Driving characteristic determination device

Abstract

To enable a driver to easily grasp driving characteristics including his / her own driving feeling.SOLUTION: A driver model generation unit 24 generates a plurality of driver models 18 for a plurality of drivers, based on the traveling data and driving sensation parameters for each of the plurality of drivers. The generated driver models 18 are models representing driving characteristics including driving feelings of the respective drivers. Classification processing unit 26 classifies the plurality of driver models 18 into a plurality of clusters CL according to the driving characteristics represented by the respective driver models 18. A cluster identification unit 28 identifies a cluster CL to which the driver model 18 of a target driver generated by the driver model generation unit 24 belongs. A notification processing unit 30 notifies the target driver of the driving characteristics including the driving sensation represented by the cluster CL to which the driver model 18 of the target driver belongs.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] This specification discloses an improvement in an operation characteristic determination device.

Background Art

[0002] Conventionally, a driver model representing the driving characteristics of a driver who drives a vehicle has been generated, and the driving characteristics of the driver have been determined based on the generated driver model. For example, in Patent Document 1, a driver model is generated based on the actually measured values of the operation of the vehicle (such as yawing angular velocity), the operation amount of the driver's operation (for example, the steering angle of the steering wheel), and the operation target value targeted by the driver, and the generated driver model is compared with other driver models to determine the driving characteristics of the current driver. An operation characteristic determination device is disclosed.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] During the driving of a vehicle, a driver performs various operations on the vehicle (such as steering operation, accelerator operation, brake operation, etc.). Here, the behavior of the vehicle expected by the driver may vary depending on the operation on the vehicle. In this specification, the relationship between the operation on the vehicle and the behavior of the vehicle expected by the driver accordingly is described as the "driving feeling" of the driver.

[0005] As an example of a driver's driving feeling, there is understeer / oversteer. Understeer as used in this specification means that when the driver turns the steering wheel by a certain steering angle amount during vehicle travel, the vehicle does not turn as much as the driver expects for that steering angle amount. Also, oversteer as used in this specification means that when the driver turns the steering wheel by a certain steering angle amount during vehicle travel, the vehicle turns more than the driver expects for that steering angle amount. Note that understeer originally means that when the vehicle is accelerating in a state where the vehicle is traveling on a circular track while maintaining a certain steering angle (in a steady circular turn), the vehicle deviates outside the circular track. And oversteer originally means that when the vehicle is accelerating in a state of steady circular turning, the vehicle deviates inside the circular track. However, in this specification, the terms understeer and oversteer are used in the above meanings.

[0006] Conventionally, a driver model of the driver has been generated and the driving characteristics of the driver have been determined based on the generated driver model. However, the driver's driving feeling has not been considered in the conventional driver model. Therefore, in the determination of the driver's driving characteristics based on the conventional driver model, the driver's driving feeling could not be determined.

[0007] The object of the driving characteristic determination device disclosed in this specification is to enable the driver to easily grasp the driving characteristics including their own driving feeling.

Means for Solving the Problem

[0008] The driving characteristic determination device disclosed in this specification includes a driver model generation unit that generates a driver model representing driving characteristics including the driving sense of the driver based on the driving data of the vehicle when the driver drives the vehicle and the driving sense parameters indicating the driving sense of the driver, a classification processing unit that classifies the plurality of driver models related to a plurality of drivers generated by the driver model generation unit into a plurality of clusters according to the driving characteristics, a cluster identification unit that identifies the cluster to which the driver model of the target driver generated by the driver model generation unit belongs among the plurality of clusters, and a notification processing unit that notifies the target driver of the driving characteristics indicated by the cluster to which the driver model of the target driver belongs.

[0009] The notification processing unit may propose to the target driver the content of the driving operation or the vehicle settings suitable for the driving characteristics indicated by the cluster to which the driver model of the target driver belongs.

Advantages of the Invention

[0010] According to the driving characteristic determination device disclosed in this specification, the driver can easily grasp the driving characteristics including his own driving sense.

Brief Description of the Drawings

[0011]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Modes for Carrying Out the Invention

[0012] Figure 1 is a schematic configuration diagram of the driving characteristic determination device 10 according to the present embodiment. The driving characteristic determination device 10 is configured by, for example, a server computer or a personal computer. Note that the driving characteristic determination device 10 may be any device as long as it can perform the functions described below.

[0013] The communication interface 12 is composed of, for example, a network adapter or the like. The communication interface 12 exhibits a function of communicating with other devices. For example, the communication interface 12 receives vehicle data obtained by a driver driving a vehicle, and various data such as driving sensation parameters. Details of these data will be described later.

[0014] The display 14 is composed of, for example, a liquid crystal display or an organic EL (Electro Luminescence) display. A screen corresponding to an instruction from the processor 22 described later is displayed on the display 14.

[0015] The memory 16 includes an HDD (Hard Disk Drive), an eMMC (embedded Multi Media Card), a ROM (Read Only Memory), a RAM (Random Access Memory), or the like. A driving characteristic determination program for operating each part of the driving characteristic determination device 10 is stored in the memory 16. Note that the driving characteristic determination program can also be stored in a computer-readable non-temporary storage medium such as a USB (Universal Serial Bus) memory or an SD card. The driving characteristic determination device 10 can read and execute the driving characteristic determination program from such a storage medium. Further, as shown in FIG. 1, a driver model 18 and a learning model 20 are stored in the memory 16.

[0016] The driver model 18 is a model that represents the driving characteristics of a driver. In particular, the driver model 18 in the present embodiment is a model in which the driving sensation of the driver is considered, in other words, a model that represents the driving characteristics including the driving sensation of the driver. The driver model 18 according to the present embodiment represents the driving characteristics of a racing driver in a circuit, but the driver model 18 may represent the driving characteristics of other drivers. The driver model 18 is generated by a driver model generation unit 24 described later and stored in the memory 16. The memory 16 stores a plurality of driver models 18 related to a plurality of drivers. Details of the driver model 18 will be described later together with the details of the driver model generation unit 24.

[0017] The learning model 20 is a model that performs clustering for dividing a plurality of driver models 18 into a plurality of clusters according to driving characteristics. The learning model 20 may be, for example, an unsupervised learning model such as a model using the k-means method, but the learning model 20 may be any model as long as the clustering is possible. Details of the learning model 20 will be described later together with the details of the classification processing unit 26.

[0018] The processor 22 is composed of, for example, a CPU (Central Processing Unit). The processor 22 is communicably connected to the communication interface 12, the display 14, and the memory 16 via a data bus. The processor 22 functions as a driver model generation unit 24, a classification processing unit 26, a cluster identification unit 28, and a notification processing unit 30 according to an operation characteristic determination program stored in the memory 16.

[0019] FIG. 2 is a conceptual diagram showing the generation process of the driver model 18. As shown in FIG. 2, the driver model generation unit 24 generates the driver model 18 based on the running data and the driving sensation parameters.

[0020] Running data refers to the running data of a vehicle when a driver drives the vehicle, and includes various data such as the average vehicle speed and centripetal acceleration at each corner. The running data can be obtained by, for example, various sensors mounted on the vehicle.

[0021] The driving feeling parameter is a parameter obtained by quantifying the driver's driving feeling. Although not limited thereto, here, the driving feeling parameter regarding understeer / oversteer as the driver's driving feeling will be described. The driving feeling parameter regarding understeer / oversteer is represented by an index K. The index K is represented by the difference between the theoretical steering angle and the actual steering angle of the steering wheel. The theoretical steering angle means the steering angle of the steering wheel when the vehicle is turning without lateral slip of the tires. The theoretical steering angle can be calculated based on the slip angle of the tire obtained from the turning radius of the vehicle and the steering gear ratio.

[0022] Generally, when K = 0 (the theoretical steering angle and the actual steering angle are the same), it means that the vehicle is turning as expected by the driver, and this is called neutral steer. However, since the expected turning manner of the vehicle varies among drivers with respect to the actual steering angle of the steering wheel, the value of the index K at which a driver feels neutral steer varies among drivers. For example, a certain driver feels neutral steer when K>0 (for example, K = 10), and another driver feels neutral steer when K<0 (for example, K = -5).

[0023] The value of the index K at which a driver feels neutral steer is not constant and varies during the running of the vehicle. Therefore, in the present embodiment, first, the driver is allowed to drive the vehicle for a certain distance (or a certain time), and the driver's comment on whether understeer or oversteer is felt at each corner of the circuit and the index K during running are obtained. Then, the average value of the index K when the driver feels understeer is defined as K US and the average value of the index K when the driver feels oversteer is defined as K OSLet it be so. The neutral steer for the driver is considered to be the intermediate value between K US and K OS So, if this value is represented as Δk, Δk is expressed by the following formula (1).

Equation

Equation

[0024] In this embodiment, the driver model 18 is represented by the t-dimensional model parameter x. The driver model 18 is generated by simulation. Specifically, the driver model generation unit 24 compares the driving data actually obtained by the driver driving the vehicle with the simulation driving data when the vehicle is driven by the driver model 18 represented by the t-dimensional model parameter x in the simulation. Then, the model parameter x is adjusted so that the difference becomes small.

[0025] Such a method for generating the driver model 18 can be represented, for example, by the following formula (3). Note that the driving data and the simulation driving data include data of a number of items such as, for example, the average vehicle speed and the centripetal acceleration, but formula (3) is a formula for the data of one of the items.

Equation

[0026] The driver model generation unit 24 adjusts the model parameter x so that δ1 in Equation (3) is minimized. The driver model generation unit 24 adjusts the model parameter x so that δ1 is minimized for each item included in the driving data and the simulated driving data. The sufficiently adjusted model parameter x represents the driver model 18 that represents the driving characteristics of the driver.

[0027] In the present embodiment, as described below, the driver model generation unit 24 further adjusts the model parameter x while considering the driving feeling parameters.

[0028] Specifically, the driver model generation unit 24 adjusts the model parameter x so that the difference between the simulated driving data f i (x) by the driver model 18 and the driving feeling parameter is minimized. Thereby, the model parameter x is adjusted to a value considering the driving feeling parameter, and the driver model 18 becomes a model representing the driving characteristics including the driving feeling of the driver. For example, when the driving feeling parameter is the above index K', the driver model generation unit 24 adjusts the model parameter x so that δ2 in the following Equation (4) is minimized.

Equation

[0029] The driver model generation unit 24 generates a plurality of driver models 18 for a plurality of drivers as described above. The generated plurality of driver models 18 are stored in the memory 16.

[0030] The classification processing unit 26 classifies a plurality of driver models 18 related to a plurality of drivers, generated by the driver model generation unit 24, into a plurality of clusters according to the driving characteristics represented by each driver model 18. Specifically, the classification processing unit 26 classifies the plurality of driver models 18 into a plurality of clusters based on a plurality of model parameters x included in each driver model 18. Since the driver model 18 in the present embodiment represents driving characteristics including the driving feeling of the driver, a plurality of drivers corresponding to the plurality of driver models 18 classified into the same cluster are drivers whose driving characteristics including the driving feeling are similar to each other.

[0031] In the present embodiment, the classification processing unit 26 classifies the plurality of driver models 18 into a plurality of clusters by inputting the plurality of model parameters x included in each driver model 18 into the learning model 20.

[0032] From the viewpoint of reducing the amount of calculation or calculation time of the classification processing, the classification processing unit 26 may perform the classification processing after reducing the dimensionality of the model parameter x. As a method for dimensionality reduction, for example, conventional techniques such as variational auto-encoder (VAE) and UMAP (Uniform Manifold Approximation and Projection) can be used.

[0033] FIG. 3 is a conceptual diagram showing clusters CL of the driver model 18. In FIG. 3, for simplicity, a plurality of driver models 18 (circles, triangles, and squares in FIG. 3) are plotted in a two-dimensional space, but the plurality of driver models 18 may be classified in a multi-dimensional space. In the example of FIG. 3, the plurality of driver models 18 are classified into three clusters: cluster CL1, cluster CL2, and cluster CL3.

[0034] After the classification process by the classification processing unit 26, the driver model generation unit 24 generates a driver model 18 for the target driver based on the driving data and driving sensation parameters for the target driver. The target driver is the driver who is the target of the notification by the notification processing unit 30 described later.

[0035] The cluster identification unit 28 identifies the cluster CL to which the driver model 18 of the target driver belongs among the plurality of clusters CL classified by the classification processing unit 26. Specifically, the cluster identification unit 28 determines the cluster CL to which the driver model 18 of the target driver belongs by comparing the plurality of model parameters x of the driver models 18 classified in each cluster CL with the model parameters x of the driver model 18 of the target driver.

[0036] In the present embodiment, the cluster identification unit 28 identifies the cluster CL to which the driver model 18 of the target driver belongs by inputting the plurality of model parameters x of the driver model 18 of the target driver into the learning model 20. As described above, since the plurality of drivers corresponding to the plurality of driver models 18 classified in the same cluster CL have similar driving characteristics including driving sensation, identifying the cluster CL to which the driver model 18 of the target driver belongs means identifying the drivers whose driving characteristics including driving sensation are similar to those of the target driver.

[0037] The notification processing unit 30 notifies the target driver of the driving characteristics including the driving sensation indicated by the cluster CL to which the driver model 18 of the target driver belongs. In the present embodiment, the notification processing unit 30 notifies the target driver by causing a message or an image indicating the driving characteristics including the driving sensation indicated by the cluster CL to which the driver model 18 of the target driver belongs to be displayed on the display 14. Thereby, the target driver can easily grasp the driving characteristics including his / her own driving sensation.

[0038] In addition, the notification processing unit 30 may propose to the target driver the content of the driving operation or the vehicle settings suitable for the driving characteristics including the driving feeling indicated by the cluster CL to which the driver model 18 of the target driver belongs. Specifically, for example, an administrator of the driving characteristic determination device 10 stores in advance in the memory 16 proposal information in which each cluster CL is associated with information indicating the content of the proposal to the driver corresponding to the driver model 18 belonging to the cluster CL (not shown in FIG. 1). Then, the notification processing unit 30 refers to the proposal information and proposes to the target driver the content of the proposal associated with the cluster CL to which the driver model 18 of the target driver specified by the cluster specifying unit 28 belongs. Examples of the content of the proposal include the strengths or weaknesses of the driver, recommended driving methods, or recommended vehicle settings.

[0039] For example, the notification processing unit 30 "Since you have a fast steering operation, you are good at circuits that require a fast steering operation." "With your driving style, the load on the front tires is large, so it is considered better to be more cautious with the accelerator in TGR corners." "For drivers with a fast steering operation, it is recommended to set the vehicle as follows: 〇〇." displays messages such as on the display 14.

[0040] Furthermore, the notification processing unit 30 may notify the target driver of the driving characteristics including the driving feeling indicated by the cluster CL to which the driver model 18 of a driver specified by the target driver (for example, the driver targeted by the target driver) belongs. Thereby, the target driver can grasp the driving characteristics including the driving feeling of the target driver.

[0041] The outline of the driving characteristic determination device 10 according to the present embodiment is as described above. Hereinafter, the processing flow of the driving characteristic determination device 10 according to the present embodiment will be described according to the flowcharts shown in FIGS. 4 and 5.

[0042] The flowchart of FIG. 4 is a flowchart showing the flow of a process of classifying a plurality of driver models 18 for a plurality of drivers into a plurality of clusters CL.

[0043] In step S10, the driver model generation unit 24 generates a plurality of driver models 18 for a plurality of drivers based on the driving data and driving feeling parameters for each of the plurality of drivers. The generated driver model 18 is a model representing driving characteristics including the driving feeling of the driver.

[0044] In step S12, the classification processing unit 26 classifies the plurality of driver models 18 for the plurality of drivers generated in step S10 into a plurality of clusters CL according to the driving characteristics represented by each driver model 18.

[0045] The flowchart of FIG. 5 is a flowchart showing the flow of a process of performing a notification according to the identified cluster CL after identifying the cluster CL to which the driver model 18 of the target driver belongs. At the start of the flowchart of FIG. 5, it is assumed that the generation of a plurality of driver models 18 for a plurality of drivers and the classification process of the plurality of driver models 18 into a plurality of clusters CL are completed.

[0046] In step S20, the driver model generation unit 24 generates a driver model 18 for the target driver based on the driving data and driving feeling parameters for the target driver.

[0047] In step S22, the cluster identification unit 28 identifies the cluster CL to which the driver model 18 of the target driver generated in step S20 belongs.

[0048] In step S24, the notification processing unit 30 notifies the target driver of the driving characteristics including the driving feeling indicated by the cluster CL identified in step S22 (that is, the cluster CL to which the driver model 18 of the target driver belongs).

[0049] The embodiments of the driving characteristic determination device according to the present disclosure have been described above. However, the driving characteristic determination device according to the present disclosure is not limited to the above embodiments, and various modifications are possible without departing from the spirit thereof.

Description of Reference Numerals

[0050] 10 Driving characteristic determination device, 12 Communication interface, 14 Display, 16 Memory, 18 Driver model, 20 Learning model, 22 Processor, 24 Driver model generation unit, 26 Classification processing unit, 28 Cluster identification unit, 30 Notification processing unit.

Claims

1. A driver model generation unit that generates a driver model representing driving characteristics including the driving sensation of the driver based on the driving data of the vehicle when the driver drives the vehicle and the driving sensation parameters indicating the driving sensation of the driver; A classification processing unit that classifies the plurality of driver models related to a plurality of drivers generated by the driver model generation unit into a plurality of clusters according to the driving characteristics; A cluster identification unit that identifies the cluster to which the driver model of the target driver generated by the driver model generation unit belongs among the plurality of clusters; A notification processing unit that notifies the target driver of the driving characteristics indicated by the cluster to which the driver model of the target driver belongs; A driving characteristic determination device, characterized by comprising the above components.

2. The notification processing unit proposes to the target driver the content of the driving operation or the vehicle setting suitable for the driving characteristics indicated by the cluster to which the driver model of the target driver belongs. The driving characteristic determination device according to claim 1, characterized by the above.

Images