value of the driver's driving style, so the vehicle driving data at the current moment is determined as valid data that affects the driving style.

[0088] Step 1024: In response to the acceleration norm being less than or equal to the preset threshold, determining that the vehicle driving data at the current moment is not valid data that affects the driving style.

[0089] In this step, if the acceleration norm is less than or equal to the preset threshold, it can be determined that the vehicle driving data at the current moment cannot reflect the driver's driving style, and it can be determined that it is not valid data that affects the driving style.

[0090] In some embodiments, the inventors considered the following: First, acceleration, a vehicle driving parameter, is crucial for identifying driving style. Aggressive drivers typically maximize and continuously exploit the tire's maximum adhesion, manifesting as aggressive acceleration and deceleration. Conservative drivers, on the other hand, typically control the vehicle in a more gradual manner. Therefore, the acceleration norm can directly reflect driving style. Second, longitudinal acceleration and lateral acceleration also have different weights in reflecting driving style. Third, vehicle speed is the most important and intuitive indicator of vehicle movement, and its changes can also reflect the driver's driving style to a certain extent. Based on these three aspects, the inventors proposed that the acceleration norm can be used as the core indicator for driving style identification, and that the acceleration norm can be modified by using vehicle speed and acceleration phase to assist in identifying driving style using more diverse information, thereby improving the overall effectiveness of driving style identification.

[0091] Specifically, the vehicle driving data also includes vehicle speed.

[0092] refer to Figure 4 , step 103 may include:

[0093] Step 1031: Determine the maximum vehicle speed value in the operating level corresponding to the vehicle speed, where the operating level indicates the level corresponding to the vehicle speed interval in which the vehicle speed is located.

[0094] Although vehicle speed may be correlated with a driver's driving style, traffic regulations may have different maximum speed limits for different driving areas, so vehicle speed cannot be directly used to judge a driver's driving style. For example, if a driver's average speed on a highway is 110 km / h, and another driver's average speed on a city road is 70 km / h, it cannot be considered that the former driver's driving style is more "aggressive" than the latter. Conversely, the driving style of a driver who averages 70 km / h on a city road is likely to be more "aggressive" than that of a driver who averages 60 km / h on the same city road. As can be seen, the connotation of the vehicle speed indicator is very rich.

[0095] In this embodiment, vehicle speeds are divided into multiple intervals, each corresponding to a specific operating level. The operating levels are used to improve comparability between different speeds. Specifically, the speeds can be divided into multiple intervals based on multiple maximum speed limits specified in traffic regulations. For example, if traffic regulations specify speed limits on different types of roads: 30, 50, 60, 70, 80, 100, 110, and 120, the speeds can be divided into [0, 30), [30, 50), [50, 60), [60, 70), [70, 80), [80, 100), [100, 120), and [120, ∞), corresponding to operating levels 1 to 8, respectively. The interval divisions and operating levels described here are merely examples and can be adjusted in practice based on actual needs. For example, based on typical urban roads and highways, the speeds can be divided into two operating levels: 0-90 km / h and 90 km / h-120 km / h. Alternatively, a large amount of raw speed data and driving style data derived through annotation can be used for cluster analysis to derive multiple speed interval thresholds. These thresholds can then be used to classify speed intervals and operating conditions. Speeds at different operating conditions require processing before they can be used for driving style recognition. However, within the same operating condition, speed directly reflects driving style.

[0096] On this basis, in step 1031, the maximum vehicle speed value in the operating condition level corresponding to the current vehicle speed can be determined, that is, the maximum vehicle speed value in the speed range in which the current vehicle speed is located can be determined, so as to use the maximum vehicle speed value to correct the vehicle speed, so that the vehicle speeds under different operating conditions can be compared with each other.

[0097] Step 1032: Correct the vehicle speed based on the maximum vehicle speed value to obtain a corrected vehicle speed.

[0098] In this step, the following formula can be used to correct the vehicle speed:

[0099] v'=v / v max-n , where v represents the vehicle speed, v' is the corrected vehicle speed obtained by correcting the vehicle speed v, and v max-n Represents the maximum vehicle speed value in the nth operating condition level, where n is the operating condition level corresponding to the speed range in which the current vehicle speed is located.

[0100] In this step, the vehicle speed is corrected so that the vehicle speed indicator becomes an indicator that can be compared with each other and more accurately reflects the driving style.

[0101] Step 1033: Calculate the acceleration phase based on the longitudinal acceleration and the lateral acceleration.

[0102] In this step, the acceleration phase can be calculated using the following formula:

[0103] β=tan -1 (|a x | / |a y |), where β is the acceleration phase.

[0104] After the acceleration phase is calculated in this step, the acceleration norm can be corrected with the help of the acceleration phase so that the driving style index value can reflect the different degrees of influence of longitudinal acceleration and lateral acceleration on driving style.

[0105] Step 1034: Based on the corrected vehicle speed and acceleration phase, the acceleration norm is corrected, and the corrected acceleration norm is used as the driving style index value at the current moment.

[0106] In this step, the corrected acceleration norm is used as the driving style index value at the current moment, thereby quantifying the driving style. Furthermore, the acceleration norm is corrected by correcting the vehicle speed and acceleration phase, thereby ensuring the accuracy of the driving style index value.

[0107] Optionally, step 1034 may include: correcting the acceleration norm according to the following formula:

[0108] DS(k)=a xy ×sin(β)×v',

[0109] Among them, DS(k) is the modified acceleration norm, a xy is the acceleration norm, β is the acceleration phase, and v' is the corrected vehicle speed obtained by correcting the vehicle speed v.

[0110] Because different drivers, under the same operating level and acceleration norm, have a more aggressive driving style as the value of β increases (this has been confirmed in related art), this embodiment introduces β to correct the acceleration norm so that the corrected acceleration norm more accurately represents the driver's driving style. It should be noted that, from a mathematical perspective, while the introduction of sin(β) changes the true value of the acceleration norm, the acceleration norm indicators for all driver groups are corrected. Driving style identification is a comparison of the relative trends of these indicators. Therefore, the deviations introduced by this correction will offset each other during the driving style identification process. The introduction of sin(β) does not reduce the accuracy of the driving style identification results, but rather improves them.

[0111] In some embodiments, step 104 includes:

[0112] The driving style index value at the current moment and the driving style index value at the historical moment are averaged, and the calculated average is used as the quantified driving style value at the current moment.

[0113] In this embodiment, the average of the current driving style index value and the historical driving style index values ​​is used as the current driving style quantification value. This ensures that the driving style quantification value is influenced not only by the current driving style index value but also by the historical driving style index values. This allows the driver's driving style to be identified based on their comprehensive performance over a longer period of time, thereby improving the reliability of the driving style identification results.

[0114] Optionally, averaging the current driving style index value and the historical driving style index values ​​is performed, including:

[0115] The current driving style index value and the most recent S-1 driving style index values ​​in history are averaged, where S is an integer greater than or equal to 2.

[0116] Specifically, only the current driving style index value and the most recent S-1 driving style index values ​​in history are obtained to calculate the current driving style quantization value. This ensures that driving style recognition is always based on a finite number of recent driving style index values, ensuring efficient calculation of the driving style quantization value and the effectiveness and near-real-time performance of the driving recognition results.

[0117] In some embodiments, step 105 includes:

[0118] In response to the vehicle driving data at the current moment being invalid, the vehicle driving data at the current moment is discarded, and the most recent S driving style index values ​​in history are averaged, and the calculated average is used as the driving style quantization value at the current moment, where S is an integer greater than or equal to 2.

[0119] Specifically, if the current vehicle driving data is invalid, the driver's current operating behavior is deemed inadequate to reflect their driving style and the data is discarded. Furthermore, the current quantified driving style value is assumed to be unchanged from the previous value, so the average of the S most recent driving style index values ​​is used as the current quantified driving style value.

[0120] That is, in this embodiment, a moving window function with a width of S is used to filter out the most recent S driving style quantization values ​​for averaging, and the calculated average is used as the driving style quantization value at the current moment, that is: Among them, DS(t) is the quantitative value of driving style at the current moment, The driving style index value corresponding to the kth valid data point is used as the latest driving style index value within the moving window, and is combined with the driving style index values ​​corresponding to the previous S-1 valid data points to obtain the average index value. DS(i) is the driving style index value corresponding to the i-th valid data point, and the value of i ranges from [kS, k]. If the current vehicle driving data is valid, the window function covers the current driving style index value and the most recent S-1 driving style index values ​​in history. If the current vehicle driving data is not valid, the window function covers the most recent S driving style index values ​​in history.

[0121] In some embodiments, step 106 includes:

[0122] According to the current driving style quantization value, a pre-stored mapping relationship between the driving style quantization value and the driving style is queried to obtain the current driving style.

[0123] A mapping relationship between the quantized driving style value and the driving style can be pre-stored in the vehicle controller or memory. This mapping relationship can be established based on experience, calibrated, or determined using other effective methods. After determining the quantized driving style value at the current moment, the controller can directly determine the current driving style based on this mapping relationship and output the current driving style, or determine corresponding response logic based on the current driving style and respond to user operations using this response logic.

[0124] In summary, the driving style recognition method provided by the embodiment of the present application can be used for recognition based on vehicle driving data such as acceleration and speed, so that the method can be applied to any vehicle. It only requires the vehicle to have sensors for detecting acceleration and speed, which is convenient for deployment on commercial vehicles. For the obtained vehicle driving data, a low-pass filter is used to remove noise information irrelevant to driving style recognition, thereby improving the recognition efficiency and the accuracy of the recognition results. The acceleration norm is used as an indicator to identify the validity of the vehicle driving data, and the valid data is used to identify the driving style, thereby further improving the efficiency and accuracy of driving style recognition. The vehicle speed is also used for The proposed method distinguishes between different driving conditions and corrects the vehicle speed for different driving conditions. The acceleration norm is modified by correcting the vehicle speed and acceleration phase, so that the driving style quantization value can more accurately represent the driving style. More parameters are introduced to identify driving style, ensuring comprehensive information coverage and making the recognition result closer to the true global driving style. In addition, a moving window is used to average the S most recent driving style index values, so that the driving style quantization value more accurately reflects the most recent driver's driving style. The current driving style can be output in real time, achieving quasi-real-time operation of driving style recognition, which is convenient for application in on-board electronic control systems.

[0125] On the other hand, an embodiment of the present application provides a driving style recognition device. Figure 5 , the device comprises:

[0126] An acquisition module 200 is used to acquire vehicle driving data at the current moment;

[0127] The determination module 300 is configured to:

[0128] Determining whether the current vehicle driving data is valid data that affects the driving style;

[0129] In response to the vehicle driving data at the current moment being valid data, determining a driving style index value at the current moment based on the vehicle driving data at the current moment, the driving style index value indicating a tendency of the driving style;

[0130] determining a quantified driving style value at the current moment based on a driving style index value at the current moment and a driving style index value at a historical moment, wherein the driving style index value at the historical moment is a driving style index value determined based on historical valid data;

[0131] In response to the vehicle driving data at the current moment being not valid data, determining a driving style quantization value at the current moment based on driving style index values ​​at historical moments;

[0132] The current driving style is determined based on the current driving style quantification value.

[0133] Optionally, the acquisition module 200 is further configured to:

[0134] Get the original vehicle driving data at the current moment;

[0135] A low-pass filtering process is performed on the driving parameter signal containing the original vehicle driving data at the current moment to obtain the vehicle driving parameter at the current moment after the filtering process.

[0136] Optionally, the vehicle driving data includes longitudinal acceleration and lateral acceleration; the determination module 300 is further configured to:

[0137] Calculate the acceleration norm based on the longitudinal acceleration and the lateral acceleration;

[0138] In response to the acceleration norm being greater than a preset threshold, determining that the vehicle driving data at the current moment is valid data affecting the driving style;

[0139] In response to the acceleration norm being less than or equal to a preset threshold, it is determined that the vehicle driving data at the current moment is not valid data affecting the driving style.

[0140] Optionally, the vehicle driving data also includes vehicle speed; the determination module 300 is further configured to:

[0141] Determine a maximum vehicle speed value in a working condition level corresponding to the vehicle speed, where the working condition level indicates the level corresponding to the vehicle speed interval within which the vehicle speed is located;

[0142] Correcting the vehicle speed based on the maximum vehicle speed value to obtain a corrected vehicle speed;

[0143] Calculate the acceleration phase based on the longitudinal acceleration and the lateral acceleration;

[0144] Based on the corrected vehicle speed and acceleration phase, the acceleration norm is corrected, and the corrected acceleration norm is used as the driving style index value at the current moment.

[0145] Optionally, the determination module 300 is further configured to correct the acceleration norm according to the following formula:

[0146] DS(k)=a xy ×sin(β)×v',

[0147] Among them, DS(k) is the modified acceleration norm, a xy is the acceleration norm, β is the acceleration phase, and v' is the corrected vehicle speed obtained by correcting the vehicle speed v.

[0148] Optionally, the determining module 300 is further configured to:

[0149] The driving style index value at the current moment and the driving style index value at the historical moment are averaged, and the calculated average is used as the quantified driving style value at the current moment.

[0150] Optionally, the determining module 300 is further configured to:

[0151] The current driving style index value and the most recent S-1 driving style index values ​​in history are averaged, where S is an integer greater than or equal to 2.

[0152] Optionally, the determining module 300 is further configured to:

[0153] In response to the vehicle driving data at the current moment being invalid, the vehicle driving data at the current moment is discarded, and the most recent S driving style index values ​​in history are averaged, and the calculated average is used as the driving style quantization value at the current moment, where S is an integer greater than or equal to 2.

[0154] Optionally, the determining module 300 is further configured to:

[0155] According to the current driving style quantization value, a pre-stored mapping relationship between the driving style quantization value and the driving style is queried to obtain the current driving style.

[0156] It should be noted that the apparatus provided in this embodiment corresponds to the aforementioned method embodiment, and each module can be used to implement each step in the aforementioned method embodiment. The details of each step will not be repeated here. Furthermore, the module division is merely exemplary. In some embodiments, the driving style recognition apparatus can be further divided into: a data preprocessing module, an effective operating condition detection module, an indicator extraction module, and a driving style recognition module. Each module performs a portion of the steps in the aforementioned method embodiment to collectively achieve driving style recognition. For example, the data preprocessing module uses a low-pass filter to remove invalid mid- and high-frequency information and high-frequency noise information that are irrelevant to driving style recognition from the original vehicle driving parameter signal, and performs filtering and smoothing on the vehicle driving parameter signal; the effective working condition detection module detects data segments that are effective for driving style recognition from the filtered vehicle driving data, which are called effective data; the indicator extraction module determines the working condition level based on the vehicle speed, corrects the vehicle speed based on the working condition level, and integrates the acceleration norm indicator, acceleration phase indicator, and vehicle speed indicator into a unified driving style quantification indicator DS; the driving style module sets a moving window and detects the driving style indicator value corresponding to the effective data within the window to perform driving style recognition, thereby achieving quasi-real-time operation of driving style recognition.

[0157] In summary, an embodiment of the present application provides a driving style recognition device. In this device, the acquired vehicle driving data at the current moment is judged to be valid data that affects the driving style. Only when the data is valid, the device determines the current moment's driving style index value based on the current moment's vehicle driving data. The current moment's driving style is comprehensively determined based on the current moment's driving style index value and the driving style index values ​​at historical moments. If the current moment's vehicle driving data is invalid, the current moment's driving style is directly determined based on the driving style index values ​​at historical moments. This avoids the negative impact of invalid data on the recognition results, while improving the accuracy of the recognition results and the efficiency of recognition. Furthermore, the current moment's driving style is always determined based on the current moment's driving style quantization value, thereby achieving quasi-real-time detection of driving style.

[0158] In this application, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance. The term "plurality" refers to two or more than two, unless expressly limited otherwise.

[0159] Those skilled in the art will readily appreciate other embodiments of the present invention after considering the specification and practicing the present invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present invention that follow the general principles of the present invention and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only.

[0160] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A driving style recognition method, characterized in that: The method comprises: Get the current vehicle driving data; determining whether the vehicle driving data at the current moment is valid data that affects the driving style; In response to the vehicle driving data at the current moment being valid data, determining a driving style index value at the current moment based on the vehicle driving data at the current moment, the driving style index value indicating a tendency of a driving style; determining a quantified driving style value at the current moment based on the driving style index value at the current moment and driving style index values ​​at historical moments, wherein the driving style index values ​​at historical moments are driving style index values ​​determined based on historical valid data; In response to the vehicle driving data at the current moment being not valid data, determining a driving style quantization value at the current moment based on the driving style index value at the historical moment; determining the driving style at the current moment based on the quantified value of the driving style at the current moment; Wherein, the vehicle driving data includes longitudinal acceleration, lateral acceleration and vehicle speed; The determining of the driving style index value at the current moment based on the vehicle driving data at the current moment includes: determining a maximum vehicle speed value in an operating condition level corresponding to the vehicle speed, the operating condition level indicating a level corresponding to a vehicle speed interval within which the vehicle speed lies; Correcting the vehicle speed based on the maximum vehicle speed value to obtain a corrected vehicle speed; Based on the longitudinal acceleration and the lateral acceleration, an acceleration phase and an acceleration norm are calculated, wherein the acceleration phase is calculated using the following formula: ,in, is the acceleration phase, is the longitudinal acceleration, is the lateral acceleration, The acceleration norm is calculated by the following formula: ,in, is the acceleration norm; The acceleration norm is corrected based on the corrected vehicle speed and the acceleration phase, and the corrected acceleration norm is used as the driving style index value at the current moment, wherein the acceleration norm is corrected according to the following formula: ， in, is the modified acceleration norm, is the acceleration norm, is the acceleration phase, For vehicle speed The corrected vehicle speed obtained after correction.

2. The method according to claim 1, characterized in that The obtaining of the vehicle driving data at the current moment includes: Get the original vehicle driving data at the current moment; Low-pass filtering is performed on the driving parameter signal containing the original vehicle driving data at the current moment to obtain filtered vehicle driving data at the current moment.

3. The method according to claim 1, characterized in that The determining whether the vehicle driving data at the current moment is valid data that affects the driving style includes: Calculating the acceleration norm based on the longitudinal acceleration and the lateral acceleration; In response to the acceleration norm being greater than a preset threshold, determining that the vehicle driving data at the current moment is valid data affecting the driving style; In response to the acceleration norm being less than or equal to the preset threshold, it is determined that the vehicle driving data at the current moment is not valid data affecting the driving style.

4. The method according to claim 1, wherein The determining of the current driving style quantization value based on the current driving style index value and the historical driving style index values ​​includes: An average processing is performed on the driving style index value at the current moment and the driving style index values ​​at the historical moments, and the calculated average is used as the quantized value of the driving style at the current moment.

5. The method according to claim 4, characterized in that The averaging of the driving style index value at the current moment and the driving style index value at the historical moment includes: An average is calculated for the current driving style index value and the most recent S-1 driving style index values ​​in history, where S is an integer greater than or equal to 2.

6. The method according to claim 1, characterized in that In response to the vehicle driving data at the current moment being not valid data, determining the driving style quantization value at the current moment based on the driving style index value at the historical moment includes: In response to the vehicle driving data at the current moment being invalid, the vehicle driving data at the current moment is discarded, and an average of S most recent driving style index values ​​in history is calculated, and the calculated average is used as the driving style quantization value at the current moment, where S is an integer greater than or equal to 2.

7. The method according to claim 1, characterized in that The determining the driving style at the current moment based on the quantified value of the driving style at the current moment includes: According to the quantized value of the driving style at the current moment, a pre-stored mapping relationship between the quantized value of the driving style and the driving style is searched to obtain the driving style at the current moment.

8. A driving style recognition device, characterized in that: The device comprises: The acquisition module is used to obtain the vehicle driving data at the current moment; Identify modules for: determining whether the vehicle driving data at the current moment is valid data that affects the driving style; In response to the vehicle driving data at the current moment being valid data, determining a driving style index value at the current moment based on the vehicle driving data at the current moment, the driving style index value indicating a tendency of a driving style; determining a quantified driving style value at the current moment based on the driving style index value at the current moment and driving style index values ​​at historical moments, wherein the driving style index values ​​at historical moments are driving style index values ​​determined based on historical valid data; In response to the vehicle driving data at the current moment being not valid data, determining a driving style quantization value at the current moment based on the driving style index value at the historical moment; determining the driving style at the current moment based on the quantified value of the driving style at the current moment; Wherein, the vehicle driving data includes longitudinal acceleration, lateral acceleration and vehicle speed; The determining of the driving style index value at the current moment based on the vehicle driving data at the current moment includes: determining a maximum vehicle speed value in an operating condition level corresponding to the vehicle speed, the operating condition level indicating a level corresponding to a vehicle speed interval within which the vehicle speed lies; Correcting the vehicle speed based on the maximum vehicle speed value to obtain a corrected vehicle speed; Based on the longitudinal acceleration and the lateral acceleration, an acceleration phase and an acceleration norm are calculated, wherein the acceleration phase is calculated using the following formula: ,in, is the acceleration phase, is the longitudinal acceleration, is the lateral acceleration, The acceleration norm is calculated by the following formula: ,in, is the acceleration norm; The acceleration norm is corrected based on the corrected vehicle speed and the acceleration phase, and the corrected acceleration norm is used as the driving style index value at the current moment, wherein the acceleration norm is corrected according to the following formula: ， in, is the modified acceleration norm, is the acceleration norm, is the acceleration phase, For vehicle speed The corrected vehicle speed obtained after correction.

Images