Steering control device

Abstract

To provide a steering control device that can learn a control parameter value when a vehicle speed is greater than zero.SOLUTION: If the magnitude of a steering angle is a minimum learning value θL or more and a maximum learning value θH or less when the steering direction is a right turning direction, PU samples a turning angle estimated from a yaw rate and a vehicle speed and the steering angle θh. PU calculates a right turning linear line L1 based on the sampled points. If the magnitude of a steering angle is the minimum learning value θL or more and the maximum learning value θH or less when the steering direction is a left turning direction, PU samples a turning angle θt estimated from the yaw rate and the vehicle speed and the steering angle θh. PU calculates a left turning linear line L2 based on the sampled points. PU substitutes a difference between intercepts of the right turning linear line L1 and the left turning linear line L2 into a dead zone width.SELECTED DRAWING: Figure 6

Description

[Technical field]

[0001] The present invention relates to a steering control device. [Background technology]

[0002] For example, the following Patent Document 1 describes a device that learns the amount of play, which is the maximum amount by which the steering angle of the steered wheels does not change even when the steering angle changes. This device rotates an electric power steering motor left and right when the key switch is turned on. Then, based on an increase in the current flowing through the motor, this device detects the steering angle at which the steered wheels start to turn. Then, this device substitutes the difference between the detected pair of steering angles for the amount of play. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent Publication No. 2022-68056 Summary of the Invention [Problem to be solved by the invention]

[0004] The amount of play varies depending on the vehicle speed, but the above device cannot calculate the amount of play when the vehicle speed is greater than zero. [Means for solving the problem]

[0005] Means for solving the above problems and their effects will be described below. 1. A steering control device which controls a steering device, the steering device having a steering shaft and steered wheels of a vehicle which turn in conjunction with the rotation of the steering shaft, and configured to execute a sampling process, a steering angle calculation process, and a control constant learning process, the sampling process being a process of synchronously sampling a steering angle and a yaw rate, the steering angle being the rotation angle of the steering shaft, the steering angle calculation process being a process in which the yaw rate and vehicle speed are input and a turning angle of the steered wheels is calculated, the control constant learning process being a process of learning a control constant based on multiple pairs of the steering angle sampled by the sampling process and the steering angle corresponding to the steering angle, and the control constant being a constant which indicates the rotational displacement of the steered wheels relative to the rotational displacement of the steering shaft.

[0006] In the above configuration, the turning angle is calculated according to the yaw rate and the vehicle speed. This makes it possible to obtain the relationship between the turning angle and the steering angle when the vehicle speed is greater than zero, and therefore to learn the value of the control constant.

[0007] 2. A steering control device as described in 1 above, wherein the control constant includes a stiffness coefficient which is a ratio of the steering angle to the steering angle, and the control constant learning process includes a process of learning, as the stiffness coefficient, the slope of a straight line determined from multiple pairs of the steering angle and the steering angle corresponding to the steering angle.

[0008] There is a proportional relationship between the steering angle and the turning angle. With the above configuration, it is possible to learn the stiffness coefficient that indicates this proportional relationship. 3. A steering control device as described in 1 or 2 above, wherein the control constant includes a dead zone width, and the dead zone width is the amount of rotation of the steering shaft after the rotation direction of the steering shaft is switched from either the right turning direction or the left turning direction to the other until the steered wheels are turned, and the control constant learning process includes a right turning straight line calculation process, a left turning straight line calculation process, and a dead zone width calculation process, and the right turning straight line calculation process is a process of calculating a right turning straight line from a plurality of pairs of the sampled steering angle and the steering angle corresponding to the steering angle when the rotation direction is a right turning direction, the left turning straight line calculation process is a process of calculating a left turning straight line from a plurality of pairs of the sampled steering angle and the steering angle corresponding to the steering angle when the rotation direction is a left turning direction, and the dead zone width calculation process is a process of calculating a difference between the intercepts of the right turning straight line and the left turning straight line as the dead zone width.

[0009] In the above configuration, the dead zone width can be calculated by calculating the right turning straight line and the left turning straight line. 4. A steering control device described in any one of 1 to 3 above, wherein the control constant learning process is a process for calculating the control constant for each of the different vehicle speeds based on multiple pairs of the steering angle sampled by the sampling process for each of the different vehicle speeds and the turning angle corresponding to the steering angle.

[0010] The control constants have a tendency to change according to the vehicle speed. Therefore, in the above configuration, the control constants are learned for each vehicle speed. This makes it possible to obtain appropriate control constants according to the vehicle speed. 5. The steering control device according to any one of 1 to 4 above, wherein the sampling process is executed when the magnitude of the steering angle is equal to or greater than a minimum value for learning.

[0011] When the magnitude of the steering angle is excessively small, the yaw rate becomes excessively small, and the accuracy of estimating the steering angle using the yaw rate decreases. In contrast, in the above configuration, the control constant is learned using the yaw rate sampled when the magnitude of the steering angle is equal to or greater than the minimum value for learning. This makes it possible to improve the learning accuracy of the control constant compared to the case where the yaw rate sampled when the magnitude of the steering angle is less than the minimum value for learning is used.

[0012] 6. The steering control device according to any one of 1 to 5 above, wherein the sampling process is executed when the magnitude of the steering angle is equal to or smaller than a maximum value for learning. When the steering angle is excessively large, the vehicle behavior is likely to become unstable, and the estimation accuracy of the steering angle using the yaw rate decreases. In contrast, in the above configuration, the control constants are learned using the yaw rate sampled when the steering angle is equal to or smaller than the maximum learning value. This makes it possible to improve the learning accuracy of the control constants compared to when the yaw rate sampled when the steering angle is larger than the maximum learning value is used.

[0013] 7. The steering control device according to any one of 1 to 6 above, wherein the sampling process is executed when the magnitude of the speed of change of the steering angle is equal to or less than a specified speed. When the steering angle changes, the yaw rate changes accordingly. Therefore, if the yaw rate is sampled when the change in the steering angle is large, the yaw rate is not stable, and the estimation accuracy of the steering angle using the yaw rate decreases. In contrast, in the above configuration, the control constant is learned using the yaw rate sampled when the magnitude of the speed of change of the steering angle is equal to or less than a specified speed. This makes it possible to improve the learning accuracy of the control constant compared to the case where the yaw rate sampled when the magnitude of the speed of change of the steering angle is greater than the specified speed is used.

[0014] 8. A steering control device described in any one of 1 to 7 above, wherein the sampling process is executed when the pair of the steering angle and the turning angle is not within a dead zone, and the dead zone is a zone in which the turning angle does not change in response to a change in the steering angle.

[0015] When the yaw rate is sampled when the set of the steering angle and the turning angle is in the dead zone, the relationship between the turning angle and the steering angle grasped from the yaw rate is affected by the dead zone. On the other hand, the dead zone is not a region that is uniquely determined according to the steering angle and the turning angle, but is a region that changes according to the steering angle history. Therefore, the relationship between the turning angle and the steering angle affected by the dead zone is a relationship specific to the situation at the time of sampling. In contrast, in the above configuration, by using the yaw rate sampled when the set of the steering angle and the turning angle is not in the dead zone, it is possible to learn a control constant that expresses a universal relationship that is unrelated to the driving history of each time.

[0016] 9. The steering control device according to any one of 1 to 8 above, wherein the sampling process is executed when the magnitude of the rate of change of the steering angle in a predetermined immediately preceding period is equal to or smaller than a threshold value.

[0017] If the change in steering angle during the predetermined period immediately prior to sampling is large, the vehicle behavior at the time of sampling may not be stable. In other words, the vehicle behavior at the time of sampling may be affected by the change in steering angle during the predetermined period. In such a case, the accuracy of the steering angle calculated from the sampled yaw rate decreases. In contrast, in the above configuration, sampling is performed when the magnitude of the change rate of the steering angle during the immediately prior predetermined period is equal to or less than a threshold value, so that the estimation accuracy of the steering angle can be improved.

[0018] 10. A steering control device described in any one of 1 to 9 above, wherein the sampling process is executed in each of a first state, a second state, a third state, and a fourth state, the first state being a state in which the steering angle is in a right turning region when the steering angle is changing in a right turning direction, the second state being a state in which the steering angle is in a left turning region when the steering angle is changing in a right turning direction, the third state being a state in which the steering angle is in a right turning region when the steering angle is changing in a left turning direction, and the fourth state being a state in which the steering angle is in a left turning region when the steering angle is changing in a left turning direction.

[0019] In the above configuration, the yaw rates sampled in each of the first state, the second state, the third state, and the fourth state are sampled, so that yaw rates corresponding to various steering situations can be sampled. Therefore, it is easy to learn control constants corresponding to various steering situations.

[0020] 11. A steering control device as described in any one of 2 to 10 (excluding those not dependent on 2) above, configured to execute a steering angle estimation process, the steering angle estimation process being a process of calculating an estimated value of the steering angle using a detection value of the steering angle and a stiffness coefficient learned by the control constant learning process.

[0021] In the above configuration, since the estimated value is calculated using the stiffness coefficient learned by the control constant learning process, it is possible to use a stiffness coefficient that is more appropriate for the steering device than when a default stiffness coefficient is used, thereby improving the accuracy of the estimated value.

[0022] 12. The steering device is a steering control device as described in claim 11, which is equipped with a motor that rotates the steering shaft and is configured to perform a steering process, and the steering process is a process of manipulating the torque of the motor in accordance with an operation amount of feedback control in which the estimated value is a control amount and a target value of the steering angle is a target value of the control amount.

[0023] In the above configuration, by using the estimated value, it is possible to execute feedback control using the steering angle as a control amount without relying on a sensor for detecting the steering angle. 13. The steering device is equipped with a motor that rotates the steering shaft and is configured to perform a steering process, the steering process being a process in which a dead zone amount calculated according to the dead zone width learned by the control constant learning process is input and the torque of the motor is manipulated, and the dead zone amount is the maximum amount by which the steering angle does not change in response to a change in the steering angle in the steering direction, a steering control device as described in 3 to 12 (excluding those not dependent on 3 above).

[0024] In the above configuration, the input for the process of manipulating the torque of the motor is the dead zone amount calculated according to the dead zone width learned by the control constant learning process. Therefore, a more accurate dead zone width can be used compared to the case where a default dead zone width is used. Therefore, in the above configuration, the torque of the motor can be manipulated according to a highly accurate dead zone amount. [Brief description of the drawings]

[0025] [Figure 1] FIG. 1 is a diagram showing a configuration of a steering system according to an embodiment. [Diagram 2] 2 is a diagram showing the relationship between the steering angle and the turning angle in the steering system shown in FIG. 1. [Diagram 3] 2 is a block diagram showing a process executed by the steering control device shown in FIG. 1. [Figure 4] 2 is a flowchart showing a procedure of a process executed by the steering control device shown in FIG. [Diagram 5] 2 is a flowchart showing a procedure of a process executed by the steering control device shown in FIG. [Figure 6] FIG. 6 is a diagram illustrating a part of the process shown in FIG. 5. [Figure 7] 2 is a flowchart showing a procedure of a process executed by the steering control device shown in FIG. [Figure 8] 2 is a flowchart showing a procedure of a process executed by the steering control device shown in FIG. [Figure 9] 2 is a flowchart showing a procedure of a process executed by the steering control device shown in FIG. [Figure 10] 2 is a time chart illustrating a process executed by the steering control device shown in FIG. 1; DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0026] Hereinafter, an embodiment will be described with reference to the drawings. "Prerequisite configuration" The steering device 10 shown in FIG. 1 includes a steering wheel 12. A steering shaft 14 is connected to the steering wheel 12. The end of the steering shaft 14 opposite to the steering wheel 12 is connected to a steering gear box 20. A sector shaft of the steering gear box 20 is connected to one end of a pitman arm 22. The other end of the pitman arm 22 is connected to one end of a drag link 24. The other end of the drag link 24 is connected to one end of a knuckle arm 26. The other end of the knuckle arm 26 is connected to a kingpin shaft 28 of a right steered wheel 40(R). The kingpin shaft 28 of the right steered wheel 40(R) and the kingpin shaft 28 of the left steered wheel 40(L) are connected by a tie rod arm 30 and a tie rod 32.

[0027] The rotational power of the motor 50 is transmitted to the steering shaft 14. The motor 50 is, for example, a synchronous motor. The output voltage of an inverter 52 is applied to the terminals of the motor 50.

[0028] The steering control device 60 includes a PU 62 and a storage device 64. The PU 62 is a software processing device such as a CPU, a GPU, or a TPU. The storage device 64 may be an electrically non-rewritable non-volatile memory. The storage device 64 may also be an electrically rewritable non-volatile memory or a storage medium such as a disk medium. The steering control device 60 executes a process for controlling a control target by the PU 62 executing a program stored in the storage device 64.

[0029] The control target of the steering control device 60 is the steering device 10. To control the control target, the steering control device 60 refers to the rotation angle θm of the motor 50 detected by the rotation angle sensor 70. The steering control device 60 also refers to the yaw rate γ detected by the yaw rate sensor 71. The steering control device 60 also refers to the currents iu, iv, iw flowing through each terminal of the motor 50. The currents iu, iv, iw may be detected as voltage drops across shunt resistors provided in each leg of the inverter 52, for example. The steering control device 60 refers to the vehicle speed V detected by the vehicle speed sensor 84 via the network 72.

[0030] The steering control device 60 is capable of communicating with a host ECU 80 via a network 72. The host ECU 80 executes a process of generating a command for intervening in the steering of the vehicle, independent of a steering command by the operation of the steering wheel 12. In other words, the host ECU 80 executes the automatic steering process. In order to execute the automatic steering process, the host ECU 80 acquires image data of the surroundings of the vehicle captured by a camera 82. The host ECU 80 also grasps an indication of intent, input by the driver via an interface 86, as to whether or not to execute the automatic steering process.

[0031] "Characteristics of steering device 10" The above-described steering device 10 has a complex link mechanism, which results in a large amount of play in which the steered wheels 40 do not rotate relative to the rotation of the steering shaft 14.

[0032] FIG. 2 shows the relationship between the steering angle θh and the turning angle θt. Here, the steering angle θh is the rotation angle of the steering shaft 14. Meanwhile, the turning angle θt is the turning angle of the tires with respect to the steered wheels 40. The neutral position O shown in FIG. 2 is a point where both the steering angle θh and the turning angle θt are zero. This means that both the steering angle θh and the turning angle θt indicate the straight ahead direction. In the following, the rotation angle in the right turning direction is taken as positive, and the rotation angle in the left turning direction is taken as negative.

[0033] As shown in Fig. 2, even if the steering angle θh is changed in the right turning direction from the neutral position O, the turning angle θt does not change until it reaches point A. Then, when the steering angle θh becomes a value larger than point A, the turning angle θt increases.

[0034] Furthermore, even if the steering angle θh is reduced at point B, the turning angle θt does not change. In other words, even if the steering direction is switched to the left at point B, the turning angle θt does not change. Note that the steering direction is the rotation direction of steering shaft 14. Then, when the steering angle θh exceeds point C and becomes even smaller, the turning angle θt decreases.

[0035] Then, at point D, the steering angle θt becomes zero. Point E is a value obtained by further displacing the steering angle θh in the left steering direction even after the steering angle θt becomes zero. Even if the steering angle θh is increased at point E, the steering angle θt does not change until the steering angle θh reaches point F. When the steering angle θh becomes even larger beyond point F, the steering angle θt increases.

[0036] In this way, for example, when the set of steering angle θh and turning angle θt is located at point B, the turning angle θt does not change even if the steering angle θh changes between point B and point C. Also, for example, when the set of steering angle θh and turning angle θt is located at point E, the turning angle θt does not change even if the steering angle θh changes between point E and point F. In this way, due to play in the steering device 10, there exists a dead zone area in which the turning angle θt does not change with respect to changes in the steering angle θh. However, the dead zone area is not an area having a fixed value of the steering angle θh, but an area that can change depending on the history of changes in the steering angle θh.

[0037] The change in steering angle θh from point B to point C occurs when steering back from a right turn. The change in steering angle θh from point E to point F occurs when steering back from a left turn. The dead zone width α shown in FIG. 2 is the length of the dead zone area where the steered angle θt does not change in response to the change in steering angle θh when steering back.

[0038] "Overview of Steering Control Device Control" Fig. 3 shows a part of the process executed by the steering control device 60. The process shown in Fig. 3 is realized by the PU 62 repeatedly executing a program stored in the storage device 64, for example, at a predetermined interval.

[0039] Feedback processing M10 is processing in which the target steering angle θt* and the steering angle θt are input and the feedback control input Tfb is output. The feedback control input Tfb is a feedback control input in which the steering angle θt is a control amount and the target steering angle θt* is a target value of the control amount. The target steering angle θt* is calculated by the host ECU 80. That is, when the automatic driving mode is selected by operating the interface 86, the host ECU 80 calculates a target trajectory on which the vehicle should travel based on image data of the surroundings of the vehicle captured by the camera 82. The host ECU 80 then calculates the target steering angle θt* to bring the actual driving route of the vehicle closer to the target trajectory.

[0040] The steering angle estimation process M12 is a process in which the steering angle θh is input and the steering angle θt is output. The steering angle θt is an estimated value. The steering angle θh is calculated by the PU 62 through an integration process of the rotation angle θm.

[0041] The dead-zone compensation torque calculation process M14 is a process in which the dead-zone compensation torque Tdz, which is the manipulated variable of the feedforward control in the dead-zone region, is output. The addition process M16 is a process of substituting the sum of the feedback operation amount Tfb and the dead zone compensation torque Tdz into the torque command value Trq*. The torque command value Trq* is a command value for the torque of the motor 50. The operation signal generation process M18 is a process of generating an operation signal MS for controlling the torque of the motor 50 to approach the torque command value Trq* and outputting it to the inverter 52. Note that the operation signal generation process M18 may be a process of bringing the currents iu, iv, and iw closer to a current command value corresponding to the torque command value Trq*.

[0042] The dead zone position calculation process M20 is a process for calculating the dead zone position Δα, which is a variable indicating a position within the dead zone region shown in Fig. 2. Fig. 2 illustrates an example of the dead zone position Δα when the set of the steering angle θh and the turning angle θt is located at point P.

[0043] The control constant learning process M22 is a process for learning a control constant which is a constant indicating the rotational displacement of the steered wheels 40 relative to the rotational displacement of the steering shaft 14. "Control constant learning process M20" Fig. 4 shows the procedure of pre-processing for the control constant learning process M22. The series of processes shown in Fig. 4 is realized by the PU 62 repeatedly executing a program stored in the storage device 64, for example, at a predetermined interval. In the following, the step number of each process is represented by a number preceded by "S".

[0044] In the series of processes shown in FIG. 4, the PU 62 first acquires the steering angle θh and the vehicle speed V (S10). Next, the PU 62 determines whether the absolute value of the steering angle θh is equal to or greater than a learning minimum value θL and equal to or less than a learning maximum value θH (S12). Here, the learning minimum value θL is set based on a lower limit value of the steering angle θh that allows the learning process described later to be performed with good accuracy. The learning maximum value θH is set based on an upper limit value of the steering angle θh that allows the learning process described later to be performed with good accuracy. In setting the learning maximum value θH, it is taken into consideration that if the steering angle θh is excessively large, the vehicle will not operate stably.

[0045] When the PU 62 determines that the absolute value is equal to or greater than the learning minimum value θL and equal to or less than the learning maximum value θH (S12: YES), the PU 62 determines whether or not a condition is established that the absolute value of the steering angular velocity ωh is equal to or less than a threshold value ωth1 (S14). The steering angular velocity ωh is the rate of change of the steering angle θh. The steering angular velocity ωh is calculated by the PU 62 using time-series data of the steering angle θh. The condition that the absolute value of the steering angular velocity ωh is equal to or less than a threshold value ωth1 is one of the execution conditions for the learning process described later. This condition is a condition that the steering angle θh is stable.

[0046] When PU 62 determines that the absolute value of steering angular velocity ωh is equal to or less than threshold value ωth1 (S14: YES), PU 62 determines whether or not the set of steering angle θh and turning angle θt is outside the dead zone (S16). The input for this process is the dead zone position Δα output by the dead zone position calculation process M20. When the absolute value of the dead zone position Δα is less than "1 / 2" of the dead zone width αi described later, PU 62 determines that the set of steering angle θh and turning angle θt is located in the dead zone.

[0047] When PU 62 determines that the set of steering angle θh and turning angle θt is not within the dead zone (S16: YES), PU 62 determines whether the following condition (A) is satisfied (S18). Condition (A): The condition that the absolute value of the steering angular velocity ωh from a timing that is a predetermined period back from the current execution timing of the series of processes in FIG. 4 to the present is equal to or less than a threshold value ωth2.

[0048] When it is determined that the condition (A) is satisfied (S18: YES), the PU 62 acquires the yaw rate γ (S20). Next, the PU 62 multiplies the yaw rate γ by "(1+A·V·V) / V / I" and assigns the result to the steering angle θt (S22). In this process, the PU 62 estimates the steering angle θt based on the yaw rate γ. Here, the stability factor A and the wheelbase I are used.

[0049] Next, the PU 62 stores the steering angle θt estimated by the process of S22 in the storage device 64 in association with the turning direction, the vehicle speed V, and the steering angle θh (S24). When the process of S24 is completed or when a negative determination is made in the processes of S12, S14, S16, or S18, the PU 62 temporarily ends the series of processes shown in FIG.

[0050] The procedure of the control constant learning process M22 is shown in Fig. 5. The series of processes shown in Fig. 5 is realized by the PU 62 repeatedly executing a program stored in the storage device 64, for example, at a predetermined interval.

[0051] 5, the PU 62 judges whether the data stored in the storage device 64 by the process of S24 is sufficient to satisfy the learning condition (S30). As an example, the learning condition is that the number of values ​​stored by the process of S24 in each of the following states 1 to 4 is a predetermined number. Here, the predetermined number may be more than one.

[0052] State 1: When the steering direction is a right turning direction, the steering angle θh is equal to or larger than the learning minimum value θL and equal to or smaller than the learning maximum value θH. State 2: When the steering direction is a right turning direction, the steering angle θh is equal to or larger than "-θH" and equal to or smaller than "-θL".

[0053] State 3: When the steering direction is a left turning direction, the steering angle θh is equal to or greater than the learning minimum value θL and equal to or less than the learning maximum value θH. State 4: When the steering direction is a left turning direction, the steering angle θh is equal to or larger than "-θH" and equal to or smaller than "-θL".

[0054] When the PU 62 determines that the data has been secured in an amount sufficient to satisfy the learning conditions (S30: YES), it assigns "1" to the variable i that identifies the data (S32). Next, PU 62 reads out data in which the vehicle speed V falls within region Ai from among the sampling data stored in storage device 64 by the process of S24 (S34). Region Ai indicates one of regions A1, A2, ... according to variable i. Regions A1, A2, ... are regions of different vehicle speeds V. For example, region A3 is 30 to 40 km / h, region A4 is 40 to 50 km / h, etc. Then, PU 62 calculates stiffness coefficient ki (S36). The stiffness coefficient is a coefficient that determines the ratio of the amount of change in steering angle θt to the amount of change in steering angle θh when the set of steering angle θh and steering angle θt is not in the dead zone region.

[0055] The black circles in Fig. 6 indicate pairs of steering angle θh and turning angle θt stored by the process of S24. As an example, PU 62 uses the least squares method to find a straight line that minimizes the deviation from the sampling points consisting of the pairs of steering angle θh and turning angle θt. Then, PU 62 substitutes the slope of the straight line for stiffness coefficient ki. Stiffness coefficient ki is a learned value when vehicle speed V is a value within region Ai.

[0056] Returning to FIG. 5, the PU 62 extracts data when the steering direction is a right turning direction from the data read out by the process of S34 (S38). Then, the PU 62 calculates a right turning straight line L1 shown in FIG. 6 based on the data extracted by the process of S38 (S40). The right turning straight line L1 is a straight line whose slope is the stiffness coefficient ki calculated by the process of S36 and has an intercept b1. The process of S40 is a process of obtaining an intercept b1 that minimizes the difference between the extracted data and the right turning straight line L1. The right turning straight line L1 is an approximation equation that approximates the relationship between the steering angle θh and the steered angle θt when the steering direction is a right turning direction.

[0057] Next, the PU 62 extracts data when the steering direction is a left turning direction from the data read out by the process of S34 (S42). The PU 62 then calculates a left turning straight line L2 based on the data extracted by the process of S42 (S44). The left turning straight line L2 is a straight line whose slope is the stiffness coefficient ki calculated by the process of S42 and has an intercept b2. The process of S44 is a process of obtaining the intercept b2 that minimizes the difference between the extracted data and the left turning straight line L2. The left turning straight line L2 is an approximation equation that approximates the relationship between the steering angle θh and the steered angle θt when the steering direction is a left turning direction.

[0058] Next, the PU 62 substitutes the absolute value of the difference between the intercepts b1 and b2 into the dead zone width αi (S46). The dead zone width αi is a learned value when the vehicle speed V is within the region Ai. Then, the PU 62 judges whether or not learning of the stiffness coefficients k1, k2, ... and the dead-band widths α1, α2, ... has been completed in all the regions A1, A2, ... related to the vehicle speed V (S48). If the PU 62 judges that there is a region for which learning has not yet been completed (S48: NO), the PU 62 increments the variable i (S50) and returns to the process of S34.

[0059] On the other hand, when the PU 62 determines that learning has been completed for all areas (S48: YES), it stores the learned data in a predetermined area of ​​the storage device 64 (S52). When the process of S52 is completed or when a negative determination is made in the process of S30, the PU 62 temporarily ends the series of processes shown in FIG.

[0060] Incidentally, until the above learning is completed, the PU 62 uses predetermined default values ​​as the stiffness coefficient ki and the dead zone width αi. The default values ​​are stored in the storage device 64 in advance. Also, for example, a region A0 where the vehicle speed V is zero and is not subject to learning by the process of FIG. 6 may be provided. In this case, the dead zone width α0 in the region A0 may be learned by using a separately known technique.

[0061] "Steering angle estimation processing" The procedure of the steering angle estimation process M12 is shown in Fig. 7. The process shown in Fig. 7 is realized by the PU 62 repeatedly executing a program stored in the storage device 64, for example, at a predetermined interval.

[0062] 7, the PU 62 first acquires the steering angle θh and the vehicle speed V (S60). Next, the PU 62 selects the stiffness coefficient ki based on the vehicle speed V acquired by the process of S60 (S62). For example, when the vehicle speed V is a value within the region A2, the PU 62 selects the stiffness coefficient k2.

[0063] Next, PU 62 assigns a value obtained by dividing steering angle θh by stiffness coefficient ki to steering angle θt0 (S64). Next, PU 62 calculates steering angle θt by appropriately correcting steering angle θt0 (S66). That is, when steering angle θt0 is located in the dead zone, PU 62 assigns the previous value of steering angle θt to steering angle θt. On the other hand, when steering angle θt0 is not located in the dead zone, PU 62 assigns a value obtained by adding the difference between the current value of steering angle θt0 and the previous value of steering angle θt to the previous value of steering angle θt to the current value of steering angle θt. Note that whether steering angle θt0 is located in the dead zone is determined by whether the magnitude of inside-dead zone position Δα calculated by inside-dead zone position calculation process M20 is smaller than "1 / 2" of dead zone width αi. That is, the process of S66 uses the dead zone position Δα as an input.

[0064] When the PU 62 completes the process of S66, the PU 62 temporarily ends the series of processes shown in FIG. "Dead zone position calculation process M20" The procedure of the dead zone position calculation process M20 is shown in Fig. 8. The process shown in Fig. 8 is realized by the PU 62 repeatedly executing a program stored in the storage device 64, for example, at a predetermined interval.

[0065] 8, the PU 62 first acquires the vehicle speed V (S70). Next, the PU 62 selects the dead zone width αi corresponding to the vehicle speed V acquired by the process of S70 (S72). For example, when the vehicle speed V is a value within the area A2, the PU 62 selects the dead zone width α2.

[0066] Next, PU 62 assigns a value obtained by adding a value obtained by subtracting the previous value "θh(n-1)" of steering angle θh from the current value "θh(n)" of steering angle θh to the previous value "Δα(n-1)" of the dead-zone position Δα, to a tentative position X (S74). Next, PU 62 determines whether or not tentative position X is equal to or greater than "(-1)·αi / 2" and equal to or less than "αi / 2" (S76). This process is a process for determining whether or not the set of steering angle θh and turning angle θt is in the dead-zone region. When PU 62 determines that tentative position X is equal to or greater than "(-1)·αi / 2" and equal to or less than "αi / 2" (S76: YES), it assigns tentative position X to the dead-zone position Δα (S77).

[0067] On the other hand, if the PU 62 makes a negative determination in the process of S76, it determines whether or not the tentative position X is greater than "αi / 2" (S78). If the PU 62 determines that the tentative position X is greater than "αi / 2" (S78: YES), it assigns "αi / 2" to the in-dead-zone position Δα (S80). On the other hand, if the PU 62 determines that the tentative position X is smaller than "αi / 2" (S78: NO), it assigns "(-1)·αi / 2" to the in-dead-zone position Δα (S82).

[0068] When the PU 62 completes the processes of S77, S80, and S82, it temporarily ends the series of processes shown in FIG. "Dead zone compensation torque calculation process M14" The procedure of the dead-zone compensation torque calculation process M14 is shown in Fig. 9. The process shown in Fig. 9 is realized by the PU 62 repeatedly executing a program stored in the storage device 64, for example, at a predetermined interval.

[0069] 9, the PU 62 first acquires the vehicle speed V (S90). Next, the PU 62 selects the dead-band width αi corresponding to the vehicle speed V selected by the process of S90 (S92).

[0070] PU 62 determines whether the value obtained by subtracting the previous value "θt*(n-1)" of target steering angle θt* from the current value "θt*(n)" is positive (S94). This process is to determine whether the steering direction is a right turning direction. If PU 62 determines that the above subtracted value is positive (S94: YES), it assigns a value obtained by subtracting the dead-band position Δα from "1 / 2" of the dead-band width αi selected by the process of S92 to the dead-band amount Dα (S96). The dead-band amount Dα by the process of S96 is the maximum amount by which the steering angle θt does not change when the steering angle θh is changed in the right turning direction.

[0071] Next, the PU 62 judges whether the dead-band amount Dα is greater than zero (S98). When the PU 62 judges that the dead-band amount Dα is greater than zero (S98: YES), the PU 62 assigns the sum of the offset amount TdzB and the value obtained by dividing the dead-band amount Dα by the dead-band width αi and adding the specified torque TdzA to the value, to the dead-band compensation torque Tdz (S100).

[0072] On the other hand, when the PU 62 makes a negative determination in the process of S94, it determines whether or not a value obtained by subtracting the previous value "θt*(n-1)" of the target steering angle θt* from the current value "θt*(n)" is negative (S102). This process is a process for determining whether or not the steering direction is a left turning direction. When the PU 62 determines that the subtracted value is negative (S102: YES), it assigns a value obtained by subtracting the inside-dead-band position Δα from "-1 / 2" of the dead-band width αi selected in the process of S92 to the dead-band amount Dα (S104). The absolute value of the dead-band amount Dα in the process of S104 is the maximum amount by which the steering angle θt does not change when the steering angle θh is changed in the left turning direction, and the maximum amount means that the absolute value is maximum. However, the dead-band amount Dα here is equal to or less than zero.

[0073] When the PU 62 determines that the dead band amount Dα is smaller than zero (S106: YES), the PU 62 assigns a value obtained by subtracting an offset amount TdzB from a value obtained by dividing the dead band amount Dα by the dead band width αi and adding the specified torque TdzA to the value, to the dead band compensation torque Tdz (S108).

[0074] On the other hand, if a negative determination is made in the processes of S98 and S106, the PU 62 assigns "0" to the dead zone compensation torque Tdz (S110). When the processes of S100, S108, and S110 are completed, or when a negative determination is made in the process of S102, the PU 62 temporarily ends the series of processes shown in FIG.

[0075] FIG. 10 illustrates the transition of the dead zone compensation torque Tdz according to this embodiment. As shown in FIG. 10, the PU 62 sets the dead zone compensation torque Tdz to a torque for turning left at time t1 when it is determined that the steering direction has changed from a right turning direction to a left turning direction based on the target turning angle θt*. In FIG. 10, it is assumed that the turning angle θt is not in the dead zone area immediately before time t1. Therefore, the dead zone amount Dα at time t1 is "-αi". Therefore, the PU 62 changes the dead zone compensation torque Tdz in a stepwise manner from "0" to "-TdzA-TdzB" at time t1. Then, the PU 62 gradually reduces the magnitude of the dead zone compensation torque Tdz until the dead zone amount Dα becomes zero. At time t2 when the dead-band amount Dα becomes zero, the PU 62 changes the dead-band compensation torque Tdz from "-TdzB" to "0" in a stepped manner.

[0076] Thereafter, at time t3 when it is determined that the steering direction has changed from a left turning direction to a right turning direction based on the target steering angle θt*, the PU 62 sets the dead zone compensation torque Tdz to a torque for turning right. The dead zone amount Dα at time t3 is "αi". Therefore, at time t3, the PU 62 changes the dead zone compensation torque Tdz in a stepwise manner from "0" to "TdzA+TdzB". Then, the PU 62 gradually decreases the magnitude of the dead zone compensation torque Tdz until the dead zone amount Dα becomes zero. At time t4 when the dead zone amount Dα becomes zero, the PU 62 changes the dead zone compensation torque Tdz in a stepwise manner from "TdzB" to "0".

[0077] In this way, when the steering angle θt falls within the dead zone, PU 62 uses the dead zone compensation torque Tdz as the manipulated variable for feedforward control to control the torque of motor 50. Therefore, the responsiveness of the steering angle θt can be improved compared to the case where the torque of motor 50 is controlled only by the feedback manipulated variable Tfb.

[0078] "Actions and Effects of the Present Embodiment" The PU 62 acquires the yaw rate γ in synchronization with the steering angle θh. The PU 62 then associates the steering angle θt calculated from the yaw rate γ and the vehicle speed V with the corresponding steering angle θh and stores them in the storage device 64. When a sufficient number of sampled values, which are pairs of the steering angle θh and the steering angle θt, are stored in the storage device 64, the PU 62 calculates an approximation equation indicated by these sampled values. The PU 62 calculates the slope of the approximation equation as the stiffness coefficient ki. The PU 62 also substitutes the difference between the intercepts of the approximation equation determined from the sampled values ​​in the right turning direction and the approximation equation determined from the sampled values ​​in the left turning direction into the dead zone width αi. This makes it possible to learn the stiffness coefficient ki and the dead zone width αi specific to the steering device 10.

[0079] According to the present embodiment described above, the following actions and effects can be obtained. (1-1) The PU 62 learns the dead-band width αi according to the vehicle speed V. This makes it possible to calculate the dead-band position Δα with higher accuracy than when the dead-band width αi is set to a value independent of the vehicle speed V.

[0080] (1-2) The PU 62 learns the stiffness coefficient ki according to the vehicle speed V. This allows the steering angle θt to be estimated with higher accuracy than when the stiffness coefficient ki is set to a value independent of the vehicle speed V.

[0081] (1-3) PU 62 estimated the turning angle θt using the yaw rate γ when the magnitude of the steering angle θh was equal to or greater than the learning minimum value θL. If the magnitude of the steering angle θh is excessively small, the yaw rate γ becomes excessively small, and the estimation accuracy of the turning angle using the yaw rate γ decreases. In contrast, in this embodiment, the turning angle θt is estimated using the yaw rate γ sampled when the magnitude of the steering angle θh is equal to or greater than the learning minimum value θL. This makes it possible to improve the learning accuracy based on the turning angle θt, compared to the case where the yaw rate γ sampled when the magnitude of the steering angle θh is less than the learning minimum value θL is used.

[0082] (1-4) PU 62 estimated the turning angle θt using the yaw rate γ when the magnitude of the steering angle θh was equal to or less than the learning maximum value θH. If the magnitude of the steering angle θh is excessively large, the vehicle behavior is likely to become unstable, and the estimation accuracy of the turning angle θt using the yaw rate γ decreases. In contrast, in this embodiment, the turning angle θt is estimated using the yaw rate γ sampled when the magnitude of the steering angle θh is equal to or less than the learning maximum value θH. This makes it possible to improve the learning accuracy based on the turning angle θt, compared to the case where the yaw rate γ sampled when the magnitude of the steering angle θh is greater than the learning maximum value θH is used.

[0083] (1-5) PU 62 estimated the turning angle θt using the yaw rate γ when the magnitude of the steering angular velocity ωh was equal to or less than the threshold value ωth1. When the steering angle θh changes, the yaw rate γ is not stable, and the estimation accuracy of the turning angle θt using the yaw rate γ decreases. In contrast, in this embodiment, the turning angle θt is estimated using the yaw rate γ sampled when the magnitude of the steering angular velocity ωh is equal to or less than the threshold value ωth1. This makes it possible to improve the learning accuracy based on the turning angle θt compared to the case where the yaw rate γ sampled when the magnitude of the steering angular velocity ωh is greater than the threshold value ωth1 is used.

[0084] (1-6) PU 62 estimated the turning angle θt using the yaw rate γ when the set of steering angle θh and turning angle θt was not in the dead zone. When the yaw rate γ is sampled when the set of steering angle θh and turning angle θt is in the dead zone, the relationship between the turning angle θt and the steering angle θh grasped from the yaw rate γ is affected by the dead zone. On the other hand, the dead zone is not a region that is uniquely determined according to the steering angle θh and the turning angle θt, but is a region that changes according to the history of the steering angle θh. Therefore, the relationship between the turning angle θt and the steering angle θh that is affected by the dead zone is a relationship that is specific to the situation at the time of sampling. In contrast, in this embodiment, the turning angle θt is estimated using the yaw rate γ sampled when the set of steering angle θh and turning angle θt is not in the dead zone. This allows the PU 62 to grasp a universal relationship between the steering angle θh and the turning angle θt that is independent of the driving history of each time.

[0085] (1-7) PU 62 estimated the turning angle θt using the yaw rate γ when the magnitude of the steering angular velocity ωh in the immediately preceding predetermined period was equal to or less than the threshold value ωth2. If the change in the steering angle in the immediately preceding predetermined period is large, the behavior of the vehicle at the time of sampling may not be stable. In other words, the behavior of the vehicle at the time of sampling may be affected by the change in the steering angle in the predetermined period. In that case, the accuracy of the turning angle θt calculated from the sampled yaw rate γ decreases. In contrast, in this embodiment, sampling is performed when the magnitude of the steering angular velocity ωh in the immediately preceding predetermined period is equal to or less than the threshold value ωth2, so that the estimation accuracy of the turning angle θt can be improved.

[0086] (1-8) PU 62 learned the dead zone width αi and the stiffness coefficient ki using the steering angle θt estimated from the yaw rate γ sampled in each of the first, second, third, and fourth states described above. This means the following, compared to the case where only the steering angle θt estimated from the yaw rate γ sampled in each of the first and third states is used. That is, it means that the right turning straight line L1 and the left turning straight line L2 are determined using a wider range of sampling points of the steering angle θh and the steering angle θt. Therefore, the accuracy with which the right turning straight line L1 and the left turning straight line L2 approximate the relationship between the steering angle θh and the steering angle θt during right turning and left turning in actual steering is improved.

[0087] (1-9) The PU 62 samples an equal number of yaw rates γ to be used for learning in each of the above-mentioned states 1, 2, 3, and 4. This increases the accuracy with which the right turning straight line L1 and the left turning straight line L2 approximate the relationship between the steering angle θh and the turning angle θt during right turning and left turning in actual steering.

[0088] (1-10) The stiffness coefficient ki is calculated using all of the sampling points for calculating the right turning straight line L1 and the left turning straight line L2. Therefore, the number of sampling points for calculating the stiffness coefficient ki can be maximized. Therefore, compared to the case where only a portion of the sampling points is used, the stiffness coefficient ki can be calculated with higher accuracy.

[0089] (1-11) The PU 62 adopts the stiffness coefficient ki as the inclination of the right turning straight line L1 and the left turning straight line L2. This makes it easier to suppress deviations in the inclinations of the right turning straight line L1 and the left turning straight line L2 compared to the case where the inclinations of the right turning straight line L1 and the left turning straight line L2 are calculated separately.

[0090] <Correspondence> The correspondence between the matters in the above embodiment and the matters described in the above "Means for solving the problem" column is as follows. Below, the correspondence is shown for each number of the solving means described in the "Means for solving the problem" column. [1, 2, 5, 6, 11] Sampling process corresponds to the process of S20, S24. Steering angle calculation process corresponds to the process of S22. Control constant learning process corresponds to the process of FIG. 5. [3] Right turn straight line calculation process corresponds to the process of S38, S40. Left turn straight line calculation process corresponds to the process of S42, S44. Dead zone width calculation process corresponds to the process of S46. [4] Corresponds to the extraction of data in the area Ai determined for each vehicle speed by the process of S34. [7] The specified speed corresponds to the threshold value ωth1. [8] Corresponds to the process of S16. [9] Corresponds to the process of S18.

[10] Corresponds to the sampling point illustrated in FIG. 6.

[12] The steering process corresponds to the feedback process M10, the dead-zone compensation torque calculation process M14, the addition process M16, and the operation signal generation process M18.

[13] In the dead-zone compensation torque calculation process M14, the dead-zone compensation torque Tdz is calculated in accordance with the dead-zone amount Dα in the processes of S96 and S104.

[0091] <Other embodiments> This embodiment can be modified as follows: This embodiment and the following modifications can be combined with each other to the extent that no technical contradiction occurs.

[0092] "About sampling process" The method of sampling in the sampling process is not limited to the example shown in Fig. 6. That is, it is not essential to perform sampling in the right-turning direction and the left-turning direction in each of the regions where the steering angle θh is equal to or greater than the learning minimum value θL and equal to or less than the learning maximum value θH, and the region where the steering angle θh is equal to or greater than "-θH" and equal to or less than "-θL". For example, sampling may be performed only in each of the right-turning direction and the left-turning direction in the region where the steering angle θh is equal to or greater than the learning minimum value θL and equal to or less than the learning maximum value θH. Also, for example, sampling may be performed only in each of the right-turning direction and the left-turning direction in the region where the steering angle θh is equal to or greater than the learning minimum value θL and equal to or less than "-θH"

[0093] The condition for executing the sampling process does not necessarily have to be a condition that results in a positive determination in all of the processes from S12 to S18. "About the steering angle calculation process" In the process of S22, the PU 62 calculates the steering angle θt using the stability factor A and the wheel base I, but this is not limited to the above. For example, the steering angle θt may be calculated by the PU 62 using a map in a state where map data is stored in the storage device 64. Here, the map data is data in which the yaw rate γ and the vehicle speed V are input variables and the steering angle θt is an output variable.

[0094] "About control constant learning process" 6, the number of sampling points used by the control constant learning process in each of the first state, the second state, the third state, and the fourth state is three, but this is not limited to this. For example, the number of sampling points used by the control constant learning process in each state may be two. Also, for example, the number of sampling points used by the control constant learning process in each state may be four. Also, it is not essential that the number of sampling points used by the control constant learning process in each state is equal to each other.

[0095] The sampling points used in learning the stiffness coefficient ki are not limited to the points exemplified in FIG. 6. That is, the sampling points used in the control constant learning process are not limited to the points sampled in each of the first to fourth states. For example, the sampling points used in learning the stiffness coefficient ki may be only points sampled in any three states among the four states of the first to fourth states. Also, for example, the sampling points used in learning the stiffness coefficient ki may be only points sampled in any two states among the four states of the first to fourth states. Also, for example, the sampling points used in learning the stiffness coefficient ki may be only points sampled in any one state among the four states of the first to fourth states. In this case, the number of sampling points may be two or more.

[0096] The sampling points used to learn the dead zone width αi are not limited to the points sampled in each of the first to fourth states. For example, the sampling points used to learn the dead zone width αi may be only the points sampled in the first and third states. In this case, the number of sampling points in each state may be two or more. Also, for example, the sampling points used to learn the dead zone width αi may be only the points sampled in the second and fourth states. In this case, the number of sampling points in each state may be two or more.

[0097] It is not essential that the sampling points used for learning the stiffness coefficient ki and the sampling points used for learning the dead zone width αi are the same. It is not essential that the learning process of the dead zone width αi is executed after the learning process of the stiffness coefficient ki. For example, the right turning straight line L1 and the left turning straight line L2 may be calculated prior to learning the stiffness coefficient ki. Even in this case, the dead zone width αi can be learned by the process of S46. In this case, the stiffness coefficient ki may be, for example, the average value of the slope of the right turning straight line L1 and the slope of the left turning straight line L2.

[0098] The control constant learning process does not necessarily have to be a process for learning different dead zone widths αi at different vehicle speeds V. For example, the control constant learning process may be a process for learning the dead zone width αi at a single vehicle speed V that is greater than zero.

[0099] The control constant learning process does not necessarily have to be a process for learning different stiffness coefficients ki at different vehicle speeds V. For example, the control constant learning process may be a process for learning the stiffness coefficient ki at a single vehicle speed V.

[0100] "About steering processing" The process of manipulating the torque of the motor 50 using the dead-zone amount Dα calculated according to the dead-zone width αi as an input is not limited to the process using the dead-zone compensation torque Tdz. For example, the process may be a process of correcting the target steering angle θt* according to the dead-zone amount Dα.

[0101] "Motor torque control" It is not essential that the drive circuit of motor 50 is operated according to the manipulated variable of control in which the steering angle is the controlled variable and the target value of the steering angle is the target value of the controlled variable. For example, the drive circuit of motor 50 may be operated according to the manipulated variable of feedforward control according to the target value. In that case, in the dead zone, PU 62 may superimpose the dead zone compensation torque Tdz on the manipulated variable of feedforward control that is calculated even outside the dead zone region.

[0102] "About autonomous driving mode" In the autonomous driving mode, the sensing member used to acquire information about the outside of the vehicle, including the lane, is not limited to the camera 82. For example, it may be a LIDAR.

[0103] "About driving modes" The driving mode in which the dead zone compensation torque Tdz is superimposed on the torque of the motor 50 is not limited to the automatic driving mode. For example, it may be a manual driving mode in which the driver performs steering. In that case, for example, in the processing of S94 and S102, the steering angle θh may be used instead of the target steering angle θt*.

[0104] "About steering control devices" The host ECU 80 and the steering control device 60 may be integrated into one unit. The steering control device is not limited to one that executes software processing. For example, it may be equipped with a dedicated hardware circuit such as an ASIC that executes at least a part of the processing executed in the above embodiment. That is, the steering control device may include a processing circuit having any of the following configurations (a) to (c). (a) A processing circuit equipped with a processing device that executes all of the above processing according to a program, and a program storage device such as a memory device that stores the program. (b) A processing circuit equipped with a processing device and a program storage device that executes part of the above processing according to a program, and a dedicated hardware circuit that executes the remaining processing. (c) A processing circuit equipped with a dedicated hardware circuit that executes all of the above processing. Here, there may be multiple software execution devices equipped with a processing device and a program storage device. Also, there may be multiple dedicated hardware circuits. [Explanation of symbols]

[0105] 10...Steering gear 12…Steering wheel 14…Steering shaft 20…Steering gear box 22…Pitman arm 24…Drag Link 26…Knuckle arm 28…Kingpin axis 30…Tie rod arm 32…Tie rod 40…Steering wheel 50…Motor 52…Inverter 60...Steering control device

Claims

1. A steering control device that controls a steering device, The steering device is A steering shaft; a steered wheel of a vehicle that is steered in accordance with the rotation of the steering shaft; Equipped with A sampling process, a steering angle calculation process, and a control constant learning process are executed. the sampling process is a process of sampling a steering angle and a yaw rate in synchronization with each other, the steering angle is a rotation angle of the steering shaft, the steering angle calculation process is a process of calculating a steering angle of the steered wheels using the yaw rate and the vehicle speed as inputs, the control constant learning process is a process of learning a control constant based on a plurality of pairs of the steering angle sampled by the sampling process and the turning angle corresponding to the steering angle, A steering control device, wherein the control constant is a constant indicating a rotational displacement of the steered wheels relative to a rotational displacement of the steering shaft.

2. The control constant includes a stiffness coefficient that is a ratio of a turning angle to the steering angle, 2. The steering control device according to claim 1, wherein the control constant learning process includes a process of learning, as the stiffness coefficient, a gradient of a straight line determined from a plurality of pairs of the steering angle and the turning angle corresponding to the steering angle.

3. The control constant includes a dead band width, the dead zone width is a rotation amount of the steering shaft from when the rotation direction of the steering shaft is switched from one of the right turning direction and the left turning direction to the other direction until the steered wheels are turned, The control constant learning process includes a right-turn straight line calculation process, a left-turn straight line calculation process, and a dead zone width calculation process, the right-turning straight line calculation process is a process of calculating a right-turning straight line from a plurality of pairs of the sampled steering angle and the steered angle corresponding to the steering angle when the rotation direction is a right-turning direction, the left turning straight line calculation process is a process of calculating a left turning straight line from a plurality of pairs of the sampled steering angle and the steered angle corresponding to the steering angle when the rotation direction is a left turning direction, 2. The steering control device according to claim 1, wherein the dead zone width calculation process calculates a difference between intercepts of the right turning straight line and the left turning straight line as the dead zone width.

4. 2. The steering control device according to claim 1, wherein the control constant learning process is a process for calculating the control constant for each of the different vehicle speeds based on a plurality of pairs of the steering angle sampled by the sampling process for each of the different vehicle speeds and the turning angle corresponding to the steering angle.

5. 2. The steering control device according to claim 1, wherein the sampling process is executed when the magnitude of the steering angle is equal to or greater than a minimum value for learning.

6. 2. The steering control device according to claim 1, wherein the sampling process is executed when the magnitude of the steering angle is equal to or smaller than a maximum value for learning.

7. 2. The steering control device according to claim 1, wherein the sampling process is executed when the magnitude of the speed of change of the steering angle is equal to or smaller than a specified speed.

8. The sampling process is executed when the set of the steering angle and the turning angle is not within a dead zone; 2. The steering control device according to claim 1, wherein the dead zone is a region in which the turning angle does not change in response to a change in the steering angle.

9. The steering control device according to claim 1 , wherein the sampling process is executed when a magnitude of a rate of change of the steering angle in a predetermined immediately preceding period is equal to or smaller than a threshold value.

10. the sampling process is performed in each of a first state, a second state, a third state, and a fourth state; the first state is a state in which the steering angle is in a right turning region when the steering angle is changing in a right turning direction, the second state is a state in which the steering angle is in a left turning region when the steering angle is changing in a right turning direction, the third state is a state in which the steering angle is in a right turning region when the steering angle is changing in a left turning direction, 2. The steering control device according to claim 1, wherein the fourth state is a state in which the steering angle is in a left turning area when the steering angle is changing in a left turning direction.

11. A steering angle estimation process is performed.

3. The steering control device according to claim 2, wherein the steering angle estimation process calculates an estimated value of the steering angle by using the detected value of the steering angle and a stiffness coefficient learned by the control constant learning process.

12. The steering device includes a motor that rotates the steering shaft, configured to execute a steering process; 12. The steering control device according to claim 11, wherein the steering process is a process of manipulating a torque of the motor in response to an operation amount of a feedback control in which the estimated value is a control amount and a target value of the steering angle is a target value of the control amount.

13. The steering device includes a motor that rotates the steering shaft, configured to execute a steering process; the steering process is a process in which a dead-zone amount calculated in accordance with the dead-zone width learned by the control constant learning process is input, and a torque of the motor is manipulated using the dead-zone amount, 4. The steering control device according to claim 3, wherein the dead band amount is a maximum amount by which the steered angle does not change with respect to a change in the steering angle in the steering direction.

Images