Longitudinal acceleration intention recognition method considering driving habits of driver with exercise intensity

Abstract

The invention belongs to the technical field of automobiles, and particularly relates to a longitudinal acceleration intention recognition method considering the exercise intensity driving habit of a driver, which comprises a probability model construction module, an operation input judgment module and an acceleration intention analysis module, the probability model construction module is used for constructing a driver accelerator pedal operation habit probability model and a motion intensity driving habit probability model, the operation input judgment module is used for receiving accelerator pedal opening degree information for judgment, and the acceleration intention analysis module is used for analyzing the acceleration intention according to the current state. And analyzing the longitudinal acceleration conforming to the habits of the driver. Operation habits and exercise intensity habits of a driver are utilized. High-degree matching of the longitudinal personalized acceleration intention of the driver is achieved, and a basis is provided for customizing the longitudinal personalized driving experience of the driver.

Description

Technical Field

[0001] The present invention relates to the field of automobile technology, and in particular to a method for recognizing longitudinal acceleration intention taking into account the driver's sports intensity and driving habits. Background Art

[0002] The rapid development of the automobile industry has brought not only improved automobile performance, but also changes in consumer consumption concepts. When consumers choose a car, they no longer rely on travel needs, but are more concerned about whether the driving experience fits their preferences. Therefore, many manufacturers and scholars have added driving style considerations to automobile-related control strategies to provide different driving experiences for drivers with different styles, in order to improve drivers' driving satisfaction.

[0003] Invention patent CN118753302A discloses a driving style recognition method and a human-machine shared steering control method. It uses the KHM algorithm to cluster the collected parameter data representing the driving style, and establishes a driving style recognition model based on the BP neural network. Then, a human-vehicle-road system model is established, and the LMT / H∞ control method is used for torque sharing control. According to the different driving styles of drivers, the particle swarm optimization algorithm is used to optimize the controller authority. While reducing the steering load of the driver, the human-machine interaction authority of different driving styles is personalized to enhance the driver's driving experience.

[0004] The method of dividing drivers by driving style and providing personalized design for drivers with different driving styles has the following problems:

[0005] First of all, the accuracy of driving style recognition cannot be guaranteed. Most methods of driving style recognition are to select relevant feature parameters, then use clustering methods to divide and use machine learning or non-machine learning methods for recognition. However, there is no accurate conclusion on what parameters should be used to identify driving style. If too few feature parameters are selected, the differences between different styles may not be captured. If too many are selected, the complexity of the model may increase. In addition, if a driver with a driving style between two styles is encountered, it will be inaccurate no matter which style is identified. Therefore, the accuracy of driving style recognition cannot be guaranteed.

[0006] In addition, due to the differences between individual drivers, even drivers with the same style may have different preferences for driving experience. Therefore, using the same control strategy for drivers with the same style cannot well meet the personalized needs of all drivers.

[0007] In summary, the driving style method can obtain a relatively satisfactory result for most drivers when processing a large number of driver samples. However, if it is targeted at a specific driver (target driver), the effect cannot be guaranteed.

[0008] For a single driver (target driver), by collecting his driving habit characteristics and using them to identify his longitudinal acceleration intention, it will be more in line with his driving motion intensity expectations and more closely match his desired personalized experience.

[0009] Therefore, we propose a longitudinal acceleration intention recognition method that takes into account the driver's sports intensity and driving habits to solve the above problems. Summary of the invention

[0010] 1. Technical issues to be resolved

[0011] In view of the deficiencies in the prior art, the present invention provides a method for identifying longitudinal acceleration intention taking into account the driver's driving habits of sports intensity, thereby solving the problems raised in the above-mentioned background technology.

[0012] (II) Technical solution

[0013] In order to achieve the above-mentioned purpose, the present invention specifically adopts the following technical solutions:

[0014] A longitudinal acceleration intention recognition method considering a driver's driving habit of sport intensity includes a probability model building module, an operation input judgment module and an acceleration intention analysis module. The probability model building module is used to build a probability model of the driver's accelerator pedal operation habit and a probability model of the driving habit of sport intensity. The operation input judgment module is used to receive accelerator pedal opening information for judgment. The acceleration intention analysis module is used to analyze the longitudinal acceleration that conforms to the driver's habit according to the current state. The longitudinal acceleration intention recognition method also includes the following steps:

[0015] S1. Under ideal driving conditions, the longitudinal driving data of the driver is collected. The probability model of driving input habits is constructed using normal distribution, and the opening of the accelerator pedal is selected to characterize the longitudinal driving operation input of the driver. Assuming that the sample data of the accelerator pedal collected in the experiment is N in total, the opening of the accelerator pedal x conforms to the following normal distribution:

[0016]

[0017] First, using the sample data x i Calculate the sample mean and sample variance S 2 :

[0018]

[0019] Then, using the moment estimation method, we establish an equation based on the population moment being equal to the sample moment:

[0020]

[0021] Combining the expectation and variance formulas of the normal distribution:

[0022] E(X)=μ1

[0023]

[0024] We can calculate μ1, Moment estimator

[0025]

[0026] Substitute the estimated values ​​of mean and standard deviation to obtain the operating habit probability model of driving input;

[0027] S2. Select longitudinal acceleration to represent the longitudinal motion intensity. Based on the acceleration data collected under ideal driving conditions, the probability model of driving motion intensity habits is obtained by processing in the same way:

[0028]

[0029] S3. The probability model of driving input shows the probability distribution of each accelerator pedal opening interval. The larger the probability density of the accelerator pedal opening interval, the more often the driver uses the corresponding accelerator pedal opening to operate; conversely, it means that the driver does not often apply the corresponding accelerator pedal opening, which well describes the driver's longitudinal operation habits. The probability model of motion intensity shows the probability of different longitudinal accelerations during the driver's driving process, which well describes the driver's driving motion intensity habits. The driver's most frequently used operation input is mapped to the longitudinal acceleration with the highest probability when driving, and the infrequently used operation input is mapped to the longitudinal acceleration with a low probability. Therefore, according to the driver's longitudinal operation input, the longitudinal motion intensity that the driver is more inclined to can be obtained;

[0030] S4. Use the same probability method to achieve the mapping of operation habits and exercise intensity habits, as follows:

[0031] Since the mapping is based on the driver's habits, the low-probability events at both ends correspond to emergency or infrequent situations, which are not suitable for characterizing the driver's habits. Therefore, most of the center of the normal distribution is selected for mapping, and 95% of the normal distribution area centered on the mean μ1 is selected for mapping, that is:

[0032] P(|x-μ1|≤xlim )=0.95

[0033] 2μ1-x lim ≤x≤x lim

[0034] For (2μ1-x lim ,x lim ) range, and establish a mapping relationship between it and the longitudinal acceleration of the corresponding probability;

[0035] Assume that the amount of operation to be mapped is x cur , firstly, according to the probability calculation formula of normal distribution:

[0036]

[0037] Calculate the corresponding probability P(x≤x cur ), after obtaining the probability, the motion intensity of the mapping is calculated using the following formula

[0038]

[0039] Furthermore, the specific construction method of the probability model construction module is as follows:

[0040] Step 1: Under ideal driving conditions, the driver's longitudinal accelerator pedal opening information is collected, and the accelerator pedal opening is used to represent the driver's longitudinal operation input. The probability model of driving input habits is constructed using normal distribution. The mean and standard deviation estimators are calculated based on sample data using moment estimation method to complete the parameter analysis of the model, and finally the probability model of operation habits is obtained;

[0041] Step 2: Under ideal driving conditions, the longitudinal acceleration data of the driver is collected, and the longitudinal motion intensity is represented by the longitudinal acceleration. The probability model of motion intensity habits is constructed using the normal distribution. The corresponding mean and standard deviation estimators are calculated using the moment estimation method based on the sample data to complete the parameter analysis of the model, and then the probability model of driving motion intensity habits is obtained.

[0042] Furthermore, if the current accelerator pedal opening x cur In (2μ1-x lim ,x lim ), it is determined that the driver has the intention to accelerate at this time, and the current pedal opening information is input into the longitudinal acceleration intention analysis module; otherwise, it is determined that the driver has no intention to accelerate at this time, and no operation is performed.

[0043] Furthermore, the acceleration intention analysis module is composed of an operation habit probability model and a driving motion intensity habit probability model.

[0044] Further, for (2μ1-x lim ,x lim ) range, and establish a mapping relationship between it and the longitudinal acceleration of the corresponding probability in the motion intensity probability model. According to the mapping relationship between the two, the corresponding longitudinal acceleration can be analyzed through the driver's accelerator pedal opening input in the current state.

[0045] (III) Beneficial effects

[0046] Compared with the prior art, the present invention provides a longitudinal acceleration intention recognition method taking into account the driver's sports intensity driving habits, which has the following beneficial effects:

[0047] The present invention constructs a probability model of operating habits and driving motion intensity habits based on the longitudinal driving data of the driver under ideal driving conditions, and establishes a connection between the two. According to the actual operation input of the driver, the driving motion intensity that meets the driver's habits at this time, that is, the expected longitudinal acceleration, can be obtained. Since the data for constructing the model comes from the driving data of the driver under ideal driving conditions, it can well reflect the driver's natural driving habits, so the driver will feel more comfortable under this motion intensity, which provides a good basis for customizing the driver's personalized driving experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a schematic diagram of the system principle of the present invention;

[0049] Figure 2 A schematic diagram of the process of constructing the probability model of the present invention;

[0050] Figure 3 Schematic diagram of the mapping relationship of the present invention. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] Example

[0053] like Figure 1-3As shown, an embodiment of the present invention proposes a longitudinal acceleration intention recognition method that takes into account the driver's motion intensity driving habits, using the driver's operating habits and motion intensity habits. A high degree of matching of the driver's longitudinal personalized acceleration intention is achieved, providing a basis for customizing his longitudinal personalized driving experience. For a certain driver, under ideal driving conditions, his driving behavior tends to his natural driving habits. Corresponding data is collected through experiments, and the data is analyzed to extract the driver's driving habit characteristics. The accelerator pedal opening and longitudinal acceleration are selected respectively to characterize the driver's longitudinal operation input and the vehicle's longitudinal motion intensity. The collected data is analyzed to extract habit features, thereby constructing the driver's operating habit model and the driver's driving motion intensity habit model. Afterwards, the connection between the two is established, and according to the driver's operation input, the information of the habit model is integrated to obtain the habitual driving motion intensity under the current driving input.

[0054] The purpose of the present invention is to solve the problem that the driving style method cannot accurately match the personalized driving experience requirements of a designated driver (target driver). For the longitudinal acceleration part in the personalized driving experience, the present invention proposes a longitudinal acceleration intention recognition method that takes into account movement intensity habits, which is conducive to further providing the target driver with a personalized driving experience that is more in line with his preferences.

[0055] A longitudinal acceleration intention recognition method considering a driver's driving habit of sport intensity includes a probability model building module, an operation input judgment module and an acceleration intention analysis module. The probability model building module is used to build a probability model of the driver's accelerator pedal operation habit and a probability model of the driving habit of sport intensity. The operation input judgment module is used to receive accelerator pedal opening information for judgment. The acceleration intention analysis module is used to analyze the longitudinal acceleration that conforms to the driver's habit according to the current state. The longitudinal acceleration intention recognition method also includes the following steps:

[0056] S1. Under ideal driving conditions, collect longitudinal driving data of a driver. Use normal distribution to build a probability model of driving input habits. Select the opening of the accelerator pedal to characterize the longitudinal driving operation input of the driver. Assume that the sample data of the accelerator pedal collected in the experiment is N in total, and the opening of the accelerator pedal x conforms to the following normal distribution:

[0057]

[0058] First, using the sample data x i Calculate the sample mean and sample variance S 2 :

[0059]

[0060] Then, using the moment estimation method, we establish an equation based on the population moment being equal to the sample moment:

[0061]

[0062] Combining the expectation and variance formulas of the normal distribution:

[0063] E(X)=μ1

[0064]

[0065] We can calculate μ1, Moment estimator

[0066]

[0067] By substituting the estimated values ​​of the mean and standard deviation, we can obtain the operating habit probability model of driving input (accelerator pedal opening);

[0068] S2. Select longitudinal acceleration to represent the longitudinal motion intensity. Based on the acceleration data collected under ideal driving conditions, the probability model of driving motion intensity habits is obtained by processing in the same way:

[0069]

[0070] S3. The probability model of driving input shows the probability distribution of each accelerator pedal opening range. The larger the probability density of a certain accelerator pedal opening range, the more often the driver uses the corresponding accelerator pedal opening to operate; conversely, it means that the driver rarely applies the corresponding accelerator pedal opening, which well describes the driver's longitudinal operation habits. Similarly, the probability model of motion intensity shows the probability of different longitudinal accelerations occurring during the driver's driving process, which well describes the driver's driving motion intensity habits. The driver's most frequently used operation input is mapped to the longitudinal acceleration with the highest probability when driving, and the infrequently used operation input is mapped to the longitudinal acceleration with a lower probability. In this way, the longitudinal motion intensity (i.e., longitudinal acceleration) that the driver tends to prefer can be obtained based on the driver's longitudinal operation input.

[0071] S4. Use the same probability method to achieve the mapping of operation habits and exercise intensity habits, as follows:

[0072] Since the mapping is performed based on the driver's habits, the low-probability events at the two ends may correspond to emergency or special situations, which are not suitable for characterizing the driver's habits, so the vast majority of the center of the normal distribution is selected for mapping. The present invention selects 95% of the normal distribution area centered on the mean μ1 for mapping, that is:

[0073] P(|x-μ1|≤x lim )=0.95

[0074] 2μ1-x lim ≤x≤x lim

[0075] For (2μ1-x lim ,x lim ) range, and establish a mapping relationship between it and the longitudinal acceleration of the corresponding probability.

[0076] Assume that the amount of operation to be mapped is x cur , firstly, according to the probability calculation formula of normal distribution:

[0077]

[0078] Calculate the corresponding probability P(x≤x cur ), after obtaining the probability, the motion intensity of the mapping is calculated using the following formula

[0079]

[0080] In some embodiments, the specific construction method of the probability model construction module is as follows:

[0081] Step 1: Under ideal driving conditions, collect the accelerator pedal opening information of a driver during longitudinal driving. Use the accelerator pedal opening to represent the driver's longitudinal operation input, and use the normal distribution to build a probability model of driving input habits. Based on the sample data, use the moment estimation method to calculate the mean and standard deviation estimators to complete the parameter analysis of the model, and finally obtain the probability model of operation habits;

[0082] Step 2: Under ideal driving conditions, collect the longitudinal acceleration data of a driver during driving. Use the longitudinal acceleration to characterize the longitudinal motion intensity, and use the normal distribution to build a probability model of motion intensity habits. Based on the sample data, use the moment estimation method to calculate the corresponding mean and standard deviation estimators to complete the parameter analysis of the model, and then obtain the probability model of driving motion intensity habits.

[0083] In some embodiments, if the current accelerator pedal opening x cur In (2μ1-x lim ,x lim ), it is determined that the driver has the intention to accelerate at this time, and the current pedal opening information is input into the longitudinal acceleration intention analysis module; otherwise, it is determined that the driver has no intention to accelerate at this time, and no operation is performed.

[0084] In some embodiments, the acceleration intention analysis module is composed of an operation habit probability model and a driving motion intensity habit probability model.

[0085] In some embodiments, for (2μ1-x lim ,x lim ) range, and establish a mapping relationship between it and the longitudinal acceleration of the corresponding probability in the motion intensity probability model. According to the mapping relationship between the two, the corresponding longitudinal acceleration can be analyzed by inputting the driver's accelerator pedal opening in the current state.

[0086] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A longitudinal acceleration intention recognition method considering the driver's driving habit of sports intensity, comprising a probability model building module, an operation input judgment module and an acceleration intention analysis module, characterized in that: The probability model building module is used to build a probability model of the driver's accelerator pedal operation habit and a probability model of the driving habit of sports intensity. The operation input judgment module is used to receive the accelerator pedal opening information for judgment. The acceleration intention parsing module is used to parse the longitudinal acceleration that conforms to the driver's habits according to the current state. The longitudinal acceleration intention recognition method also includes the following steps: S1. Under ideal driving conditions, the longitudinal driving data of the driver is collected and the probability model of driving input habits is constructed using normal distribution. The opening of the accelerator pedal is selected to characterize the longitudinal driving operation input of the driver. Assume that the sample data of the accelerator pedal collected in the experiment is N in total, and the opening of the accelerator pedal x conforms to the following normal distribution: First, using the sample data x i Calculate the sample mean and sample variance S 2 : Then, using the moment estimation method, we establish an equation based on the population moment being equal to the sample moment: Combining the expectation and variance formulas of the normal distribution: E(X)=μ1 We can calculate μ1, Moment estimator Substitute the estimated values ​​of mean and standard deviation to obtain the operating habit probability model of driving input; S2. Select longitudinal acceleration to represent the longitudinal motion intensity. Based on the acceleration data collected under ideal driving conditions, the probability model of driving motion intensity habits is obtained by processing in the same way: S3. The probability model of driving input shows the probability distribution of each accelerator pedal opening interval. The larger the probability density of the accelerator pedal opening interval, the more often the driver uses the corresponding accelerator pedal opening to operate; conversely, it means that the driver does not often apply the corresponding accelerator pedal opening, which well describes the driver's longitudinal operation habits. The probability model of motion intensity shows the probability of different longitudinal accelerations during the driver's driving process, which well describes the driver's driving motion intensity habits. The driver's most frequently used operation input is mapped to the longitudinal acceleration with the highest probability when driving, and the infrequently used operation input is mapped to the longitudinal acceleration with a low probability. Therefore, according to the driver's longitudinal operation input, the longitudinal motion intensity that the driver is more inclined to can be obtained; S4. Use the same probability method to achieve the mapping of operation habits and exercise intensity habits, as follows: Since the mapping is based on the driver's habits, the low-probability events at both ends correspond to emergency or infrequent situations, which are not suitable for characterizing the driver's habits. Therefore, most of the center of the normal distribution is selected for mapping, and 95% of the normal distribution area centered on the mean μ1 is selected for mapping, that is: P(|x-μ1|≤x lim )=0.95 2μ1-x lim ≤x≤x lim For (2μ1-x lim ,x lim ) range, and establish a mapping relationship between it and the longitudinal acceleration of the corresponding probability; Assume that the amount of operation to be mapped is x cur , firstly, according to the probability calculation formula of normal distribution: Calculate the corresponding probability P(x≤x cur ), after obtaining the probability, the motion intensity of the mapping is calculated using the following formula The specific construction method of the probability model construction module is as follows: Step 1: Under ideal driving conditions, the driver's longitudinal accelerator pedal opening information is collected, and the accelerator pedal opening is used to represent the driver's longitudinal operation input. The probability model of driving input habits is constructed using normal distribution. The mean and standard deviation estimators are calculated based on sample data using moment estimation method to complete the parameter analysis of the model, and finally the probability model of operation habits is obtained; Step 2: Under ideal driving conditions, the longitudinal acceleration data of the driver is collected, and the longitudinal motion intensity is represented by the longitudinal acceleration. The probability model of motion intensity habits is constructed using the normal distribution. The corresponding mean and standard deviation estimators are calculated using the moment estimation method based on the sample data to complete the parameter analysis of the model, and then the probability model of driving motion intensity habits is obtained. If the current accelerator pedal opening x cur In (2μ1-x lim ,x lim ), it is determined that the driver has the intention to accelerate at this time, and the current pedal opening information is input into the longitudinal acceleration intention analysis module; otherwise, it is determined that the driver has no intention to accelerate at this time, and no operation is performed.

2. The longitudinal acceleration intention recognition method considering the driver's driving habit of sports intensity according to claim 1 is characterized by: The acceleration intention analysis module is composed of an operation habit probability model and a driving motion intensity habit probability model.

3. The longitudinal acceleration intention recognition method considering the driver's driving habit of sports intensity according to claim 2 is characterized by: For (2μ1-x lim ,x lim ) range, and establish a mapping relationship between it and the longitudinal acceleration of the corresponding probability in the motion intensity probability model. According to the mapping relationship between the two, the corresponding longitudinal acceleration can be analyzed through the driver's accelerator pedal opening input in the current state.

Images