Travel control system and travel control method

Abstract

Travel control system and travel control method. A travel control system for a vehicle provided with a drive source, a wheel having a wheel body connected to the drive source via a power transmission member and having a tire mounted on the wheel body, and a brake device for braking the wheel, the travel control system including: an estimation unit configured to estimate a tire torsional stiffness and a road surface friction coefficient based on at least a rotational speed of the drive source, a rotational speed of the wheel body, a vehicle body speed, and a torque applied to the wheel body; and a control unit configured to control at least one of the drive source and the brake device to cause the tire not to exceed an adhesion limit derived from the tire torsional stiffness and the road surface friction coefficient.

Description

TECHNICAL FIELD

[0001] The present application relates to a traveling control system and a traveling control method. BACKGROUND

[0002] When the tire is gripping, the drive shaft torsional vibration occurs, and when the tire is slipping, the tire slip causes the drive shaft torsion to be released and the drive shaft torsional vibration to disappear. JP2018-155696 discloses a road surface determining device that determines the road surface by observing the vibration mode of a selected portion, on the basis of the fact that, when the drive shaft torsional vibration occurs, a vibration mode in which the engine set forward-backward resonance and the drive shaft torsion resonance are combined appears on the engine set and the sprung mass (vehicle body), and, when the drive shaft torsional vibration disappears, a simple vibration in which the engine forward-backward resonance appears.

[0003] JP2019-31112A discloses a traveling control method that detects a rotational fluctuation of a differential device and a rotational fluctuation of a wheel body connected to a drive shaft via the differential device, sets a slip recognition amount on the basis of an amplitude ratio and a phase delay of a rotational fluctuation amplitude of the wheel body with respect to a rotational fluctuation amplitude of the differential device, and controls a driving force of the tire so that the slip recognition amount does not exceed a slip recognition amount threshold value corresponding to an elastic slip limit of the tire to the road surface.

[0004] In JP2018-155696A, it is determined that the tire is mainly slipping when the disappearance of the drive shaft torsional vibration is confirmed by vibration measurement using an acceleration sensor. However, when the drive shaft torsional vibration disappears, the tire is already in a slip state (road surface μ maximum state), and the behavior of the vehicle has already begun to become unstable. The ideal control to stabilize the behavior of the vehicle is to take measures to suppress the slip immediately before the slip, but in the case where it is determined whether the tire is slipping after the drive shaft torsional vibration disappears, such ideal control cannot be achieved. JP2018-155696A also proposes a method based on vibration measurement using an acceleration sensor for avoiding a decrease in detection accuracy of the wheel speed at low speed. However, the vibration of the engine set and the sprung mass (vehicle body) is generated by cooperation of many components such as a suspension and a chassis, and its measurement error is often affected by variations and / or deterioration of the components. Furthermore, since the vibration is also affected by the mass of the sprung mass (vehicle body), the use conditions such as the number of occupants and the load amount can be a cause of the error.

[0005] In JP 2019-31112 A, since a slip identification quantity capable of determining an elastic slip limit is introduced, it is possible to predict an adhesion limit (elastic slip limit) below which tire slip (slip slip) does not occur, before slip occurs. However, the slip identification quantity is an index normalized by the slip speed of the tire, and although it is possible to predict the tire slip speed corresponding to the adhesion limit, it is not possible to directly predict the tire driving force at that time. In actual control, the driving force of the tire is controlled to indirectly control the slip speed, and thus the tire can momentarily be in a slip slip state. Therefore, the travel control method of JP 2019-31112 A has room for improvement. Specifically, if the tire driving force corresponding to the adhesion limit can be directly estimated, it is possible to keep the tire in an elastic slip state.

[0006] In addition, each of JP 2018-155696 A and JP 2019-31112 A uses some vibration or rotational fluctuation, and when no vibration is generated or the vibration is very small with respect to the noise of the sensor, tire slip cannot be determined. SUMMARY

[0007] In view of the above background, a main object of the present application is to provide a vehicle control system and a vehicle control method capable of estimating the driving force of a tire corresponding to an adhesion limit with high precision without taking into account the deterioration and / or change of component differences depending on the use of the vehicle, and keeping the tire in an elastic slip state. Another object of the present application is to provide a vehicle control system and a vehicle control method applicable even when no specific vibration or rotational fluctuation is generated.

[0008] To achieve the above object, one aspect of the present application provides a travel control system for a vehicle 1 provided with a driving source 5, a wheel 3 having a wheel body W connected to the driving source via a power transmission member 6, and having a tire T mounted on the wheel body, and a brake device 8 for braking the wheel, the system including: a first rotation sensor 12C configured to acquire a rotation speed of the driving source; a second rotation sensor 12A configured to acquire a rotation speed of the wheel body; a vehicle body speed acquisition unit 12B, 12D configured to acquire information related to a vehicle body speed; a torque acquisition unit 14C configured to acquire a torque applied to the wheel body; an estimation unit 14A configured to estimate a tire torsional stiffness as a stiffness of the tire and a road surface friction coefficient as a friction characteristic between the tire and the road surface, based on at least the rotation speed of the driving source, the rotation speed of the wheel body, the vehicle body speed, and the torque applied to the wheel body; and a control unit 14B configured to control at least one of the driving source and the brake device to cause the tire not to exceed an adhesion limit derived from the tire torsional stiffness and the road surface friction coefficient.

[0009] According to this aspect, the tire can be kept in the elastic slip state. Since the tire is kept in the elastic slip state, the vehicle behavior matches the steering operation, and energy loss can be reduced.

[0010] In the above aspect, preferably, the estimation unit estimates an adhesion limit driving force corresponding to the adhesion limit of the tire on the basis of the tire torsional stiffness and the road surface friction coefficient.

[0011] According to this aspect, the estimation unit estimates the adhesion limit driving force on the basis of the tire torsional stiffness and the road surface friction coefficient that have been estimated. Therefore, the estimation unit can obtain an accurate adhesion limit driving force.

[0012] In the above aspect, preferably, the control unit controls at least one of the driving source and the braking device so that the absolute value of the driving force of the tire does not exceed the adhesion limit driving force.

[0013] According to this aspect, since the control unit controls at least one of the driving source and the braking device on the basis of the adhesion limit driving force, the tire is kept in the elastic slip state.

[0014] In the above aspect, preferably, the vehicle body speed acquisition unit includes a non-driven wheel rotation sensor 12B configured to acquire a rotational speed of a non-driven wheel, and an acceleration sensor 12D configured to acquire a front-rear acceleration of the vehicle, and the estimation unit estimates the vehicle body speed by using a weighted average of a first speed acquired by the non-driven wheel rotation sensor and a second speed acquired by integrating the front-rear acceleration acquired by the acceleration sensor, and changes a weight of the weighted average to maximize a likelihood of state quantities (e.g., the tire torsional stiffness and the road surface friction coefficient) estimated by the estimation unit.

[0015] According to this aspect, the vehicle body speed can be accurately estimated even when a slip slip occurs in the non-driven wheel.

[0016] In the above aspect, preferably, the estimation unit determines whether the road surface is rough, and when it is determined that the road surface is rough, changes the weight to reduce an influence of the first speed on the vehicle body speed.

[0017] According to this aspect, the vehicle body speed can be accurately estimated even when the vehicle is traveling on a rough road. When the vehicle is traveling on a rough road, rotational fluctuations occur in the non-driven wheels due to irregularities in the road surface, and since the timing of passing over the irregularities in the road surface differs between the driven wheels and the non-driven wheels, such rotational fluctuations become observation noise in obtaining the vehicle body speed. Therefore, when the vehicle is traveling on a rough road, it is preferable to reduce the influence of the first speed of the non-driven wheels on the vehicle body speed. On the other hand, when it is determined that the vehicle is traveling on a flat road, it is possible to increase the influence of the first speed of the non-driven wheels on the vehicle body speed, thereby improving responsiveness of the estimated value when the road surface friction coefficient suddenly changes. In this way, by determining whether the road surface is rough and changing the influence of the first speed of the non-driven wheels on the vehicle body speed accordingly, it is possible to achieve both robustness when traveling on a rough road and high responsiveness of the estimated value.

[0018] In the above aspect, preferably, the estimation unit predicts the wheel load based on the front-rear acceleration and the lateral acceleration, and determines the wheel load so as to maximize a likelihood of at least one of the wheel load, the tire torsional stiffness, and the road surface friction coefficient.

[0019] According to this aspect, it is possible to estimate the wheel load while taking into account changes in the sprung mass (vehicle body) attitude such as pitch and roll, and changes in the number of passengers and / or the load. The estimation unit corrects changes in the sprung mass (vehicle body) attitude (which is a transient change in the wheel load) based on the front-rear acceleration and the lateral acceleration, and determines the wheel load so as to be a maximum value of the likelihood described above, thereby correcting steady-state changes in the wheel load due to changes in the number of passengers and / or the load. By using the estimated wheel load, it is possible to correctly estimate the tire torsional stiffness and the road surface friction coefficient.

[0020] In the above aspect, preferably, the estimation unit calculates an adhesion limit slip ratio corresponding to the adhesion limit of the tire with respect to the road surface based on the tire torsional stiffness and the road surface friction coefficient, and the control unit controls at least one of the drive source and the brake device so that the absolute value of the slip ratio of the tire is smaller than or equal to the adhesion limit slip ratio.

[0021] According to this aspect, it is possible to return the tire to the elastic slip state based on the slip ratio and the adhesion limit slip ratio. In the sliding slip state, the driving force of the tire decreases as the slip ratio increases. Therefore, in the sliding slip state, it is possible to maintain the sliding slip state even if the driving force of the tire is smaller than or equal to the adhesion limit driving force. In this case, it is possible to return the tire to the elastic slip state by limiting the driving force of the tire so that the slip ratio is smaller than or equal to the adhesion limit slip ratio.

[0022] Another aspect of the present application provides a travel control method of a vehicle 1 provided with a drive source 5, a wheel 3 having a wheel body W connected to the drive source via a power transmission member 6 and having a tire T mounted on the wheel body W, a brake device 8 for braking the wheel, and a control device 14, the method being executed by the control device and including: estimating a tire torsional stiffness that is a stiffness of the tire and a road surface friction coefficient that is a friction characteristic between the tire and a road surface, based on at least a rotational speed of the drive source, a rotational speed of the wheel, and a vehicle body speed; estimating an adhesion limit driving force corresponding to an adhesion limit of the tire, based on the tire torsional stiffness and the road surface friction coefficient; and controlling at least one of the drive source and the brake device so that an absolute value of a driving force of the tire is kept smaller than or equal to the adhesion limit driving force.

[0023] According to this aspect, since the tire torsional stiffness and the road surface friction coefficient are estimated, and the adhesion limit driving force is estimated based on the tire torsional stiffness and the road surface friction coefficient that have been estimated, an accurate adhesion limit driving force can be obtained. By controlling the drive source and the brake device based on the adhesion limit driving force, the tire is kept in an elastic slip state. Since the tire is kept in the elastic slip state, the vehicle behavior matches the steering operation, and energy loss can be reduced.

[0024] According to the above-described configuration, it is possible to provide a vehicle control system and a vehicle control method that can estimate a driving force of a tire corresponding to an adhesion limit and keep the tire in an elastic slip state. BRIEF DESCRIPTION OF DRAWINGS

[0025] Figure 1 is a configuration diagram of a vehicle on which a vehicle control system is mounted;

[0026] Figure 2 is a graph showing a relationship between a slip ratio and a tire driving force;

[0027] Figure 3 is an explanatory diagram showing a dynamic model of a driving wheel;

[0028] Figure 4A is a graph showing a rotational fluctuation transmission characteristic between a differential device and a driving wheel;

[0029] Figure 4B is an explanatory diagram showing a relationship between a frequency and a vibration mode;

[0030] Figure 5 is a graph showing root loci of an elastic slip mode and a sliding slip mode;

[0031] Figure 6is a graph showing a relationship between tire torsional stiffness, a road surface friction coefficient, and an attachment limit slip ratio;

[0032] Figure 7 is a flowchart showing a procedure of a travel control method executed by the control device;

[0033] Figure 8 is a graph showing lateral acceleration with respect to a front wheel steering angle in the embodiment and the comparative example; and

[0034] Figure 9 is an explanatory view showing a path and a theoretical path in the embodiment and the comparative example. DETAILED DESCRIPTION

[0035] Hereinafter, a travel control system and a vehicle control method according to one embodiment of the present application will be described with reference to the drawings. As shown in Figure 1 The vehicle 1 is a four-wheel automobile and has a vehicle body 2 and four wheels 3 provided on the vehicle body 2. The wheels 3 include two front wheels 3F as drive wheels and two rear wheels 3R as non-drive wheels. Each wheel 3 has a wheel body W and a tire T mounted on the wheel body W.

[0036] The vehicle 1 has a drive source 5 for driving the front wheels 3F. The drive source 5 can be an internal combustion engine or an electric motor. The drive source 5 can include a speed reducer and a differential device. In the present embodiment, the drive source 5 is configured by an internal combustion engine 5A, a speed reducer 5B, and a differential device 5C (DN). The differential device 5C of the drive source 5 is connected to each front wheel 3F via a power transmission member 6. The power transmission member 6 can be a drive shaft.

[0037] The vehicle 1 has a brake device 8 for braking each wheel 3. Each brake device 8 includes a hydraulic pressure supply device 8A and a disc brake 8B provided in the wheel body W of each wheel 3 to be actuated by hydraulic pressure from the hydraulic pressure supply device 8A.

[0038] The vehicle 1 has a travel control system 10 that controls the drive source 5 and the brake device 8. The travel control system 10 includes a control device 14 that controls the drive source 5 and the brake device 8 based on signals from a driving operation element 11 and a vehicle sensor 12. The driving operation element 11 includes a steering wheel 11A for receiving a steering operation of a driver, an accelerator pedal 11B for receiving an acceleration operation of the driver, and a brake pedal 11C for receiving a deceleration operation of the driver.

[0039] The vehicle sensors 12 include: left and right front wheel speed sensors 12A (first rotation sensors) that detect the rotational speeds of the left and right front wheels, respectively; left and right rear wheel speed sensors 12B (non-driven wheel rotation sensors) that detect the rotational speeds of the left and right rear wheels, respectively; a drive source rotational speed sensor 12C (second rotation sensor) that detects the rotational speed of the output end of the drive source 5; and an acceleration sensor 12D that detects the front-rear and lateral accelerations of the vehicle body 2. The front wheel speed sensors 12A and the rear wheel speed sensors 12B each detect the rotational speed of the corresponding wheel body W. The left and right rear wheel speed sensors 12B and the acceleration sensor 12D function as a vehicle body speed acquisition unit that acquires information related to the vehicle body speed.

[0040] The drive source rotational speed sensor 12C detects the rotational speed of the main drive gear of the differential device of the drive source 5. The vehicle sensors 12 further include a steering angle sensor 12E that detects the steering angle of the steering wheel 11A, an accelerator pedal sensor 12F that detects the operation amount of the accelerator pedal 11B, a brake pedal sensor 12G that detects the operation amount of the brake pedal 11C, and an engine rotational speed sensor 12H that detects the rotational speed of the internal combustion engine 5A. In addition, the vehicle sensors 12 include a vertical acceleration sensor 12K that detects the vertical acceleration of the vehicle body 2. It is preferable that a vertical acceleration sensor 12K be provided for each wheel 3. The vertical acceleration sensor 12K can be provided on a suspension arm (not shown in the drawings) that supports each wheel 3. The acceleration sensor 12D and the vertical acceleration sensor 12K can be configured as a common 3-axis or 6-axis acceleration sensor. The output torque of the internal combustion engine 5A is estimated by the control device 14, as described later.

[0041] The control device 14 is an electronic control unit (ECU) composed of a CPU, a ROM, a RAM, and the like. The control device 14 performs various vehicle controls by performing calculation processing in accordance with the program of the CPU. The control device 14 includes an estimation unit 14A, a control unit 14B, and a torque acquisition unit 14C (torque acquisition mechanism). The estimation unit 14A estimates the tire torsional stiffness (i.e., the torsional stiffness of the tire T) and the road surface friction coefficient (i.e., the frictional characteristics between the tire T and the road surface) based on at least the rotational speed of the drive source 5, the rotational speed of the wheel body W, the vehicle body speed, and the torque applied to the wheel body W, and estimates the adhesion limit driving force corresponding to the adhesion limit of the tire based on the tire torsional stiffness and the road surface friction coefficient. The control unit 14B controls at least one of the drive source 5 and the brake device 8 so that the tire does not exceed the adhesion limit derived from the tire torsional stiffness and the road surface friction coefficient. In addition, the control unit 14B controls at least one of the drive source 5 and the brake device 8 so that the driving force of the tire is less than or equal to the adhesion limit driving force.

[0042] The torque acquisition unit 14C acquires the output torque of the internal combustion engine 5A. For example, the torque acquisition unit 14C preferably estimates the output torque of the internal combustion engine 5A based on the intake air amount and the negative pressure in the intake manifold. In addition, when the drive source 5 is an electric motor, the output torque of the electric motor is preferably estimated based on the phase current supplied to the electric motor. Note that in another embodiment, a torque sensor for detecting the output torque can be provided on the internal combustion engine 5A or the electric motor. In addition, the torque acquisition unit 14C estimates the brake torque applied to the wheel body W based on the control amount of the brake device 8 by the control unit 14B.

[0043] Hereinafter, the estimation method of the estimation unit 14A for the tire torsional stiffness, the road surface friction coefficient, and the adhesion limit driving force will be described. The estimation unit 14A performs the estimation by executing a program constructed based on the theory shown below.

[0044] Since the wheel body W is made of a metal such as aluminum or steel, and its stiffness is sufficiently high compared to the tire T made of rubber. When the driving torque is applied to the wheel body W, the sidewall portion and the tread portion of the tire T are elastically deformed. Therefore, it is assumed here that the wheel body W and the tread of the tire T are represented by rigid body masses, and a spring force acts in a direction that suppresses the torsion between them. At the contact portion between the tire T and the road surface, the tire T is deformed due to the mass of the vehicle 1, so that the tire T is in contact with the road surface (ground contact surface) with a certain constant width (ground contact width). At the ground contact surface, a friction force F acts between the tire and the road surface, and this friction force F is expressed by the following equation.

[0045] F = μN (1)

[0046] Here, μ is the road surface friction coefficient, i.e., the friction coefficient between the tire T and the road surface; N is the wheel load, i.e., the ground contact load of the tire T. The change in the road surface friction coefficient μ depends on the air pressure and the degree of aging of the tire T, the road surface, the weather, the climate, and the like. The magnitude of the friction force F needs to match the magnitude of the driving force, which is the force that makes the vehicle 1 travel against the travel resistance (acceleration, deceleration, or constant speed travel).

[0047] At the moment the driving torque is applied to the wheel body W, the torque has not yet been transmitted to the tire T, and the tire T has not yet rolled. At this time, the tire T undergoes elastic deformation, and a torsional angle is generated between the wheel body W and the tire T. In this state, the tire T is in a state of stationary torsion, where a torsional angle proportional to the driving torque of the wheel body W is generated. When the torsional angle is generated, the torque is transmitted to the tire T as its reaction force, and the tire T begins to roll. As the tire T rolls, one element of the tire T that undergoes elastic deformation leaves the ground contact surface, and the elastic strain is released. At this time, the reaction force used to transmit the driving torque of the wheel body W becomes insufficient, the amount of which is corresponding to the magnitude of the released elastic strain, so the rolling of the tire T will temporarily stop. However, instead of one element of the tire T leaving the ground contact surface, a new element of the tire T contacts the road surface and generates elastic strain, thereby recovering the lost reaction force, and the tire T rolls again. The boundary conditions for individual elements are not unique for each element and move with the movement of the elements; such a situation is specifically called a moving boundary. As the actual tire T continues to roll, the above phenomena occur successively. Therefore, the roll angle of tire T decreases at a constant rate relative to the rotation angle of wheel W. Since the rotation angle of wheel W per unit time is proportional to its rotational speed (angular velocity), the roll angle of tire T per unit time also decreases proportionally to the rotational speed of wheel W, resulting in a constant rotational transmission loss. This phenomenon is called elastic slip, because apparent slip occurs between wheel W and the road surface due to elastic deformation. Since the amount of elastic slip is generated at a constant rate relative to the rotational speed of wheel W, the rotational speed loss Δω caused by slip is related to the rotational speed ω of wheel W. 轮 The ratio between them is considered to be the slip velocity ratio S r .

[0048] Sr = Δω / ω 轮 (2)

[0049] The elastic slip characteristics of tire T are as follows Figure 2 As shown. Because the friction between tire T and the road surface has a limit, as the driving torque of wheel W increases, the contact surface between tire T and the road surface begins to slip. This is called slip-slip, to distinguish it from elastic slip. Therefore, as the driving torque of wheel W increases, the elastic slip state transitions to the slip-slip state. The boundary between the elastic slip state and the slip-slip state is called the elastic slip limit or adhesion limit, and the driving force (torque) corresponding to the adhesion limit is called the adhesion limit driving force (torque).

[0050] In the elastic slip state, when a torsional angle occurs between the wheel body W and the tire T due to elastic deformation... Furthermore, when the ground contact surface moves by the ground contact length, the strain energy generated due to elastic deformation... is stored in the ground contact surface before rolling, and this strain energy is released by rolling. This strain energy does not have any effect on the travel of the vehicle 1, and thus it can be considered that, in this state, the driving energy from the wheel body W is dissipated by the cycle of accumulation and release of the strain. In understanding that this energy dissipation occurs due to apparent slip (elastic slip), the following equation can be obtained using the frictional force F acting on the ground contact surface.

[0051]

[0052] That is, the energy dissipation can be replaced by the virtual work calculated from the frictional force and the apparent slip, as shown in Equation 3, where k T is the torsional stiffness of the tire T [Nm / rad], R is the dynamic radius of the tire T [m], and T f is the frictional torque generated at the ground contact surface [Nm]. If the tire T rolls while the torsion angle is included in the rotation angle of the wheel body W is included in the rotation angle of the wheel body W Then, based on the geometric relationship, the slip velocity ratio S r is represented by the following Equation 4.

[0053]

[0054] According to Equations 2 and 4, is represented by the following Equation 5.

[0055]

[0056] By substituting this into Equation 3, the following Equation 6 can be obtained.

[0057]

[0058] As expressed in Equation 6, the frictional torque T f is represented by the viscous resistance, which is proportional to the slip (loss of rotational speed) Δω that occurs between the wheel body W and the road surface. Here, c T is the frictional damping coefficient between the tire and the road surface [Nm / (rad / s)], which corresponds to the viscous coefficient, and is proportional to the tire torsional stiffness k T .

[0059] The dynamic model from the driving source 5 to the contact surface can be represented as shown in Figure 3 Based on this model, the state equation is represented by the following Equation 7. This Equation 7 is derived for one of the left and right front wheels of the vehicle 1, which constitutes an FF vehicle in which an internal combustion engine is installed in the front portion thereof to drive the front wheels via a transmission.

[0060]

[0061] Here, θ DN is the rotational angle fluctuation [rad] of the main drive gear of the differential DN (output shaft of the drive source 5), θ W is the rotational angle fluctuation [rad] of the wheel body, θ T is the rotational angle fluctuation [rad] of the tire, I w is the moment of inertia [kgm 2 ] of the wheel body, I T is the moment of inertia [kgm 2 ] of the tire, and k D is the torsional stiffness [Nm / rad] of the power transmission member 6 (drive shaft).

[0062] By making formula 7 dimensionless using the following formula 8, a state variable (vector) represented by formula 9 can be represented by formula 10.

[0063]

[0064] The frequency response of the rotational fluctuation of the wheel body W to the rotational fluctuation of the differential DN obtained from formula 10 can be shown as in Figure 4A . Figure 4A Shown is, with respect to frequency, the amplification ratio (amplitude ratio m) of the rotational fluctuation amplitude of the wheel body W to the rotational fluctuation amplitude of the differential DN and the phase delay (phase delay Ψ1) of the rotational fluctuation of the wheel body W to the rotational fluctuation of the differential DN.

[0065] According to formula 6, as the frictional damping coefficient c T value decreases, the slip state approaches the sliding slip state. In Figure 4A , (a) represents the response in the elastic slip state, and (c) represents the response in the sliding slip state. Also, (b) indicates the boundary (adhesion limit) between the two slip states. When Figure 4A the graphs (a) and (c) of the amplitude ratio are compared with each other, it can be seen that, when entering the sliding slip state, a new peak appears on the low frequency side, and the peak on the high frequency side moves toward the higher frequency side. The vibration mode corresponding to the peak on the high frequency side will be referred to as the elastic slip mode, and the vibration mode corresponding to the peak on the low frequency side will be referred to as the sliding slip mode.

[0066] Figure 4B The existence ranges of the elastic slip mode and the sliding slip mode with respect to frequency and the frictional damping coefficient c T are shown in Figure 4B . In Figure 4B , the existence ranges of the elastic slip mode and the sliding slip mode are indicated by solid lines.

[0067] In the elastic slip mode, since the driving force is transmitted to the road surface due to the elastic deformation of the tire T, the elastic force generated by the tire torsional stiffness k T acts as a reaction force on the wheel body W. Therefore, the wheel body W receives the combination of the elastic force generated by the driving shaft stiffness k D and the tire torsional stiffness k T , and thus the wheel body W vibrates. In Figure 4A and Figure 4B , the elastic slip mode is seen on the high frequency side. As shown in Figure 4B , as the frictional damping coefficient c T decreases, the elastic slip mode shifts to the higher frequency side, i.e., the slip state approaches the sliding slip state from the elastic slip state. This corresponds to the phenomenon in the graph of the amplitude ratio shown in Figure 4A , i.e., the peak on the high frequency side moves to the higher frequency side when the state transitions to the sliding slip state.

[0068] In the sliding slip mode, since the tire T and the road surface dynamically slip, the elastic force generated by the tire torsional stiffness k T is released due to the slip, and the reaction force acting on the wheel body W also disappears. Therefore, the wheel body W and the tire T become one, and only receive the elastic force generated by the driving shaft stiffness k D , and thus they vibrate in phase. In Figure 4A and Figure 4B , the sliding slip mode is seen on the low frequency side. As shown in Figure 4B , when the frictional damping coefficient c T is smaller than the constant (i.e., when the slip state becomes the sliding slip state), the sliding slip mode appears, which does not appear in the elastic slip state. This corresponds to the phenomenon in the graph of the amplitude ratio shown in Figure 4A , i.e., a new peak appears on the low frequency side when the slip state becomes the sliding slip state.

[0069] As described above, when the slip state transitions from the elastic slip state to the sliding slip state, the sliding slip mode appears. Therefore, it is possible to determine the adhesion limit by monitoring the appearance of the sliding slip mode. However, from Figure 4Athe peak on the low frequency side cannot be confirmed at the limit of adhesion. When the peak on the low frequency side is clearly confirmed as in the road surface determining device described in JP 2018-155696 A, only the state in which the slip slide has been progressing can be determined (in JP 2018-155696 A, since the observation position of the vibration is different, it looks like the pattern disappears with the slip slide, but they are the same phenomenon). That is, by simply observing the vibration waveform, the appearance of the slip slide pattern cannot be strictly determined. First, the limit of adhesion cannot be predicted from the elastic slip state. Therefore, attention is focused on the dimensionless quantity ζ2 which represents the damping state of the system. As shown in Equation 8, the dimensionless quantity ζ2 is a dimensionless quantity composed of the frictional damping coefficient c T and the tire torsional stiffness k T and uniquely represents the damping state of the system without being affected by various factor variations. If the current dimensionless quantity ζ2 can be estimated, by comparing it with a threshold value corresponding to the limit of adhesion, the occurrence of the slip slide can be strictly determined. In addition, since the deviation between the dimensionless quantity ζ2 and the aforementioned threshold value can serve as a basis for determining the margin before the occurrence of the slip slide, it is useful to know the dimensionless quantity ζ2. Hereinafter, first, the method of acquiring the dimensionless quantity ζ2 will be described.

[0070] Torque fluctuations often occur in the internal combustion engine that is the drive source 5 of the vehicle 1, and such torque fluctuations are also transmitted from the differential device DN to the tires. As a cause of the torque fluctuations, in the case of the internal combustion engine, there is a fluctuation in the pressure inside the cylinder, and in the case of the electric motor, there is a cogging torque due to the number of poles. In the differential device DN, a rotational fluctuation due to the input torque fluctuation occurs at the same time. Here, the rotational fluctuation of the differential device DN is expressed by the following Equation 11.

[0071]

[0072] Equation 11 can be considered as a forced excitation under a boundary condition. A1 is the rotational fluctuation amplitude of the differential device DN [m], Ω is the angular frequency of the excitation force (torque fluctuation of the internal combustion engine E) [rad / s], and t is time [s]. In this forced excitation state, the state equation represented by Equation 10 becomes the following equation.

[0073]

[0074] In Equation 12, B represents an external force (excitation input), and a natural vibration mode (hereinafter referred to as a natural mode) that the original system has is determined by the Jacobian matrix A. The parameters that determine the Jacobian matrix A are p, ω1, ω2, and ζ2, of which p and ω1 are design specifications (known values). Therefore, once the dimensionless quantity ω2 and the dimensionless quantity ζ2 corresponding to the slip identification quantity are known, the natural mode is known. In Equation 7, there are two dominant equations, and there are two unknown dimensionless quantities (i.e., ω2 and ζ2) at the same time, and therefore, it should be possible to uniquely determine ω2, ζ2. Note that because the dimensionless quantity ω2 is obtained from the tire torsional stiffness k T , and the dimensionless quantity ζ2 is obtained from the frictional damping coefficient c T and the tire torsional stiffness k T , it is possible to determine the dimensionless quantities ω2, ζ2, which correspond to determining the frictional damping coefficient c T and the tire torsional stiffness k T .

[0075] It is assumed that the periodic solution of Equation 12 is expressed as follows.

[0076]

[0077] By substituting the periodic solution of Equation 13 into Equation 12 and performing coefficient determination based on the Galerkin method, the following relational expression is obtained.

[0078]

[0079] Here, m is an amplification ratio (amplitude ratio) of the rotational fluctuation amplitude of the wheel body to the rotational fluctuation amplitude of the differential device DN, and Ψ1 is a phase delay of the rotational fluctuation of the wheel body with respect to the rotational fluctuation of the differential device DN. Therefore, by measuring the rotational fluctuation of the differential device DN and the rotational fluctuation of the wheel body, it is possible to obtain the dimensionless quantities ω2, ζ2 according to Equation 14.

[0080] Next, if the current dimensionless quantities ω2, ζ2 have been obtained according to Equation 14, a method of acquiring the relationship between the dimensionless quantity ζ2 and the natural mode will be described. The dimensionless quantity ω2 reflects the change in the tire torsional stiffness k T , but since there is no significant change under the same conditions, the relationship between the dimensionless quantity ζ2 and the natural mode will be described under the assumption that the tire torsional stiffness k T is constant. Therefore, the dimensionless quantity ζ2 uniquely corresponds to the frictional damping coefficient c T . The behavior of the natural mode can be described by obtaining the eigenvalue λ of the Jacobian matrix A. Figure 5 The behavior (root locus) of the eigenvalue λ corresponding to the above-described slip slip mode is shown. Figure 5 (a) to (b) of FIG. 9 correspond toFigure 4A (a) to (c). Note that if the tire torsional stiffness k T If the frequency of the vibration mode changes, then the frequency of the vibration mode will also change. Figure 5 The scale of the root locus will also change, but the main characteristics described below remain unchanged. Furthermore, in this case, the dimensionless quantity ω² is already known (and therefore the tire torsional stiffness k is already known). T There is no control problem.

[0081] exist Figure 5 In the diagram, the horizontal axis represents the real axis, the vertical axis represents the imaginary axis, and the imaginary part represents the vibrational solution. Under the elastic slip state (see...), Figure 5 (a) has a pair of roots on the real axis, which means there is no vibrational solution. That is, no vibration corresponding to the slip-slip mode is generated. On the other hand, when the slip state becomes a slip-slip state (see...) Figure 5 (c)), the root has an imaginary part, which indicates that vibration has occurred. That is, it can be understood that when the dimensionless quantity ζ² becomes less than ζ... C Time (see) Figure 5 (c) indicates that a sliding mode occurs. Therefore, based on the dimensionless quantity ζ C The value of can be used to determine the slip state as follows:

[0082] When the dimensionless quantity ζ2>ζ C At this time, the slip state is the elastic slip state;

[0083] When the dimensionless quantity ζ2=ζ C At that time, the slip state is the adhesion limit; and

[0084] When the dimensionless quantity ζ2 < ζ C At that time, the slip state is a sliding slip state.

[0085] Where ζ C These are values ​​that have been changed according to design specifications. Figure 5 In the middle, ζ2 and the friction damping coefficient c T The value is in ζ C This is illustrated exemplarily when the value is 0.86. Once the dimensionless quantity ζ is known... C Then, the friction damping coefficient c when the slip state becomes the elastic slip limit can be obtained according to Formula 8. Tc .

[0086] However, in order to estimate the dimensionless quantities ω2, ζ2 based on the above theory, specific vibrations (for example, torque fluctuations of the internal combustion engine as the drive source 5) are necessary. That is, there is a problem in that the tire slip cannot be determined when no vibrations are generated or the vibrations are very small compared to sensor noise. Therefore, attention is focused on the fact that the natural modes that the original system has are determined by the Jacobian matrix A described above. That is, by identifying a dynamic model from the drive source 5 to the contact surface between the tire T and the road surface (as shown in Figure 3 , and evaluating the Jacobian matrix A of the identified model, the dimensionless quantities ω2, ζ2 (friction damping coefficient c T and tire torsional stiffness k T ) can be estimated even in the case where no vibrations are generated.

[0087] A method in which the estimation unit 14A identifies a dynamic model and the tire model shown in Figure 3 will be described below. In this model identification, mainly the tire torsional stiffness k T and the road surface friction coefficient λ μx are estimated as model parameters. For example, the estimation unit 14A preferably estimates the tire torsional stiffness k T and the road surface friction coefficient λ μx by using a known Kalman filter or observer. In the present embodiment, one example of an estimation method using a Kalman filter will be described. A state equation based on a dynamic model from the drive source 5 to the contact surface (as shown in Figure 3 can be expressed by the following equation 15. A left front wheel of the vehicle 1 configured as an FF vehicle in which an internal combustion engine is installed in the front of the vehicle 1 to drive the front wheels via a transmission will be exemplarily described below. By appropriately changing or adjusting the torsional stiffness (drive shaft stiffness k D ) of the power transmission member 6 (drive shaft), a load movement formula, and the like, estimation can be similarly performed with respect to other wheels.

[0088]

[0089] Here, θ DN is a rotational angle disturbance [rad] of a main drive gear of the differential DN (an output shaft of the drive source 5), θ W is a rotational angle disturbance [rad] of a wheel body, θ T is a rotational angle disturbance [rad] of a tire, k T is a tire torsional stiffness [Nm / rad], λ μx is a road surface friction coefficient in the front-rear direction [-] (a friction coefficient between the tire and the road surface), Fz is a wheel load [N], a DN is a torque fluctuation amplitude [Nm], is a phase of the torque fluctuation [rad], and Vx is the front-rear ground speed [m / s] of the vehicle center of gravity, I DN is the moment of inertia [kgm 2 ] of the main drive gear of the differential device DN (output shaft of the drive source 5), k D is the torsional rigidity [Nm / rad] of the power transmission member 6 (drive shaft), I w is the moment of inertia [kgm 2 ] of the wheel body, I T is the moment of inertia [kgm 2 ] of the tire, R e is the dynamic radius [m] of the tire, F x is the driving force [N], V cxfl is the ground speed [m / s] in the longitudinal direction of the wheel (left front wheel), a ^ f ( ^ denotes the hat operator) is the tire side slip angle [deg] of the front wheel. γ^ fl is the camber angle [deg] of the wheel (left front wheel), k f is the front roll rigidity [Nm / rad], k r is the rear roll rigidity [Nm / rad], h is the center of gravity height [m], d f is the front tire width [m], m is the vehicle weight [kg], a y is the lateral acceleration [m / s 2 ], a x is the front-rear acceleration [m / s 2 ], T DNo is the average torque [Nm] of the main drive gear of the differential device DN (output shaft of the drive source 5), T brk is the braking torque [Nm] applied to the wheel body by the brake device 8, N e is the engine rotational speed [rpm], and v is a coefficient corresponding to the type of the internal combustion engine, and is 2 in the case of an in-line 4-cylinder 4-stroke engine. The average torque T DNo of the differential device DN is obtained from the estimated output torque of the drive source 5 and the reduction ratio of the transmission. The estimated output torque of the drive source 5 can be generally estimated from the amount of air flowing in or the negative pressure in the intake manifold when the drive source 5 is an internal combustion engine, and can be estimated from the phase current when the drive source 5 is an electric motor. The braking torque T brk can be generally estimated on the basis of the hydraulic pressure from the hydraulic supply device 8A. In addition, the superscript “^ ^ (hat)” indicates that the value is an estimated value. The tire side slip angle a^ f and the camber angle γ^ fl.

[0090] Generally, when the drive source 5 is an internal combustion engine, a torque fluctuation occurs based on its ignition cycle, and thus, it is necessary to take into account the periodic torque fluctuation as the output torque of the transmission. Therefore, it is assumed that the torque fluctuation is transmitted to the wheels via the drive shaft. Since the torque fluctuation occurs every ignition cycle in the internal combustion engine, the frequency of the torque fluctuation is proportional to the rotational speed of the internal combustion engine. That is, the angular frequency of the torque fluctuation (“·” represents a point) is expressed by the following equation 16. The superscript point (“·”) is a derivative operator, and represents the derivative of the phase angle (i.e., the angular frequency). Note that, when the drive source 5 is an internal combustion engine, in order to further improve the estimation accuracy, the periodic torque fluctuation is taken into account in the model, but the following Kalman filter works even in the case where such a torque fluctuation or a specific vibration is not generated.

[0091]

[0092] The wheel load F z in equation 15 is a vertical load acting on the left and right front wheels and the left and right rear wheels, and can be expressed by the following equation in consideration of the load movement due to acceleration and deceleration of the vehicle 1 and turning. Equation 15 describes the left front wheel load F zfl .

[0093]

[0094] R e (F z ) is a tire dynamic radius and depends on the wheel load F z . F x is a frictional force (driving force) in the tire longitudinal direction occurring between the tire and the road surface. F x is represented by a function with k T , λ μx , F z , θ W , V cxfl , α ^ , γ ^ as parameters, using a tire model based on the Magic formula (Pacejka), which will be described in detail below.

[0095] If the wheel load and / or the slip between the tire and the road surface change under the usage conditions of the tire, they will affect the driving force generated in the tire. These factors of influence are implemented as correction coefficients in the tire model based on the following equations 18 to 21.

[0096]

[0097] Here, F zo is a standard value of the wheel load assumed for the tire used [N]. In this embodiment, the left front wheel load F zf corresponds to F z .

[0098]

[0099] Here, a is a tire side slip angle [deg], V cx is a ground speed in the wheel longitudinal direction [m / s], and V cy is a lateral speed [m / s]. In this embodiment, the tire side slip angle (estimated value) a^ f of the front wheel is substituted into a. In addition, the ground speed V cxfl in the longitudinal direction of the left front wheel (to be described later (Equation 38)) is substituted into V cx .

[0100] γ * = sin γ (20)

[0101] Here, γ is a camber angle of the wheel [deg], and the camber angle (estimated value) γ^ fl of the left front wheel is substituted into it.

[0102]

[0103] Here, κ is a slip ratio, which is an index representing slip in the direction of the driving force.

[0104] When the vehicle 1 is in a straight running state, that is, when the side slip angle is 0 (pure slip), the tire torsional stiffness kT, the road surface friction coefficient λ μx , and the wheel load F z determine the main friction characteristics, and the driving force F xo is expressed by the following equation.

[0105] F xo = D x sin {C x tan -1 [B x κ x -E x (B x κ x -tan- 1 B x κ x )]} (22)

[0106] κ x = κ (23)

[0107] C x =p Cx1 (>0) (24)

[0108] D x =μ x F z (>0) (25)

[0109] μ x =(p Dx1 +p Dx2 df z (1-p) Dx3 γ 2 )λ μx (26)

[0110] E x =(p Ex1 +p Ex2 df z +p Ex3 df Z 2 )[1-p Ex4 sgn(μ x (27)

[0111]

[0112] B x =K xκ / (C x D x +ε x (29)

[0113] Here, p cx1 p Dx1 p Dx2 p Dx3 p Ex1 p Ex2 p Ex3 p Ex4 and ε x It is a constant. ε x It is a sufficiently small value, set to avoid division by 0, and has no physical meaning.

[0114] When a sideslip angle (combined slip) occurs, the sideslip angle contributes to frictional saturation; therefore, the driving force F x It is expressed by the following formulas 30 to 37.

[0115] F x =G xα F xo (30)

[0116] G xα =cos{G xαtan -1 [B xα α S -E xα (B xα α S -tan -1 B xα α S )]} / G xαo (>0) (31)

[0117] G xαo =cos{C xα tan -1 [B xα S Hxα -E xα (B xα S Hxα -tan -1 B xα S Hxα )]} (32)

[0118] α S =α * +S Hxα (33)

[0119] B xα =(r Bx1 +r Bx3 γ *2 )cos(tan -1 r Bx2 κ)·λ xα (>0) (34)

[0120] C xα =r Cx1 (35)

[0121] E xα =r Ex1 +r Ex2 df z (≤1) (36)

[0122] S Hxα =r Hx1 (37)

[0123] Here, r Bx1 , r Bx2 , r Bx3 , λ xa , r Cx1 , r Ex1 and r Ex2 are constants.

[0124] Wheel speed (rotational speed of the wheel body or θ WThe derivative of the equation and the ground velocity V in the longitudinal direction of the drive wheel (left front wheel). cxfl This is the variable used to obtain the slip ratio κ. The ground velocity V in the longitudinal direction of the drive wheel (left front wheel) is... cxf1 The following formula is derived from the ground velocity V at the vehicle's center of gravity in the longitudinal direction. x (Vehicle speed), front wheel steering angle δ f [degrees] and yaw rate γ [degrees / s].

[0125]

[0126] Here, l is the wheelbase [m].

[0127] Vehicle speed V x It is the forward and backward acceleration a x * is obtained by integrating the equation. The acceleration a is obtained by integrating the equation. x * can simply be equal to a x However, if a six-axis inertial sensor or similar device can be used, it is preferable to set the value after tilt correction (planar projection).

[0128] For the state equation of Equation 15, the observation equation is expressed by the following formula.

[0129]

[0130] λ μ λ is a pseudo-observable of the road surface friction coefficient and is used to constrain the estimated value of the road surface friction coefficient, such that the estimated value is greater than or equal to 0 and less than or equal to 1. Preferably, λ μ It is set to a value greater than or equal to 0 and less than or equal to 1. For example, preferably, λ μ The value is set to 0 when the estimated coefficient of road friction is less than 0, and to 1 when the estimated coefficient of road friction is greater than 1. V rr This is the wheel speed of the rear wheel, which is the non-driving wheel. When it is assumed that the rear wheel does not slip, the rear wheel speed V is... rr This becomes equal to the vehicle's speed VX. Based on the wheel speed V of the left rear wheel. Wrl and the wheel speed V of the right rear wheel Wrr The observed value is represented by the following formula 40.

[0131]

[0132] Formula 15 is the state equation representing continuous time, and the estimation unit 14A at the observed value y = t The calculation is performed within each sampling interval of (y1, y2, y3, y4). The discrete-time representation of Equations 15 and 39 can be expressed by the following formula.

[0133]

[0134] Here, k is the discrete time of each sampling interval, y(k) is a four-dimensional time series, x(k) is a twelve-dimensional state vector, and u(k) is a twelve-dimensional system input vector. x(k) and u(k) are expressed by the following formulas.

[0135]

[0136] In addition, v(k) is a twelve-dimensional system noise vector with a mean vector of 0 and a covariance matrix expressed by Q, and w(k) is a four-dimensional observation noise vector with a mean vector of 0 and a covariance matrix expressed by R.

[0137] Assume that v(k) and w(k) are independent Gaussian white noises, which are expressed by Equation 43.

[0138]

[0139] f(x,u) is based on the discrete-time integral (forward Euler method) of Equation 15 and is expressed as a twelve-dimensional function by Equation 44 below. h(x) is a four-dimensional function, represented by Equation 45 below. Δt is the discrete time interval (sampling interval).

[0140]

[0141]

[0142] The following section describes how to calculate the state estimate x using an extended Kalman filter (EKF). ^ (k)( ^ The program representing the hat operator, EKF is a nonlinear Kalman filter.

[0143] State estimate x ^ The initial value of (k) is x ^ (0)( ^ The hat operator is assumed to be a Gaussian probability vector following N(x0,Σ0) and is represented by the following formula 46.

[0144]

[0145] For k = 1, 2, ..., the prior state estimate x ^- (k) is represented by the following formula 47.

[0146]

[0147] As long as Equations 48 and 49 are derived through linear approximation, the prior error covariance matrix as shown in Equation 50 can be obtained.

[0148]

[0149] P(k)=A(k-1)P(k-1)AT(k-1)+·Q(50)

[0150] Therefore, the Kalman gain matrix G(k) can be expressed by the following formula 51.

[0151] G(k)=P(k)C(k)( ( k )P -(k)C(k)+R(k))- 1 (51)

[0152] State estimate x ^ (k)( ^ The hat operator is represented by the following formula 52.

[0153]

[0154] The posterior error covariance matrix P(k) is expressed by the following formula 53.

[0155] P(k)=(IG(k)CT(k))P(k)(53)

[0156] Based on the above, obtain the state estimate x. ^ (k)(Formula 52), based on this, the estimated value of tire torsional stiffness k ^ T As x ^ The seventh element of (k) is obtained, and the estimated value of the road surface friction coefficient λ is obtained. ^ μx As x ^ The eighth element of (k) is obtained. In the method using the Kalman filter, it is not necessary to explicitly obtain the amplitude and phase through frequency analysis, and it is easy to apply even in the instantaneous state.

[0157]

[0158] Estimated vehicle speed V ^ x It is obtained as a weighted average of the first speed obtained by the front wheel speed sensor 12A and the second speed obtained by the integration of the front and rear accelerations obtained by the acceleration sensor 12D. In addition, the weights are set based on the Kalman gain to maximize the likelihood of the state variable x(k).

[0159] By reflecting the state estimate x in the tire model ^ The tire parameters in (k), and the friction damping coefficient c between the tire and the road surface. T([Nm(rad / s)]) can be represented by the following equation 55.

[0160]

[0161] Here, p x , B x , C x , D x , E x is expressed by the following equation 56.

[0162]

[0163] Here, p cx1 , p Dx1 , p Dx2 , p Dx3 , p Ex1 , p Ex2 , p Ex3 , and p Ex4 are constants.

[0164] On the other hand, as described above, once the dimensionless quantity ζ C is known, the frictional damping coefficient c Tc can be obtained according to equation 8 when the slip state becomes the adhesion limit. x The slip ratio κ c in equation 55 when the slip state becomes the adhesion limit is expressed as κ T The frictional damping coefficient c Tc represented by equation 55 is equal to the frictional damping coefficient c C when the slip state becomes the adhesion limit, which is obtained from the dimensionless quantity ζ c , and thus the following equation 57 is obtained.

[0165]

[0166] According to this equation, the adhesion limit slip ratio κ T , which is the slip ratio when the slip state becomes the adhesion limit, can be obtained. Preferably, the adhesion limit slip ratio is calculated off-line in advance for mapping. For example, it is preferable to set the adhesion limit slip ratio using the graph shown in FIG. 9 based on the tire torsional stiffness k μx and the road surface friction coefficient λ Figure 6 .

[0167] Once the adhesion limit slip ratio is determined, the adhesion limit driving force F xc , which is the tire driving force F x corresponding to the adhesion limit, is determined based on the tire model represented by equations 22 to 37. The adhesion limit driving force F xcwhich can be converted into an attachment limit torque T xc , which is a driving torque of the tire corresponding to the attachment limit x . The estimation unit 14A can set the attachment limit driving force F T based on the tire torsional stiffness k μx and the road surface friction coefficient λ xc by using a map representing a relationship between the attachment limit driving force F xc (or the attachment limit torque T T ) and the tire torsional stiffness k μx and the road surface friction coefficient λ xc .

[0168] Next, with reference to Figure 7 , a control program executed by the control device 14 will be described. The control device 14 executes the program shown in FIG. 10 at every prescribed time interval. Figure 7

[0169] First, the estimation unit 14A determines whether the road surface on which the vehicle 1 travels is rough (S1). In the present embodiment, the estimation unit 14A acquires the vertical acceleration of the vehicle body 2 based on a signal from a vertical acceleration sensor provided on the vehicle body 2, and determines whether the absolute value of the vertical acceleration is less than or equal to a prescribed determination value. When the absolute value of the vertical acceleration is greater than the determination value, the estimation unit 14A determines that the road surface is rough. In another embodiment, the estimation unit 14A can determine whether the road surface is rough based on a road surface image captured by a camera, instead of making the determination of the rough road surface based on the vertical acceleration. In addition, the estimation unit 14A can detect the vertical travel of the wheel 3 with respect to the vehicle body 2 with a travel sensor, and determine whether the road surface is rough based on the amount of change and / or the rate of change of the vertical travel.

[0170] When the road surface is rough (YES in determination in S1), the estimation unit 14A reduces the influence of the rear wheel speed, which is the non-driving wheel speed, in obtaining the vehicle body speed Vx (the front-rear ground speed of the vehicle center of gravity). As described above, the vehicle body speed V x is acquired as a weighted average of the rear wheel speed acquired by the rear wheel speed sensor 12B and the integral of the front-rear acceleration acquired by the acceleration sensor 12D. When the rear wheel 3R does not slip with respect to the road surface, the rear wheel speed matches the vehicle body speed. However, when the road surface is rough, rotational fluctuations occur in the non-driving wheel due to the irregularity of the road surface, and such rotational fluctuations become observation noise in obtaining the vehicle body speed due to the difference in timing of passing over the irregular road surface between the driving wheel and the non-driving wheel. In order to reduce the error caused by such observation noise, the estimation unit 14A reduces the influence of the rear wheel speed on the vehicle body speed V x ​of the influence of the rough road surface. In the present embodiment, in step S2, the estimation unit 14A changes the parameter of the fourth row of the fourth column in the covariance matrix R for the observation error to a value corresponding to the rough road surface. Thus, the vehicle body speed V x of the correction amount is reduced.

[0171] When the determination result in step S1 is NO or after the process of step S2 is executed, the estimation unit 14A acquires the state estimate value x ^ (k) including the tire parameters by using the above-described Kalman filter (S3). The estimation unit 14A inputs the rotation speed of the differential main gear acquired by the drive source rotation speed sensor 12C, the rotation speed of the wheel body W acquired by the front wheel speed sensor 12A, the front-rear acceleration acquired by the acceleration sensor 12D, the engine rotation speed acquired by the engine rotation speed sensor 12H, the engine torque acquired by the torque acquisition unit 14C, and the brake torque of the wheel body W to the Kalman filter, and acquires the state estimate value x ^ (k) as the output of the Kalman filter.

[0172] Subsequently, based on the tire torsional stiffness k T and the road surface friction coefficient λ μx included in the state estimate value x ^ (k), the estimation unit 14A sets the adhesion limit slip ratio κ C , which is the slip ratio κ x corresponding to the adhesion limit (S4). As described above, the estimation unit 14A preferably sets the adhesion limit slip ratio κ C by using the graph shown in FIG. 9 based on the tire torsional stiffness k T and the road surface friction coefficient λ μx . Figure 6

[0173] Then, based on the adhesion limit slip ratio κ C , the estimation unit 14A sets the adhesion limit driving force F xc , which is the tire driving force F x corresponding to the adhesion limit (S5). As described above, the estimation unit 14A preferably sets the adhesion limit driving force F xc based on the adhesion limit slip ratio κ C by using the tire model represented by the formulas 22 to 37. The estimation unit 14A can set the adhesion limit driving force F xc based on a graph that describes the relationship between the adhesion limit slip ratio κ C and the adhesion limit driving force F xc . In another embodiment, the estimation unit 14A can set the adhesion limit driving force F T based on the tire torsional stiffness k xc .and the road surface friction coefficient λ μx Instead of steps S4 and S5, the attachment limit driving force F xc .

[0174] Subsequently, the control unit 14B determines whether the required driving force F t of the driving source 5 is less than or equal to the attachment limit driving force F xc (S6). Preferably, the control unit 14B acquires the operation amount of the accelerator pedal 1 IB on the basis of the signal from the accelerator pedal sensor 12F and sets the required driving force F t of the driving source 5 on the basis of the operation amount of the accelerator pedal 1 IB. In addition, the control unit 14B can set the required driving force F t on the basis of the rotational speed of the internal combustion engine 5A in addition to the operation amount of the accelerator pedal 1 IB.

[0175] When the required driving force F t is less than or equal to the attachment limit driving force F xc (the determination result in step S6 is YES), the control unit 14B determines whether the slip ratio K x is less than or equal to the attachment limit slip ratio K C (S7).

[0176] When the determination result in step S6 is NO or when the determination result in step S7 is NO, the control unit 14B limits the output of the internal combustion engine 5A or activates the brake device 8 so that the driving force F x of the tire T becomes less than or equal to the attachment limit driving force F xc and the slip ratio K x becomes less than or equal to the attachment limit slip ratio K C (S8). For example, the control unit 14B can make the driving force F x of the tire T less than or equal to the attachment limit driving force F xc or make the slip ratio K x less than or equal to the attachment limit slip ratio K C by limiting the required driving force to be less than or equal to a prescribed threshold value. In addition, the control unit 14B can make the driving force F x of the tire T less than or equal to the attachment limit driving force F xc or make the slip ratio K x less than or equal to the attachment limit slip ratio K C by limiting the output of the internal combustion engine 5A to be less than or equal to a prescribed threshold value.

[0177] In the above-described embodiment, the estimation unit 14A estimates the tire torsional stiffness and the road surface friction coefficient on the basis of the pattern, and estimates the adhesion limit driving force on the basis of the tire torsional stiffness and the road surface friction coefficient that have been estimated. Since the estimation unit 14A estimates the adhesion limit driving force on the basis of a dynamic model from the driving source 5 to the tire contact surface configured by the differential device DN, the wheel body W, and the tire T, the estimation is less likely to be affected by a change and / or deterioration of components. In addition, since the estimation unit 14A estimates the adhesion limit driving force on the basis of the tire torsional stiffness and the road surface friction coefficient, the estimation is less likely to be affected by a use condition such as the number of occupants and the load. x Thus, the tire T can be appropriately kept in the elastic slip state.

[0178] The control unit 14B keeps the driving force F x of the tire T smaller than or equal to the adhesion limit driving force F xc by controlling at least one of the driving source 5 and the brake device 8. Thus, the tire T is kept in the elastic slip state. Thus, as Figure 8 shown in the embodiment, even when the steering angle is large, the tire T is kept in the elastic slip state and can generate a lateral acceleration corresponding to the steering angle. On the other hand, in the comparative example in which no control is performed to keep the driving force F x of the tire T smaller than or equal to the adhesion limit driving force F xc , as the steering angle increases, the slip ratio increases, and the rate of increase of the lateral acceleration with respect to the steering angle decreases. Thus, as Figure 9 shown in the comparative example, in the present embodiment, it is possible to make the travel path of the vehicle closer to the theoretical path determined by the steering angle than in the comparative example.

[0179] Because the control method performed by the control device 14 to keep the driving force F x of the tire T smaller than or equal to the adhesion limit driving force F xc maintains a low slip ratio, it is possible to reduce a loss of energy. Thus, fuel economy can be improved.

[0180] In addition, since the tire T is prevented from becoming the sliding slip state, the behavior of the vehicle caused by steering does not greatly change depending on the road surface, and thus, the occupant can obtain a sense of security regarding the behavior of the vehicle 1.

[0181] Since the estimation unit 14A estimates the tire torsional stiffness on the basis of a dynamic model from the driving source 5 to the tire and a tire model, the travel control system 10 does not need an additional sensor (such as a strain sensor of the tire T or the like) to directly acquire the state of the tire T.

[0182] The control device 14 limits the required driving force F t of the driving source 5 to be less than or equal to the attachment limit driving force F xc . For this reason, the tire can be kept in the elastic slip state. Since the required driving force due to the driver's operation and / or the vehicle control is limited, the vehicle behavior is prevented from changing rapidly. Furthermore, the control is also relatively easy.

[0183] The specific embodiments have been described above, but the present application is not limited to the above-described embodiments and can be modified or changed in various ways. For example, as the control program executed by the control device 14, an acceleration-side program that sets the required driving force F t of the driving source 5 based on the operation amount of the accelerator pedal 1 IB is described, but the control program executed by the control device 14 can be a deceleration-side program that sets the required driving force F t based on the brake pedal. In the deceleration side, the driving force F x and the slip ratio K x are defined by negative values, and therefore, at least one of the driving source 5 and the brake device 8 should be controlled so that the absolute values of the driving force F x and the slip ratio K x do not exceed the attachment limit driving force F xc and the attachment limit slip ratio K C , respectively. In this case, the driving source 5 controls the engine brake (regeneration in the case of an electric motor), and the brake device 8 controls the brake torque, thereby enabling the tire to be kept in the elastic slip state also in the deceleration side.

Claims

1. A travel control system for a vehicle provided with a drive source, a wheel having a wheel body connected to the drive source via a power transmission member and having a tire mounted on the wheel body, and a brake device for braking the wheel, the travel control system comprising: a first rotation sensor configured to acquire a rotational speed of the drive source; a second rotation sensor configured to acquire a rotational speed of the wheel body; a vehicle body speed acquisition unit configured to acquire information related to a vehicle body speed; a torque acquisition unit configured to acquire a torque applied to the wheel body; an estimation unit configured to estimate a tire torsional stiffness that is a stiffness of the tire and a road surface friction coefficient that is a friction characteristic between the tire and a road surface, based on at least the rotational speed of the drive source, the rotational speed of the wheel body, the vehicle body speed, and the torque applied to the wheel body, and estimate an attachment limit driving force corresponding to an attachment limit of the tire, based on the tire torsional stiffness and the road surface friction coefficient, by using a map of a predefined relationship of the attachment limit driving force, the tire torsional stiffness, and the road surface friction coefficient; and a control unit configured to control at least one of the drive source and the brake device so that an absolute value of the driving force of the tire does not exceed the attachment limit driving force, wherein the vehicle body speed acquisition unit includes a non-drive wheel rotation sensor configured to acquire a rotational speed of a non-drive wheel and an acceleration sensor configured to acquire a fore-aft acceleration of the vehicle, the estimation unit estimates the vehicle body speed by using a weighted average of a first speed acquired by the non-drive wheel rotation sensor and a second speed acquired by integrating the fore-aft acceleration acquired by the acceleration sensor, and changes a weight of the weighted average to maximize a likelihood of a state quantity estimated by the estimation unit, and the estimation unit determines whether the road surface is rough by comparing a determination value with an absolute value of a vertical acceleration of a vehicle body, and changes the weight to reduce an influence of the first speed on the vehicle body speed when it is determined that the road surface is rough. the estimation unit predicts a wheel load based on a fore-aft acceleration and a lateral acceleration, and determines the wheel load to maximize a likelihood of a state quantity value of at least one of the wheel load, the tire torsional stiffness, and the road surface friction coefficient.

2. The travel control system according to claim 1, wherein the estimation unit calculates an attachment limit slip ratio corresponding to the attachment limit of the tire to the road surface based on the tire torsional stiffness and the road surface friction coefficient, and 3. The travel control system according to claim 1, wherein the control unit controls at least one of the drive source and the brake device so that an absolute value of a slip ratio of the tire is smaller than or equal to the attachment limit slip ratio. ​

Images