Estimation device and method for road friction coefficient based on electric vehicles

Abstract

The electric vehicle-based road surface friction coefficient estimation device and method according to the present invention input a road condition image captured by an electric vehicle in motion into a road surface friction coefficient estimation model to estimate the road surface friction coefficient, and the electric vehicle controls the torque of a drive motor based on the estimated road surface friction coefficient. Furthermore, the road surface friction coefficient estimation model is trained using training data in which road condition images captured for different roads with varying pavement or wet conditions are matched with the road surface friction coefficient for each road condition image. The model is updated based on the road surface friction coefficient directly calculated by the electric vehicle according to the conditions at which wheel slip occurs, and the road condition image at the time of the wheel slip.

Description

Technology Field

[0001] The present invention relates to an apparatus and method for estimating the coefficient of friction of a road surface in real time using an electric vehicle. Background Technology

[0002] Recently, the use of electric vehicles has been increasing. Since these electric vehicles can generate greater torque compared to internal combustion engines, it is known that the likelihood of wheel slip—a phenomenon where the wheels spin freely due to excessive driving force exceeding the adhesion between the tire and the road surface on low-friction roads—is higher than that of internal combustion engine vehicles.

[0003] In other words, unlike conventional internal combustion engines, electric vehicles use a motor instead of an engine and a battery instead of fuel. Due to the characteristics of motor control, they have the advantage of maximizing initial launch torque compared to internal combustion engines. However, because the motor can generate maximum torque when accelerating with full throttle (Wide Open Throttle), the vehicle is more susceptible to wheel slip than internal combustion vehicles during starting or acceleration.

[0004] Meanwhile, since electric vehicles are driven by motors, the exact wheel torque being applied can be known, enabling accurate slip control, and when slip occurs, the wheel torque information obtained can be used to accurately estimate the coefficient of friction of the road surface.

[0005] In addition, as the utilization rate of connected vehicles has recently increased, road surface condition information captured by the vehicle can be transmitted to a server, making it possible to estimate the road surface friction coefficient using this information.

[0006] In this regard, the inventor proposed Registered Patent No. 10-2649360 (Title of Invention: Wheel Slip Control Method for Electric Vehicles). This invention relates to a wheel slip control method for electric vehicles, and more specifically, it utilizes machine learning techniques to more accurately estimate the coefficient of friction of the road surface and calculates an upper limit for motor torque based on this. By controlling the motor driver according to the calculated upper limit, it enables more accurate control of wheel slip in electric vehicles, which are more likely to experience wheel slip on low-friction roads compared to internal combustion engine vehicles.

[0007] More specifically, the main features include the step of acquiring real-time weather information including ambient temperature, humidity, and precipitation using connected car technology, acquiring driving vehicle information including wheel torque, wiper usage, and wheel speed measured through sensors equipped in the vehicle, and calculating the weight of the vehicle using the acquired vehicle information and vehicle specifications.

[0008] However, there may be cases where road surface conditions cannot be accurately estimated based solely on weather information. Since it is now possible to additionally obtain driving video information, we propose a new method to estimate the road surface friction coefficient using this data. Prior art literature

[0009] Republic of Korea Published Patent No. 10-2649360 (Title of Invention: Wheel Slip Control Method for Electric Vehicles). The problem to be solved

[0010] In order to solve the problems of the prior art, the present invention proposes an electric vehicle-based road surface friction coefficient estimation device and method capable of estimating the road surface friction coefficient using driving video information of an electric vehicle.

[0011] The problems of the present invention are not limited to those mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the description below. means of solving the problem

[0012] As a technical means for achieving the above-mentioned technical problem, an electric vehicle-based road surface friction coefficient estimation device according to an embodiment of the present invention comprises: a communication module; a memory storing a road surface friction coefficient estimation program; and a processor executing the road surface friction coefficient estimation program. The road surface friction coefficient estimation program inputs a road condition image captured by an electric vehicle in motion into a road surface friction coefficient estimation model to estimate the road surface friction coefficient, and causes the electric vehicle to control the torque of a drive motor based on the estimated road surface friction coefficient. The road surface friction coefficient estimation model is trained using training data in which road condition images captured for each road with different pavement conditions or wetness conditions are matched with the road surface friction coefficient for each road condition image, and is trained to estimate the road surface friction coefficient of the corresponding road when a road condition image is input.

[0013] In addition, the operation method of the road surface friction coefficient model performed by the electric vehicle-based road surface friction coefficient estimation device includes the step of receiving road condition images captured by the electric vehicle and the step of inputting the road condition images into the road surface friction coefficient estimation model to estimate the road surface friction coefficient. In this case, the road surface friction coefficient estimation model is trained using training data in which road condition images captured for roads with different pavement conditions or wetness conditions are matched with the road surface friction coefficient for each road condition image. Effects of the invention

[0014] The present invention addresses the limitation of conventional methods where road surface conditions during electric vehicle driving could not be accurately determined solely by weather information; by additionally utilizing road condition images captured during driving, it estimates the road surface friction coefficient in real time and can reflect this in the driving of the electric vehicle. Brief explanation of the drawing

[0015] Figure 1 shows the configuration of a system according to an embodiment of the present invention. Figure 2 shows the server configuration of an electric vehicle-based road surface friction coefficient estimation device according to an embodiment of the present invention. FIG. 3 shows the basic configuration of an electric vehicle according to an embodiment of the present invention. FIG. 4 shows an example of a road surface friction coefficient estimation model according to an embodiment of the present invention. FIG. 5 shows a flowchart of a method for estimating the road surface friction coefficient according to an embodiment of the present invention. FIG. 6 illustrates the operation of a road surface friction coefficient estimation model according to an embodiment of the present invention. FIG. 7 is a flowchart of the operation of a road surface friction coefficient estimation model according to an embodiment of the present invention. FIG. 8 is a graph for estimating the road surface friction coefficient according to an embodiment of the present invention. FIG. 9 shows the structure of an electric vehicle according to an embodiment of the present invention. Specific details for implementing the invention

[0016] Embodiments of the present invention are described in detail below with reference to the attached drawings so that those skilled in the art can easily implement the invention. However, the present invention may be embodied in various different forms and is not limited to the embodiments described herein. Furthermore, in order to clearly explain the invention in the drawings, parts unrelated to the explanation have been omitted, and similar parts throughout the specification are denoted by similar reference numerals.

[0017] Throughout the specification of the present invention, when a part is described as being "connected" to another part, this includes not only cases where they are "directly connected," but also cases where they are "electrically connected" with other elements interposed between them.

[0018] Throughout the specification of the present invention, when a member is described as being located “on” another member, this includes not only cases where a member is in contact with another member, but also cases where another member exists between the two members.

[0019] Preferred embodiments of the present invention will be described in detail below with reference to the attached drawings and the contents described below. However, the present invention is not limited to the embodiments described herein and may be embodied in other forms. Throughout the specification, the same reference numerals indicate the same components.

[0020] Hereinafter, an electric vehicle-based road surface friction coefficient estimation device and method according to an embodiment of the present invention will be described.

[0021] Figure 1 illustrates the configuration of the system (10) of the present invention.

[0022] A system (10) according to one embodiment of the present invention includes at least one electric vehicle (200) and a server (100) that receives various information through data communication with the electric vehicles. The electric vehicle-based road surface friction coefficient estimation device of the present invention may be implemented in the form of a server (100) or in a form executed in an electric vehicle (200).

[0023] Such a road surface friction coefficient estimation device is implemented in the form of a computing device, and when the road surface friction coefficient estimation device is implemented as a server (100), it can operate in a cloud computing service model such as SaaS (Software as a Service), PaaS (Platform as a Service), or IaaS (Infrastructure as a Service). In addition, the road surface friction coefficient estimation device can be built in the form of a private cloud, a public cloud, or a hybrid cloud.

[0024] As shown in FIG. 2, the server (100) of the electric vehicle-based road surface friction coefficient estimation device of the present invention includes a communication module (110), a memory (120), and a processor (130).

[0025] The communication module (110) may be a device including hardware and software necessary to transmit and receive signals, such as control signals or data signals, through wired or wireless connections with other network devices. The communication module (110) can perform data communication with a plurality of electric vehicles (200) to receive road condition images captured by each electric vehicle (200). Additionally, the communication module (110) can transmit road surface friction coefficient estimation results to each electric vehicle (200).

[0026] The memory (120) may contain a road surface friction coefficient estimation program. The road surface friction coefficient estimation program inputs a road condition image captured by an electric vehicle in motion into a road surface friction coefficient estimation model to estimate the road surface friction coefficient, and causes the electric vehicle (200) to control the driving motor torque based on the estimated road surface friction coefficient.

[0027] At this time, memory (120) should be interpreted as a general term for a non-volatile storage device that retains stored information even when power is not supplied, and a volatile storage device that requires power to retain stored information. Memory (120) can perform the function of temporarily or permanently storing data processed by the processor (130). Memory (120) may include magnetic storage media or flash storage media in addition to a volatile storage device that requires power to retain stored information, but the scope of the present invention is not limited thereto.

[0028] The processor (130) executes a road surface friction coefficient estimation program stored in memory (120) and inputs a road condition image captured by the electric vehicle (200) in motion into the road surface friction coefficient estimation model to estimate the road surface friction coefficient. In this way, the estimated road surface friction coefficient is transmitted to each electric vehicle (200) so that the electric vehicle (200) controls the torque of the drive motor.

[0029] A processor (130) may refer to a data processing device embedded in hardware having a physically structured circuit to perform a function expressed by code or instructions included in a program, for example. Examples of such data processing devices embedded in hardware may include a microprocessor, a central processing unit (CPU), a processor core, a multiprocessor, an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), etc., but the scope of the present invention is not limited thereto.

[0030] Additionally, the server (100) can estimate the road surface friction coefficient using information received from the electric vehicle (200) and build or update a road surface friction coefficient estimation model.

[0031] FIG. 3 illustrates the basic configuration of the electric vehicle (200) of the present invention.

[0032] The electric vehicle (200) includes a camera (210), a drive motor (220), a battery (230), an interface device (240), a control unit (250), and a communication module (260).

[0033] In particular, the electric vehicle (200) can directly estimate the road surface friction coefficient and build or update a road surface friction coefficient estimation model by means of a program executed in the control unit (250). Additionally, when wheel slip occurs, the control unit (250) directly calculates the road surface friction coefficient and transmits the road condition image at the time of wheel slip and the directly calculated road surface friction coefficient to the road surface friction coefficient estimation model so that the road surface friction coefficient estimation model is updated.

[0034] As previously explained, the road surface friction coefficient estimation device of the present invention can be implemented in the form of a server (100) or a computing device running on at least one electric vehicle (200).

[0035] In addition, the road surface friction coefficient estimation device estimates the road surface friction coefficient in real time using a driving video received from an electric vehicle (200) using a road surface friction coefficient estimation model learned using a deep neural network as shown in FIG. 4. In addition, the road surface friction coefficient estimation model can be updated based on information regarding wheel slip.

[0036] Here, the road surface friction coefficient estimation model is trained based on a deep neural network and estimates the road surface friction coefficient based on driving footage of an electric vehicle. Specifically, a learning model is constructed to estimate the road surface friction coefficient using training data that includes road condition footage captured under various driving environments of the electric vehicle and road surface friction coefficients matched to those footages. For example, the learning model can be constructed based on a Convolutional Neural Network (CNN) model that extracts feature information from the footage.

[0037] In addition, the road condition images may include multiple road images captured with various road pavement conditions (asphalt pavement, concrete pavement, unpaved road, etc.) or various road wetness conditions (road condition images with different degrees of dryness or wetness, road condition images with different degrees of snowfall, ice formation, etc.), and the pre-measured road surface friction coefficients for each road image are matched to constitute the training data.

[0038] In addition, the road condition images are classified by time unit for each vehicle, capturing footage of a pre-set Region of Interest (ROI) facing the road in front or behind the vehicle, and each image is utilized as training data.

[0039] Meanwhile, the road surface friction coefficient estimation model built on the server (100) or electric vehicle (200) may be prepared differently for each vehicle, or may be built by additionally including vehicle information in the training data. That is, as described below, since the road surface friction coefficient may vary depending on the mass of the vehicle and the force of the tires, the road surface friction coefficient estimation model may be built in a form that estimates the road surface friction coefficient by additionally considering these factors.

[0040] FIG. 5 is a flowchart of the operation method of a road surface friction coefficient model performed by a road surface friction coefficient estimation device according to one embodiment of the present invention.

[0041] First, the road surface friction coefficient estimation model receives a road condition image captured by the electric vehicle (200) (S110).

[0042] In this case, the road surface friction coefficient estimation model was trained using training data in which road condition images captured for roads with different pavement or wet conditions were matched with the road surface friction coefficient for each road condition image.

[0043] Next, the road surface friction coefficient estimation model inputs the received road condition image into the road surface friction coefficient estimation model to estimate the road surface friction coefficient (S120).

[0044] Specifically, when the road surface friction coefficient estimation model is primarily learned, the operation of the road surface friction coefficient estimation device changes depending on whether wheel slip occurs in the road condition image received while the electric vehicle (200) is driving, as shown in FIGS. 6 and 7.

[0045] At this time, whether wheel slip occurs can be determined using the vehicle's driving speed and the wheel's circumferential speed as shown in the formula below.

[0046]

[0047] If the slip ratio calculated according to the above formula is detected to be greater than a specific threshold value, it can be determined that wheel slip has occurred. This threshold is defined as a tunable parameter and typically has a value of around 0.2.

[0048] The process of determining whether wheel slip occurs based on the slip ratio described above and estimating the road surface friction coefficient based on whether wheel slip occurs is as follows.

[0049] First, if wheel slip does not occur, the road condition images captured by each vehicle are transmitted to the road surface friction coefficient estimation model of the server (100) or electric vehicle (200) (S200).

[0050] Next, the road surface friction coefficient estimation model estimates the road surface friction coefficient using the received image (S210).

[0051] At this time, the estimated road surface friction coefficient is transmitted to the control unit (250) of the electric vehicle (200), and the control unit (250) uses the road surface friction coefficient to adjust the torque of the electric vehicle (200) so that wheel slip does not occur.

[0052] Finally, the control unit (250) may output a notification warning the driver when the road surface friction coefficient is below a threshold value, or limit the vehicle speed by taking into account the road surface friction coefficient (S220).

[0053] Next, when wheel slip occurs, the electric vehicle (200) directly calculates the road surface friction coefficient in that state, and transmits the road condition image at the time of wheel slip and the directly calculated road surface friction coefficient to a road surface friction coefficient estimation model executed in the electric vehicle (200) or a road surface friction coefficient estimation model executed in the server (100) so that the road surface friction coefficient estimation model is updated. More specifically, we will examine the process of calculating the road surface friction coefficient when wheel slip occurs.

[0054] First, the mass and tire force of the electric vehicle (200) are calculated. (S300)

[0055] Referring to Fig. 8, the coefficient of friction between the tire and the ground surface cannot be determined during constant speed driving. The coefficient of friction between the tire and the ground surface can be estimated by utilizing the nonlinear characteristics of the maximum static friction coefficient between the tire and the ground surface when the tire force reaches near saturation.

[0056] In addition, Figure 8 illustrates the relationship between the tire slip ratio and the road surface friction coefficient under various road surface conditions (dry asphalt, dry concrete, snowy road, icy road).

[0057] Since the electric vehicle (200) is composed of a motor (220) capable of accurately determining most of the torque for braking and driving, it is possible to accurately estimate the friction coefficient when the tire force is saturated. Therefore, a conventional road surface friction coefficient estimator can estimate the road surface friction coefficient using the tire force when the slope of the graph becomes 0.

[0058] Meanwhile, to estimate the coefficient of friction, the weight of the vehicle can be estimated as follows. Using the geometric structure of the vehicle and the formula for the force acting on the vehicle shown in FIG. 9, the mass (m) of the moving vehicle can be calculated by the following equation (1).

[0059] (1) Here, , , , Each represents the driving force of the electric vehicle (200) in motion, that is, the force applied to the wheels, and the air resistance, rolling resistance, and gradient resistance acting on the electric vehicle (200) in motion, and can be calculated by the following equations (2) to (5).

[0060] (2)

[0061] (3)

[0062] (4)

[0063] (5)

[0064] At this time, the variables used in the above equations (2) to (5) are all information that can be obtained from the vehicle information obtained in the vehicle information acquisition step and from geometric relationships including the vehicle specifications and the slope angle (θ) of the road surface. Since the air resistance, rolling resistance, and gradient resistance acting on the vehicle corresponding to the above equations (3) to (5) correspond to already known calculation formulas, a detailed explanation is omitted.

[0065] In addition, the used in the above formula (2) represents the torque input to the transmission (TM), in the case of the EV mode applied to the electric vehicle (200). It can be represented as follows.

[0066] and, class represents the gear ratios of the transmission and differential, respectively. represents the efficiency of the electric vehicle (200), and is the radius of the wheel.

[0067] And, using the mass estimation results described above, the force of the tire can be estimated as follows.

[0068] ( : Shaft torque, Brake torque, : Wheel moment of inertia, :Cloud resistance, : Effective radius, : Tire power)

[0069] Next, the coefficient of friction of the road surface is calculated using the tire force and mass values ​​obtained from the above-described formula. (S310)

[0070] At this time, the coefficient of friction can be defined as the maximum value obtained by dividing the tire force from 2 to 3 seconds before the occurrence of wheel slip until the point of slip occurrence by the vertical load value, as shown in the formula below.

[0071]

[0072] Next, when wheel slip occurs, the road condition image captured by the electric vehicle (200) and the road surface friction coefficient calculated in the above process are transmitted as new training data to the server (100) or the road surface friction coefficient estimation model (S320).

[0073] Finally, the road surface friction coefficient estimation model is updated based on the transmitted image and the road surface friction coefficient. (S330)

[0074] In this way, the electric vehicle (200) directly calculates the road surface friction coefficient using its own information when wheel slip occurs, and transmits the road condition image taken when wheel slip occurs and the directly calculated road surface friction coefficient as new training data to the road surface friction coefficient model so that the road surface friction coefficient model is updated.

[0075] Additionally, the step (S120) of inputting the received road condition image into a road surface friction coefficient estimation model to estimate the road surface friction coefficient may include the step of the road surface friction coefficient estimation model outputting the estimated road surface friction coefficient to the corresponding server (100) or electric vehicle (200).

[0076] One embodiment of the present invention may also be implemented in the form of a recording medium comprising computer-executable instructions, such as program modules executed by a computer. A computer-readable medium may be any available medium accessible by a computer and includes both volatile and non-volatile media, and both removable and non-removable media. Additionally, a computer-readable medium may include a computer storage medium. A computer storage medium includes both volatile and non-volatile, removable and non-removable media implemented by any method or technique for storing information, such as computer-readable instructions, data structures, program modules, or other data.

[0077] Although the method and system of the present invention have been described in relation to specific embodiments, some or all of their components or operations may be implemented using a computer system having a general-purpose hardware architecture.

[0078] The foregoing description of the present invention is for illustrative purposes only, and those skilled in the art will understand that other specific forms can be easily modified without altering the technical concept or essential features of the present invention. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. For example, each component described as a single unit may be implemented in a distributed manner, and components described as distributed may likewise be implemented in a combined form.

[0079] A person skilled in the art to which the present invention pertains will understand that, based on the foregoing description, other specific forms can be easily modified without altering the technical spirit or essential features of the present invention. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. The scope of the present invention is defined by the claims set forth below, and all modifications or variations derived from the meaning and scope of the claims and equivalent concepts should be interpreted as being included within the scope of the present invention. Explanation of the symbols

[0080] 10: System 100: Server 110: Communication module 120: Memory 130: Processor 200: Electric Vehicle 250: Control unit

Claims

Claim 1 A road surface friction coefficient estimation device based on an electric vehicle comprises: a communication module; a memory storing a road surface friction coefficient estimation program; and a processor executing the road surface friction coefficient estimation program, wherein the road surface friction coefficient estimation program inputs a road condition image captured by an electric vehicle in motion into a road surface friction coefficient estimation model to estimate the road surface friction coefficient, and causes the electric vehicle to control the torque of a drive motor based on the estimated road surface friction coefficient, wherein the road surface friction coefficient estimation model is trained using training data in which road condition images captured for each road with different pavement conditions or wet conditions of the road and road surface friction coefficients for each road condition image are matched, and is trained to estimate the road surface friction coefficient of the corresponding road when the road condition image is input. Claim 2 A road surface friction coefficient estimation device according to claim 1, wherein the road surface friction coefficient estimation program updates the road surface friction coefficient estimation model based on the road surface friction coefficient directly calculated by the electric vehicle according to the conditions at the time of wheel slip occurrence and the road condition image at the time of wheel slip occurrence when wheel slip occurs in the electric vehicle while driving. Claim 3 In claim 1, the road surface friction coefficient estimation device is implemented in the form of a server that communicates with at least one electric vehicle. Claim 4 In claim 1, the road surface friction coefficient estimation device is implemented in the form of a computing device executed in the electric vehicle. Claim 5 A method of operation of a road surface friction coefficient model performed by an electric vehicle-based road surface friction coefficient estimation device, wherein the road surface friction coefficient estimation model is trained using training data in which road surface friction coefficients for each road surface friction coefficient are matched with road condition images taken for each road with different pavement conditions or wet conditions of the road, and the method of operation of a road surface friction coefficient model comprises the steps of receiving a road condition image taken by an electric vehicle and inputting the road condition image into the road surface friction coefficient estimation model to estimate the road surface friction coefficient. Claim 6 A method of operation of a road surface friction coefficient model according to claim 5, further comprising the step of updating the road surface friction coefficient estimation model based on a road surface friction coefficient directly calculated by the electric vehicle according to conditions when wheel slip occurs and an image of the road condition when wheel slip occurs. Claim 7 A method of operation of a road surface friction coefficient model according to claim 5, further comprising the step of outputting the road surface friction coefficient estimated in the above step to the electric vehicle.

Images