阅读说明：本技术 转向系统 (Steering system ) 是由 东真康 三木大辅 罗伯特·富克斯 于 2020-03-18 设计创作，主要内容包括：提供了一种转向系统(1),其包括对电动马达(18)进行控制的控制器(12)。控制器(12)具有对合成了在包括第一传动装置的各传动装置中产生的摩擦扭矩的合成摩擦扭矩进行估算的合成摩擦扭矩估算单元(48)。合成摩擦扭矩估算单元(48)具有：基于电动马达(18)的角速度计算第一传动装置的滑移速度的滑移速度计算部(81)；基于滑移速度计算第一传动装置的摩擦系数的摩擦系数计算部(82)；计算作用在第一传动装置的齿面上的法向力的齿面法向力计算部(84、86)；以及利用摩擦系数、齿面法向力和预先设定的一个或更多个修正因子计算合成摩擦扭矩的摩擦扭矩计算部(83、85、87、88、89、90、91)。(A steering system (1) is provided that includes a controller (12) that controls an electric motor (18). The controller (12) has a synthetic friction torque estimation unit (48) that estimates a synthetic friction torque that synthesizes friction torques generated in the respective transmission devices including the first transmission device. The synthetic friction torque estimation unit (48) has: a slip speed calculation unit (81) that calculates the slip speed of the first transmission device on the basis of the angular speed of the electric motor (18); a friction coefficient calculation unit (82) that calculates the friction coefficient of the first transmission device based on the slip speed; a tooth surface normal force calculation section (84, 86) that calculates a normal force acting on a tooth surface of the first transmission; and a friction torque calculation section (83, 85, 87, 88, 89, 90, 91) that calculates a resultant friction torque using the friction coefficient, the tooth surface normal force, and one or more preset correction factors.)

1. A steering system (1), characterized by comprising:

an electric motor (18);

a wheel steering shaft (14), the wheel steering shaft (14) being moved in an axial direction by a torque of the electric motor (18) to thereby steer a wheel (3) to be steered;

a plurality of transmissions;

an angular velocity detection unit (25), the angular velocity detection unit (25) detecting or estimating an angular velocity of the electric motor (18); and

a controller (12), the controller (12) controlling the electric motor (18), wherein:

one of the transmissions is a first transmission that outputs rotation of the electric motor (18) at a reduced speed;

the controller (12) has a synthetic friction torque estimation unit (48), the synthetic friction torque estimation unit (48) estimating a synthetic friction torque that synthesizes friction torques generated in the respective transmissions; and

the synthetic friction torque estimation unit (48) has: a slip speed calculation unit (81) that calculates the slip speed of the first transmission device on the basis of the angular speed; a friction coefficient calculation unit (82) that calculates a friction coefficient of the first transmission device based on the slip speed; a tooth surface normal force calculation section (84, 86) that calculates a normal force acting on a tooth surface of the first transmission; and a friction torque calculation section (83, 85, 87, 88, 89, 90, 91) that calculates the resultant friction torque using the friction coefficient of the first transmission, the tooth surface normal force of the first transmission, and one or more preset correction factors.

2. The steering system (1) according to claim 1, wherein the friction torque calculation section (83, 85, 87, 88, 89, 90, 91) is configured to calculate the synthetic friction torque based on a synthetic friction coefficient that is a value obtained by multiplying the friction coefficient by a predetermined first correction factor and a synthetic tooth surface normal force that is a value obtained by multiplying the tooth surface normal force by a predetermined second correction factor.

3. The steering system (1) according to claim 1, characterized in that the friction torque calculation section (83, 85, 87, 88, 89, 90, 91) is configured to calculate the resultant friction torque by calculating a first friction torque generated in the first transmission based on the friction coefficient and the tooth surface normal force, and then multiplying the obtained first friction torque by a predetermined third correction factor.

4. The steering system (1) according to any one of claims 1 to 3, characterized in that the steering system (1) further comprises:

a steering member (2); and

a steering shaft (6), the steering shaft (6) rotating integrally with the steering member (2), wherein:

the first transmission is a transmission that outputs the torque of the electric motor (18) to the steering shaft (6) or the wheel steering shaft (14); and

the other of the transmissions is a second transmission that converts the rotation of the steering shaft (6) into an axial movement of the wheel steering shaft (14).

5. The steering system (1) according to claim 4, characterized in that the steering system (1) further includes a torque detection unit (11), the torque detection unit (11) detecting a steering torque input from the steering member (2), wherein the controller (12) has an axial force estimation unit (46), the axial force estimation unit (46) estimating an axial force acting on the wheel steering shaft (14) based on the steering torque, the torque of the electric motor (18), the resultant friction torque, and the angle of the electric motor.

6. The steering system (1) according to claim 4, characterized in that the steering system (1) further comprises:

a torque detection unit (11), the torque detection unit (11) detecting a steering torque input from the steering member (2); and

an axial force estimation unit (46), the axial force estimation unit (46) detecting or estimating an axial force acting on the wheel steering shaft (14), wherein,

the tooth surface normal force calculation section (84, 86) sets a first contact force calculated based on the torque of the electric motor (18), the steering torque, and the axial force as the tooth surface normal force when the first contact force is greater than a predetermined value; and when the first contact force is equal to or less than the predetermined value, setting the predetermined value as the tooth surface normal force.

7. The steering system (1) according to claim 1, characterized in that the steering system (1) further comprises:

a steering member (2);

a steering shaft (6), the steering shaft (6) rotating integrally with the steering member (2);

a torque detection unit (11), the torque detection unit (11) detecting a steering torque input from the steering member (2); and

an axial force estimation unit (46), the axial force estimation unit (46) detecting or estimating an axial force acting on the wheel steering shaft (14), wherein:

the first transmission is a transmission that outputs the torque of the electric motor (18) to the steering shaft (6) or the wheel steering shaft (14);

another of the transmissions is a second transmission that converts rotation of the steering shaft (6) into axial movement of the wheel steering shaft (14);

the tooth flank normal force calculation section (84, 86) is configured to calculate a first tooth flank normal force based on the torque of the electric motor (18), the steering torque, and the axial force, and set a second tooth flank normal force; wherein a first flank normal force is a normal force acting on a flank of the first transmission in a first contact state, a second flank normal force is a normal force acting on the flank of the first transmission in a second contact state, and

the friction torque calculation section (83, 85, 87, 88, 89, 90, 91) is configured to calculate a composite first tooth surface normal force by multiplying the first tooth surface normal force by a predetermined fourth correction factor, calculate a composite second tooth surface normal force by multiplying the second tooth surface normal force by a predetermined fifth correction factor, calculate a composite friction coefficient by multiplying the friction coefficient by a predetermined sixth correction factor, and calculate the composite friction torque based on the composite friction coefficient and one of the composite first tooth surface normal force and the composite second tooth surface normal force having a larger absolute value.

8. Steering system (1) according to claim 7, characterized in that:

the tooth surface normal force calculation section (84, 86) has a first tooth surface normal force correction section (91, 92), and the first tooth surface normal force correction section (91, 92) corrects the first tooth surface normal force based on the previous synthetic frictional torque calculated by the frictional torque calculation section; and

the friction torque calculating section (83, 85, 87, 88, 89, 90, 91) is configured to calculate the synthetic first tooth surface normal force by multiplying the corrected first tooth surface normal force by the fourth correction factor, calculate the synthetic second tooth surface normal force by multiplying the second tooth surface normal force by the fifth correction factor, calculate the synthetic friction coefficient by multiplying the friction coefficient by the sixth correction factor, and calculate the synthetic friction torque based on the synthetic friction coefficient and one of the synthetic first tooth surface normal force and the synthetic second tooth surface normal force having a larger absolute value.

9. The steering system (1) according to any one of claims 1, 2, 3, 7 and 8, characterized in that one of the transmissions is a third transmission arranged on a power transmission path leading from the first transmission to the wheel steering shaft (14).

Technical Field

The present invention relates to a steering system.

Background

In the field of controlling assist torque in an electric power steering system (EPS) and controlling reaction torque in a steer-by-wire system, a technology has been developed that estimates reaction force from a road surface and rack axial force using signals from sensors installed in the EPS and a vehicle to transmit information about the road surface to a driver, thereby improving steering performance. For example, japanese patent application publication No. 2017-226318(JP2017-226318A) discloses a technique of estimating a rack axial force using information from a sensor mounted in an EPS (motor current, motor angle, and steering torque) and information from a sensor mounted in a vehicle (vehicle speed).

Disclosure of Invention

The technique described in JP2017-226318A has the following disadvantages: it cannot accurately estimate the frictional torque and therefore has low accuracy in estimating the rack axial force from the condition of the road surface or the condition of the tire. The invention provides a steering system capable of accurately estimating a friction torque generated in the steering system.

A steering system according to an aspect of the present invention includes: an electric motor; a wheel turning shaft that is moved in an axial direction by a torque of the electric motor to turn a wheel to be turned; a plurality of transmissions; an angular velocity detection unit that detects or estimates an angular velocity of the electric motor; and a controller that controls the electric motor. One of the transmissions is a first transmission that outputs rotation of the electric motor at a reduced speed. The controller has a synthetic friction torque estimation unit that estimates a synthetic friction torque that synthesizes friction torques generated in the respective transmission devices. The synthetic friction torque estimation unit has: a slip speed calculation unit that calculates a slip speed of the first transmission device based on the angular speed; a friction coefficient calculation unit that calculates a friction coefficient of the first transmission device based on the slip speed; a tooth surface normal force calculation section that calculates a normal force acting on a tooth surface of the first transmission; and a friction torque calculation section that calculates a resultant friction torque using the friction coefficient of the first transmission, the tooth surface normal force of the first transmission, and one or more correction factors set in advance.

In this aspect, the resultant friction torque is calculated using a tooth surface normal force calculation section that calculates a normal force acting on a tooth surface of the first transmission, a friction coefficient and a tooth surface normal force of the first transmission, and one or more correction factors set in advance. Therefore, the frictional torque generated in the steering system can be accurately estimated. Further, the calculation of the composite friction torque is simplified as compared with the case where the friction torques generated in the respective transmission devices are calculated separately using the respective meshing models and then the calculated respective friction torques are synthesized.

In the above aspect, the friction torque calculating section may be configured to calculate the synthetic friction torque based on the synthetic friction coefficient and the synthetic tooth surface normal force. The resultant friction coefficient is a value obtained by multiplying the friction coefficient by a predetermined first correction factor, and the resultant tooth surface normal force is a value obtained by multiplying the tooth surface normal force by a predetermined second correction factor. In the above-described aspect, the friction torque calculating section may be configured to calculate the resultant friction torque by calculating a first friction torque generated in the first transmission based on the friction coefficient and the tooth surface normal force, and then multiplying the obtained one friction torque by a predetermined third correction factor.

In the above aspect, the steering system may further include a steering member and a steering shaft that rotates integrally with the steering member. The first transmission may be a transmission that outputs the torque of the electric motor to the steering shaft or the wheel steering shaft. Another of the transmissions may be a second transmission that converts rotation of the steering shaft into axial movement of the wheel steering shaft.

In the above structure, the steering system may further include a torque detection unit that detects steering torque input from the steering member. The controller may have an axial force estimation unit that estimates an axial force acting on the wheel steering shaft based on the steering torque, the torque of the electric motor, the resultant friction torque, and the angle of the electric motor. In the above structure, the steering system may further include: a torque detection unit that detects a steering torque input from the steering member; and an axial force estimation unit that detects or estimates an axial force acting on the wheel steering shaft. The tooth surface normal force calculation section may set a first contact force calculated based on a torque, a steering torque, and an axial force of the electric motor as the tooth surface normal force when the first contact force is greater than a predetermined value; and when the first contact force is equal to or less than a predetermined value, the predetermined value is set as the tooth surface normal force.

In the above aspect, the steering system may further include: a steering member; a steering shaft that rotates integrally with the steering member; a torque detection unit that detects a steering torque input from the steering member; and an axial force estimation unit that detects or estimates an axial force acting on the wheel steering shaft. The first transmission may be a transmission that outputs the torque of the electric motor to the steering shaft or the wheel steering shaft. Another of the transmissions may be a second transmission that converts rotation of the steering shaft into axial movement of the wheel steering shaft. The tooth surface normal force calculation section may be configured to calculate a first tooth surface normal force that is a normal force acting on the tooth surface of the first transmission in the first contact state, and set a second tooth surface normal force that is a normal force acting on the tooth surface of the first transmission in the second contact state, based on the torque, the steering torque, and the axial force of the electric motor. The friction torque calculating part may be configured to calculate a composite first tooth surface normal force by multiplying the first tooth surface normal force by a predetermined fourth correction factor, calculate a composite second tooth surface normal force by multiplying the second tooth surface normal force by a predetermined fifth correction factor, calculate a composite friction coefficient by multiplying the friction coefficient by a predetermined sixth correction factor, and calculate the composite friction torque based on the composite friction coefficient and one of the composite first tooth surface normal force and the composite second tooth surface normal force having a larger absolute value.

In the above-described configuration, the tooth surface normal force calculation section may have a first tooth surface normal force correction section that corrects the first tooth surface normal force based on the resultant friction torque previously calculated by the friction torque calculation section. The friction torque calculating part may be configured to calculate a composite first tooth surface normal force by multiplying the corrected first tooth surface normal force by a fourth correction factor, calculate a composite second tooth surface normal force by multiplying the second tooth surface normal force by a fifth correction factor, calculate a composite friction coefficient by multiplying the friction coefficient by a sixth correction factor, and calculate a composite friction torque based on the composite friction coefficient and one of the composite first tooth surface normal force and the composite second tooth surface normal force having a larger absolute value.

In the above aspect, one of the transmissions may be a third transmission that is disposed on a power transmission path leading from the first transmission to the wheel rotation shaft.

Drawings

Features, advantages, and technical and industrial significance of exemplary embodiments of the present invention will be described below with reference to the accompanying drawings, in which like reference numerals represent like elements, and wherein:

fig. 1 is a schematic diagram showing a schematic structure of an electric power steering system to which a steering system according to an embodiment of the present invention is applied;

FIG. 2 is a block diagram showing an electrical structure of the ECU;

FIG. 3 is a schematic diagram showing an inertial model of an electric power steering system;

FIG. 4 is a block diagram showing the structure of an extended state observer;

fig. 5 is a graph illustrating a correlation between a meshing friction torque of a worm wheel (worm) and a worm (worm gear) and a meshing friction torque of a rack and a pinion;

FIG. 6 is a block diagram showing the structure of a friction torque estimating unit;

fig. 7 is a schematic diagram showing a model of engagement between a worm wheel and a worm.

FIG. 8 is a graph showing the estimated value of rack axial force ^ F when the steering wheel is repeatedly steered and returned_rA graph of changes over time;

FIG. 9 is a graph showing the measured value of the rack axial force and the estimated value ^ F of the rack axial force_rEtc. of the relationship between the same; and

fig. 10 is a block diagram showing a configuration of a modification of the friction torque estimating means.

Detailed Description

Embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Fig. 1 is a schematic diagram showing a schematic configuration of an electric power steering system to which a steering system according to an embodiment of the present invention is applied. This electric power steering apparatus (steering system) 1 is a column assist type electric power steering system (hereinafter referred to as "column EPS") in which an electric motor and a speed reducer are provided in a column portion.

The column type EPS1 includes a steering wheel 2 as a steering member for steering the vehicle, a wheel steering mechanism 4 that steers wheels 3 to be steered in conjunction with rotation of the steering wheel 2, and a steering assist mechanism 5 that assists the driver in steering. The steering wheel 2 and the wheel steering mechanism 4 are mechanically coupled by a steering shaft 6, a first universal joint 28, an intermediate shaft 7, and a second universal joint 29.

The steering shaft 6 includes a first shaft 8 coupled to the steering wheel 2 and a second shaft 9 coupled to the intermediate shaft 7 through a first universal joint 28. The first shaft 8 and the second shaft 9 are coupled to each other so as to be rotatable relative to each other by a torsion bar 10. A torque sensor 11 is provided around the steering shaft 6. The torque sensor 11 detects a torsion bar torque T applied to the torsion bar 10 based on the rotational positions of the first shaft 8 and the second shaft 9 relative to each other_tb. Torsion bar torque T detected by torque sensor 11_tbIs input into an Electronic Control Unit (ECU) 12.

The wheel steering mechanism 4 is formed of a rack-and-pinion mechanism including a pinion shaft 13 and a rack shaft 14 as a wheel steering shaft. Each end of the rack shaft 14 is linked to the wheel 3 to be steered by a tie rod 15 and a knuckle arm (not shown). The pinion shaft 13 is coupled to the intermediate shaft 7 by a second universal joint 29. A pinion 16 is coupled to a tip end of the pinion shaft 13.

The rack shaft 14 extends linearly in the left-right direction of the vehicle. A rack 17 that meshes with the pinion 16 is formed at an intermediate portion of the rack shaft 14 in the axial direction. The pinion 16 and the rack 17 constitute a rack-and-pinion mechanism that converts rotation of the pinion shaft 13 into axial movement of the rack shaft 14. When the steering wheel 2 is manipulated (rotated), the rotation is transmitted to the pinion shaft 13 through the steering shaft 6 and the intermediate shaft 7. Then, the rotation of the pinion shaft 13 is converted into the axial movement of the rack shaft 14 by the pinion 16 and the rack 17. The wheel 3 to be steered is steered by this axial movement.

The steering assist mechanism 5 includes an electric motor 18 and a speed reducer 19, the electric motor 18 generating a steering assist force, and the speed reducer 19 transmitting an output torque of the electric motor 18 to the wheel steering mechanism 4 after amplification. In this embodiment, the electric motor 18 is a three-phase brushless motor. The speed reducer 19 is constituted by a worm mechanism including a worm 20 and a worm wheel 21 meshing with the worm 20. The speed reducer 19 is accommodated in the gear housing 22. Hereinafter, the reduction ratio (gear ratio) of the reduction gear 19 is represented by i_wwAnd (4) showing. Reduction ratio i_wwA worm angle θ defined as a rotation angle of the worm 20_wgAnd a worm wheel angle theta which is a rotation angle of the worm wheel 21_wwRatio of (a)_wg/θ_ww)。

The worm 20 is driven to rotate by the electric motor 18. The worm wheel 21 is connected to the second shaft 9 so as to be rotatable integrally with the second shaft 9. The worm wheel 21 is driven to rotate by the worm 20. The electric motor 18 is driven according to the steering state of the driver and a command from an external controller such as an automatic steering system, and the worm 20 is driven to rotate by the electric motor 18. Thus, the worm wheel 21 is driven to rotate, and the motor torque is applied to the steering shaft 6, rotating the steering shaft 6 (second shaft 9). Then, the rotation of the steering shaft 6 is transmitted to the pinion shaft 13 through the intermediate shaft 7.

The rotation of the pinion shaft 13 is converted into the axial movement of the rack shaft 14. The wheel 3 to be steered is steered by this axial movement. Thus, the EPS1 may provide steering assist by the electric motor 18 as the worm 20 is driven to rotate by the electric motor 18. The rotation angle of the rotor of the electric motor 18 is detected by a rotation angle sensor 25 such as a resolver. The vehicle speed V is detected by a vehicle speed sensor 26. The output signal of the rotation angle sensor 25 and the vehicle speed V detected by the vehicle speed sensor 26 are input to the ECU 12. The electric motor 18 is controlled by the ECU 12.

Fig. 2 is a block diagram showing an electrical structure of the ECU 12. The ECU12 includes a microcomputer 40, a drive circuit (three-phase inverter circuit) 31 that is controlled by the microcomputer 40 and supplies power to the electric motor 18, and a current detection unit 32 that detects a current flowing through the electric motor 18 (hereinafter referred to as "motor current").

The microcomputer 40 includes a CPU and a memory (ROM, RAM, nonvolatile memory, and the like), and functions as a plurality of functional processing units by executing a predetermined program. These function processing units include a motor control unit 41, a first multiplying unit 42, a rotation angle calculating unit 43, a second multiplying unit 44, a differential calculating unit 45, a rack axial force estimating unit 46, a third multiplying unit 47, and a friction torque estimating unit 48.

The first multiplying unit 42 multiplies the motor current I detected by the current detecting unit 32_mMultiplied by the torque constant K of the electric motor 18_TAnd reduction ratio i of reduction gear 19_wwBy the motor torque T of the electric motor 18 obtained_m(＝K_T.I_m) The torque acting on the second shaft 9 (worm wheel 21) (hereinafter referred to as "drive torque i") is calculated_ww.T_m"). Rotation angle calculation unit 43 calculates rotor rotation angle θ of electric motor 18 based on the output signal of rotation angle sensor 25_m. The second multiplying unit 44 multiplies the rotor rotation angle θ_mMultiplying the reduction ratio i of the reduction gear 19_wwTo the inverse of to rotate the rotor by the angle theta_mConverted into a rotation angle of the second shaft 9 (worm wheel 21) (hereinafter referred to as "worm wheel angle θ_ww"). Differential calculation unit 45 calculates worm wheel angle θ_wwDifferentiating with respect to time to calculate the angular velocity d theta of the worm gear_ww/dt。

The motor control unit 41 is based on, for example, the vehicle speed V detected by the vehicle speed sensor 26 and the torsion bar torque T detected by the torque sensor 11_tbAnd a motor current I detected by the current detection unit 32_mmAnd the rotor rotation calculated by the rotation angle calculation unit 43Angle theta_mThe drive circuit 31 is subjected to drive control. Specifically, the motor control unit 41 is based on the torsion bar torque T_tbAnd a vehicle speed V set current command value. The current command value is a motor current I flowing through the electric motor 18_mThe target value of (2). The current command value corresponds to a target value of steering assist force (assist torque) according to the vehicle state and the steering condition. The motor control unit 41 controls the driving of the drive circuit 31 so that the motor current detected by the current detection unit 32 approaches the current command value. Therefore, it is possible to provide appropriate steering assist according to the vehicle state and the steering condition. The current instruction value is sometimes set according to a command from an external controller such as an automatic driving system.

Rack axial force estimation unit 46 is based on worm wheel angle θ_wwDriving torque i_ww.T_mTorsion bar torque T_tbAnd the friction torque (synthetic friction torque) T estimated by the friction torque estimation unit 48_fEstimating rack axial force F_r. Hereinafter, will be defined by ^ F_rRepresenting rack axial force F_rAn estimate of (a). The third multiplication unit 47 multiplies the rack axial force ^ F_rMultiplied by the gear ratio i of the rack-and- pinion mechanism 16, 17_rpTo calculate the axial force F passing through the rack_rTorque acting on the second shaft 9 (worm wheel 21) (hereinafter referred to as "torque-converted rack axial force i)_rp.^F_r”)。

Friction torque estimation unit 48 is based on worm wheel angular velocity d θ_wwDt, drive torque i_ww.T_mTorsion bar torque T_tbAnd a torque-converted rack axial force i estimated by the rack axial force estimation unit 46_rp.^F_rA composite value (composite friction torque) T is estimated, which is a composite of the friction torque generated in the speed reducer 19 and the friction torque generated in the rack-and- pinion mechanism 16, 17_f。

Next, the rack axial force estimation unit 46 and the friction torque estimation unit 48 will be described in detail. First, the rack axial force estimation unit 46 will be described. As shown in fig. 2, the rack axial force estimation unit 46 includes an adder 51 and an extended state observer 52. The adder 51 adds the driving torque i_ww.T_mTorsion bar torque T_tbAnd the friction torque T estimated by the friction torque estimation unit 48_fAnd (4) adding. Hereinafter, i_ww.T_m、T_tbAnd T_fSum of (i)_ww.T_m+T_tb+T_f) Will be composed of_inAnd (4) showing.

Extended state observer 52 is based on worm wheel angle θ_wwAddition result T of the sum adder 51_inEstimating rack axial force F_rAnd the like. The extended state observer 52 will be described in detail later. The extended state observer 52 measures the worm wheel angle θ by using an inertial model 101 of the electric power steering system 1 shown in fig. 3, for example_wwAngular speed of worm gear d theta_wwTorque converted rack axial force i_rp.F_rAnd rack axial force F_rAn estimation is performed. Hereinafter, the worm wheel angle θ_wwEstimated value of (d θ), worm wheel angular velocity_wwRack axial force i converted from estimated value and torque of dt_rp.F_rValue of and rack axial force F_rThe values of the estimates are respectively defined by ^ theta_ww、^dθ_ww/dt、i_rp.^F_rAnd ^ F_rAnd (4) showing.

The inertial model 101 is an inertial model of the device 102 assuming that the respective parts are connected by a rigid body, and adding up the inertia of the second shaft 9 and the worm wheel 21, the inertia of the rotor of the electric motor 18, the inertia of the worm 20, the inertia of the intermediate shaft (the first universal joint 28, the second universal joint 29, and the intermediate shaft 7), the inertia of the pinion shaft 13, and the mass of the rack shaft 14 (converted into the inertia of the pinion shaft 13). Torsion bar torque T_tbApplied from the steering wheel 2 to the device 102 through the torsion bar 10 and the torque converted rack axial force i_rp.F_rFrom the wheel 3 to be steered to the device 102.

Drive torque i_ww.T_mAlso applied to the device 102 through the worm gear 20. Further, a resultant torque (resultant friction torque) T in which the friction torque generated in the speed reducer 19 and the friction torque generated in the rack-and- pinion mechanism 16, 17 are combined is used_fTo the device 102. When the inertia of the device 102 is J_sThen, the equation of motion of an inertial model 101 is expressed by the following expression (1):

due to i_ww·T_m+T_tb+T_f＝T_inThus for obtaining rack axial force F_rTorque conversion value of i_rp.F_rThe expression of (c) is the following expression (2):

the state space model of the extended state observer is represented by the following expression (3):

in the expression (3), x_eIs a state variable vector, u1 is a known input vector, y is an output vector (measured value), A_eIs a system matrix, B_eIs an input matrix, C_e1Is a first output matrix, and D_eIs a feed-through matrix. x is the number of_eU1 and y are represented by the following expression (4):

A_e、B_e、C_e1and D_eExpressed by the following expression (5):

by applying the Luenberger state observer to the extended state model, the torque-converted rack axial force i can be converted as in the ordinary state observer_rp.F_r(rack axial force F_r) An estimation is performed. The observer model is expressed by the following expression (6):

in expression (6), Λ x_eRepresents x_eAn estimate of (a). L is the observer gain matrix. Y represents the estimated value of y. The observer gain matrix L is represented by the following expression (7):

in expression (7), L1, L2, and L3 are the first observer gain, the second observer gain, and the third observer gain, respectively, and ω [ rad/sec [ ]]Is the pole frequency. The pole frequency ω is set according to the load to be compensated by the observer. Rack axial force F_r(estimated value) is determined by using a state variable vector ^ x_eThe following expression (8):

in the expression (8), C_e2Is a second output matrix, and is represented by the following expression (9):

fig. 4 is a block diagram showing the structure of the extended state observer 52. The extended state observer 52 includes A_eMultiplication units 61, B_eMultiplication unit 62, C_e1Multiplication units 63 and C_e2Multiplication unit 64, D_eA multiplier 65, a first adder 66, a second adder 67, an L multiplier 68, a third adder 69, and an integrator 70. Addition result T of adder 51 of fig. 2_in(＝i_ww.T_m+T_tb+T_f) An input vector u1 equivalent to expression (6) and supplied to B_eMultiplication unit 62 and D_eAnd a multiplication unit 65. The worm wheel angle θ calculated by the second multiplying unit 44 of fig. 2_wwAn output vector (measured value) corresponding to expression (6) and is obtainedAnd supplied to the second adding section 67.

The result of the calculation by the integrating part 70 is included in the state variable vector x_eIs estimated value ax_eThe estimated value of the middle worm wheel angle ^ theta_wwEstimated value of angular speed of worm wheel ^ d theta_wwDt and torque-converted rack axial force estimate i_rp·∧F_r. At the start of the calculation, these estimated values ^ theta_ww、∧dθ_ww(dt and i)_rp·∧F_rIs for example zero. C_e1The multiplier 63 multiplies the ^ x calculated by the integrator 70_eMultiplying by C_e1To calculate C of expression (6)_e1.∧x_e. In this embodiment, C_e1.∧x_eIs the estimated value of the worm wheel angle ^ theta_ww。

C_e2The multiplier 64 multiplies the ^ x calculated by the integrator 70_eMultiplying by C_e2To calculate C of expression (8)_e2.∧x_e. In this embodiment, C_e2.∧x_eIs the estimated value of axial force of the rack ^ F_rAnd the estimated value of the axial force of the rack is inverted V F_rIs the output of the extended state observer 52. A. the_eThe multiplier 61 multiplies the ^ x calculated by the integrator 70_eMultiplying by A_eTo calculate A of expression (6)_e.∧x_e。B_eThe multiplying unit 62 multiplies T_inMultiplying by B_eTo calculate B of expression (6)_e.u1。D_eThe multiplying unit 65 multiplies T_inMultiplying by D_eTo calculate D of expression (6)_e.u1。

The first adding portion 66 will be formed by D_eD calculated by the multiplying unit 65_eU1 and C_elC calculated by the multiplying unit 63_el.∧x_e(＝∧θ_ww) And added to calculate the estimated value a y of the output vector of expression (6). In this embodiment, D _e0, and hence Λ y θ_ww. The second addition portion 67 adds the estimated value Λ y of the output vector calculated by the first addition portion 66 from the measured output vector y (═ θ y)_ww) Subtracted to calculate the difference (y-y).

The L multiplier 68 multiplies the calculation result (y-y) of the second adder 63 by L) The observer gain L is multiplied to calculate L (y-a y) of expression (6). The third adder 69 adds A to A_eCalculation result a of the multiplier 61_e.∧x_e、B_eCalculation result B of the multiplier 62_eU1, the calculation result L (y-A y) of the L multiplication section 68 is summed up to calculate d A x of expression (6)_eAnd/dt. Integral 70 for d ^ x_eThe/dt is integrated to calculate ^ x of expression (6)_e。

Next, the friction torque estimation unit 48 will be described in detail. First, the basic concept of the friction torque estimation unit 48 will be described. As will be described later, the meshing friction torque of the worm wheel 21 and the worm 20 (hereinafter referred to as a first friction torque T) can be estimated by using a meshing model of the worm wheel and the worm and a friction coefficient estimation model (LuGre model or the like)_fw&w) An estimation is performed. Similarly, the meshing friction torque of the rack 17 and the pinion 16 (hereinafter referred to as the second friction torque T) can be estimated using a rack-and-pinion meshing model and a friction coefficient estimation model_fR&P)。

First friction torque T estimated in this way_fW&WAnd a second friction torque T_fR&PAnd a composite friction torque T obtained by combining these friction torques_fW&W+T_fR&PMay be illustrated in a graph as shown in fig. 5, where motor torque is plotted on the horizontal axis and friction torque is plotted on the vertical axis. In fig. 5, W1 denotes the first friction torque T in the first contact state_fW&WThe range of (1). W2 denotes the first friction torque T in the second contact state_fW&WA range of (d); r1 represents the second friction torque T in the first contact state_fR&PA range of (d); r2 represents the second friction torque T in the second contact state_fR&PThe range of (1).

As can be seen from fig. 5, the first friction torque T in the first contact state_fW&WSecond friction torque T in contact with the first_fR&PHas a correlation therebetween, the first friction torque T in the second contact state_fW&WSecond friction torque T in second contact state_fR&PHave a correlation therebetween. Thus, it can be seen that in the first friction torqueMoment T_fW&WWith resultant friction torque T_fW&W+T_fR&PThere is also a correlation between them.

By using this correlation, the friction torque estimation unit 48 estimates the first friction torque T_fW&WThen from the estimated first friction torque T_fW&WEstimating a first friction torque T_fW&WAnd a second friction torque T_fR&PResultant friction torque T_f. Alternatively, the friction torque estimation unit 48 estimates the first friction torque T for calculation_fW&WAnd from the estimated first friction torque T_fW&WIs used to calculate the resultant friction torque T_fAnd then estimating a synthetic friction torque T using the estimated plurality of synthetic calculation elements_f。

In this embodiment, the frictional torque estimation unit 48 uses a worm wheel and worm gear mesh model to estimate the tooth surface normal force F in a state where the worm wheel 21 is in one-point contact with the worm 20_N1The resultant tooth surface normal force F in the one-point contact state of the worm wheel 21 and the worm 20 is calculated by multiplying (an example of "first tooth surface normal force" in the claims) by a predetermined one-point contact time correction factor (an example of "fourth correction factor" in the claims)_N1com(example of "synthetic first-tooth-surface normal force" in the claims).

Further, the frictional torque estimation unit 48 estimates the normal force F of the tooth surface by the two-point contact state of the worm wheel 21 and the worm 20_N2The resultant tooth surface normal force F in the two-point contact state of the worm wheel 21 and the worm 20 is calculated by multiplying (an example of "second tooth surface normal force" in the claims) by a predetermined two-point contact time correction factor (an example of "fifth correction factor" in the claims)_N2com(example of "synthetic second flank normal force" in the claims). Then, the frictional torque estimation unit 48 applies the resultant tooth surface normal force F in the one-point contact state_N1comResultant tooth surface normal force F in two-point contact state_N2comThe one with the larger absolute value is calculated as the resultant tooth surface normal force F_Ncom。

Further, the friction torque estimation unit 48 calculates the slip velocity v of the worm wheel 21 and the worm 20_sAnd by using the obtained slip velocity v_sCalculating the friction coefficient mu between the worm wheel 21 and the worm 20_W&W. Then, the friction torque estimation unit 48 estimates the friction coefficient μ_W&WMultiplying by a predetermined friction coefficient correction factor (an example of "first correction factor" or "sixth correction factor" of the claims) to calculate a composite friction coefficient μ_com. Then, the frictional torque estimation unit 48 estimates the frictional torque by using the resultant tooth surface normal force F_NcomAnd the resultant coefficient of friction mu_comCalculating the resultant Friction Torque T_f。

Fig. 6 is a block diagram showing the structure of the friction torque estimating unit 48. The friction torque estimation unit 48 includes a slip speed calculation section 81, a friction coefficient calculation section 82, a friction coefficient correction section 83, a two-point contact tooth surface normal force calculation section 84, a two-point contact tooth surface normal force correction section 85, a one-point contact tooth surface normal force calculation section 86, a one-point contact tooth surface normal force correction section 87, a maximum value selection section 88, a multiplication section 89, and a multiplication section 90.

First, the two-point-contact-tooth-surface normal-force calculation section 84, the two-point-contact-tooth-surface normal-force correction section 85, the one-point-contact-tooth-surface normal-force calculation section 86, the one-point-contact-tooth-surface normal-force correction section 87, and the maximum-value selection section 88 will be described. The two-point contact tooth surface normal force calculation section 84 and the one-point contact tooth surface normal force calculation section 86 set the normal force acting on the tooth surface in the two-point contact state and the normal force acting on the tooth surface in the one-point contact state, respectively, by using the worm wheel-worm mesh model.

FIG. 7 is a schematic diagram showing a meshing model of a worm wheel and worm, in FIG. 7, the suffix ww denotes the worm wheel, the suffix wg denotes the worm, the x-axis and the y-axis are tangent to respective meshing points on the pitch circles of the worm and worm wheel, the z-axis is in a direction along a common radial direction for the gears_wwAnd remain constant. It is also assumed that the frictional torque on the tooth surfaces is along the lead angle γ of the worm gear_wwOfActing in opposite directions.

When the system is stopped, due to the pre-pressure F₀The teeth of the worm engaged with the worm wheel are in contact with the worm wheel at two points above and below the worm wheel. This state will be referred to as a two-point contact state. Interaction force F between worm wheel and worm_c,ww、F_c,wgBy the normal force N of the tooth flanks occurring at the two contact points i-1, 2_i,xx(xx: ww, wg) and friction torque F_i,xxAnd (4) forming. Tooth surface normal force N_i,xxFrom modulus k_cStrain of the material represented by the spring (b) of (a).

When the compression amount of the upper spring or the lower spring becomes zero, the contact point is lost. A state in which one of the two contact points is lost is referred to as a one-point contact state. Friction torque T on gear tooth surface_fW&WExpressed by the following expression (10):

in expression (10), μ_W&WIs the coefficient of friction, r_wwRadius of the worm wheel, F_NIs the tooth flank normal force. Hereinafter, the tooth surface normal force F will be described_NThe method of (3). The following expression (11) represents the tooth-surface contact force Fc, which is the tooth-surface contact force F_cTo take account of the pre-pressure F₀Contact force between tooth surfaces in the case of (2):

in expression (11), J_wwIs the inertia of the worm gear, J_wgIs the inertia of the worm, J_mIs the inertia of the motor, and J_cThese inertias are converted to the sum of the inertias on the steering column shaft. T is_mIs motor torque, T_wwIs an external torque, i, acting on the second shaft 9_wwIs the gear ratio of the speed reducer 19. External torque T_wwIs torsion bar torque T_tbRack axial force i converted from torque_rp.∧F_rAnd (T)_tb+i_rp.∧F_r)。i_rpIs the gear ratio of the rack-and- pinion mechanism 16, 17.

When the contact state is a two-point contact state, the tooth surface contact force F_cEqual to or less than a predetermined value F₀/sin(β_ww)(F_c≤F₀/sin(β_ww)). In this case, the tooth surface normal force F is set based on the following expression (12a)_N. On the other hand, when the contact state is the one-point contact state, the tooth surface contact force F_cGreater than a predetermined value F₀/sin(β_ww)(F_c>F₀/sin(β_ww)). In this case, the tooth surface normal force F is set based on the following expression (12b)_N：

It is known that the tooth surface normal force F calculated based on the expression (12a) when the contact state is the two-point contact state_NIs larger than the tooth surface normal force F calculated based on the expression (12b)_NOn the contrary, when the contact state is a one-point contact state, the tooth surface normal force F calculated based on the expression (12b)_NIs larger than the tooth surface normal force F calculated based on the expression (12a)_NAbsolute value of (a). Therefore, the tooth surface normal force F calculated based on the expression (12a)_NAnd a tooth surface normal force F calculated based on the expression (12b)_NOne of the larger absolute values of is the tooth surface normal force F_N。

Referring back to fig. 6, the two-point-contact tooth surface normal force calculation section 84 converts the tooth surface normal force F represented by expression (12a)_NTooth surface normal force F set to the two-point contact state_N2. The two-point contact tooth surface normal force correcting section 85 sets the tooth surface normal force F set by the two-point contact tooth surface normal force calculating section 84_N2Multiplying the correction factor when two points contact to calculate the resultant tooth surface normal force F under the two-point contact state_N2com。

The one-point-contact tooth surface normal force calculation section 86 calculates the tooth surface normal force F to be expressed by expression (12b)_NSet to the tooth surface normal force F in the one-point contact state_N1. The one-point-contact-tooth-surface normal-force correcting section 87 sets the tooth-surface normal force F set by the one-point-contact-tooth-surface normal-force calculating section 86_N1Multiplying by a correction factor at one point contact to calculate a resultant tooth flank normal force F at one point contact_N1com。

The maximum value selecting section 88 selects the resultant tooth surface normal force F in the first contact state_N1comAnd resultant tooth flank normal force F in the second contact state_N2comOne of them having a larger absolute value as the final resultant tooth surface normal force F_NcomAnd supplies the selected one to the multiplication section 89. Next, the slip speed calculation portion 81, the friction coefficient calculation portion 82, and the friction coefficient correction portion 83 will be described. The slip speed calculation section 81 and the friction coefficient calculation section 82 calculate the friction coefficient μ of the meshing portion of the worm wheel and the worm by using the LuGre model_W&WAn estimation is performed. Coefficient of friction, mu, calculated by LuGre model_W&WBy using the slip velocity v of two objects_sAnd the following expression (13) for the flexural state variable p of the bristles:

here,. mu._cIs the coulomb friction coefficient. Mu.s_baIs the static friction coefficient. v. of_stbIs the Stribeck (Stribeck) velocity coefficient. Sigma₀Is the stiffness coefficient of the bristles. Sigma₁Is the damping coefficient of the bristles. Sigma₂Is the viscous friction coefficient. These six parameters were obtained experimentally. Slip velocity v as input to LuGre model_sCalculated based on the following expression (14):

the slip speed calculation section 81 calculates the slip speed v based on expression (14)_s. SubstitutionIllustratively, the estimated value of the angular velocity of the worm wheel ^ d θ calculated by the extended state observer 52_wwThe angular speed d θ of the worm wheel can be used as the expression (14)/dt_wwAnd/dt. In this case, the differential calculation unit 45 of fig. 2 may be omitted. The friction coefficient calculation unit 82 uses the slip speed v calculated by the slip speed calculation unit 81_sCalculating the friction coefficient mu based on expression (13)_W&W. The friction coefficient correction unit 83 adjusts the friction coefficient mu_W&WMultiplying by a predetermined coefficient of friction correction factor to calculate a resultant coefficient of friction mu_com. The synthetic friction coefficient μ calculated by the friction coefficient correction unit 83_comIs supplied to the multiplication section 89.

The multiplying portion 89 combines the tooth surface normal forces F_NcomMultiplication by the resulting coefficient of friction mu_com. The multiplier 90 multiplies the resultant friction force μ which is the result of the multiplication by the multiplier 89_com.F_NcomMultiplied by r_ww/sin(γ_ww) To calculate the resultant friction torque T_f. In this embodiment, the first friction torque T generated in the speed reducer 19 is estimated by the friction torque estimation unit 48_fW&WAnd the second friction torque T generated in the rack-and- pinion mechanism 16, 17_fR&PAnd thus the friction torque generated in the electric power steering system 1 can be correctly estimated.

In this embodiment, the first friction torque T_fW&WAnd a second friction torque T_fR&PThe resultant friction torque T of_fFriction coefficient mu based on speed reducer 19_W&WNormal force F of tooth surface in two-point contact state_N2Normal force F of tooth surface in one-point contact state_N1A preset friction coefficient correction factor, a two-point contact time correction factor and a one-point contact time correction factor. Therefore, the first friction torque T is calculated separately from using a separate engagement model_fW&WAnd a second friction torque T_fR&PThe calculated friction torques are then compared to the resultant friction torque T_fThe calculation of (2) is simplified.

In this embodiment, the normal force F of the tooth surface in the two-point contact state is obtained by_N2A value obtained by multiplying the correction factor at the time of two-point contact and a tooth surface normal force F in a state of one-point contact_N1One of the values obtained by multiplying the correction factor at one point contact with a larger absolute value is calculated as the resultant tooth surface normal force F_Ncom. Alternatively, the tooth surface normal force F in the two-point contact state_N2Tooth surface normal force F in one-point contact state_N1May be calculated as the tooth surface normal force F of the speed reducer 19_NThe normal force F of the tooth surface of the speed reducer 19 can be obtained_NThe resultant tooth surface normal force F is calculated by multiplying a predetermined normal force correction factor (an example of "second correction factor" in the claims)_Ncom。

Tooth surface normal force F in two-point contact state_N2Tooth surface normal force F in one-point contact state_N1May be calculated as the tooth surface normal force F of the speed reducer 19_NNormal force F of tooth surface of reduction gear 19 obtained_NCan be multiplied by the friction coefficient mu of the speed reducer 19_W&WThe multiplication result mu_W&W.F_NA predetermined friction torque correction factor (an example of "third correction factor" in the claims) may be multiplied to calculate the synthetic friction torque T_f。

FIG. 8 is a graph showing the estimated value of rack axial force ^ F estimated by the rack axial force estimation unit 46 when the steering wheel is repeatedly steered and returned_rEtc. over time. FIG. 9 is a graph showing the measured value of the rack axial force and the estimated value ^ F of the rack axial force_rEtc. of the relationship between the same. In fig. 8 and 9, a curve S1 represents the driving torque i_ww·T_mAnd torsion bar torque T_tbAnd (i)_ww·T_m+T_tb) A graph of values obtained converted into forces in the rack axial direction. The curve S2 represents the estimated value Λ F of the rack axial force estimated by the rack axial force estimation unit 46_rThe figure (a). The curve S3 is a graph representing the measured value of the rack axial force. The measured value of the rack axial force is plotted on the abscissa in fig. 9.

As can be seen from fig. 8 and 9, in this embodiment, friction in the electric power steering system 1 is greatly compensated. Furthermore, as can be seen in fig. 8, the error in the estimated rack axial force is greater during steering wheel return than during steering wheel turn. Possible causes for this are: a difference in friction torque between when the steering wheel is steered by applying a force from the motor side and when the steering wheel is returned from the steering column shaft side. Therefore, the present inventors invented a friction torque estimation unit (hereinafter referred to as a modification of the friction torque estimation unit) that can reduce a rack axial force estimation error as compared with the friction torque estimation unit 48 of fig. 6.

Fig. 10 is a block diagram showing a configuration of a modification of the friction torque estimating means. Those portions in fig. 10 that correspond to the portions of fig. 6 described above may be denoted by the same reference numerals as those in fig. 6. The friction torque estimation unit 48A differs from the friction torque estimation unit 48 of fig. 6 in that a correction term calculation section 91 and an addition section 92 (an example of "first tooth surface normal force correction section" in the claims) are added.

The correction term calculation unit 91 calculates the previous calculation result (μ) of the multiplication unit 89_com·F_Ncom)_n-1Multiplying by a predetermined correction gain A to calculate a correction term A (mu)_com·F_Ncom)_n-1. The addition unit 92 adds the tooth surface normal force F in the one-point contact state set by the one-point contact tooth surface normal force calculation unit 86_N1Adding a correction term A (mu)_com·F_Ncom)_n-1To correct the normal force F of the tooth surface in one point contact state_N1. Tooth surface normal force (F) in one-point contact state after correction_N1+A·(μ_com·F_Ncom)_n-1) Is provided to the one-point-contact-tooth-surface normal-force modification portion 87.

Correction term A (mu)_com·F_Ncom)_n-1Is set so that the tooth surface normal force F in a one-point contact state during steering of the steering wheel_N1Becomes smaller and the tooth surface normal force F in the one-point contact state during the return of the steering wheel_N1Becomes larger in absolute value. Thus, friction is synthesized during steering of the steering wheelTorque T_fBecomes smaller and the resultant friction torque T during the return of the steering wheel_fBecomes larger in absolute value. Therefore, the rack axial force estimation error during steering and returning of the steering wheel can be reduced.

Also in this modification, the tooth surface normal force F in the two-point contact state_N2Tooth surface normal force (F) in one-point contact state after correction_N1+A·(μ_com·F_Ncom)_n-1) May be calculated as the tooth surface normal force F of the speed reducer 19_NNormal force F of tooth surface of reduction gear 19 obtained_NThe resultant tooth flank normal force F may be calculated by multiplying a predetermined normal force correction factor_Ncom。

Tooth surface normal force F in two-point contact state_N2Tooth surface normal force (F) in one-point contact state after correction_N1+A·(μ_com·F_Ncom)_n-1) May be calculated as the tooth surface normal force F of the speed reducer 19_NNormal force F of tooth surface of reduction gear 19 obtained_NCan be multiplied by the friction coefficient mu of the speed reducer 19_W&WThe multiplication result mu can then be calculated_W&W·F_NMultiplying a predetermined friction torque correction factor to calculate a resultant friction torque T_f。

In the above embodiment, the friction torque T is synthesized_fBy using the torque for calculating the engagement friction torque (first friction torque T) of the worm wheel 21 and the worm 20_fW&W) And a plurality of correction factors. Alternatively, the meshing friction torque (second friction torque T) of the rack 17 and the pinion 16 may be calculated by using the calculation result_fR&P) And a plurality of correction factors estimate the resultant friction torque T_f. Furthermore, the resultant friction torque T_fIt is also possible to estimate the second friction torque T_fR&PAnd using the estimated second friction torque T_fR&PAnd a correction factor.

In the above embodiments, the example in which the present invention is applied to the column type EPS has been shown. However, the invention may also be applied to any other EPS having a plurality of transmission devices than the column type EPS, including a rack assist type EPS in which the output of an electric motor is applied to a rack shaft, and the like. One example of such an EPS is a rack parallel type EPS in which a rotational force of an electric motor disposed parallel to a rack shaft is transmitted to a ball screw mechanism mounted on the rack shaft through a pulley and a belt, and the rack shaft is thus moved. The EPS has two transmission devices, one is a transmission mechanism composed of a belt wheel and a belt, and the other is a ball screw mechanism.

Another example is a double-pinion type EPS in which a second pinion shaft that is not coupled to a steering shaft is provided in addition to the pinion shaft (hereinafter referred to as a first pinion shaft) of fig. 1, and an electric motor is coupled to the second pinion shaft through a speed reducer. The EPS has three transmissions, one being a speed reducer that transmits the rotational force of an electric motor to a second pinion shaft, and the others being two rack-and-pinion mechanisms. A rack-and-pinion mechanism composed of a second pinion shaft and a rack shaft serves as a third transmission device that is disposed on a power transmission path leading from the speed reducer to the rack shaft.

The present invention can also be applied to a steer-by-wire system. Various other design changes may be made to the present invention within the scope of the matters recited in the claims.