Control device, electric power steering device, and control method

Abstract

A control device, an electric power steering device, and a control method, the control device having: a reaction force control section that generates an input torque input to a control object and controls a reaction force transmitted to a steering operator; and an assist control section that generates a correction torque that corrects the input torque based on an output of the control object and a nominal model. The assist control section has: a high-pass filter having a first cutoff frequency; a low-pass filter having a second cutoff frequency higher than the first cutoff frequency; and a disturbance compensation value calculation section that, when a transfer function of the low-pass filter is Q(s) and a transfer function of the high-pass filter is HPF(s), restricts a transfer function of the control object to a transfer function of the nominal model in a frequency band in which a gain of a gain characteristic of Q(s)·HPF(s) is 1. The disturbance compensation value calculation section calculates a disturbance compensation value that compensates for at least a part of an automatic return torque generated in the control object. The correction torque includes the disturbance compensation value.

Description

Technical Field

[0001] This invention relates to a control device, an electric power steering system, and a control method. Background Technology

[0002] Electric power steering systems installed in vehicles are known. For example, the electric power steering system described in Patent Document 1 has a motor control device that includes an interference observer for estimating interference torque.

[0003] Patent Document 1: Japanese Patent Application Publication No. 2018-183046

[0004] In electric power steering systems like those described above, there is a need to improve the steering feel experienced by the driver steer the vehicle. However, in the control system of an electric power steering system, stability, disturbance suppression characteristics, and responsiveness are all intertwined. Therefore, it is difficult to adjust these factors effectively, making it challenging to improve the steering feel experienced by the driver. Summary of the Invention

[0005] In view of the above, one of the objects of the present invention is to provide a control device that can improve the steering feel felt by the helmsman, an electric power steering device having such a control device, and a control method that can improve the steering feel felt by the helmsman.

[0006] One embodiment of the control device of the present invention controls at least a portion of a steering mechanism, including a motor, as the controlled object. The steering mechanism has: an input shaft connected to a steering wheel operated by a helmsman; an output shaft connected to the input shaft via a torsion bar; and the motor connected to the output shaft. The control device further includes: a reaction force control unit that generates an input torque to be input to the controlled object based on a torsion bar torque to be generated in the torsion bar, and controls the reaction force transmitted from the steering wheel to the helmsman; and an auxiliary control unit that generates a correction torque to correct the input torque based on the output and nominal model of the controlled object. The auxiliary control unit includes: a high-pass filter having a first cutoff frequency; a low-pass filter having a second cutoff frequency higher than the first cutoff frequency; and an interference compensation value calculation unit. The auxiliary control unit is configured such that, when the transfer function of the low-pass filter is set to Q(s) and the transfer function of the high-pass filter is set to HPF(s), in the frequency band where the gain is 1 in the gain characteristic of Q(s)·HPF(s), the transfer function of the controlled object is constrained to the transfer function of the nominal model. The interference compensation value calculation unit calculates an interference compensation value to compensate for at least a portion of the self-correcting torque generated by the controlled object. The correction torque includes the interference compensation value.

[0007] One embodiment of the electric power steering device of the present invention includes the control device and the steering mechanism described above.

[0008] One aspect of the control method of the present invention controls at least a portion of a steering mechanism, including a motor, as the controlled object. The steering mechanism has: an input shaft connected to a steering wheel operated by a helmsman; an output shaft connected to the input shaft via a torsion bar; and the motor connected to the output shaft. The control method includes: generating an input torque to be input to the controlled object based on a torsion bar torque to be generated by the torsion bar, and controlling the reaction force transmitted from the steering wheel to the helmsman; and generating a correction torque to correct the input torque based on the output and nominal model of the controlled object. Generating the corrected torque includes the following: when the transfer function of a high-pass filter having a first cutoff frequency is set to HPF(s), and the transfer function of a low-pass filter having a second cutoff frequency higher than the first cutoff frequency is set to Q(s), in the frequency band where the gain is 1 in the gain characteristic of Q(s)·HPF(s), constraining the transfer function of the controlled object to the transfer function of the nominal model; and calculating an interference compensation value to compensate for at least a portion of the auto-centering torque generated by the controlled object. The corrected torque includes the interference compensation value.

[0009] According to one aspect of the present invention, it is possible to improve the steering feel experienced by a helmsman operating the steering wheel of a vehicle equipped with an electric power steering system. Attached Figure Description

[0010] Figure 1 This is a schematic diagram illustrating the electric power steering device of the first embodiment.

[0011] Figure 2 This is a block diagram illustrating the configuration of a control device according to one embodiment.

[0012] Figure 3 This is a functional block diagram illustrating the function of a processor in a control device according to one embodiment.

[0013] Figure 4 This is a graph illustrating the gain characteristics of the complementary sensitivity function and the gain characteristics of the reciprocal of the modeling error between the transfer function of the controlled object and the transfer function of the nominal model.

[0014] Figure 5 This is a graph showing an example of the measured results of steering angle and torsional torque without applying model follow control.

[0015] Figure 6This is a graph showing an example of the measured results of steering angle and torsional torque when model follow control is applied.

[0016] Figure 7 This is another example of a graph showing the measured results of steering angle and torsional torque without applying model follow control.

[0017] Figure 8 This is another example of a graph showing the measured results of steering angle and torsional torque when model follow control is applied.

[0018] Figure 9 This is another example of a graph showing the measured results of steering angle and torsional torque without the application of model follow control.

[0019] Figure 10 This is another example of a graph showing the measured results of steering angle and torsional torque under the application of model follow control.

[0020] Figure 11 This is another example of a graph showing the measured results of steering angle and torsional torque without applying model follow control.

[0021] Figure 12 This is another example of a graph showing the measured results of steering angle and torsional torque under the application of model follow control.

[0022] Label Explanation

[0023] 100: Control device; 210: Reaction force control unit; 230: Auxiliary control unit; 232: Low-pass filter; 233: High-pass filter; 250: Friction compensation value calculation unit; 260: Interference compensation value calculation unit; 280: Status feedback unit; 521: Steering wheel; 524a: Input shaft; 524b: Output shaft; 530: Steering mechanism; 543: Motor; 546: Torque bar; 560: Controlled object; 1000: Electric power steering device; Cf1: First cutoff frequency; Cf2: Second cutoff frequency; P(s): Transfer function of the controlled object; P n (s): Transfer function of the nominal model; T(s): Complementary sensitivity function; T a Differential torque; T f Correction torque; T h Steering torque (torsion bar torque); T p Torque; T r Input torque; V d Interference compensation value; V f Friction compensation value; V s : State compensation value; θa: Rotation angle. Detailed Implementation

[0024] Hereinafter, embodiments of the control device, electric power steering device, and control method of the present invention will be described with reference to the accompanying drawings. However, sometimes unnecessary detailed descriptions are omitted. For example, detailed descriptions of matters already well known and repetitive descriptions of substantially the same structures are sometimes omitted. This is to avoid making the following description unnecessarily lengthy and to facilitate understanding by those skilled in the art.

[0025] The following embodiments are illustrative, and the control device, electric power steering device, and control method of the present invention are not limited to the following embodiments. For example, the values, steps, and order of steps shown in the following embodiments are merely examples, and various changes can be made as long as there is no technical contradiction. The embodiments or examples described below are merely illustrative, and various combinations can be made as long as there is no technical contradiction.

[0026] Figure 1 The electric power steering device 1000 of this embodiment shown is mounted in a vehicle. For example... Figure 1 As shown, the electric power steering system 1000 includes a steering mechanism 530 and a control device 100. The steering mechanism 530 includes a steering mechanism section 520 and an auxiliary mechanism section 540. The electric power steering system 1000 controls the auxiliary mechanism section 540 via the control device 100, thereby generating an auxiliary torque that assists the steering torque T generated in the steering mechanism section 520 by the driver steering the steering wheel 521. h This auxiliary torque reduces the driver's workload when operating the steering wheel 521. The driver of the vehicle is the one who controls the steering wheel 521.

[0027] The steering mechanism 520 includes a steering wheel 521, a steering shaft 522, universal couplings 523A and 523B, an input shaft 524a, an output shaft 524b, a rack and pinion mechanism 525, a rack shaft 526, left and right ball joints 552A and 552B, tie rods 527A and 527B, steering knuckles 528A and 528B, and left and right steering wheels 529A and 529B. That is, the steering mechanism 530 includes a steering wheel 521, a steering shaft 522, universal couplings 523A and 523B, an input shaft 524a, an output shaft 524b, a rack and pinion mechanism 525, a rack shaft 526, left and right ball joints 552A and 552B, tie rods 527A and 527B, steering knuckles 528A and 528B, and left and right steering wheels 529A and 529B.

[0028] The steering shaft 522 is an extension of the steering wheel 521 used by the helmsman for steering. One end of the input shaft 524a is connected via universal couplings 523A and 523B to the end of the steering shaft 522 opposite to the side connected to the steering wheel 521. Thus, the steering wheel 521 is connected to the input shaft 524a via universal couplings 523A and 523B and the steering shaft 522. The output shaft 524b is connected to the input shaft 524a via a torsion bar 546, described later. More specifically, one end of the output shaft 524b is connected to the other end of the input shaft 524a via the torsion bar 546. The other end of the output shaft 524b is connected to the rack shaft 526 via a rack and pinion mechanism 525.

[0029] The input shaft 524a and the output shaft 524b are coaxially configured. The input shaft 524a and the output shaft 524b are rotatable about the same central axis. The input shaft 524a and the output shaft 524b are rotatable relative to each other within the range of torsion of the torsion bar 546 described later.

[0030] The auxiliary mechanism 540 includes a steering torque sensor 541, a rudder angle sensor 542, a motor 543, a reduction gear 544, an inverter 545, and a torsion bar 546. That is, the steering mechanism 530 includes the steering torque sensor 541, rudder angle sensor 542, motor 543, reduction gear 544, inverter 545, and torsion bar 546. The torsion bar 546 connects the input shaft 524a and the output shaft 524b. The torsion bar 546 is coaxially arranged with the input shaft 524a and the output shaft 524b. In the following description, the imaginary axis passing through the common central axis of the input shaft 524a, the output shaft 524b, and the torsion bar 546 is referred to as the rotation axis R. The torsion bar 546 is capable of twisting about the rotation axis R.

[0031] The steering torque sensor 541 detects the steering torque T in the steering mechanism section 520 by detecting the amount of torsion of the torsion bar 546 about the rotation axis R. h Steering torque T h The torque generated by the torsion bar 546 is the torsional moment about the rotation axis R. The steering angle sensor 542 can detect the rotation angle θa of the input shaft 524a about the rotation axis R. The rotation angle θa of the input shaft 524a is equal to the steering angle of the steering wheel 521. That is, the steering angle sensor 542 can detect the steering angle of the steering wheel 521 by detecting the rotation angle θa of the input shaft 524a. Based on the steering torque sensor 541 and the steering angle sensor 542, the rotation angle θb of the output shaft 524b can be detected.

[0032] Inverter 545 converts DC power into analog sine waves (U-phase, V-phase, and W-phase) three-phase AC power, based on the motor drive signal input from control device 100, and supplies it to motor 543. Motor 543 is connected to output shaft 524b via reduction gear 544. Three-phase AC power is supplied from inverter 545 to motor 543. Motor 543 is, for example, an embedded magnet synchronous motor (IPMSM), a surface magnet synchronous motor (SPMSM), or a switched reluctance motor (SRM). Motor 543 generates a steering torque T by receiving three-phase AC power from inverter 545. h The corresponding auxiliary torque. The motor 543 transmits the generated auxiliary torque to the output shaft 524b via the reduction mechanism 544.

[0033] The control device 100 controls a control object 560 in the steering mechanism 530, which has at least an input shaft 524a, an output shaft 524b, and a motor 543. In this embodiment, the control object 560 includes a steering wheel 521, universal couplings 523A and 523B, an input shaft 524a, an output shaft 524b, a torsion bar 546, a motor 543, and a reduction mechanism 544. Since the control object 560 includes an input shaft 524a and an output shaft 524b that can rotate relative to each other via the torsion bar 546, the motion of the control object 560 cannot be described by a simple equation of motion for a single inertial system. The control object 560 varies between one and two inertial systems depending on the strength with which the helmsman grips the steering wheel 521. The harder the helmsman grips the steering wheel 521, the closer the control object 560 is to a single inertial system. The weaker the helmsman grips the steering wheel 521, the closer the controlled object 560 is to two inertial systems.

[0034] The control device 100 is electrically connected to the inverter 545. Based on detection signals from the steering torque sensor 541, the rudder angle sensor 542, and the vehicle speed sensor 300 mounted on the vehicle, the control device 100 generates a motor drive signal and outputs it to the inverter 545. The control device 100 controls the rotation of the motor 543 via the inverter 545, thereby controlling the controlled object 560. More specifically, the control device 100 controls the switching operations of multiple switching elements in the inverter 545. Specifically, the control device 100 generates control signals to control the switching operations of each switching element and outputs them to the inverter 545. Each switching element is, for example, a MOSFET. In the following description, the control signals controlling the switching operations of each switching element will be referred to as "gate control signals".

[0035] Control device 100 is based on steering torque T hThe torque command value is generated, for example, by vector control to control the torque and rotational speed of motor 543. Vector control is a method of decomposing the current flowing in motor 543 into current components that contribute to torque generation and current components that contribute to magnetic flux generation, and independently controlling the mutually perpendicular current components. Control device 100 is not limited to vector control and can perform other closed-loop control. The rotational speed of motor 543 is represented, for example, by the rotational speed of the rotor in 1 minute [rpm] or the rotational speed of the rotor in 1 second [rps].

[0036] Alternatively, the steering torque T can be directly input to the control device 100 from the steering torque sensor 541. h The control device 100 can also calculate the steering torque T based on the output value of the steering torque sensor 541. h The steering angle of the steering wheel 521 can be directly input to the control device 100 from the steering angle sensor 542, or the control device 100 can calculate the steering angle value based on the output value of the steering angle sensor 542.

[0037] Furthermore, the control device 100 and the motor 543 are modularized and manufactured and sold as a motor module. The motor module, comprising the motor 543 and the control device 100, is suitable for use in the electric power steering system 1000. Additionally, the control device 100 can be manufactured and sold independently of the motor 543 as a control device for controlling the electric power steering system 1000.

[0038] Figure 2 A typical example of the structure of the control device 100 in this embodiment is shown. The control device 100 includes, for example, a power supply circuit 111, an angle sensor 112, an input circuit 113, a communication I / F 114, a drive circuit 115, a ROM 116, and a processor 200. The control device 100 can be implemented as a printed circuit board (PCB) on which these electronic components are mounted.

[0039] Regarding the processor 200, the vehicle speed sensor 300, steering torque sensor 541, and rudder angle sensor 542, all mounted on the vehicle, are communicatively connected to the processor 200. The vehicle speed sensor 300 transmits vehicle speed to the processor 200. The steering torque T is transmitted from the steering torque sensor 541 to the processor 200. h The steering angle is sent from the rudder angle sensor 542 to the processor 200.

[0040] Processor 200 is a semiconductor integrated circuit, also known as a central processing unit (CPU) or microprocessor. Processor 200 sequentially executes a computer program stored in ROM 116, which describes a set of commands for controlling the motor drive, to achieve the desired processing. Control device 100, in addition to or replacing processor 200, may include an FPGA (Field Programmable Gate Array), GPU (Graphics Processing Unit), ASIC (Application Specific Integrated Circuit), ASSP (Application Specific Integrated Circuit), or a combination of two or more circuits selected from these circuits, all equipped with a CPU. Processor 200 generates a PWM (Pulse Width Modulation) signal based on the actual current value and the set current command value such as the rotation angle of the motor 543 rotor, and outputs the PWM signal to drive circuit 115.

[0041] The power supply circuit 111 is connected to an external power source (not shown). The power supply circuit 111 generates the DC voltage required by various parts of the control device 100. The DC voltage generated in the power supply circuit 111 is, for example, 3V or 5V.

[0042] Angle sensor 112 detects the rotation angle of the rotor of motor 543 and outputs it to processor 200. Angle sensor 112 can be a rotary transformer, a Hall element such as a Hall IC, or an MR sensor with a magnetoresistive element. Processor 200 can calculate the angular velocity ω [rad / s] of motor 543 based on the electrical angle θm of motor 543 obtained from angle sensor 112. Alternatively, control device 100 can replace angle sensor 112 with a speed sensor capable of detecting the rotational angular velocity of motor 543 and an acceleration sensor capable of detecting the rotational angular acceleration of motor 543.

[0043] The motor current value detected by a current sensor (not shown) is input to input circuit 113. In the following description, the motor current value detected by the current sensor (not shown) will be referred to as the "actual current value". Input circuit 113 converts the level of the input actual current value to the input level of processor 200 as needed, and outputs the actual current value to processor 200. A typical example of input circuit 113 is an analog-to-digital converter circuit.

[0044] The communication I / F 114 is, for example, an input / output interface for transmitting and receiving data based on the vehicle's control area network (CAN).

[0045] Typically, the drive circuit 115 is a gate driver or pre-driver. The drive circuit 115 generates gate control signals based on the PWM signal and provides these signals to the gates of the multiple switching elements in the inverter 545. For example, when the motor 543 being driven is a motor capable of being driven at low voltage, the drive circuit 115 as a gate driver may not always be necessary. In this case, the gate driver function in the drive circuit 115 can be implemented in the processor 200.

[0046] ROM 116 is electrically connected to processor 200. ROM 116 is, for example, a writable memory, a rewritable memory, or a read-only memory. Examples of writable memory include PROM (Programmable Read Only Memory). Examples of rewritable memory include flash memory and EEPROM (Electrically Erasable Programmable Read Only Memory). ROM 116 stores a control program containing a set of commands for enabling processor 200 to control motor drive. For example, the control program stored in ROM 116 is temporarily expanded into RAM (not shown) during boot.

[0047] Figure 3 An example of the function blocks of the processor 200 in this embodiment is shown. The processor 200, which functions as a computer, uses each function block to sequentially execute the processing or tasks required to control the motor 543. Figure 3 The functional blocks of the processor 200 shown can be installed in the processor 200 as software such as firmware, as hardware, or as both software and hardware. The processing of each functional block in the processor 200 is typically described in a computer program as a software module and stored in ROM 116. However, when using an FPGA or similar device, all or part of these functional blocks can be installed as a hardware accelerator. Furthermore, the control method of the control device 100 in this embodiment can be implemented by installing it in a computer and causing the computer to perform the desired action.

[0048] The processor 200 includes a reaction force control unit 210, an auxiliary control unit 230, a status feedback unit 280, a subtractor SU1, an adder AD1, and an adder AD2. That is, the control device 100 includes the reaction force control unit 210, the auxiliary control unit 230, the status feedback unit 280, the subtractor SU1, the adder AD1, and the adder AD2. In other words, the processor 200 of the control device 100 is equipped with functions corresponding to the reaction force control unit 210, the auxiliary control unit 230, the status feedback unit 280, the subtractor SU1, the adder AD1, and the adder AD2, respectively.

[0049] The steering torque T detected by steering torque sensor 541 h The input is sent to the reaction force control circuit 210. The reaction force control unit 210 is based on the steering torque T. h That is, the input torque T to be input to the controlled object 560 is generated by the torque of the torsion bar 546. r Input torque T r This is the target torque of motor 543, the torque command value. The reaction force control circuit 210 generates the input torque T. r The torque of motor 543 is controlled to control the reaction force transmitted from steering wheel 521 to the helmsman. The reaction force control circuit 210 controls the steering torque T when the steering frequency or speed is within a specified range. h Phase compensation is applied to generate the input torque T r The steering frequency is the frequency of the steering angle that changes based on the driver's operation of the steering wheel 521. The steering speed is the speed of the steering angle that changes based on the driver's operation of the steering wheel 521. Figure 3 The illustrated reaction force control circuit 210 has a basic auxiliary operation circuit 211 and a phase compensator 212.

[0050] Basic auxiliary computing unit 211 obtains steering torque T h And vehicle speed. The basic auxiliary computing unit 211 is based on steering torque T. h And the vehicle speed generates the basic auxiliary torque. For example, the basic auxiliary calculation unit 211 has a specified steering torque T. h A lookup table (LUT) relating vehicle speed and basic auxiliary torque is used. The basic auxiliary calculation unit 211 can refer to this lookup table to perform calculations based on the steering torque T. h The basic auxiliary torque is determined based on the vehicle speed and the corresponding relationship. The basic auxiliary calculation unit 211 can determine the basic auxiliary torque based on the change in the basic auxiliary torque relative to the steering torque T. h The slope of the ratio of the change in the base auxiliary gain is used to determine the base auxiliary gain.

[0051] In this embodiment, the phase compensator 212 adjusts the auxiliary gain within a range of acceptable steering frequencies when the driver operates the steering wheel 521, compensating for the stiffness of the torsion bar 546. The acceptable steering frequency range is, for example, below 5Hz. The phase compensator 212 can also adjust the steering torque T when the steering frequency is below 5Hz. h The torque of a torsion bar can be applied, for example, by first-order phase compensation. First-order phase compensation can be represented, for example, by the transfer function of equation (1).

[0052]

[0053] In equation (1), s is the Laplace transform, f1 is the frequency of the zeros of the transfer function [Hz], and f2 is the frequency of the poles of the transfer function [Hz]. A graph with gain or loop gain as the vertical axis and the logarithm of frequency as the horizontal axis is called a gain plot. In a gain plot, the zero is the intersection of the gain curve and the horizontal axis representing 0 dB, and the pole is the maximum point of the gain curve. For example, phase lead compensation can be applied by making the frequency of the pole higher than the frequency of the zero. The greater the interval between the frequency of the pole and the frequency of the zero, the greater the phase lead.

[0054] Phase compensator 212 generates input torque T based on the basic auxiliary torque output from basic auxiliary arithmetic unit 211 and basic auxiliary gain. r For example, phase compensator 212 is a stabilizing compensator capable of applying stable phase compensation to the basic auxiliary torque. Phase compensator 212 can have a transfer function of order two or higher with frequency characteristics that vary according to the basic auxiliary gain. The transfer function of order two or higher is represented using parameters of responsiveness and attenuation ratio (damping). The transfer function of order two or higher can be represented, for example, by equation (2). By setting the order of the transfer function to two, damping can be applied to the characteristics of the transfer function. By changing the damping, the phase characteristics can be adjusted.

[0055]

[0056] In equation (2), s is the Laplace transform, ω1 is the frequency of the zeros of the transfer function, ω2 is the frequency of the poles of the transfer function, ζ1 is the attenuation ratio of the zeros, and ζ2 is the attenuation ratio of the poles. The frequency of the poles ω2 is lower than the frequency of the zeros ω1.

[0057] The auxiliary control unit 230 generates a parameter for the input torque T based on the output and nominal model of the controlled object 560. r Correction torque T f In this embodiment, the correction torque T f It is fed back to the input torque T rThe feedback torque. The nominal model is the internal model used as a model of the constrained controlled object 560 when controlling the controlled object 560. The nominal model will be described in detail later. In this embodiment, the auxiliary control unit 230 is a model following controller configured to perform model following control. The specific structure of the auxiliary control unit 230 will be described in detail later.

[0058] Subtractor SU1 receives input torque T r Subtract the correction torque T output from the auxiliary control unit 230 f The output from subtractor SU1 is input to adder AD1 and auxiliary control unit 230. Adder AD1 adds the output from state feedback unit 280 to the output from subtractor SU1, and outputs the resulting value to adder AD2. Adder AD2 adds disturbance torque T to the output from adder AD1. d The obtained value is then output to control object 560.

[0059] Interference torque T d This is the difference between the actual output torque of motor 543 and the ideal output torque of motor 543. Disturbance torque T d Includes the disturbance torque applied externally to the controlled object 560. Disturbance torque T d Examples include excess torque caused by friction and loosening due to mechanical elements such as the motor 543 and the reduction gear 544, torque fluctuations generated by the motor 543, automatic return torque, and interference torque that may occur when driving on unpaved, loose roads or gravel roads. Automatic return torque refers to the torque acting on the steering wheel 521 in the return direction due to the elasticity of the tires twisted when the steering wheel 521 is turned.

[0060] In this embodiment, the auxiliary control unit 230 calculates the angular velocity ω based on the rotation angle θa of the input shaft 524a. θ Generate correction torque T f And feed back to the input torque T r Angular velocity ω θ This is equivalent to the angular velocity of motor 543 theoretically calculated based on the rotation angle θa of input shaft 524a. For example, immediately after the operator begins to rotate steering wheel 521, input shaft 524a rotates along with steering wheel 521. However, there is a situation where motor 543 has not yet started driving and output shaft 524b has not rotated. In this case, the actual angular velocity ω of motor 543 is zero, but theoretically, if input shaft 524a rotates, motor 543 will also rotate, and output shaft 524b will also rotate. Angular velocity ω θ This is a value equivalent to the angular velocity of motor 543 when motor 543 is theoretically rotating. Therefore, the angular velocity ωθ Sometimes the actual angular velocity ω of motor 543 differs from the actual angular velocity ω. Furthermore, the angular velocity ω... θ The calculated rotation angle θa can be the value detected by the rudder angle sensor 542, or it can be the value calculated based on the rotation angle θb of the output shaft 524b.

[0061] The auxiliary control unit 230 includes an inverse nominal model 231, a low-pass filter 232, a high-pass filter 233, an auxiliary adjustment unit 270, a subtractor SU2, and an adder AD3. The high-pass filter 233 has a first cutoff frequency Cf1. The first cutoff frequency Cf1 is, for example, 2Hz or higher and 10Hz or lower, preferably 5Hz or higher and 7Hz or lower.

[0062] The low-pass filter 232 has a second cutoff frequency Cf2 that is higher than the first cutoff frequency Cf1. The second cutoff frequency Cf2 is, for example, 3Hz or higher and 50Hz or lower. However, the upper limit of the second cutoff frequency Cf2 can be set to approximately 140Hz or higher and 200Hz or lower. The low-pass filter 232 has an order of 3 or higher. The low-pass filter 232 can be composed of multiple low-pass filters. The low-pass filter 232 and the high-pass filter 233 are connected in series.

[0063] The auxiliary control unit 230 is configured such that, when the transfer function of the low-pass filter 232 is set to Q(s) and the transfer function of the high-pass filter 233 is set to HPF(s), the transfer function P(s) of the controlled object 560 is constrained to the nominal model transfer function P in the frequency band where the gain is 1 in the gain characteristic of Q(s)·HPF(s). n (s). Q(s)·HPF(s) is the complementary sensitivity function T(s) of the inner loop formed by the auxiliary control unit 230. For example... Figure 4 As shown, Q(s)·HPF(s), i.e., the complementary sensitivity function T(s), has a gain of 0dB in the frequency band where the frequency f is above the first cutoff frequency Cf1 and below the second cutoff frequency Cf2, i.e., the gain in the transfer function is 1. Figure 4 The absolute value of the complementary sensitivity function T(s) is shown in the figure. Furthermore, in this specification, "the transfer function of the controlled object is constrained to be the transfer function of the nominal model" means, for example, when observing the input-output relationship, that the controlled object is controlled such that the transfer function of the controlled object appears to be the transfer function of the nominal model.

[0064] The inverse nominal model 231 is the inverse model of the prescribed nominal model (equipment model) used to constrain the controlled object 560. In this embodiment, the transfer function P of the prescribed nominal model... n (s) is represented by the following equation (3). The transfer function P of the inverse nominal model 231 n-1 (s) is represented by the following equation (4).

[0065]

[0066]

[0067] In equations (3) and (4), s is the Laplace transform, J STGn B is a parameter representing the moment of inertia of the nominal model. STGn ω is a parameter representing the viscous friction coefficient of the nominal model. 1n It is the transfer function P n The frequency of the zero point of (s), ω 2n It is the transfer function P n The frequency of the poles of (s), ζ 1n It is the transfer function P n The decay ratio at the zero point of (s), ζ 2n It is the transfer function P n The attenuation ratio of the poles of (s).

[0068] In this embodiment, the nominal model is a model with frequency characteristics between one inertial system and two inertial systems. The transfer function P of the nominal model described above is... n Equation (3) of (s) is the equation representing two inertial systems with the addition of a decay term. In equation (3) above, the decay term is 2ζ. 1n ω 1n s and 2ζ 2n ω 2n s. The expression after removing these attenuation terms from equation (3) becomes the expression representing the two inertial systems. In this embodiment, the transfer function P of the nominal model n The number of times (s) is 3.

[0069] In this embodiment, the nominal model is a model that takes into account the mechanical characteristics when the helmsman steers the steering wheel 521. As described above, for the controlled object 560, the stronger the helmsman grips the steering wheel 521, the closer it is to a single inertial system; the weaker the grip, the closer it is to a two-inertial system. Therefore, the transfer function P(s) of the controlled object 560 varies between a single inertial system and a two-inertial system depending on how much force is applied to the steering wheel 521 from the helmsman's arm. In this embodiment, by setting the nominal model to have frequency characteristics between a single inertial system and a two-inertial system, the transfer function P(s) of the nominal model can be made to remain constant regardless of whether the state of the controlled object 560 is a single inertial system or a two-inertial system. nThe modeling error Δ(s) between the transfer function P(s) of the controlled object 560 and P(s) will not become too large. Therefore, regardless of how the helmsman steers the steering wheel 521, the controlled object 560 can be appropriately controlled using the nominal model. Thus, in this embodiment, the nominal model is a model that takes into account the mechanical characteristics of the controlled object 560 based on how the helmsman holds the steering wheel 521. By having such a nominal model as its internal model, the control device 100 is able to perform appropriate control of the controlled object 560.

[0070] It should be noted that in this specification, "the nominal model is a model that takes into account the mechanical characteristics of the helmsman when he / she operates the steering wheel" means, for example, that the nominal model is capable of compensating for at least a portion of the effects on the controlled object caused by the mechanical characteristics of the helmsman operating the steering wheel. The nominal model can also be, for example, a model that directly incorporates the mechanical characteristics of the helmsman's arm movements.

[0071] like Figure 3 As shown, the output of the controlled object 560 is input into the inverse nominal model 231. Specifically, the angular velocity ω calculated based on the rotation angle θa of the input axis 524a is input into the inverse nominal model 231. θ The inverse nominal model 231 is based on the above equation (4) and the input angular velocity ω. θ Output torque T p That is, the auxiliary control unit 230 calculates the torque T using a nominal model based on the output of the controlled object 560. p When the output value of the nominal model becomes the same as the output value of the controlled object 560, the torque T p It is equal to the value of the torque input to the nominal model.

[0072] Subtractor SU2 generates the differential torque T by subtracting the output of subtractor SU1 from the output of the inverse nominal model 231. a That is, the subtractor SU2 receives torque T. p Subtract the feedback correction torque T f Then, the state compensation value V, described later, is fed back. s Previous input torque T r The differential torque T is generated. a Differential torque T a For example, the interference torque T d The estimated value. The differential torque T output from subtractor SU2. a The differential torque T, after being filtered sequentially by a series-connected low-pass filter 232 and a high-pass filter 233, is input to adder AD3. aThis results in a state where frequency components lower than the first cutoff frequency Cf1 and higher than the second cutoff frequency Cf2 have been removed. That is, the differential torque T, which has undergone filtering in the low-pass filter 232 and the high-pass filter 233, is in a state where... a The frequency component T is above the first cutoff frequency Cf1 and below the second cutoff frequency Cf2. aM .

[0073] The auxiliary adjustment unit 270 generates compensation values ​​for friction and interference, and adjusts the differential torque T. a In this embodiment, the auxiliary adjustment unit 270 adjusts the differential torque T. a The frequency component T in aM The auxiliary adjustment unit 270 is connected in parallel with the high-pass filter 233. The auxiliary adjustment unit 270 includes a friction compensation value calculation unit 250, an interference compensation value calculation unit 260, and a subtractor SU3.

[0074] Subtractor SU3 subtracts the output value from the high-pass filter 233 from the output value from the low-pass filter 232. Here, the output value from the low-pass filter 232 is from the differential torque T. a The value after removing frequency components higher than the second cutoff frequency Cf2. The output value from the high-pass filter 233 is from the differential torque T. a The value is obtained by removing frequency components higher than the second cutoff frequency Cf2 and lower than the first cutoff frequency Cf1. Therefore, the value output from the subtractor SU3 becomes the differential torque T. a The frequency component T that is lower than the first cutoff frequency Cf1 aL The output of subtractor SU3 is input to friction compensation value calculation unit 250 and interference compensation value calculation unit 260. Frequency component T aL This includes friction, self-correcting torque, disturbance torque caused by loosening of the controlled object 560, and torque fluctuations generated in the controlled object 560.

[0075] Friction compensation value calculation unit 250 calculates the differential torque T based on the differential torque T. a Calculate the friction compensation value V to compensate for at least a portion of the frictional force generated in the controlled object 560. f As described above, the value from the subtractor SU3 input to the friction compensation value calculation unit 250 becomes the differential torque T. a The frequency component T that is lower than the first cutoff frequency Cf1 aL Therefore, in this embodiment, the friction compensation value calculation unit 250 is based on the differential torque T. a The friction compensation value V is calculated using the frequency components lower than the first cutoff frequency Cf1. f .

[0076] The friction compensation value calculation unit 250 includes a limiter 252 and a gain adjuster 253. The limiter 252 restricts the output value from the subtractor SU3. If the input value exceeds an upper or lower threshold, the limiter 252 limits the input value to that threshold. The gain adjuster 253 multiplies the output value from the limiter 252 by a gain K1. The friction compensation value calculation unit 250 calculates the differential torque T... a The friction compensation value V is calculated by multiplying the frequency components lower than the first cutoff frequency Cf1 by the limit of limiter 252 and gain K1. f The threshold of limiter 252 and the value of gain K1 are predetermined, for example, based on the actual frictional force generated in the controlled object 560.

[0077] In order to adjust the correction torque T used in the model follow control in the auxiliary control unit 230 f When applying friction compensation, it is necessary to pay attention to the stability condition of the model-following control. This condition is based on the small gain theorem (described later), which restricts the gain in the gain characteristic of the transfer function of the friction compensation value calculation unit 250, which takes stability into account, to not exceed 1. This is derived from the design conditions of the low-pass filter 232. In this embodiment, the value of gain K1 in the gain adjuster 253 is set to a maximum of 1, and a subtractor SU3 is provided before the limiter 252 to apply subtraction processing so that the gain in the gain characteristic becomes 1 under this condition. In other words, the friction compensation value calculation unit 250 functions as a low-pass filter with a transfer function of 1-HPF(s).

[0078] Friction compensation value V output from friction compensation value calculation unit 250 f Differential torque T a Frequency component T aL The friction compensation value V is a value that compensates for at least a portion of the frictional force component. Typically, the controlled object 560 requires moderate friction; therefore, the friction compensation value calculation unit 250 calculates a value smaller than the actual frictional force generated on the controlled object 560 as the friction compensation value V. f Therefore, high-precision friction compensation can be achieved while maintaining a suitable residual friction force on the controlled object 560. Based on the friction compensation value V... f The objects of friction compensation include, for example, the friction of motor 543, the friction of reduction mechanism 544, and the left-right difference of friction of reduction mechanism 544.

[0079] Here, in the differential torque T a Frequency component T aLIn addition to the frictional component, the torque also includes the self-aligning torque generated by the controlled object 560, the disturbance torque caused by the loosening of the controlled object 560, and the torque fluctuation generated by the controlled object 560. Therefore, the frequency component T is controlled by the limiter 252 and the gain adjuster 253. aL The friction compensation value V obtained through processing f It also includes compensation values ​​for compensating for at least a portion of the automatic return torque generated in the controlled object 560, the disturbance torque caused by the loosening generated in the controlled object 560, and the torque fluctuation generated in the controlled object 560.

[0080] Vehicles equipped with the electric power steering system 1000 can operate in driving modes with both automatic driving and manual driving modes. In this mode, the gain K1 of the gain adjuster 253 can also be switched according to the driving mode. The greater the gain K1 of the gain adjuster 253, the greater the reduction in friction. The gain K1 in the preferred automatic driving mode is greater than the gain K1 set in the manual driving mode. Therefore, optimal friction compensation can be applied to the automatic driving mode, where further friction reduction is required.

[0081] The interference compensation calculation unit 260 calculates an interference compensation value V to compensate for at least a portion of the automatic return torque generated in the controlled object 560. d In this embodiment, the interference compensation value V d The compensation value includes compensation for at least a portion of the frictional force generated in the controlled object 560, the disturbance torque caused by loosening in the controlled object 560, and the torque fluctuation generated in the controlled object 560. The disturbance compensation value calculation unit 260 is based on the torque T output from the inverse nominal model 231. p With input torque T r The difference is the differential torque T. a To calculate the interference compensation value V d That is, the interference compensation value calculation unit 260 is based on the differential torque T. a To calculate the interference compensation value V d The differential torque T a The torque T is calculated using the nominal model based on the output of the controlled object 560. p With input torque T r The difference. As described above, the value from the subtractor SU3 input to the interference compensation value calculation unit 260 becomes the differential torque T. a The frequency components in the signal are lower than the first cutoff frequency Cf1. Therefore, in this embodiment, the interference compensation value calculation unit 260 is based on the differential torque T. a The interference compensation value V is calculated using the frequency components lower than the first cutoff frequency Cf1. d .

[0082] The interference compensation value calculation unit 260 includes a limiter 262 and a gain adjuster 263. The limiter 262 imposes a limit on the output value from the subtractor SU3. If the input value exceeds a threshold value (upper or lower limit), the limiter 262 limits the input value to that threshold value. The threshold value of the limiter 262 is, for example, different from the threshold value of the limiter 252. The gain adjuster 263 multiplies the output value from the limiter 262 by a gain K2. The transfer function P(s) of the controlled object 560 is constrained to the nominal model transfer function P. n Under the condition of (s), the maximum value of the gain K2 of the gain adjuster 263 is determined. The value of gain K2 is, for example, different from the value of gain K1. The value of gain K2 is, for example, about 0.3 or higher and about 0.8 lower. The gain K2 of the gain adjuster 263 can also be switched according to the vehicle's driving mode.

[0083] Interference compensation value V d Differential torque T a Frequency component T aL The value compensates for at least a portion of the self-correcting torque component. The disturbance compensation value calculation unit 260, for example, calculates a value approximately half the self-correcting torque actually generated in the controlled object 560 as the disturbance compensation value V. d The actual self-correcting torque generated by the controlled object 560 is determined experimentally in advance for each frequency. The threshold value of the limiter 262 and the value of the gain K2 of the interference compensation value calculation unit 260 are adjusted to approximately half the magnitude of the pre-determined self-correcting torque to calculate the interference compensation value V. d The value of interference compensation V calculated in the interference compensation value calculation unit 260. d It is the friction compensation value V calculated in the friction compensation value calculation unit 250. f Different values.

[0084] Here, in the differential torque T a Frequency component T aL In addition to the automatic return torque, the torque also includes the frictional force generated by the controlled object 560, the disturbance torque caused by the loosening of the controlled object 560, and the torque fluctuation generated by the controlled object 560. Therefore, the frequency component T is controlled by the limiter 262 and the gain adjuster 263. aL The interference compensation value V obtained through processing d It also includes compensation values ​​for compensating for at least a portion of the frictional force generated in the controlled object 560, the disturbance torque caused by the loosening of the controlled object 560, and the torque fluctuation generated in the controlled object 560.

[0085] Adder AD3 adds the output value from the high-pass filter 233 to the output value from the auxiliary adjustment unit 270. That is, adder AD3 adds the output value from the frequency component T. aM Add friction compensation value V f Interference compensation value V d The frequency component T is output from adder AD3. aM Friction compensation value V f Interference compensation value V d The correction torque T is calculated by adding them together. f The correction torque T output from adder AD3 f The input that is fed back to the controlled object 560, namely the input torque T r Thus, in this embodiment, the auxiliary control unit 230 uses the high-pass filter 233 to remove the differential torque T after removing frequency components lower than the first cutoff frequency Cf1. a That is, the frequency component T aM Add friction compensation value V f Interference compensation value V d To generate the corrected torque T f .

[0086] The status feedback unit 280, based on the output of the controlled object 560, provides feedback on the input torque T. r Feedback state compensation value V s So that the apparent transfer function of the controlled object 560 approximates the nominal model transfer function P. n (s). The apparent transfer function of the controlled object 560 is, for example, the transfer function of a portion when the part located inside the feedback loop created by the auxiliary control unit 230 is considered as a single part. Specifically, in this embodiment, the apparent transfer function of the controlled object 560 is the overall transfer function of the portion from the subtractor SU1 to the output of the controlled object 560, and is the transfer function of the portion combining the state feedback unit 280 and the controlled object 560. In this embodiment, the state feedback unit 280 is affected by the correction torque T. f The input torque T after correction and before input to the controlled object 560 r Feedback state compensation value V s .

[0087] State compensation value V s This includes a compensation value that compensates for at least a portion of the inertial force, viscous force, and frictional force generated in the controlled object 560. More specifically, the state compensation value V... sThe compensation value includes at least a portion of the inertial force, viscous force, and frictional force generated in the motor 543. In this embodiment, the state compensation value V s These are compensation values ​​for the inertial force generated by motor 543, the viscous force generated by motor 543, and the frictional force generated by motor 543, respectively.

[0088] The state feedback unit 280 includes an inertial compensator 281, a viscous compensator 282, and a friction compensator 283. The inertial compensator 281 calculates a compensation value based on the angular velocity ω of the motor 543 to compensate for at least a portion of the inertial force generated in the motor 543. The viscous compensator 282 calculates a compensation value based on the angular velocity ω of the motor 543 to compensate for at least a portion of the viscous force generated in the motor 543. The friction compensator 283 calculates a compensation value based on the angular velocity ω of the motor 543 to compensate for at least a portion of the frictional force generated in the motor 543. In this embodiment, the state compensation value V... s The compensation value is calculated by the inertial compensator 281, the viscous compensator 282, and the friction compensator 283. The compensation values ​​calculated by the inertial compensator 281, the viscous compensator 282, and the friction compensator 283 are output to the adder AD1 and compared with the compensation value calculated by the correction torque T. f Corrected input torque T r Add them together.

[0089] Next, the control of the auxiliary control unit 230 will be described in more detail. The auxiliary control unit 230 uses the inverse model of the nominal model, i.e., the inverse nominal model 231, which is an internal model, to control the controlled object 560. In this embodiment, through the feedback loop generated by the auxiliary control unit 230, compensation for torque fluctuations and the like, which depend on the angular velocity ω of the motor 543, can be performed. The signal of the angular velocity ω used for control can be corrected according to the type of motor 543, and the accuracy of the angular velocity ω signal can be improved compared to current signals, etc. As a result, high-precision torque fluctuation compensation can be applied to torque control.

[0090] The auxiliary control unit 230 is structurally similar to conventional interference estimators (interference observers), but its intended function and effect are different. Conventional interference estimators estimate interference torque by setting the inverse device model, which is an internal model, to a model close to the controlled object 560, and reduce the impact of interference by pre-adding or subtracting interference torque.

[0091] In the control of the auxiliary control unit 230 in this embodiment, the transfer function P(s) of the controlled object 560 is constrained to the nominal model transfer function P, which is an internal model, by means of a feedback loop. n The effect of (s). For example, if the nominal model is defined in a way that does not have torque fluctuations, then by model following control, the transfer function P(s) of the controlled object 560 is constrained to have the characteristic of no torque fluctuations. As a result, by applying torque fluctuation compensation, torque fluctuations can be reduced. In addition, by setting the nominal model to a low-inertia model and constraining the controlled object 560 using the nominal model, the controlled object 560 can be treated as a low-inertia model. In addition, by setting the nominal model to a low-viscosity model and constraining the controlled object 560 using the nominal model, the controlled object 560 can also be treated as a low-viscosity model. By performing model following control by the auxiliary control unit 230, in addition to compensation for torque fluctuations of the motor 543, loss torque compensation or motor inertia compensation is also performed. In the above equations (3) and (4), by appropriately setting J STGn And B STGn It can assign the desired frequency characteristics to the transfer function P(s) of the controlled object 560.

[0092] The transfer function P(s) of the controlled object 560 is compared with the transfer function P of the nominal model. n When the modeling error of (s) is set as Δ(s), the transfer function P(s) of the control object 560 is represented by the following equation (5).

[0093]

[0094] The gain characteristic of the transfer function P(s) of the controlled object 560 has peaks at two frequency values, for example. The modeling error Δ(s) appears, for example, near the peak of the higher frequency of the two peaks in the gain characteristic of the controlled object 560. Therefore, as Figure 4 As shown, the reciprocal of the modeling error Δ(s), 1 / Δ(s), has a trough in the relatively high-frequency region. Figure 4 In this modeling, the modeling error Δ(s) is expressed in absolute value. If the modeling error Δ(s) increases, the transfer function P(s) of the controlled object 560 will differ from the transfer function P(s) of the nominal model. n As the deviation from the nominal model (s) increases, the control of the controlled object 560 using the nominal model based on the auxiliary control unit 230 becomes unstable. Therefore, in the region where the modeling error Δ(s) is relatively small, the gain of the complementary sensitivity function T(s), i.e., Q(s)·HPF(s), is set to 1, so that the controlled object 560 is constrained by the nominal model, thereby enabling stable and appropriate control of the controlled object 560. The frequency characteristics of the modeling error Δ(s) can be adjusted by adjusting the transfer function P of the nominal model.n J in (s) STGn And B STGn The frequency band where the gain of Q(s)·HPF(s) is 1 can be adjusted by adjusting the first cutoff frequency Cf1 and the second cutoff frequency Cf2. Thus, in the frequency band where the modeling error Δ(s) is small, the gain of Q(s)·HPF(s) can be adjusted to 1.

[0095] exist Figure 4 In the frequency band below the second cutoff frequency Cf2, 1 / Δ(s) is relatively high, and decreases sharply in the frequency band above the second cutoff frequency Cf2. Model-following control that subjects the controlled object 560 to nominal model constraints can, for example, stably operate within a range where 1 / Δ(s) is greater than 1, i.e., greater than 0 dB. Therefore, as... Figure 4 As shown, by adjusting the frequency band where the gain of Q(s)·HPF(s) is 1 in such a way that 1 / Δ(s) is greater than 1, the controlled object 560 can be stably and appropriately controlled under the nominal model constraint when the gain of Q(s)·HPF(s) is 1.

[0096] For example, in order to expand the frequency band that allows the controlled object 560 to be stably and appropriately controlled under nominal model constraints, as long as it is within the range where 1 / Δ(s) does not become less than 1, i.e. Figure 4 In the frequency band below the frequency at which the curve representing 1 / Δ(s) intersects the horizontal axis, increasing the second cutoff frequency Cf2 is sufficient. However, if the second cutoff frequency Cf2 is too high, in the frequency band higher than the second cutoff frequency Cf2, although 1 / Δ(s) decreases, the gain of Q(s)·HPF(s) remains relatively high, and control may become unstable. In contrast, in this embodiment, since the order of the low-pass filter 232 is set to 3 or higher, the gain of Q(s)·HPF(s) can be drastically reduced in the region with frequencies higher than the second cutoff frequency Cf2. Therefore, even if the second cutoff frequency Cf2 is relatively high, the gain of Q(s)·HPF(s) can be immediately reduced in the frequency band higher than the second cutoff frequency Cf2, thus suppressing the control of the controlled object 560 from becoming unstable.

[0097] The robustness and stability of the auxiliary control unit 230 are guaranteed when the small gain theorem shown in Equation (6) holds between the complementary sensitivity function T(s) and the modeling error Δ(s).

[0098]

[0099] As described above, in order to perform model following control using the nominal model in the auxiliary control unit 230, it is sufficient for T(s) = 1. However, if robustness and stability are considered, the above equation (6) must be satisfied. It can be understood that in all frequency bands, it is not possible to simultaneously satisfy T(s) = 1 and equation (6), and it is not possible to simultaneously satisfy the suppression of interference and robustness and stability of the auxiliary control unit 230.

[0100] like Figure 4 As shown, in the region where the frequency is below the first cutoff frequency Cf1, the gain of Q(s)·HPF(s), i.e., the gain of the complementary sensitivity function T(s), is also less than 1. In the region where the gain of Q(s)·HPF(s) is less than 1, the input torque T is adjusted in the reaction force control circuit 210. r The control is used to control the controlled object 560. As described above, in the region where the frequency is higher than the second cutoff frequency Cf2, the gain of Q(s)·HPF(s) is significantly reduced, becoming the correction torque T from the auxiliary control unit 230. f The state of almost no feedback to the input of the controlled object 560. On the other hand, in the region where the frequency is lower than the first cutoff frequency Cf1, the gain of Q(s)·HPF(s) is set to a certain magnitude, and the correction torque T is adjusted. f The input is fed back to the control object 560. In the region where the frequency is lower than the first cutoff frequency Cf1, the compensation value generated in the auxiliary adjustment unit 270 is fed back to the input of the control object 560 according to the gain of Q(s)·HPF(s).

[0101] The control device 100 performs torque control in the reaction force control circuit 210 for low-frequency torque signals lower than the first cutoff frequency Cf1, and controls the angular velocity ω≒0 for high-frequency interference higher than the second cutoff frequency Cf2, thereby stabilizing steering without removing the steering wheel 521. To achieve this, the control device 100 performs high-frequency gain reduction of torque control using the reaction force control circuit 210 and constrains the transfer function P(s) of the controlled object 560 to a characteristic of decreasing high-frequency gain using the auxiliary control circuit 230. The reason for performing the latter is to prevent the controlled object 560 from reacting to interference when it is input.

[0102] The effective range of the model-following control of the auxiliary control unit 230 is the region above the first cutoff frequency Cf1 and below the second cutoff frequency Cf2. That is, the lower limit frequency of the effective range of the model-following control depends on the first cutoff frequency Cf1. Therefore, the lower limit frequency of the effective range of the model-following control is determined by adjusting the first cutoff frequency Cf1 of the high-pass filter 233 so as not to hinder the control of the reaction force control circuit 210 in the low-frequency region.

[0103] According to this embodiment, a parameter for the input torque T is generated based on the output and nominal model of the controlled object 560. r Correction torque T f The auxiliary control unit 230 is configured such that, in the frequency band where the gain is 1 in the gain characteristic of the complementary sensitivity function T(s) relative to the modeling error Δ(s) between the controlled object 560 and the nominal model, the transfer function P(s) of the controlled object 560 is constrained to be the transfer function P of the nominal model. n (s). Specifically, in the region where the modeling error Δ(s) between the nominal model and the controlled object 560 is small, the gain in the gain characteristic of Q(s)·HPF(s) is set to 1, and the difference between the output of the controlled object 560 and the nominal model is used as the correction torque T. f Feedback to input torque T r This allows the transfer function P(s) of the controlled object 560 to appear approximately close to the transfer function P of the nominal model. n (s). Therefore, for example, by setting the nominal model to a model that does not produce torque ripple, torque ripple can be removed from the output of the controlled object 560 in the frequency band where the gain is 1 in the gain characteristic of Q(s)·HPF(s), or torque ripple can be reduced. In addition, by setting the nominal model to a model with low inertia and low viscosity, the controlled object 560 can be constrained to a model with low inertia and low viscosity, and the controlled object 560 can be easily controlled.

[0104] For example, in conventional interference estimators, an internal model that closely approximates the controlled object 560 is used to compensate for interference generated in the controlled object 560. However, it is difficult to have an internal model that is exactly the same as the controlled object 560, and a modeling error Δ(s) will inevitably occur regardless. Therefore, in conventional interference estimators, in order to suppress control instability, the gain in the gain characteristic of Q(s)·HPF(s) is set to a value less than 1 in all frequency bands. Furthermore, in conventional interference estimators, since the internal model is only made close to the actual model of the controlled object 560, even if external interference applied to the controlled object 560 can be estimated, torque fluctuations and other disturbances generated within the controlled object 560 itself cannot be eliminated.

[0105] In contrast, in this embodiment, the nominal model of the control device 100, which serves as an internal model, is set as an ideal model of the controlled object 560, rather than a model intended to reproduce the actual controlled object 560. In regions where the modeling error Δ(s) is small, the gain of Q(s)·HPF(s) is set to 1. Therefore, by appropriately setting the nominal model, not only can external interference applied to the actual controlled object 560 be removed, but also torque fluctuations generated within the controlled object 560 can be eliminated. Thus, according to this embodiment, the controlled object 560 can be appropriately controlled by the control device 100, improving the steering feel perceived by the helmsman.

[0106] Furthermore, for example, the steering mechanism 530 is a structure in which the input shaft 524a and the output shaft 524b are connected by a torsion bar 546, and it is not simply a single inertial system. Therefore, if the object controlled by the control device 100 is set as, for example, a single inertial system containing only the motor 543, it is difficult to adequately protect against torque fluctuations and interference. In contrast, as in this embodiment, we consider capturing the portion including the two sides separated by the torsion bar 546 as the control object 560, but we do not capture this control object 560 as two simple inertial systems. As mentioned above, the control object 560 varies between one inertial system and two inertial systems depending on the steering method of the helmsman's steering wheel 521. Therefore, even if the control object 560 is simply set as a model of two inertial systems, it is sometimes difficult to adequately protect against torque fluctuations and interference.

[0107] In contrast, according to this embodiment, the nominal model is a model that takes into account the mechanical characteristics when the helmsman steers the steering wheel 521. Therefore, the nominal model can be appropriately matched with the characteristics of the control object 560, which vary depending on the helmsman's steering method of the steering wheel 521. Thus, by constraining the control object 560 to the nominal model through the aforementioned model-following control, torque fluctuations and disturbances can be compensated more appropriately. Therefore, the helmsman's sense of steering can be further improved.

[0108] Furthermore, according to this embodiment, the transfer function P of the nominal model n The degree of (s) is 3 or more. Here, the inventors have specified that the degree of the transfer function of the steering mechanism 530 is, for example, 6. Therefore, by using the transfer function P of the nominal model... n (s) By setting a higher-order transfer function that is closer to the transfer function of the steering mechanism 530, the steering mechanism 530 can be controlled more appropriately by the control device 100. Therefore, the steering feel perceived by the helmsman can be further improved.

[0109] Furthermore, in this embodiment, the controlled object is captured within the range of the third-order transfer function, and the nominal model's transfer function P... n The degree of (s) is also set to 3. However, for example, it is also possible to capture the controlled object within the range of transfer functions of degree 4 or higher, and to set the nominal model's transfer function P... n The degree of (s) is set to 4 or higher. This determines the degree of the transfer function of the range captured as the controlled object and the transfer function P of the nominal model. n The closer the order of (s) is to the order of the transfer function of the steering mechanism 530, i.e., 6, the more appropriate the control can be. For example, if the steering torque sensor 541 is a rotary transformer, etc., the transfer function of the steering torque sensor 541 can be considered to be 2. Therefore, for example, if the controlled object is the control object obtained by adding the steering torque sensor 541 to the control object 560 of this embodiment, if the transfer function P of the nominal model is... n Setting the number of times (s) to 5 allows for more appropriate control.

[0110] Furthermore, according to this embodiment, the nominal model is a model having frequency characteristics between one inertial system and two inertial systems. As described above, the characteristics of the controlled object 560 vary between one inertial system and two inertial systems depending on the steering method of the helmsman 521. Therefore, by setting the nominal model to have frequency characteristics between one inertial system and two inertial systems, the controlled object 560 can be controlled more appropriately using the nominal model. Therefore, the steering feel perceived by the helmsman can be further improved.

[0111] Furthermore, according to this embodiment, the transfer function P of the nominal model is... n The expression for (s) is the expression for the two inertial systems with the attenuation term added. Therefore, the transfer function P of the nominal model can be expressed as... n (s) can be appropriately and easily set up as a model with frequency characteristics between one inertial system and two inertial systems.

[0112] Furthermore, according to this embodiment, the transfer function P of the nominal model n (s) is expressed by equation (3) above. Therefore, the transfer function P of the nominal model can be made more appropriately and easily. n (s) becomes a model with frequency characteristics between one inertial system and two inertial systems.

[0113] As explained above, by capturing a width wider than that of an inertial system and setting a nominal model that matches the control object 560, the nominal model can be used as an object that can compensate for interferences such as torque fluctuations with a wider frequency band than before through feedback control. This allows for a wider frequency band for suppressing interference than before. Specifically, for example, the torque fluctuations generated by the control object 560 include torque fluctuations caused by the worm gear pair used in the reduction mechanism 544. The torque fluctuations caused by this worm gear pair can sometimes be interference at around 50Hz. In conventional interference estimators, torque fluctuations are inherently unsuppressible, and the frequency band that can be suppressed as interference is lower than 50Hz; therefore, torque fluctuations caused by the worm gear pair as described above cannot be suppressed. In contrast, according to the structure and method of this embodiment, by appropriately setting the nominal model, the auxiliary control unit 230 can also compensate for relatively high-frequency torque fluctuations caused by such worm gear pairs, and suppress these relatively high-frequency torque fluctuations.

[0114] Furthermore, according to this embodiment, the control device 100 includes a state feedback unit 280, which provides feedback on the input torque T based on the output of the controlled object 560. r Feedback state compensation value V s So that the apparent transfer function of the controlled object 560 approximates the nominal model transfer function P. n (s). Therefore, the controlled object 560, which is to be controlled by the feedback of the auxiliary control unit 230, can be made to appear close to the nominal model that is the internal model. As a result, when performing model following control based on the auxiliary control unit 230, the controlled object 560 can be regarded as a model close to the nominal model, and the modeling error Δ(s) between the controlled object 560 and the nominal model can be reduced. Therefore, by setting the gain in the gain characteristic of Q(s)·HPF(s) to 1, the transfer function P(s) that can constrain the controlled object 560 to the nominal model can be expanded. n (s) frequency band. Therefore, model following control based on the auxiliary control unit 230 can be performed in a wider frequency band, which can further improve the steering feel perceived by the helmsman.

[0115] Furthermore, according to this embodiment, the state feedback unit 280 adjusts the torque T by passing through the correction torque T. f After correction, the previous input torque T is input to the controlled object 560. r Feedback state compensation value V sTherefore, feedback from the state feedback unit 280 can be incorporated into the feedback loop of the auxiliary control unit 230. Thus, from the perspective of the auxiliary control unit 230, the state feedback unit 280 and the controlled object 560 can be considered as a single controlled object. Therefore, by considering the transfer function of this aggregated single controlled object as the apparent transfer function P(s) of the controlled object 560, the control of the auxiliary control unit 230, which uses the nominal model, can be performed more appropriately.

[0116] Furthermore, according to this embodiment, the state compensation value V s This includes compensation values ​​that compensate for at least a portion of the inertial forces, viscous forces, and frictional forces generated in the controlled object 560. Therefore, it enables the apparent transfer function of the controlled object 560 to more closely approximate the nominal model's transfer function P. n (s). In this embodiment, state compensation values ​​V, which respectively include inertial force, viscous force, and frictional force, are fed back by the inertial compensator 281, the viscous compensator 282, and the friction compensator 283. s This allows the apparent transfer function of the controlled object 560 to more appropriately approximate the nominal model's transfer function P. n (s). In addition, the gains of the inertial compensator 281, the viscous compensator 282 and the friction compensator 283 are appropriately set to make the controlled object 560 close to the value of the nominal model.

[0117] Furthermore, according to this embodiment, the state compensation value V s The compensation value includes compensation for at least a portion of the inertial force, viscous force, and frictional force generated in the motor 543. Therefore, it is possible to compensate for the inertial force, etc., generated in the motor 543 in a manner close to the nominal model. Consequently, the apparent transfer function of the controlled object 560 can be made to more appropriately approximate the nominal model's transfer function P. n (s).

[0118] Furthermore, according to this embodiment, the auxiliary control unit 230 calculates the torque T based on the output of the controlled object 560 using a nominal model. p With the correction torque T f After correction and feedback state compensation value V s Previous input torque T r The difference generates the correction torque T. f That is, it is possible to add a state compensation value V to the input of the subtractor SU2 of the auxiliary control unit 230. s Previous input torque T rTherefore, from the perspective of the auxiliary control unit 230, it is easier to consider the state feedback unit 280 and the controlled object 560 together as a single controlled object. As a result, the apparent transfer function P(S) of the controlled object 560 as observed from the auxiliary control unit 230 can more appropriately approximate the nominal model.

[0119] Furthermore, for example, in two inertial systems, it is difficult to estimate the overall motion of the two inertial systems based solely on the motion of the output-side inertial system. That is, in this embodiment, it is also difficult to estimate the motion of the nominal model between one and two inertial systems based solely on information about the rotation angle θb of the output shaft 524b, and it is difficult to properly calculate the torque T output from the inverse nominal model 231. p In contrast, according to this embodiment, the auxiliary control unit 230 generates a correction torque T based on the rotation angle θa of the input shaft 524a. f Therefore, by using the rotation angle θa of the input shaft 524a as the input side, the motion of the nominal model between one inertial system and two inertial systems can be appropriately estimated. Thus, the torque T output from the inverse nominal model 231 can be appropriately calculated. p It can appropriately generate the corrective torque T f .

[0120] Furthermore, the self-centering torque is transmitted to the helmsman as a feel when steering the steering wheel 521. Therefore, for example, consider significantly reducing the gain of the auxiliary control unit 230 in the low-frequency region containing the self-centering torque, so that the self-centering torque is not compensated. In this case, since no compensation is performed in the auxiliary control circuit 230, compensation is applied in the reaction force control circuit 210 to control the controlled object 560. However, in this case, sometimes the compensation amount in the reaction force control circuit 210 becomes large, and the gain of the reaction force control circuit 210 becomes too large. Therefore, the control of the control device 100 may become unstable.

[0121] In contrast, according to this embodiment, the auxiliary control unit 230 has an interference compensation value V that calculates at least a portion of the automatic return torque generated in the controlled object 560. d Interference compensation value calculation unit 260. Correction torque T f Includes interference compensation value V dTherefore, even in the low-frequency region containing the self-centering torque, compensation based on the auxiliary control unit 230 can be performed by at least the amount needed to compensate for the self-centering torque. This reduces the amount of compensation required in the reaction force control circuit 210, and lowers the gain in the reaction force control circuit 210. Therefore, it is possible to suppress instability in the control of the control device 100. Therefore, the steering feel perceived by the helmsman can be further improved. Furthermore, by compensating for at least a portion of the self-centering torque using the auxiliary control unit 230, the reaction force applied to the helmsman when steering the steering wheel 521 can be reduced. Therefore, the steering feel perceived by the helmsman can be further improved. In particular, in this embodiment, by compensating only a portion of the self-centering torque using the auxiliary control unit 230, a suitable feel can be provided to the helmsman when steering the steering wheel 521, and the helmsman can easily steering the steering wheel 521.

[0122] Furthermore, according to this embodiment, the interference compensation value V d The compensation value includes at least a portion of the frictional force generated in the controlled object 560, the disturbance torque caused by loosening in the controlled object 560, and the torque fluctuation generated in the controlled object 560. Therefore, the disturbance compensation value V calculated in the disturbance compensation value calculation unit 260 of the auxiliary control unit 230 is... d This not only compensates for the automatic return torque, but also for at least a portion of the disturbance torque and torque fluctuations caused by friction, looseness, and other factors. Therefore, even in the low-frequency region below the first cutoff frequency Cf1 where model-following control is not possible, at least a portion of the disturbance torque and torque fluctuations caused by friction, looseness, and other factors can be compensated. Consequently, the steering feel perceived by the helmsman can be further improved in the low-frequency region below the first cutoff frequency Cf1.

[0123] Furthermore, according to this embodiment, the interference compensation value calculation unit 260 calculates the value based on the differential torque T. a To calculate the interference compensation value V d The differential torque T a The torque T is calculated using the nominal model based on the output of the controlled object 560. p With input torque T r The difference. Therefore, it is possible to determine the differential torque T. a By appropriately estimating values ​​such as the automatic return torque, the disturbance compensation value V can be calculated appropriately. d .

[0124] Furthermore, according to this embodiment, the interference compensation value calculation unit 260 is based on the differential torque T a The frequency components that are lower than the first cutoff frequency Cf1, i.e., frequency components T aLTo calculate the interference compensation value V d Because the self-correcting torque is contained in a relatively low frequency component T. aL Therefore, based on the frequency component T aL Calculate the interference compensation value V d It can appropriately compensate for self-correcting torque. Additionally, in the frequency component T... aL It also includes the frictional force generated in the controlled object 560, the disturbance torque caused by loosening in the controlled object 560, and the torque fluctuation generated in the controlled object 560. Therefore, by analyzing the frequency component T... aL The interference compensation value V is calculated by performing processing based on limiter 262 and gain adjuster 263. d Therefore, in addition to the automatic return torque, it can also calculate and compensate for the interference compensation value V caused by friction generated in the controlled object 560, the disturbance torque caused by loosening in the controlled object 560, and the torque fluctuation generated in the controlled object 560. d .

[0125] Furthermore, according to this embodiment, the auxiliary control unit 230 includes a friction compensation value calculation unit 250, which is based on the differential torque T. a To calculate the friction compensation value V, which compensates for at least a portion of the frictional force generated in the controlled object 560. f The auxiliary control unit 230 removes frequency components T lower than the first cutoff frequency Cf1 using a high-pass filter 233. aL Differential torque T a That is, the frequency component T aM Add friction compensation value V f Interference compensation value V d To generate the corrected torque T f Therefore, the auxiliary control unit 230 can more appropriately compensate for the frictional force generated on the controlled object 560.

[0126] Furthermore, according to this embodiment, the friction compensation value calculation unit 250 is based on the differential torque T. a The frequency components that are lower than the first cutoff frequency Cf1, i.e., frequency components T aL To calculate the friction compensation value V f Therefore, even in the low-frequency region below the first cutoff frequency Cf1 where model-following control is not possible, friction can be appropriately compensated. Consequently, in the low-frequency region below the first cutoff frequency Cf1, the gain of the reaction force control circuit 210 can be further reduced, and the instability of the control of the controlled object 560 can be further suppressed.

[0127] Furthermore, in conventional friction compensation control, when the angular velocity ω of the motor 543 is near zero, in order to prevent vibration, the change in the friction compensation value relative to the angular velocity ω of the motor 543 must be gradual. As a result, high-precision friction compensation control cannot be achieved. According to the inventor's research, in order to solve this problem, it is desirable to estimate and compensate for friction sequentially. According to the friction compensation calculation unit 250 of this embodiment, friction can be estimated sequentially, and the friction compensation value V can be calculated. f Therefore, this problem can be solved.

[0128] Furthermore, for example, an assist device has been developed that identifies lane markings such as white or yellow lines when driving on highways, assisting vehicles in automatically following lane markings. It is known that in vehicles equipped with such assist devices and electric power steering, if there is a left-right difference in friction of the deceleration mechanism 544, it will affect the control of the assist device that keeps the vehicle moving straight along the center of the lane. According to the friction compensation calculation unit 250 of this embodiment, even when there is a left-right difference in friction of the deceleration mechanism 544, friction can be estimated sequentially, thus solving the aforementioned problem. Furthermore, the angular velocity ω of the motor 543, which is the output of the controlled object 560, contains information related to the left-right difference in friction of the deceleration mechanism 544.

[0129] The inventors confirmed the effectiveness of the model-following control obtained by applying the control device 100 of the above-described embodiment through actual vehicle testing. In the actual vehicle testing, comparisons were made between the case without model-following control and the case with model-following control, respectively, regarding torque fluctuation, viscous sensation, friction, and inertial sensation.

[0130] exist Figure 5 , Figure 7 , Figure 9 as well as Figure 11 The figure shows the measured results of the steering angle [deg] and torsional torque [Nm] without the application of model follow control. Figure 6 , Figure 8 , Figure 10 as well as Figure 12 The results of the steering angle [deg] and torsional torque [Nm] measurements are shown when model follower control is applied. Figures 5-12 In the various graphs, the horizontal axis represents the steering angle [deg], and the vertical axis represents the torque [Nm]. Torsional torque is the torsion bar torque, which is the steering torque T. h .

[0131] exist Figure 5 as well as Figure 6 The waveform shown is when the steering wheel 521 is operated at 90 [deg / s]. Observe... Figure 5 and Figure 6The magnified views confirmed the difference compared to the model follower control without its application. Figure 5 Compared to the waveform, model follower control was applied. Figure 6 The torque fluctuation of the waveform is reduced. This confirms that by applying model-following control, torque fluctuation when steering wheel 521 is operated can be reduced.

[0132] exist Figure 7 as well as Figure 8 The waveform in the image represents the waveform when the steering frequency is set to 2Hz and the steering wheel 521 is operated. According to... Figure 7 as well as Figure 8 It was confirmed that model adjustment control was applied. Figure 8 The torque variation D1em in waveform EM1 is less than that without model-adjusted control. Figure 7 The torque variation D1ce in waveform CE1. This confirms that by applying model-following control, the viscous sensation when steering wheel 521 is reduced.

[0133] exist Figure 9 as well as Figure 10 The waveform is shown when the steering frequency is set to 0.5Hz and the steering wheel 521 is steered at ±10deg. According to... Figure 9 and Figure 10 It was confirmed that, compared to models that did not apply model-following control... Figure 9 Compared to the width D2ce indicated by the arrow in waveform CE2, model follower control was applied. Figure 10 The width D2em, indicated by the arrow in waveform EM2, is smaller. Widths D2ce and D2em correspond to the magnitude of friction, respectively. Therefore, it is confirmed that by applying model-following control, friction when steering wheel 521 is operated can be reduced.

[0134] exist Figure 11 as well as Figure 12 The image shows the waveform when the steering frequency is set to 2Hz and steering wheel 521 is returned to its original position. Figure 11 In the diagram, the dashed ellipse Ece represents the waveform portion that causes the steering wheel 521 to return. Figure 12 In the diagram, the ellipse Eem, represented by a dashed line, depicts the waveform portion when the steering wheel 521 returns to its original position. According to... Figure 11 and Figure 12 It was confirmed that, compared to models that did not apply model-following control... Figure 11 Compared to the sticking caused by inertia during return in waveform CE3, the model follower control was applied. Figure 12 The waveform EM3 shows less sticking due to inertia during return. This confirms that by applying model follow control, the inertia felt by the steering wheel 521 during steering can be reduced.

[0135] This invention is not limited to the embodiments described above. Other structures and methods can be employed within the scope of the technical concept of this invention. In the embodiments described above, the auxiliary control unit is configured such that, in the frequency band where the gain is 1 in the gain characteristic of the complementary sensitivity function of the modeling error between the controlled object and the nominal model, the transfer function of the controlled object is constrained to the transfer function of the nominal model, but this is not a limitation. The auxiliary control unit only needs to be configured such that, in the frequency band where the gain is approximately 1 in the gain characteristic of the complementary sensitivity function of the modeling error between the controlled object and the nominal model, the transfer function of the controlled object is constrained to the transfer function of the nominal model. "Approximately 1 gain" includes, for example, a gain of 0.8 or higher and 1.2 or lower. This numerical range, for example, is the range in which the gain of the substantial interference suppression characteristic can be adjusted to 1, taking into account both the positive and negative efficiencies of the worm gear pair, when the reduction mechanism connected to the motor has a worm gear pair. Since the efficiency of the worm gear pair is approximately 0.8, the gain needs to be adjusted by ±0.2 relative to the target value of 1.

[0136] In the above embodiments, when the transfer function of the low-pass filter is set to Q(s) and the transfer function of the high-pass filter is set to HPF(s), the complementary sensitivity function is a function represented by Q(s)·HPF(s), but is not limited thereto. For example, when the transfer function of the controlled object is equal to the transfer function of the nominal model, the complementary sensitivity function can also be represented by Q(s).

[0137] The nominal model can be any model with any transfer function. For example, the controlled object can be set as an inertial system containing a motor, and the nominal model's transfer function P can be... n (s) is set as the following equation (7).

[0138]

[0139] In equation (7), s is the Laplace transform, J mn B is a parameter representing the moment of inertia of the nominal model. mn It is a parameter representing the viscous friction coefficient of the nominal model.

[0140] The correction torque generated by the auxiliary control unit can be any torque that corrects the input torque, and the input torque can be corrected arbitrarily. The auxiliary control unit can also correct the input torque using the correction torque in control systems other than feedback control, such as feedforward control.

[0141] The controlled object, as the object of the control device and control method in this invention, can be any part of the steering mechanism as long as it includes at least a motor. The controlled object can be a control object of one inertial system or a control object of two or more inertial systems.

Claims

1. A control device that controls the motor-containing portion of a steering mechanism as the controlled object. The steering mechanism has the following features: The input shaft is connected to a steering wheel operated by the helmsman; An output shaft, which is connected to the input shaft via a torsion bar; and The motor is connected to the output shaft. in, The control device has: The reaction force control unit generates an input torque to be input to the controlled object based on the torque to be generated in the torque bar, and controls the reaction force transmitted from the steering wheel to the helmsman. as well as The auxiliary control unit generates a correction torque to correct the input torque based on the output of the controlled object and a nominal model, wherein the nominal model is a model with frequency characteristics between one inertial system and two inertial systems. The auxiliary control unit has: A high-pass filter with a first cutoff frequency; A low-pass filter having a second cutoff frequency higher than the first cutoff frequency; and Interference compensation value calculation department, The auxiliary control unit is configured such that, when the transfer function of the low-pass filter is set to Q(s) and the transfer function of the high-pass filter is set to HPF(s), in the frequency band where the gain is 1 in the gain characteristic of Q(s)·HPF(s), the transfer function of the controlled object is constrained to the transfer function of the nominal model. Constraining the transfer function of the controlled object to the transfer function of the nominal model means that, when observing the input-output relationship, the controlled object is controlled so that its transfer function appears to be the transfer function of the nominal model. The interference compensation value calculation unit calculates an interference compensation value that compensates for at least a portion of the automatic return torque generated by the controlled object. The correction torque includes the interference compensation value.

2. The control device according to claim 1, wherein, The interference compensation value includes a compensation value that compensates for at least a portion of the frictional force generated in the controlled object, the interference torque caused by the loosening of the controlled object, and the torque fluctuation generated in the controlled object.

3. The control device according to claim 1 or 2, wherein, The interference compensation value calculation unit calculates the interference compensation value based on the differential torque, which is the difference between the torque calculated using the nominal model based on the output of the controlled object and the input torque.

4. The control device according to claim 3, wherein, The interference compensation value calculation unit calculates the interference compensation value based on the frequency component of the differential torque that is lower than the first cutoff frequency.

5. The control device according to claim 3, wherein, The auxiliary control unit includes a friction compensation value calculation unit, which calculates a friction compensation value based on the differential torque to compensate for at least a portion of the frictional force generated on the controlled object. Furthermore, the auxiliary control unit generates the correction torque by adding the friction compensation value and the interference compensation value to the differential torque after removing frequency components lower than the first cutoff frequency through the high-pass filter.

6. The control device according to claim 5, wherein, The friction compensation value calculation unit calculates the friction compensation value based on the frequency component of the differential torque that is lower than the first cutoff frequency.

7. The control device according to claim 1 or 2, wherein, The auxiliary control unit generates the correction torque based on the rotation angle of the input shaft.

8. The control device according to claim 1 or 2, wherein, The low-pass filter is applied at least three times.

9. An electric power steering device, comprising: The control device according to any one of claims 1 to 8; and The steering mechanism.

10. A control method that controls the motor-containing portion of a steering mechanism as the controlled object. The steering mechanism has the following features: The input shaft is connected to a steering wheel operated by the helmsman; An output shaft, which is connected to the input shaft via a torsion bar; and The motor is connected to the output shaft. in, This control method includes the following: The input torque to be input to the controlled object is generated based on the torque to be generated by the torque bar, and the reaction force transmitted from the steering wheel to the helmsman is controlled. as well as A correction torque is generated to correct the input torque based on the output of the controlled object and a nominal model, wherein the nominal model is a model with frequency characteristics between one inertial system and two inertial systems. Generating the corrected torque includes the following: When the transfer function of a high-pass filter with a first cutoff frequency is set to HPF(s), and the transfer function of a low-pass filter with a second cutoff frequency higher than the first cutoff frequency is set to Q(s), in the frequency band where the gain is 1 in the gain characteristic of Q(s)·HPF(s), the transfer function of the controlled object is constrained to the transfer function of the nominal model. Constraining the transfer function of the controlled object to the transfer function of the nominal model means that when observing the input-output relationship, the controlled object is controlled so that the transfer function of the controlled object appears to be the transfer function of the nominal model. as well as Calculate the disturbance compensation value to compensate for at least a portion of the automatic return torque generated by the controlled object. The correction torque includes the interference compensation value.

Images