Control device, motor device, electric power steering device, control method, and program

Abstract

This improves the stability of model-following control while suppressing an increase in the computational load of the control device. [Solution] The control device comprises a model-following control unit that generates a correction torque to correct the input torque input to the controlled object based on a nominal model based on the configuration of the controlled object, and a calculation unit that calculates the input value input to the model-following control unit. The model-following control unit has an inverse nominal model which is the inverse model of the nominal model and into which the input value is input, and is configured such that the transfer function of the controlled object is constrained to the transfer function of the nominal model. The calculation unit is capable of executing a first calculation process to calculate the input value. In the first calculation process, the calculation unit calculates the input value based on a first output value obtained based on a first sensor and indicating the output of the motor, and a second output value obtained based on a second sensor and indicating the output of the reduction mechanism.

Description

[Technical Field]

[0001] The present invention relates to a control device, a motor device, an electric power steering device, a control method, and a program. [Background technology]

[0002] An electric power steering system installed in a vehicle is known (for example, Patent Document 1). [Prior art documents] [Patent Documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2018-183046 [Overview of the Initiative] [Problems that the invention aims to solve]

[0004] In electric power steering systems like the one described above, model-following control is sometimes implemented, which constrains the transfer function of the controlled object to the transfer function of the nominal model. In this case, the closer the nominal model is to the controlled object and the smaller the modeling error, the more stably the model-following control can be executed. However, there was a problem in that the closer the nominal model is to the controlled object, the more complex the nominal model becomes, increasing the computational load on the control device.

[0005] In view of the above circumstances, one of the objectives of the present invention is to provide a control device, a motor device, an electric power steering device, a control method, and a program that can improve the stability of model-following control while suppressing an increase in computational load. [Means for solving the problem]

[0006] One aspect of the control device of the present invention is a control device for controlling a portion of an electric power steering system mounted on a vehicle, which has an input shaft to which a steering wheel operated by an operator is connected, an output shaft connected to the input shaft via a torsion bar, and a motor connected to the output shaft via a reduction mechanism, the control device comprising: a model-following control unit that generates a correction torque to correct the input torque input to the control device based on a nominal model based on the configuration of the control device; and a calculation unit that calculates the input value input to the model-following control unit. The model-following control unit has an inverse nominal model which is the inverse model of the nominal model and to which the input value is input, and is configured such that the transfer function of the control device is constrained to the transfer function of the nominal model in a frequency band where the complementary sensitivity gain, which is the gain in the gain characteristic of the complementary sensitivity function for the modeling error between the control device and the nominal model, is approximately 1. The calculation unit is capable of executing a first calculation process for calculating the input value. In the first calculation process, the calculation unit calculates the input value based on a first output value obtained based on the first sensor and indicating the output of the motor, and a second output value obtained based on the second sensor and indicating the output of the reduction mechanism.

[0007] One embodiment of the motor device of the present invention comprises the control device described above and the motor.

[0008] One embodiment of the electric power steering device of the present invention comprises the above-described motor device and a steering mechanism having the input shaft, the output shaft, and the torsion bar.

[0009] One aspect of the control method of the present invention is a control method for controlling a portion of an electric power steering device mounted on a vehicle, which has an input shaft to which a steering wheel operated by an operator is connected, an output shaft connected to the input shaft via a torsion bar, and a motor connected to the output shaft via a reduction mechanism, the portion including the motor and the reduction mechanism being controlled, the method comprising: performing model following control to generate a correction torque for correcting the input torque input to the control object based on a nominal model based on the configuration of the control object; constraining the transfer function of the control object to the transfer function of the nominal model in a frequency band where the complementary sensitivity gain, which is the gain in the gain characteristics of the complementary sensitivity function for the modeling error between the control object and the nominal model, is approximately 1; and performing a first calculation process to calculate the input value input in the model following control to an inverse nominal model, which is the inverse model of the nominal model. The first calculation process includes calculating the input value based on a first output value obtained based on a first sensor and indicating the output of the motor, and a second output value obtained based on a second sensor and indicating the output of the reduction mechanism.

[0010] One aspect of the program of the present invention involves causing a computer to execute the above-described control method. [Effects of the Invention]

[0011] According to one aspect of the present invention, in an electric power steering system, the stability of model-following control can be improved while suppressing an increase in the computational load of the control device. [Brief explanation of the drawing]

[0012] [Figure 1] Figure 1 is a schematic diagram showing an electric power steering system according to one embodiment. [Figure 2] Figure 2 is a block diagram showing the configuration of a control device according to one embodiment. [Figure 3] Figure 3 is a functional block diagram showing the functions of the processor in a control device according to one embodiment. [Figure 4] Figure 4 is a graph illustrating the gain characteristics of the complementary sensitivity function in one embodiment, 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. [Figure 5] Figure 5 is a graph showing an example of the relationship between steering angle and self-aligning torque. [Figure 6] Figure 6 is a block diagram showing the input section in one embodiment. [Figure 7] Figure 7 is a flowchart showing a part of the processing of the calculation unit in one embodiment. [Figure 8] Figure 8 is a block diagram showing the functions of the arithmetic unit when the first arithmetic process is executed in one embodiment. [Figure 9] Figure 9 is a block diagram showing the functions of the arithmetic unit when the second arithmetic process is executed in one embodiment. [Figure 10] Figure 10 is a graph showing an example of the gain of the high-frequency output value and the gain of the low-frequency output value in one embodiment. [Modes for carrying out the invention]

[0013] The electric power steering system 1000 of this embodiment, shown in Figure 1, is mounted on a vehicle. As shown in Figure 1, the electric power steering system 1000 comprises a steering mechanism 530 and a control device 100. The steering mechanism 530 has a steering mechanism section 520 and an auxiliary mechanism section 540. The electric power steering system 1000 controls the auxiliary mechanism section 540 by the control device 100, thereby generating a steering torque T in the steering mechanism section 520 when the driver operating the vehicle steers the steering wheel 521. h It generates an auxiliary torque to assist the driver. This auxiliary torque reduces the burden on the driver when operating the steering wheel 521. The driver of the vehicle is the helmsman who steers the vehicle's steering wheel 521.

[0014] The steering mechanism 520 includes a steering wheel 521, a steering shaft 522, universal joints 523A, 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, 552B, tie rods 527A, 527B, knuckles 528A, 528B, and left and right tires 529A, 529B. In other words, the steering mechanism 530 includes a steering wheel 521, a steering shaft 522, universal joints 523A, 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, 552B, tie rods 527A, 527B, knuckles 528A, 528B, and left and right tires 529A, 529B.

[0015] The steering shaft 522 is a shaft that extends from the steering wheel 521, which is operated by the driver. One end of the input shaft 524a is connected to the end of the steering shaft 522 opposite to the side connected to the steering wheel 521, via universal joints 523A and 523B. Thus, the steering wheel 521 is connected to the input shaft 524a via the universal joints 523A and 523B and the steering shaft 522. The output shaft 524b is connected to the input shaft 524a via a torsion bar 546, which will be 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.

[0016] The input shaft 524a and the output shaft 524b are arranged coaxially. The input shaft 524a and the output shaft 524b are rotatable around the same central axis. The input shaft 524a and the output shaft 524b are rotatable relative to each other within the range in which the torsion bar 546, described later, can twist.

[0017] The auxiliary mechanism 540 includes a motor 543, a reduction mechanism 544, an inverter 545, a torsion bar 546, a first sensor 410, and a second sensor 420. In other words, the steering mechanism 530 includes a motor 543, a reduction mechanism 544, an inverter 545, a torsion bar 546, a first sensor 410, and a second sensor 420. The torsion bar 546 connects the input shaft 524a and the output shaft 524b. The torsion bar 546 is arranged coaxially with the input shaft 524a and the output shaft 524b. In the following description, the virtual 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 around the rotation axis R.

[0018] The first sensor 410 detects the rotation angle θ of the rotor of the motor 543. m The first sensor 410 detects this and outputs it to the processor 200, which will be described later. The first sensor 410 may be a resolver, a Hall element such as a Hall IC, or an MR sensor having a magnetoresistive element. The control device 100 determines the rotation angle θ of the motor 543 based on the first sensor 410. m The first output value θ1 is obtained as a value indicating the rotation angle θ. m This is the output of motor 543. In other words, the first output value θ1 indicates the output of motor 543. As shown in Figure 2, in this embodiment, two first sensors 410 are provided. The two first sensors 410 may be provided together as one sensor device, or they may be provided as separate sensor devices. When the two first sensors 410 are provided together as one sensor device, the sensor device measures the rotation angle θ of motor 543. m It outputs two first output values ​​θ1, each representing a different outcome.

[0019] In this embodiment, the second sensor 420 is composed of a steering torque sensor 541 and a steering angle sensor 542. The steering torque sensor 541 detects the amount of twist around the rotation axis R of the torsion bar 546, thereby determining the steering torque T in the steering mechanism 520. h It detects the steering torque T.h is the torsional bar torque generated in the torsional bar 546 and is the torsional moment around the rotation axis R. The rudder angle sensor 542 can detect the rotation angle θ around the rotation axis R of the input shaft 524a a . The rotation angle θ of the input shaft 524a a is equal to the steering angle θ of the steering wheel 521 h . That is, the rudder angle sensor 542 can detect the steering angle θ of the steering wheel 521 by detecting the rotation angle θ of the input shaft 524a a . Based on the steering torque sensor 541 and the rudder angle sensor 542, it is possible to detect the rotation angle θ of the output shaft 524b h . The rotation angle θ of the output shaft 524b b is the steering angle θ b . s

[0020] In this embodiment, the control device 100 obtains the second output value θ2 as a value indicating the steering angle θ based on the second sensor 420. In this embodiment, the steering angle θ s is the output of the speed reduction mechanism 544. That is, the second output value θ2 indicates the output of the speed reduction mechanism 544. Note that the second sensor 420 may be a sensor composed of only one sensor that directly detects the rotation angle θ of the output shaft 524b s . b

[0021] ​​The inverter 545 shown in Figure 1 converts DC power into three-phase AC power, which is a pseudo-sine wave of U-phase, V-phase, and W-phase, according to the motor drive signal input from the control device 100, and supplies it to the motor 543. The motor 543 is connected to the output shaft 524b via a reduction mechanism 544. The motor 543 is supplied with three-phase AC power from the inverter 545. The motor 543 is, for example, an Interior Permanent Magnet Synchronous Motor (IPMSM), a Surface Mounted Permanent Magnet Synchronous Motor (SPMSM), or a Switched Reluctance Motor (SRM). The motor 543, supplied with three-phase AC power from the inverter 545, generates steering torque T h It generates an auxiliary torque corresponding to the current. The motor 543 transmits the generated auxiliary torque to the output shaft 524b via the reduction mechanism 544.

[0022] The control device 100 controls the portion of the steering mechanism 530 mounted on the vehicle that includes at least the motor 543 and the reduction mechanism 544, treating this portion as the controlled object 560. In this embodiment, the controlled object 560 includes a steering mechanism 520, a torsion bar 546, a motor 543, and a reduction mechanism 544. Since the controlled 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 controlled object 560 cannot be described by the equations of motion of a simple one-inertial frame alone. The controlled object 560 changes between one-inertial frames and two-inertial frames depending on how hard the driver grips the steering wheel 521. The harder the driver grips the steering wheel 521, the closer the controlled object 560 is to one-inertial frames. The softer the driver grips the steering wheel 521, the closer the controlled object 560 is to two-inertial frames. Thus, the controlled object 560 is composed of two inertial frames.

[0023] The control device 100 is electrically connected to the inverter 545. Based on detection signals detected by the steering torque sensor 541, the steering 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 controlled object 560 by controlling the rotation of the motor 543 via the inverter 545. More specifically, the control device 100 controls the switching operation of multiple switching elements in the inverter 545. Specifically, the control device 100 generates a control signal to control the switching operation of each switching element and outputs it to the inverter 545. Each switching element is, for example, a MOSFET (Metal-Oxide-Semiconductor Field-Effect Transistor). In the following description, the control signal that controls the switching operation of each switching element will be called a "gate control signal".

[0024] The control device 100 controls the steering torque T h A torque command value is generated based on these factors, and the torque and rotational speed of the motor 543 are controlled, for example, by vector control. Vector control is a method of decomposing the current flowing through the motor 543 into a current component that contributes to torque generation and a current component that contributes to magnetic flux generation, and independently controlling each mutually orthogonal current component. The control device 100 is not limited to vector control and can perform other closed-loop control methods. The rotational speed of the motor 543 is expressed, for example, as the number of revolutions per minute [rpm (revolutions per minute)] or the number of revolutions per second [rps (revolutions per second)].

[0025] Furthermore, the control device 100 receives the steering torque T directly from the steering torque sensor 541. h The value may be input, or the control device 100 may input the steering torque T from the output value of the steering torque sensor 541. h The value of may be calculated. The control device 100 directly receives the steering angle θ of the steering wheel 521 from the steering angle sensor 542. hThe value may be input, or the control device 100 may input the steering angle θ from the output value of the steering angle sensor 542. h You may also calculate the value of .

[0026] In this embodiment, the electric power steering system 1000 includes a motor unit 100a. The motor unit 100a includes a control device 100, a motor 543, and an inverter 545. The motor unit 100a can be manufactured and sold independently of the other parts of the electric power steering system 1000. Furthermore, the control device 100 can be manufactured and sold independently of the other parts of the motor unit 100a as a control device for controlling the electric power steering system 1000.

[0027] Figure 2 shows a typical configuration of the control device 100 in this embodiment. The control device 100 includes, for example, a power supply circuit 111, two first sensors 410, an input circuit 113, a communication interface 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.

[0028] The vehicle speed sensor 300, steering torque sensor 541, and steering angle sensor 542 mounted on the vehicle are connected to the processor 200 so that they can input signals to the processor 200. The vehicle speed is input to the processor 200 from the vehicle speed sensor 300. The steering torque T is input to the processor 200 from the steering torque sensor 541. h The following is input: The processor 200 receives the steering angle θ from the steering angle sensor 542. h The following is entered.

[0029] The processor 200 is a semiconductor integrated circuit, also known as a central processing unit (CPU) or microprocessor. The processor 200 sequentially executes a computer program stored in the ROM 116, which contains instructions for controlling the motor drive, to achieve the desired processing. In addition to the processor 200, or in place of the processor 200, the control device 100 may have an FPGA (Field Programmable Gate Array) equipped with a CPU, a GPU (Graphics Processing Unit), an ASIC (Application Specific Integrated Circuit), an ASSP (Application Specific Standard Product), or a combination of two or more circuits selected from these circuits. The processor 200 sets a current command value according to the actual current value and the rotation angle of the motor 543 rotor, generates a PWM (Pulse Width Modulation) signal, and outputs the PWM signal to the drive circuit 115.

[0030] The power supply circuit 111 is connected to an external power supply (not shown). The power supply circuit 111 generates the DC voltage required for each part of the control device 100. The DC voltage generated by the power supply circuit 111 is, for example, 3V or 5V.

[0031] The processor 200 can calculate the angular velocity ω [rad / s] of the motor 543 based on the electrical angle of the motor 543 obtained from the first sensor 410. In addition, the control device 100 may be equipped with a speed sensor capable of detecting the rotational angular velocity of the motor 543 and an acceleration sensor capable of detecting the rotational angular acceleration of the motor 543, separate from the first sensor 410.

[0032] The input circuit 113 receives the motor current value detected by a current sensor (not shown). In the following description, the motor current value detected by the current sensor (not shown) will be referred to as the "actual current value". The input circuit 113 converts the level of the input actual current value to the input level of the processor 200 as needed, and outputs the actual current value to the processor 200. A typical example of the input circuit 113 is an analog-to-digital conversion circuit.

[0033] Communication I / F114 is an input / output interface for transmitting and receiving data in accordance with, for example, an in-vehicle control area network (CAN: Controller Area Network).

[0034] The drive circuit 115 is typically a gate driver or pre-driver. The drive circuit 115 generates gate control signals according to the PWM signal and applies gate control signals to the gates of the multiple switching elements of the inverter 545. For example, when the motor 543 to be driven is a motor that can be driven at a low voltage, the drive circuit 115 as a gate driver may not be necessary. In that case, the gate driver function of the drive circuit 115 can be implemented in the processor 200.

[0035] ROM116 is electrically connected to processor 200. ROM116 is, for example, writable memory, rewritable memory, or 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). ROM116 stores a control program containing instructions for processor 200 to control motor drive. For example, the control program stored in ROM116 is temporarily loaded into RAM (not shown) during boot-up.

[0036] Figure 3 shows an example of the functional blocks of the processor 200 in this embodiment. The processor 200, which is a computer, sequentially executes the processing or tasks necessary for controlling the motor 543 using each functional block. Each functional block of the processor 200 shown in Figure 3 may be implemented in the processor 200 as software such as firmware, as hardware, or as both software and hardware. Typically, the processing of each functional block in the processor 200 is described in a computer program in software module units and stored in the ROM 116. However, when using an FPGA or the like, all or part of these functional blocks may be implemented as hardware accelerators. Furthermore, the control method for the controlled object 560 in this embodiment is executed by the processor 200, which is a computer, executing a program stored in the control device 100. In other words, the program of this embodiment stored in the control device 100 causes the processor 200, which is a computer, to execute the control method for the controlled object 560 in this embodiment.

[0037] The processor 200 has a controller 200a. The controller 200a has an assist control unit 210, a model following control unit 230, a state feedback unit 280, an input unit 290, and a subtractor SU1. In other words, the control device 100 is equipped with an assist control unit 210, a model following control unit 230, a state feedback unit 280, an input unit 290, and a subtractor SU1. To put it another way, the processor 200 of the control device 100 has functions implemented that correspond to the assist control unit 210, the model following control unit 230, the state feedback unit 280, the input unit 290, and the subtractor SU1, respectively.

[0038] The assist control unit 210 receives the steering torque T detected by the steering torque sensor 541. h The input torque T is input to the controlled object 560. The assist control unit 210 receives the input torque T from the controlled object 560. r steering torque T hIn other words, it is generated based on the torsion bar torque generated in the torsion bar 546. To put it another way, the control method of the controlled object 560 is the input torque T input to the controlled object 560. r steering torque T h This includes generating based on the input torque T. r This is the target torque of motor 543 and is the torque command value. The assist control unit 210 receives the input torque T r The assist control unit 210 controls the reaction force transmitted to the helmsman from the steering wheel 521 by generating a steering torque T when the steering frequency or steering speed is within a predetermined range. h By applying phase compensation to the input torque T r The steering frequency is the frequency of the steering angle that changes based on the steering wheel 521's operation by the steering wheel operator. The steering speed is the speed of the steering angle that changes based on the steering wheel 521's operation by the steering wheel operator. The assist control unit 210 illustrated in Figure 3 includes a base assist calculation unit 211 and a phase compensator 212.

[0039] The base assist calculation unit 211 calculates the steering torque T h The system also obtains the vehicle speed. The vehicle speed is the speed of the vehicle. The base assist calculation unit 211 calculates the steering torque T h And it generates base assist torque based on vehicle speed. For example, the base assist calculation unit 211 calculates the steering torque T h It has a lookup table (LUT) that defines the relationship between vehicle speed and base assist torque. The base assist calculation unit 211 refers to the lookup table and calculates the steering torque T h Based on the vehicle speed, the corresponding base assist torque can be determined. The base assist calculation unit 211 calculates the steering torque T h The base assist gain can be determined based on the slope defined by the ratio of the change in base assist torque to the change in the amount of fluctuation.

[0040] In this embodiment, the phase compensator 212 adjusts the base assist gain within the range of steering frequencies that the helmsman can take when operating the steering wheel 521, and compensates for the rigidity of the torsion bar 546. The range of steering frequencies that can take is, for example, 5 Hz or less. When the steering frequency is 5 Hz or less, the phase compensator 212 adjusts the steering torque T h In other words, a first-order phase compensation may be applied to the torsion bar torque. The first-order phase compensation is represented, for example, by the transfer function of equation (1).

[0041]

number

[0042] In equation (1), s is a Laplace transformer, f1 is the zero frequency [Hz] of the transfer function, and f2 is the pole frequency [Hz] of the transfer function. A graph with gain, or loop gain, on the vertical axis and the logarithm of frequency on the horizontal axis is called a gain diagram. In a gain diagram, the zero point represents the intersection of the gain curve and the horizontal axis representing 0 dB, and the pole represents the maximum point of the gain curve. For example, phase lead compensation can be applied by setting the pole frequency higher than the zero frequency. The larger the distance between the pole frequency and the zero frequency, the greater the phase lead.

[0043] The phase compensator 212 uses the base assist torque and base assist gain output from the base assist calculation unit 211 to determine the input torque T r This generates a stable phase compensator. For example, the phase compensator 212 is a stabilization compensator, and stabilization phase compensation can be applied to the base assist torque. The phase compensator 212 may have a transfer function of order 2 or higher whose frequency characteristics are variable according to the base assist gain. A transfer function of order 2 or higher is expressed using a responsiveness parameter and a damping ratio parameter. A transfer function of order 2 or higher can be expressed by equation (2), for example. By making the order of the transfer function 2, damping can be applied to the characteristics of the transfer function. By changing the damping, it is possible to adjust the phase characteristics.

[0044]

number

[0045] In equation (2), s is a Laplace transformer, ω1 is the zero frequency of the transfer function, ω2 is the pole frequency of the transfer function, ζ1 is the zero attenuation ratio, and ζ2 is the pole attenuation ratio. The pole frequency ω2 is lower than the zero frequency ω1.

[0046] The model following control unit 230 controls the input torque T r Correction torque T to compensate for f This is generated based on a nominal model based on the configuration of the controlled object 560. In this embodiment, the corrected torque T f The input torque T r This is the feedback torque that is fed back to the control target 560. The nominal model is an internal model used as a constraint model for the control target 560 when controlling the control target 560. The nominal model will be described in detail later. The model following control unit 230 is a controller configured to perform model following control. The control method for the control target 560 is the input torque T r Correction torque T to compensate for f This includes performing model-following control, which generates a model based on a nominal model derived from the configuration of the controlled object 560. The specific configuration of the model-following control unit 230 will be described in detail later.

[0047] The subtractor SU1 receives the input torque T output from the assist control unit 210. r Therefore, the corrected torque T output from the model following control unit 230. f Subtract the value. The output from subtractor SU1 is input to adder AD1 and model following control unit 230. Adder AD1 outputs the value obtained by adding the output from state feedback unit 280 to the output from subtractor SU1 to adder AD2. Adder AD2 adds the disturbance torque T to the output from adder AD1. d The value obtained by adding this value is output to the controlled device 560.

[0048] Disturbance Torque T d This is the difference between the output torque of the ideal motor 543 and the actual output torque of the motor 543. Disturbance torque T d This includes disturbance torques applied externally to the controlled object 560. Disturbance torque T d This includes, for example, excess torque caused by friction and play resulting from mechanical elements such as the motor 543 and reduction mechanism 544, torque ripple in the motor 543, and self-aligning torque T. SAT , external disturbance torque that may occur when driving on unpaved, bumpy or gravel roads, and steering torque T h This includes, among others.

[0049] In this embodiment, the model following control unit 230 receives the input value θ output from the input unit 290. i Based on the corrected torque T f Generates input torque T r The model following control unit 230 includes an inverse nominal model 231, a first filter 232a, a second filter 232b, an assist adjustment unit 270, a subtractor SU2, and an adder AD3. In this embodiment, the first filter 232a is a high-pass filter. The first filter 232a has a first cutoff frequency Cf1. The first cutoff frequency Cf1 is, for example, 2Hz or more and 10Hz or less. In this embodiment, the first cutoff frequency Cf1 is higher than 5Hz and lower than 10Hz.

[0050] In this embodiment, the second filter 232b is a low-pass filter. The second filter 232b has a second cutoff frequency Cf2 that is higher than the first cutoff frequency Cf1. The second cutoff frequency Cf2 is, for example, 3 Hz or more and 50 Hz or less. However, the upper limit of the second cutoff frequency Cf2 may be set to a range of approximately 140 Hz or more and 200 Hz or less. The order of the second filter 232b is third order or higher. The second filter 232b may be composed of, for example, multiple low-pass filters. The first filter 232a and the second filter 232b are connected in series.

[0051] The model-following control unit 230 ensures that in the frequency band where the complementary sensitivity gain GT, which is the gain in the gain characteristic of the complementary sensitivity function T(s) for the modeling error between the controlled object 560 and the nominal model, is approximately 1, the transfer function P(s) of the controlled object 560 is equal to the transfer function P of the nominal model. n It is configured to be constrained to (s). In other words, the control method of this embodiment uses model following control to determine the transfer function P(s) of the controlled object 560 in a frequency band where the complementary sensitivity gain GT is approximately 1, compared to the transfer function P of the nominal model. n This includes being constrained to (s). "The complementary sensitivity gain GT is approximately 1" includes not only the case where the complementary sensitivity gain GT is 1, but also, for example, the case where the complementary sensitivity gain GT is 0.8 or more and 1.2 or less. This numerical range is, for example, the range in which the gain of the effective disturbance suppression characteristic can be adjusted to 1, taking into account the positive and negative efficiency of the worm gear when the reduction mechanism 544 connected to the motor 543 has a worm gear. Since the efficiency of the worm gear is about 0.8, it is necessary to adjust the gain by ±0.2 relative to the target value of 1.

[0052] The complementary sensitivity function T(s) is the complementary sensitivity function of the inner loop composed of the model following control unit 230. Figure 4 shows the complementary sensitivity gain GT in the complementary sensitivity function T(s). The complementary sensitivity gain GT is the gain of the complementary sensitivity function T(s) as a transfer function, and is the absolute value of the complementary sensitivity function T(s). In the graph of Figure 4, the horizontal axis represents frequency f [Hz], and the vertical axis represents the complementary sensitivity gain GT. As shown in Figure 4, the complementary sensitivity function T(s) has a gain of approximately 0 dB, i.e., the complementary sensitivity gain GT in the transfer function is approximately 1, in at least a portion of the frequency band where the frequency f is above the first cutoff frequency Cf1 and below the second cutoff frequency Cf2. In the example in Figure 4, the complementary sensitivity gain GT is 1 in the frequency band where the frequency f1a is higher than the first cutoff frequency Cf1 and below the second cutoff frequency Cf2 and below the second cutoff frequency Cf2. Frequency f1a is lower than frequency f2a. In the frequency band between frequency f1a and frequency f2a, the complementary sensitivity gain GT may be, for example, a value of 0.95 or more and less than 1. The complementary sensitivity gain GT at the first cutoff frequency Cf1 is smaller than the complementary sensitivity gain GT at frequency f1a. The complementary sensitivity gain GT at the second cutoff frequency Cf2 is smaller than the complementary sensitivity gain GT at frequency f2a. In this embodiment, the frequency band in which the complementary sensitivity gain GT is approximately 1 is the frequency band between frequency f1b and frequency f2b. Frequency f1b is higher than the first cutoff frequency Cf1 and lower than frequency f1a. Frequency f2b is lower than the second cutoff frequency Cf2 and higher than frequency f2a. In the frequency band between frequency f1b and frequency f2b, the complementary sensitivity gain GT is, for example, 0.8 or more and less and less than 1. In this specification, "the transfer function of the controlled object is constrained to the transfer function of the nominal model" means, for example, that the controlled object is controlled in such a way that, when looking at the input-output relationship, its transfer function appears to be the transfer function of the nominal model.

[0053] The transfer function P(s) of the controlled object 560 is a plant characteristic on which model-following control is performed. The transfer function P(s) of the controlled object 560 is expressed, for example, by the following equation (3).

[0054]

number

[0055] The inverse nominal model 231 is the inverse model of a given nominal model used to constrain the controlled object 560. The transfer function P of the nominal model. n(s) is represented by, for example, the following formula (4). The transfer function P of the inverse nominal model 231 n -1 (s) is represented by, for example, the following formula (5).

[0056] [Number]

[0057] [Number]

[0058] In formulas (4) and (5), s is the Laplace transform variable, and J n is a parameter representing the moment of inertia of the nominal model, and B n is a parameter representing the viscous friction coefficient of the nominal model. Note that the transfer function P n (s) and the transfer function P n -1 (s) of the inverse nominal model 231 are not limited to the examples shown in formulas (4) and (5), and are not particularly limited.

[0059] As shown in FIG. 3, the input value θ i , which is the output from the input unit 290, is input to the inverse nominal model 231. That is, in the present embodiment, the input value θ i is input to the model following control unit 230. The inverse nominal model 231 outputs the torque T i based on the above formula (5) and the input value θ p input. That is, the model following control unit 230 calculates the torque T p using the nominal model based on the output of the input unit 290. The torque T p is equal to the value of the torque input to the nominal model when the output value of the nominal model is the same as the output value of the control target 560.

[0060] The subtractor SU2 subtracts the output from the subtractor SU1 from the output of the inverse nominal model 231 to obtain the differential torque Ta It generates the corrected torque T. The output from subtractor SU1 is the corrected torque T. f Input torque T after subtraction r Therefore, the model following control unit 230 has a corrected torque T f Input torque T after subtraction r The following is input. In this embodiment, the subtractor SU2 receives the corrected torque T f After the feedback is received, the state compensation value V, which will be described later, s Input torque T before it is fed back r Torque T p Subtracting from the difference torque T a Generates the differential torque T. a For example, disturbance torque T d This is an estimated value. The differential torque T output from subtractor SU2. a The signal is input to the second filter 232b, where it undergoes low-pass filtering, and then input to the first filter 232a, where it undergoes high-pass filtering. The differential torque T after filtering by the first filter 232a and the second filter 232b is then input to the first filter 232a and the second filter 232b. a This is input to adder AD3. The torque input from the first filter 232a to adder AD3 is the difference torque T output from subtractor SU2. a The signal is filtered by the first filter 232a and the second filter 232b, removing frequency components lower than the first cutoff frequency Cf1 and frequency components higher than the second cutoff frequency Cf2. In other words, the torque input from the first filter 232a to the adder AD3 is the difference torque T output from the subtractor SU2. a Among these, frequency components T with a first cutoff frequency of Cf1 or higher and a second cutoff frequency of Cf2 or lower. aM That is the case.

[0061] The assist adjustment unit 270 generates compensation values ​​for friction and disturbances, and adjusts the differential torque T a Adjusts the differential torque T. In this embodiment, the assist adjustment unit 270 adjusts the differential torque T. a Among them, frequency component T aMThe assist adjustment unit 270 is coupled in parallel to the first filter 232a. The assist adjustment unit 270 includes a friction compensation value calculation unit 250, a disturbance compensation value calculation unit 260, and a subtractor SU3.

[0062] The subtractor SU3 subtracts the output value from the first filter 232a from the output value from the second filter 232b. Here, the output value from the second filter 232b is the differential torque T. a The value obtained by removing frequency components higher than the second cutoff frequency Cf2, i.e., frequency component T. aML Therefore, the output value from the first filter 232a is the differential torque T. a The value obtained by removing frequency components higher than the second cutoff frequency Cf2 and frequency components lower than the first cutoff frequency Cf1, i.e., frequency component T aM Therefore, the value output from the subtractor SU3 is the differential torque T. a Among these, the frequency component T is lower than the first cutoff frequency Cf1. aL The output of the subtractor SU3 is input to the friction compensation value calculation unit 250 and the disturbance compensation value calculation unit 260. Frequency component T aL This includes frictional force and self-aligning torque T SAT This includes disturbance torque caused by play in the controlled object 560, and torque ripple occurring in the controlled object 560.

[0063] The friction compensation value calculation unit 250 calculates a friction compensation value V that compensates for at least a portion of the friction force generated in the controlled object 560. f The difference torque T a The calculation is based on the following: As described above, the value from the subtractor SU3 input to the friction compensation value calculation unit 250 is the differential torque T a Among these, the frequency component T is lower than the first cutoff frequency Cf1. aL Therefore, in this embodiment, the friction compensation value calculation unit 250 calculates the differential torque T a Based on the component with a frequency lower than the first cutoff frequency Cf1, the friction compensation value V f Calculate.

[0064] The friction compensation value calculation unit 250 includes a limiter 252 and a gain adjuster 253. The limiter 252 limits the output value from the subtractor SU3. If the input value exceeds an upper or lower threshold, the limiter 252 clips the input value to the upper or lower 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 For the components with frequencies lower than the first cutoff frequency Cf1, the limiter 252 and the gain K1 are applied to obtain a friction compensation value V. f The threshold value of the limiter 252 and the gain K1 value are predetermined, for example, based on the frictional force actually generated on the controlled object 560.

[0065] Friction compensation value V output from friction compensation value calculation unit 250 f is the differential torque T a The frequency component T aL This value compensates for at least a portion of the frictional force component included in the control object 560. Generally, a moderate amount of friction is required for the control object 560, so the friction compensation value calculation unit 250 sets the friction compensation value V to a value smaller than the frictional force actually generated in the control object 560. f This is calculated as follows. This makes it possible to achieve highly accurate friction compensation while leaving an appropriate amount of frictional force on the controlled object 560. Friction compensation value V f The friction compensation provided by this system includes, for example, the friction of the motor 543, the friction of the reduction mechanism 544, and the difference in friction between the left and right sides of the reduction mechanism 544.

[0066] A vehicle equipped with an electric power steering system 1000 can be driven according to driving modes that include an automatic driving mode and a manual driving mode. In this case, the gain K1 of the gain adjuster 253 may be switched according to the driving mode. The larger the gain K1 of the gain adjuster 253, the greater the degree of friction reduction. It is preferable that the gain K1 in automatic driving mode is larger than the gain K1 set in manual driving mode. This makes it possible to apply optimal friction compensation to the automatic driving mode, where friction reduction is more important.

[0067] The disturbance compensation value calculation unit 260 calculates the self-aligning torque T generated in the controlled object 560. SAT Disturbance compensation value V compensates for at least a portion of it. d The disturbance compensation value V is calculated in this embodiment. d This includes a compensation value that compensates for at least a portion of the frictional force generated in the controlled object 560, the disturbance torque caused by the play in the controlled object 560, and the torque ripple generated in the controlled object 560. The disturbance compensation value calculation unit 260 calculates the torque T output from the inverse nominal model 231. p and input torque T r The difference is the differential torque T. a Based on this, the disturbance compensation value V d The disturbance compensation value calculation unit 260 calculates the torque T based on the output of the controlled object 560 using a nominal model. p and input torque T r The difference is the differential torque T. a Based on this, the disturbance compensation value V d The differential torque T is calculated. As described above, the value from the subtractor SU3 input to the disturbance compensation value calculation unit 260 is the differential torque T. a Among these, the frequency component is lower than the first cutoff frequency Cf1. Therefore, in this embodiment, the disturbance compensation value calculation unit 260 calculates the differential torque T a The disturbance compensation value V is based on the component with a frequency lower than the first cutoff frequency Cf1. d Calculate.

[0068] The disturbance compensation value calculation unit 260 includes a limiter 262 and a gain adjuster 263. The limiter 262 limits the output value from the subtractor SU3. If the input value exceeds an upper or lower threshold, the limiter 262 clips the input value to the upper or lower threshold. The threshold of the limiter 262 is different from, for example, the threshold 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 the transfer function P of the nominal model. nThe maximum value of the gain K2 of the gain adjuster 263 is determined under the constraint of (s). The value of gain K2 is different from, for example, the value of gain K1. The value of gain K2 is, for example, between 0.1 and 0.8. The gain K2 of the gain adjuster 263 may be switched according to the vehicle's driving mode.

[0069] Disturbance compensation value V d is the differential torque T a The frequency component T aL This value compensates for at least a portion of the self-aligning torque component included in the control. The disturbance compensation value calculation unit 260 calculates, for example, the self-aligning torque T that actually occurs in the control target 560. SAT The disturbance compensation value V is approximately half of the value of the disturbance compensation value. d The self-aligning torque T that actually occurs in the controlled object 560 is calculated as follows. SAT For example, the threshold value and gain K2 of the limiter 262 of the disturbance compensation value calculation unit 260 are determined in advance by the self-aligning torque T SAT The disturbance compensation value V is set to a value between 0.1 and 0.8 times the magnitude of the disturbance. d The value is adjusted to the value at which the disturbance compensation value V is calculated in the disturbance compensation value calculation unit 260. d This is the friction compensation value V calculated in the friction compensation value calculation unit 250. f This is a different value.

[0070] Here, the differential torque T a The frequency component T aL This includes the frictional force generated in the controlled object 560 and the self-aligning torque T generated in the controlled object 560. SAT This includes disturbance torque caused by play in the controlled object 560, and torque ripple occurring in the controlled object 560. Therefore, the frequency component T aL The friction compensation value V obtained by processing with limiter 252 and gain adjuster 253 f This includes disturbances other than friction, namely the self-aligning torque T generated in the controlled object 560. SATThis also includes compensation values ​​that compensate for at least a portion of the disturbance torque caused by play in the controlled object 560, and the torque ripple that occurs in the controlled object 560. Furthermore, the frequency component T aL The disturbance compensation value V obtained by processing with limiter 262 and gain adjuster 263 is obtained. d Self-aligning torque T SAT The system also includes compensation values ​​that compensate for other disturbances, namely frictional forces occurring in the controlled object 560, disturbance torques resulting from play in the controlled object 560, and at least a portion of the torque ripple occurring in the controlled object 560.

[0071] Correction torque T used for model following control in the model following control unit 230 f In order to apply friction compensation and disturbance compensation performed in the assist adjustment unit 270, it is necessary to pay attention to the stability conditions of the model following control. This condition, according to the small gain theorem described later, is that the gain in the gain characteristics of the transfer function of the assist adjustment unit 270, constrained to characteristics that consider stability, does not exceed 1. This is derived from the design conditions of the second filter 232b. In this embodiment, the values ​​of gains K1 and K2 in the gain adjusters 253 and 263 are set to a maximum of 1, and subtraction processing is applied by providing a subtractor SU3 before the limiters 252 and 262 so that the gain in the gain characteristics under these conditions becomes 1. In other words, the assist adjustment unit 270 behaves as a low-pass filter having a transfer function of 1-Q1(s). Q1(s) is the transfer function of the first filter 232a, which is a high-pass filter. The assist adjustment unit 270 applies a low-pass filter process with a transfer function of 1-Q1(s) to the torque output from the second filter 232b, and the friction compensation value calculation unit 250 and the disturbance compensation value calculation unit 260 adjust and output the processed value, respectively.

[0072] The adder AD3 adds the output value from the assist adjustment unit 270 to the output value from the first filter 232a. In other words, the adder AD3 adds the frequency component T aM Friction compensation value V f and disturbance compensation value V dAdd and . Adder AD3 outputs the frequency component T aM and friction compensation value V f and disturbance compensation value V d The corrected torque T is calculated by adding these together. f The following is output. Correction torque T is output from adder AD3. f This is the input to the controlled object 560, i.e., the input torque T. r This is fed back to the model following control unit 230. In this embodiment, the model following control unit 230 receives the differential torque T from which frequency components lower than the first cutoff frequency Cf1 have been removed by the first filter 232a, which is a high-pass filter. a , that is, frequency component T aM For this, the friction compensation value V f and disturbance compensation value V d Adding these together, the corrected torque T f Generates.

[0073] The state feedback unit 280 shown in Figure 3 calculates the apparent transfer function of the controlled object 560 based on the output of the controlled object 560, and calculates the transfer function P of the nominal model. n (s) should be approached, state compensation value V s Input torque T r Feedback is provided to the control target 560. The apparent transfer function of the control target 560 is, for example, the transfer function of a single part when the part located inside the feedback loop created by the model following control unit 230 is considered as a single part. Specifically, in this embodiment, the apparent transfer function of the control target 560 is the transfer function of the entire part from the subtractor SU1 to the output of the control target 560, and is the transfer function of the combined part of the state feedback unit 280 and the control target 560. In this embodiment, the state feedback unit 280 is corrected torque T f The input torque T after correction and before being input to the controlled object 560. r For this, the state compensation value V s Provide feedback.

[0074] State compensation value V sThis includes a compensation value that compensates for at least a portion of the inertial force, viscous force, and frictional force acting on the controlled object 560. More specifically, the 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 motor 543. In this embodiment, the state compensation value V s This is a compensation value that includes the inertial force, viscous force, and frictional force acting on the motor 543, respectively.

[0075] The state feedback unit 280 includes an inertia compensator 281, a viscosity compensator 282, and a friction compensator 283. The inertia compensator 281 controls the steering angle θ s Based on this, a compensation value is calculated to compensate for at least a portion of the inertial force generated in the motor 543. The viscous compensator 282 controls the steering angle θ s Based on this, a compensation value is calculated to compensate for at least a portion of the viscous force generated in the motor 543. The friction compensator 283 controls the steering angle θ. s Based on this, a compensation value is calculated that compensates for at least a portion of the frictional force generated in the motor 543. In this embodiment, the state compensation value V s This consists of a compensation value calculated by the inertia compensator 281, a compensation value calculated by the viscosity compensator 282, and a compensation value calculated by the friction compensator 283. The compensation value calculated by the inertia compensator 281, the compensation value calculated by the viscosity compensator 282, and the compensation value calculated by the friction compensator 283 are output to the adder AD1, and the corrected torque T f Input torque T after correction r It will be added to.

[0076] The input unit 290 inputs an input value θ to the model following control unit 230 based on the output from the controlled object 560. iThe output from the controlled object 560 is a value detected using various sensors. In this embodiment, the input unit 290 receives a first output value θ1 obtained based on the first sensor 410 and a second output value θ2 obtained based on the second sensor 420. As shown in Figure 6, in this embodiment, the input unit 290 receives the first output value θ1 as the first output value θ 1a and the first output value θ 1b Two values ​​are input: θ. First output value θ 1a This is a first output value θ1 obtained based on one of the two first sensors 410. 1b This is a first output value θ1 obtained based on the other of the two first sensors 410.

[0077] The input unit 290 includes an abnormality determination unit 291, a calculation unit 292, and a low-pass filter unit 293. In other words, the control device 100 comprises an abnormality determination unit 291, a calculation unit 292, and a low-pass filter unit 293. In this embodiment, a low-pass filter unit 293 is provided for each output value input to the input unit 290. In this embodiment, the input unit 290 has three low-pass filter units 293. The cutoff frequencies of the three low-pass filter units 293 are, for example, 20 Hz or more and 30 Hz or less. In this embodiment, the cutoff frequencies of the three low-pass filter units 293 are the same as each other. However, the cutoff frequencies of the three low-pass filter units 293 may be different from each other.

[0078] The three low-pass filter units 293 apply low-pass filtering to each of the three output values ​​input to the input unit 290. The first output value θ after low-pass filtering by the low-pass filter units 293 1a The value θ is obtained by dividing it by the reduction ratio N of the reduction mechanism 544. 3a The result is then input to the abnormality detection unit 291. The first output value θ is then subjected to low-pass filtering by the low-pass filter unit 293. 1b The value θ is obtained by dividing it by the reduction ratio N of the reduction mechanism 544. 3bThe following is input to the abnormality detection unit 291. The reduction ratio N is the value obtained by dividing the rotational angular velocity of the motor 543 by the rotational angular velocity of the output of the reduction mechanism 544. Reduction ratio N and rotational angle θ of the motor 543 m and steering angle θ s This means N=θ m / θ s The following relationship is satisfied. The second output value θ2, which has been subjected to low-pass filtering by the low-pass filter unit 293, is the value θ 3c This is input to the abnormality determination unit 291. In this way, the abnormality determination unit 291 receives two first output values ​​θ 1a ,θ 1b The three values ​​θ include two values ​​obtained by dividing each by the reduction ratio N of the reduction mechanism 544, and one second output value θ2. 3a ,θ 3b ,θ 3c The following values ​​are input. In this embodiment, the three values ​​θ are input to the abnormality determination unit 291. 3a ,θ 3b ,θ 3c These are values ​​that have been subjected to low-pass filtering.

[0079] The abnormality detection unit 291 obtains two first output values ​​θ based on each of the two first sensors 410. 1a ,θ 1b Based on the two first sensors 410 and the one second sensor 420, an abnormality determination process is performed to determine whether or not an abnormality has occurred in the two first sensors 410 and the one second sensor 420. In other words, the control method for controlling the controlled object 560 is based on the two first output values ​​θ 1a ,θ 1b The abnormality determination process includes determining whether or not an abnormality has occurred in the two first sensors 410 and the one second sensor 420 based on the three values ​​θ. 3a ,θ 3b ,θ 3c The three values ​​θ are compared with each other. The abnormality detection unit 291 uses three values ​​θ 3a ,θ 3b ,θ 3cIf one of the values ​​differs from the other two values ​​by a predetermined threshold or more, it is determined that there is an abnormality in the sensor used to acquire that one value among the three sensors, which include the two first sensors 410 and the one second sensor 420. The predetermined threshold is, for example, greater than or equal to the maximum value that is expected to be the variation in output values ​​acquired based on each sensor when each sensor is functioning normally. The predetermined threshold is, for example, the first output value θ when each sensor is functioning normally. 1a ,θ 1b The value θ obtained based on 3a ,θ 3b And the value θ obtained based on the second output value θ2 3c The error is greater than or equal to the maximum value of the expected error. The abnormality determination unit 291 uses three values ​​θ 3a ,θ 3b ,θ 3c If the difference between them is smaller than the predetermined threshold, the three values ​​θ 3a ,θ 3b ,θ 3c Since the values ​​are essentially the same, each sensor is determined to be functioning normally.

[0080] The abnormality detection unit 291 uses two values ​​θ 3b ,θ 3c The difference between them is less than a predetermined threshold, and the value θ 3a and value θ 3b The difference from, and the value θ 3a and value θ 3c If at least one of the differences between the two is greater than or equal to a predetermined threshold, then the value θ 3a The system determines that an abnormality has occurred in one of the first sensors 410 used to acquire the data. The abnormality determination unit 291 determines that an abnormality has occurred in one of the two values ​​θ 3a ,θ 3c The difference between them is less than a predetermined threshold, and the value θ 3b and value θ 3a The difference from, and the value θ 3b and value θ 3c If at least one of the differences between the two is greater than or equal to a predetermined threshold, then the value θ 3b The system determines that an abnormality has occurred in the other first sensor 410, which was used to acquire the value θ. The abnormality determination unit 291 determines that an abnormality has occurred in the other first sensor 410, which was used to acquire the value θ. 3a ,θ 3bThe difference between them is less than a predetermined threshold, and the value θ 3c and value θ 3a The difference from, and the value θ 3c and value θ 3b If at least one of the differences between the two is greater than or equal to a predetermined threshold, then the value θ 3c The second sensor 420, which was used to acquire the data, is determined to be malfunctioning. The malfunction determination unit 291 determines that there is a malfunction in the second sensor 420. 3a ,θ 3b ,θ 3c The system outputs a signal RS, indicating the determination result, to the calculation unit 292.

[0081] The calculation unit 292 has two first output values ​​θ 1a ,θ 1b Then, a second output value θ2 and a signal RS are input. The calculation unit 292 calculates the input value θ based on the input output value and signal RS. i The calculation unit 292 is capable of performing a first calculation process CP1 and a second calculation process CP2. In other words, the control method for controlling the controlled object 560 includes performing the first calculation process CP1 and performing the second calculation process CP2. The first calculation process CP1 and the second calculation process CP2 are based on the input value θ. i This is the process of calculating [the result].

[0082] If the abnormality determination unit 291 determines that no abnormality has occurred in at least one of the two first sensors 410 and the second sensor 420, the calculation unit 292 performs the first calculation process CP1 to obtain the input value θ i The calculation unit 292 calculates the input value θ using the second calculation process CP2. If the abnormality determination unit 291 determines that there is no abnormality in at least one of the two first sensors 410, and that there is an abnormality in the second sensor 420, the calculation unit 292 calculates the input value θ using the second calculation process CP2. i The first calculation process CP1 calculates the input value θ when the abnormality determination process determines that no abnormality has occurred in at least one of the two first sensors 410 and the second sensor 420. iThe second calculation process CP2 calculates the input value θ if, in the abnormality determination process, it is determined that no abnormality has occurred in at least one of the two first sensors 410, and an abnormality has occurred in the second sensor 420. i This includes calculating [something].

[0083] The calculation unit 292 determines, for example, whether to execute the first calculation process CP1 or the second calculation process CP2, according to the flowchart shown in Figure 7. Figure 7 is a flowchart illustrating an example where the abnormality determination unit 291 determines that no abnormalities have occurred in at least two of the three sensors, the two first sensors 410 and the one second sensor 420. For example, if the abnormality determination unit 291 determines that abnormalities have occurred in two or more sensors, the calculation unit 292 does not execute either the first calculation process CP1 or the second calculation process CP2, and stops the assist control by the electric power steering device 1000.

[0084] As shown in Figure 7, the calculation unit 292 determines whether or not an abnormality has occurred in any of the three sensors (step S110). In step S110, the calculation unit 292 determines whether or not an abnormality has occurred in any of the three sensors based on the signal RS input from the abnormality determination unit 291. If the calculation unit 292 determines in step S110 that no abnormality has occurred in any of the three sensors (step S110: NO), it executes the first calculation process CP1 (step S130). On the other hand, if the calculation unit 292 determines in step S110 that an abnormality has occurred in any of the three sensors (step S110: YES), it determines whether or not the abnormal sensor is the second sensor 420 (step S120). In step S120, the calculation unit 292 determines whether or not the abnormal sensor is the second sensor 420 based on the signal RS. If the calculation unit 292 determines in step S120 that the sensor experiencing the abnormality is one of the two first sensors 410 (step S120: NO), it executes the first calculation process CP1 (step S130). On the other hand, if the calculation unit 292 determines in step S120 that the sensor experiencing the abnormality is the second sensor 420 (step S120: YES), it executes the second calculation process CP2 (step S140).

[0085] Figure 8 is a block diagram showing the functions of the arithmetic unit 292 when the first arithmetic process CP1 is executed. Figure 9 is a block diagram showing the functions of the arithmetic unit 292 when the second arithmetic process CP2 is executed. As shown in Figures 8 and 9, the arithmetic unit 292 has a processing determination unit 292a. The processing determination unit 292a has two first output values ​​θ 1a ,θ 1b Then, one second output value θ2 and the signal RS are input. The process of determining the calculation process to be executed, as illustrated in Figure 7, is performed in the processing determination unit 292a. As shown in Figure 8, if the processing determination unit 292a determines to execute the first calculation process CP1, the first output value θ 1c It also outputs a second output value θ2.

[0086] First output value θ 1cThe two first output values ​​θ 1a ,θ 1b This value is calculated based on at least one of the following. In this embodiment, the first output value θ 1c This is a value used in the first arithmetic processing CP1 and the second arithmetic processing CP2 in the arithmetic unit 292. If it is determined that there is no abnormality in the two first sensors 410, the first output value θ 1c The two first output values ​​θ 1a ,θ 1b This is the average value. In other words, if the abnormality determination unit 291 determines that no abnormality has occurred in the two first sensors 410, the calculation unit 292 uses the two first output values ​​θ obtained based on the two first sensors 410 as the first output value θ1c used in the first calculation process CP1 and the second calculation process CP2. 1a ,θ 1b The average value is used. If it is determined that an abnormality has occurred in one of the first sensors 410 and that there is no abnormality in the other first sensor 410, the first output value θ 1c The two first output values ​​θ 1a ,θ 1b This is the value of the first output value θ1 obtained based on the first sensor 410 that is not abnormal. In other words, if the abnormality determination unit 291 determines that an abnormality has occurred in one of the two first sensors 410 and that there is no abnormality in the other first sensor 410, the calculation unit 292 determines the first output value θ used in the calculation process. 1c The first output value θ1 obtained based on the other first sensor 410 is used.

[0087] The first output value θ output from the processing determination unit 292a in the first calculation process CP1 1c The signal is then divided by the reduction ratio N, and then subjected to high-pass filtering in the high-pass filter section 292H to obtain the high-frequency output value θ. 1f This is the result. High-frequency output value θ 1f The first output value θ 1cThis value is obtained by applying a high-pass filter and a division process by dividing it by the reduction ratio N of the reduction mechanism 544. The second output value θ2 output from the processing determination unit 292a in the first calculation process CP1 is subjected to a low-pass filter process in the low-pass filter unit 292L to obtain the low-frequency output value θ 2f This is the result. Low-frequency output value θ 2f This value is obtained by applying a low-pass filter to the second output value θ2. In this embodiment, the cutoff frequency of the high-pass filter section 292H and the cutoff frequency of the low-pass filter section 292L are the same. That is, in this embodiment, the cutoff frequency in the high-pass filter processing performed in the first calculation process CP1 is the same as the cutoff frequency in the low-pass filter processing performed in the first calculation process CP1. In the following description, the cutoff frequency of the high-pass filter section 292H and the cutoff frequency of the low-pass filter section 292L are referred to as the cutoff frequency Cf3. The cutoff frequency Cf3 is, for example, 10Hz or more and 100Hz or less. In this embodiment, the cutoff frequency Cf3 is 15Hz or more and 50Hz or less. That is, the cutoff frequency Cf3 in the low-pass filter processing and high-pass filter processing performed in the first calculation process CP1 of this embodiment is 15Hz or more and 50Hz or less. The cutoff frequency Cf3 is more preferably 20Hz or more and 30Hz or less.

[0088] Furthermore, "the cutoff frequency in high-pass filtering is the same as the cutoff frequency in low-pass filtering" includes not only cases where the cutoff frequencies in high-pass filtering and low-pass filtering are exactly the same, but also cases where they are approximately the same. "The cutoff frequency in high-pass filtering and low-pass filtering are approximately the same" includes cases where the cutoff frequencies in high-pass filtering and low-pass filtering differ from each other within the tolerance range of manufacturing variations of the filters performing each filtering process.

[0089] High-frequency output value θ 1f The gain and low-frequency output value θ 2f The gain has a frequency characteristic as shown in the graph in Figure 10. In the graph in Figure 10, the horizontal axis represents the frequency f [Hz], and the vertical axis represents the gain of each output value. As shown in Figure 10, the frequency band below the cutoff frequency Cf3 is the first frequency band FB1. The frequency band above the cutoff frequency Cf3 is the second frequency band FB2.

[0090] High-frequency output value θ in the first frequency band FB1 1f The gain is the high-frequency output value θ in the second frequency band FB2. 1f It is smaller than the gain. High-frequency output value θ at cutoff frequency Cf3 1f The gain is less than 1. The high-frequency output value θ at the cutoff frequency Cf3. 1f The gain is, for example, 1 / √2, or -3 [dB]. The high-frequency output value θ 1f The gain is 1 in the frequency band above the cutoff frequency Cf3, specifically above the frequency fb. The high-frequency output value θ is below the frequency fb. 1f The gain decreases as the frequency f decreases.

[0091] Low-frequency output value θ in the second frequency band FB2 2f The gain is the low-frequency output value θ in the first frequency band FB1. 2f It is smaller than the gain. Low-frequency output value θ at cutoff frequency Cf3 2f The gain is less than 1. In the example in Figure 10, the low-frequency output value θ at the cutoff frequency Cf3. 2f The gain is the high-frequency output value θ at the cutoff frequency Cf3. 1f This is the same as the gain. The low-frequency output value θ at the cutoff frequency Cf3. 2f The gain is, for example, 1 / √2, or -3 [dB]. Low frequency output value θ 2fThe gain is 1 in the frequency band below the cutoff frequency Cf3, specifically below the frequency fa. In the frequency band above the frequency fa, the low-frequency output value θ is... 2f The gain of the signal decreases as the frequency f increases.

[0092] As shown in Figure 8, in the first calculation process CP1, the high-frequency output value θ 1f and low-frequency output value θ 2f These are added together by adder AD4. Adder AD4 has a high-frequency output value θ 1f and low-frequency output value θ 2f Add the two values ​​together, and the resulting sum is the input value θ. i It outputs as follows. Thus, in the first arithmetic processing CP1, the arithmetic unit 292 outputs the first output value θ 1c The high-frequency output value θ is obtained by applying a high-pass filter and a division process by the reduction ratio N to the given value. 1f Then, the low-frequency output value θ obtained by applying a low-pass filter to the second output value θ2 is obtained. 2f The input value θ is obtained by adding the two values ​​together. i The first calculation process CP1 calculates the high-frequency output value θ. 1f And the low-frequency output value θ 2f The input value θ is obtained by adding the two values ​​together. i This includes calculating the input value θ in the first calculation process CP1 based on the first output value θ1 obtained based on the first sensor 410 and indicating the output of the motor 543, and the second output value θ2 obtained based on the second sensor 420 and indicating the output of the reduction mechanism 544. i The first arithmetic process CP1 calculates the input value θ based on the first output value θ1 and the second output value θ2. i This includes calculating [the value].

[0093] As shown in Figure 9, if the processing determination unit 292a determines to execute the second arithmetic processing CP2, the first output value θ 1c The first output value θ is output from the processing determination unit 292a in the second calculation process CP2. 1c It is divided by the reduction ratio N, and the input value θ iIt is said that in the second arithmetic processing CP2, the arithmetic unit 292 calculates the first output value θ. 1c The input value θ is obtained by performing a division operation by dividing it by the reduction ratio N of the reduction mechanism 544. i In other words, the second arithmetic process CP2 processes the first output value θ. 1c The input value θ is obtained by performing a division operation by dividing it by the reduction ratio N. i This includes doing so.

[0094] Next, the control by the model-following control unit 230 will be described in more detail. The model-following control unit 230 controls the controlled object 560 using the inverse model of the nominal model it has as an internal model, i.e., the inverse nominal model 231. In this embodiment, the feedback loop created by the model-following control unit 230 makes it possible to compensate for torque ripple and other factors that depend on the angular velocity ω of the motor 543.

[0095] The model-following control unit 230 is structurally similar to a conventional disturbance estimator (disturbance observer), but its intended function and effect are different. Conventional disturbance estimators estimate disturbance torque by using an inverse plant model, which is an internal model, that closely resembles the controlled object 560, and reduce the effect of disturbances by pre-adding or subtracting the disturbance torque.

[0096] In this embodiment, the control by the model-following control unit 230 is performed by a feedback loop so that the transfer function P(s) of the controlled object 560 is the transfer function P of the nominal model which has an internal model. nThe effect of being constrained by (s) is utilized. For example, if the nominal model is defined so that there is no torque ripple, the transfer function P(s) of the controlled object 560 is constrained to the characteristic of having no torque ripple by model following control, and as a result, torque ripple can be reduced by applying torque ripple compensation. Alternatively, by making the nominal model a low-inertia model and constraining the controlled object 560 with the nominal model, the controlled object 560 can be treated as a low-inertia model. Alternatively, by making the nominal model a low-viscosity model and constraining the controlled object 560 with the nominal model, the controlled object 560 can be treated as a low-viscosity model. By executing model following control by the model following control unit 230, in addition to torque ripple compensation for the motor 543, for example, lost torque compensation or motor inertia compensation is performed. In the above-mentioned equations (4) and (5), J n and B n By appropriately setting this, the desired frequency characteristics can be imparted to the transfer function P(s) of the controlled object 560.

[0097] Transfer function P(s) of the controlled object 560 and the transfer function P of the nominal model n When the modeling error with (s) is Δ(s), the transfer function P(s) of the controlled object 560 can be expressed, for example, by the following equation (6).

[0098]

number

[0099] The gain characteristics of the transfer function P(s) of the controlled object 560 have peaks at two frequency values, for example. The modeling error Δ(s) appears, for example, near the higher frequency peak of the two peaks in the gain characteristics of the controlled object 560. Therefore, as shown in Figure 4, the reciprocal of the modeling error Δ(s), 1 / Δ(s), has a bottom in the relatively high-frequency region. In Figure 4, the modeling error Δ(s) is shown as an absolute value. As the modeling error Δ(s) increases, the transfer function P(s) of the controlled object 560 and the transfer function P of the nominal model nThe deviation from (s) becomes large, and the control of the controlled object 560 using the nominal model by the model-following 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) is set to approximately 1, and the controlled object 560 is constrained to the nominal model, thereby enabling stable and suitable control of the controlled object 560. Transfer function P of the nominal model n (s) J n and B n By adjusting the modeling error Δ(s), the frequency characteristics of the modeling error Δ(s) are adjusted. By adjusting the first cutoff frequency Cf1 and the second cutoff frequency Cf2, the frequency band in which the gain of the complementary sensitivity function T(s) is approximately 1 is adjusted. This allows the gain of the complementary sensitivity function T(s) to be approximately 1 in the frequency band in which the modeling error Δ(s) is small.

[0100] In Figure 4, 1 / Δ(s) is relatively high in the frequency band below the second cutoff frequency Cf2, and decreases sharply in the frequency band above the second cutoff frequency Cf2. Model-following control, which constrains the controlled object 560 to the nominal model, can be performed stably, for example, in the range where 1 / Δ(s) is greater than 1, i.e., greater than 0 dB. Therefore, as shown in Figure 4, by adjusting 1 / Δ(s) to be greater than 1 in the frequency band where the gain of the complementary sensitivity function T(s) is approximately 1, the controlled object 560 can be stably and suitably controlled by constraining it to the nominal model when the gain of the complementary sensitivity function T(s) is approximately 1.

[0101] For example, in order to broaden the frequency band over which the controlled object 560 can be stably and suitably controlled while constrained to the nominal model, the second cutoff frequency Cf2 should be increased within the range where 1 / Δ(s) is not less than 1, that is, within the frequency band lower than the frequency at which the curve representing 1 / Δ(s) in Figure 4 intersects with the horizontal axis. However, if the second cutoff frequency Cf2 is increased too much, the gain of the complementary sensitivity function T(s) may remain relatively high in the frequency band above the second cutoff frequency Cf2, even though 1 / Δ(s) has decreased, which may lead to unstable control. In contrast, in this embodiment, since the order of the second filter 232b, which is a low-pass filter, is set to the third order or higher, the gain of the complementary sensitivity function T(s) can be sharply reduced in the region where the frequency is higher than the second cutoff frequency Cf2. This allows the gain of the complementary sensitivity function T(s) to be quickly reduced in frequency bands higher than the second cutoff frequency Cf2, even when the second cutoff frequency Cf2 is set relatively high, thereby suppressing instability in the control of the controlled device 560.

[0102] The robust stability of the model-following control unit 230 is guaranteed when the small gain theorem shown in equation (7) below holds between the complementary sensitivity function T(s) and the modeling error Δ(s).

[0103]

number

[0104] As described above, in order to perform model following control using a nominal model in the model following control unit 230, the complementary sensitivity gain GT of the complementary sensitivity function T(s) only needs to be approximately 1. However, considering robust stability, it is necessary to satisfy equation (7) above. As can be understood from this, it is not possible to reconcile making the complementary sensitivity gain GT approximately 1 in all frequency bands with satisfying equation (7), and thus the suppression of disturbances by the model following control unit 230 and robust stability are incompatible.

[0105] As shown in Figure 4, even in the low-frequency region FA1, where the frequency f is lower than the first cutoff frequency Cf1, the complementary sensitivity gain GT of the complementary sensitivity function T(s) becomes less than 1. In the region where the complementary sensitivity gain GT of the complementary sensitivity function T(s) becomes less than 1, the assist control unit 210 controls the input torque T r The controlled object 560 is controlled by performing this control. In the high-frequency region FA2, where the frequency is higher than the second cutoff frequency Cf2, the complementary sensitivity gain GT of the complementary sensitivity function T(s) is greatly reduced, and the correction torque T from the model following control unit 230 is reduced. f This results in a state where there is almost no feedback to the input of the controlled object 560. On the other hand, in the low-frequency region FA1, the complementary sensitivity gain GT of the complementary sensitivity function T(s) is set to a certain magnitude, and the corrected torque T f This is fed back to the input of the controlled device 560. In the low-frequency region FA1, the compensation value generated in the assist adjustment unit 270 described above is fed back to the input of the controlled device 560 according to the complementary sensitivity gain GT of the complementary sensitivity function T(s). In this embodiment, the value of the complementary sensitivity gain GT in the low-frequency region FA1 is 0.5 or more. In this embodiment, the steady-state gain T(0) of the complementary sensitivity function T(s) is 0.5.

[0106] Between the motor 543 and the reduction mechanism 544, the motor 543 is easier to model, while the reduction mechanism 544 is more difficult to model. Therefore, when the controlled object 560 includes both the motor 543 and the reduction mechanism 544, as in this embodiment, it is easier to create a nominal model with a small modeling error with respect to the motor 543, while attempting to minimize the modeling error with respect to the reduction mechanism 544 tends to complicate the nominal model. This leads to a problem where the closer the nominal model is to the configuration of the controlled object 560, the greater the computational load on the control device 100. On the other hand, if the nominal model is made to a certain extent, the behavior of the reduction mechanism 544 can be simulated to some extent in low frequency bands such as the first frequency band FB1 mentioned above, but it is difficult to simulate the behavior of the reduction mechanism 544 in high frequency bands such as the second frequency band FB2 mentioned above. Therefore, if the nominal model is made to a certain extent, the modeling error Δ(s) for the controlled object 560 becomes larger in high frequency bands such as the second frequency band FB2. Therefore, when model following control is performed based solely on the output of the reduction mechanism 544, simplifying the nominal model sometimes made it difficult to stably perform model following control in high frequency bands such as the second frequency band FB2.

[0107] To address the above problem, according to this embodiment, the control device 100 receives the input value θ from the model following control unit 230. i It includes a calculation unit 292 that calculates the input value θ. The model following control unit 230 is the inverse model of the nominal model and has an input value θ i The inverse nominal model 231 is input to the controlled object 560, and in a frequency band where the complementary sensitivity gain GT, which is the gain in the gain characteristic of the complementary sensitivity function T(s) with respect to the modeling error Δ(s) between the controlled object 560 and the nominal model, is approximately 1, the transfer function P(s) of the controlled object 560 is the transfer function P of the nominal model n It is configured to be constrained by (s). The arithmetic unit 292 is configured to use the input value θ iA first calculation process CP1 can be executed to calculate the input value θ. In the first calculation process CP1, the calculation unit 292 calculates the input value θ based on a first output value θ1 obtained based on the first sensor 410 and indicating the output of the motor 543, and a second output value θ2 obtained based on the second sensor 420 and indicating the output of the reduction mechanism 544. i This calculates the input value θ that is input to the inverse nominal model 231 in model-following control. In other words, the control method for controlling the controlled object 560 is the input value θ that is input to the inverse nominal model 231 in model-following control. i This includes executing a first arithmetic process CP1 which calculates the input value θ based on the first output value θ1 and the second output value θ2. i This includes calculating the following. Since modeling the motor 543 is easier than modeling the reduction mechanism 544, the output of the motor 543 is easier to simulate even if the nominal model is simplified to some extent. Also, in the low frequency band, the behavior of the reduction mechanism 544 is easier to simulate even if the nominal model is simplified to some extent, and the output of the reduction mechanism 544 is easier to simulate. For example, in the high frequency band where the modeling error Δ(s) of the nominal model tends to be large, an input value θ such that model following control is performed using a first output value θ1 that represents the output of the motor 543, which is easier to model. i By doing so, stable model-following control can be performed even in high-frequency bands where the modeling error Δ(s) tends to be large. This makes it possible to reduce the computational load on the control device 100 by using a relatively simple nominal model while still enabling stable model-following control. Therefore, according to this embodiment, the stability of model-following control can be improved while suppressing an increase in the computational load on the control device 100. Since model-following control can be stably performed on the controlled object 560 including the reduction mechanism 544, the torque ripple of the output torque of the reduction mechanism 544 can be suitably reduced.

[0108] According to this embodiment, in the first arithmetic processing CP1, the arithmetic unit 292 calculates the first output value θ 1c The value obtained by applying a high-pass filter and a division process by the reduction ratio N of the reduction mechanism 544, i.e., the high-frequency output value θ, is the value obtained by applying a high-pass filter to it. 1fThen, the value obtained by applying a low-pass filter to the second output value θ2, i.e., the low-frequency output value θ 2f The input value θ is obtained by adding the two values ​​together. i The first calculation process CP1 calculates the high-frequency output value θ. 1f And the low-frequency output value θ 2f The input value θ is obtained by adding the two values ​​together. i This includes calculating the input value θ. i The gain is the high-frequency output value θ shown in Figure 10. 1f Gain and low-frequency output value θ 2f The gain can be made to have a frequency characteristic that is the sum of the gains. As a result, in the low first frequency band FB1, model following control can be accurately performed based on the second output value θ2 which represents the output of the reduction mechanism 544, where the modeling error tends to be small. Furthermore, in the high second frequency band FB2, model following control can be stably performed based on the first output value θ1 which represents the output of the motor 543, where the modeling error does not tend to be large even in the high frequency band. Therefore, it is possible to perform model following control stably over a wide frequency band while using a relatively simple nominal model. In addition, since there is no need to perform control such as switching the output value used depending on the frequency band, the computational load on the control device 100 can be further suppressed.

[0109] According to this embodiment, the cutoff frequency in the high-pass filter processing of the first arithmetic processing unit CP1 is the same as the cutoff frequency in the low-pass filter processing of the first arithmetic processing unit CP1. Therefore, compared to the case where the cutoff frequencies of the high-pass filter processing and the low-pass filter processing are different in the first arithmetic processing unit CP1, the design of each filter can be simplified. Also, the high-frequency output value θ 1f Gain and low-frequency output value θ 2f This suppresses the occurrence of frequency bands where both the gain and gain are smaller than the gain of each output value at the cutoff frequency Cf3. This suppresses the occurrence of frequency bands where the model following control becomes unstable.

[0110] According to this embodiment, the cutoff frequency Cf3 in the low-pass filter processing of the first arithmetic processing unit CP1 is between 15 Hz and 50 Hz. The behavior of the deceleration mechanism 544 in the frequency band between 15 Hz and 50 Hz is easy to simulate even with a simple nominal model. Therefore, by setting the cutoff frequency Cf3 in the low-pass filter processing of the first arithmetic processing unit CP1 to a frequency in this frequency band, the nominal model can be further simplified while making it easier to perform stable model-following control.

[0111] According to this embodiment, the control device 100 obtains two first output values ​​θ based on each of the two first sensors 410. 1a ,θ 1b The system includes an abnormality determination unit 291 that determines whether or not an abnormality has occurred in the two first sensors 410 and the one second sensor 420 based on the two first output values ​​θ2 obtained from the one second sensor 420. The abnormality determination unit 291 determines whether or not an abnormality has occurred in the two first output values ​​θ 1a ,θ 1b The three values ​​θ include two values ​​obtained by dividing each by the reduction ratio N of the reduction mechanism 544, and one second output value θ2. 3a ,θ 3b ,θ 3c We compare them with each other and the three values ​​θ 3a ,θ 3b ,θ 3c If one of the values ​​differs from the other two values ​​by a predetermined threshold or more, it is determined that an abnormality has occurred in the sensor used to acquire that one value among the three sensors, which include the two first sensors 410 and the one second sensor 420. In other words, the control method for controlling the controlled object 560 uses two first output values ​​θ 1a ,θ 1b The system includes an abnormality determination process that determines whether or not an abnormality has occurred in the two first sensors 410 and the one second sensor 420, based on three values ​​θ2. 3a ,θ 3b ,θ 3c We compare them with each other and the three values ​​θ 3a ,θ 3b ,θ 3cThis includes determining that an abnormality has occurred in the sensor used to acquire the value of one of the three sensors if that value differs from the other two values ​​by a predetermined threshold or more. Therefore, it is possible to detect abnormalities occurring in the three sensors, including the two first sensors 410 and the one second sensor 420. Furthermore, the transfer function P(s) of the controlled object 560 is determined by model following control, and the transfer function P of the nominal model P n Because it can be constrained to (s), the delay between the output of the reduction mechanism 544 and the output of the motor 543 can be reduced, and the error between the output value of the reduction mechanism 544 calculated from the first output value θ1 obtained based on the first sensor 410 and the second output value θ2 obtained based on the second sensor 420 can be reduced. As a result, the three values ​​θ when each sensor is normal can be reduced compared to when model following control is not performed. 3a ,θ 3b ,θ 3c The difference between them can be reduced. Therefore, the predetermined threshold used for anomaly detection can be reduced compared to when model following control is not performed. As a result, if an anomaly occurs in any one of the three sensors, the three values ​​θ 3a ,θ 3b ,θ 3c The time it takes for the difference between one of the values ​​and the other two values ​​to exceed a predetermined threshold is shortened. This allows for rapid detection of abnormalities in the sensor.

[0112] Furthermore, for example, two first output values ​​θ obtained based on two first sensors 410 1a ,θ 1b When determining an anomaly in the two first sensors 410 by comparing them, the first output value θ 1a ,θ 1bEven if the difference between the two first sensors 410 exceeds a predetermined threshold, it is not possible to determine which of the two first sensors 410 has malfunctioned. In contrast, for example, by providing three first sensors 410 and comparing the three first output values ​​θ1 output based on the three first sensors 410, it is possible to identify the first sensor 410 that has malfunctioned. However, in this case, it is necessary to add one more first sensor 410 as the third sensor, which increases the manufacturing cost of the electric power steering device 1000. In contrast, in this embodiment, since the delay between the output of the reduction mechanism 544 and the output of the motor 543 can be reduced by model following control, the second sensor 420 for detecting the output of the reduction mechanism 544 can be used as a third sensor added to the two first sensors 410. Therefore, it is possible to identify the sensor that has malfunctioned without increasing the number of first sensors 410. As a result, the increase in the manufacturing cost of the electric power steering device 1000 can be suppressed.

[0113] According to this embodiment, the three values ​​θ 3a ,θ 3b ,θ 3cThese are values ​​that have been subjected to low-pass filtering. In model-following control, by making the nominal model that constrains the controlled object 560 a model that includes the reduction mechanism 544, it is easier to reduce the error between the output value of the reduction mechanism 544 calculated from the first output value θ1 and the second output value θ2 obtained based on the second sensor 420. However, if the nominal model is a somewhat simplified model and cannot simulate the behavior of the reduction mechanism 544 in a high frequency band such as the second frequency band FB2, the modeling error Δ(s) between the nominal model and the controlled object 560 will increase in that high frequency band. Therefore, due to stability issues, it is difficult to use the second output value θ2 in model-following control in a high frequency band such as the second frequency band FB2. For this reason, in this embodiment, in a high frequency band such as the second frequency band FB2, model-following control is performed using the first output value θ1 which represents the output of the motor 543. However, in this case, the effect of making the nominal model a model that includes the reduction mechanism 544, that is, the effect of reducing the error between the output value of the reduction mechanism 544 calculated from the first output value θ1 and the second output value θ2 obtained based on the second sensor 420, cannot be obtained. Therefore, the three values ​​θ from which high-frequency components have been removed by low-pass filtering are not obtained. 3a ,θ 3b ,θ 3c By comparing these values, the predetermined threshold used to determine anomalies can be made smaller. Therefore, anomalies in the sensor can be detected more quickly. (Three values ​​θ) 3a ,θ 3b ,θ 3c The lower the cutoff frequency in the low-pass filter processing applied, the lower the error between the output value of the reduction mechanism 544 calculated from the first output value θ1 and the second output value θ2 obtained based on the second sensor 420. Therefore, the three values ​​θ 3a ,θ 3b ,θ 3c To reduce the predetermined threshold used for comparison between two values, three values ​​θ 3a ,θ 3b ,θ 3cThe cutoff frequency in the low-pass filter applied to the signal should be lowered. However, the lower the cutoff frequency, the more the three values ​​θ 3a ,θ 3b ,θ 3c The delay that occurs becomes larger, and the value θ at which the abnormality appears. 3a ,θ 3b ,θ 3c The timing of processing becomes delayed. Therefore, the time until an anomaly is detected tends to be longer. Hence, the three values ​​θ 3a ,θ 3b ,θ 3c The cutoff frequency in the low-pass filtering process applied is preferably determined based on the required fault-tolerant time interval (FTTI), etc.

[0114] According to this embodiment, the calculation unit 292 calculates the input value θ. i The second calculation process CP2, which calculates the input value θ, can be executed. If the abnormality determination unit 291 determines that no abnormality has occurred in at least one of the two first sensors 410 and the second sensor 420, the calculation unit 292 calculates the input value θ by the first calculation process CP1. i The calculation unit 292 calculates the input value θ using the second calculation process CP2. If the abnormality determination unit 291 determines that there is no abnormality in at least one of the two first sensors 410, and that there is an abnormality in the second sensor 420, the calculation unit 292 calculates the input value θ using the second calculation process CP2. i The first output value θ is calculated. In the second calculation process CP2, the calculation unit 292 calculates the first output value θ. 1c The input value θ is obtained by performing a division operation by dividing it by the reduction ratio N of the reduction mechanism 544. i In other words, the control method for controlling the controlled object 560 is the input value θ. i The second calculation process CP2 is executed to calculate the input value θ, and if the abnormality determination process determines that no abnormality has occurred in at least one of the two first sensors 410 and the second sensor 420, the first calculation process CP1 is executed to calculate the input value θ iThe second calculation process CP2 calculates the input value θ if, in the abnormality determination process, it is determined that no abnormality has occurred in at least one of the two first sensors 410, and an abnormality has occurred in the second sensor 420. i The second calculation process CP2 includes calculating the first output value θ. 1c The input value θ is obtained by performing a division operation by dividing it by the reduction ratio N of the reduction mechanism 544. i This includes the following. Therefore, even if an abnormality occurs in the second sensor 420, the assist control by the electric power steering device 1000 can be continued by using the first output value θ1 obtained based on the first sensor 410.

[0115] According to this embodiment, if the abnormality determination unit 291 determines that no abnormalities have occurred in the two first sensors 410, the calculation unit 292 determines the first output value θ used in the first calculation process CP1. 1c Two first output values ​​θ are obtained based on the two first sensors 410. 1a ,θ 1b The average value is used. In other words, the control method for controlling the controlled object 560 determines that no abnormalities have occurred in the two first sensors 410 during the abnormality determination process, and the first output value θ used in the first calculation process CP1 is used. 1c Two first output values ​​θ are obtained based on the two first sensors 410. 1a ,θ 1b This includes using the average value of the two first output values ​​θ. 1a ,θ 1b One of them is the first output value θ 1c Compared to when used as such, the first output value θ used in the first calculation process CP1 1c This can improve the accuracy of the first output value θ in the second calculation process CP2. 1a ,θ 1b The average value of the first output value θ 1c By doing so, the first output value θ used in the second calculation process CP2 is 1c This can improve accuracy.

[0116] According to this embodiment, if the abnormality determination unit 291 determines that an abnormality has occurred in one of the two first sensors 410 and that no abnormality has occurred in the other first sensor 410, the calculation unit 292 determines the first output value θ used in the first calculation process CP1. 1c The first output value θ1 obtained based on the other first sensor 410 is used as the first output value θ used in the first calculation process CP1 when the abnormality determination process determines that an abnormality has occurred in one of the two first sensors 410 and that no abnormality has occurred in the other first sensor 410. 1c This includes using a first output value θ1 obtained based on the other first sensor 410. Therefore, even if an abnormality occurs in one of the two first sensors 410, the first calculation process CP1 can be executed, and the assist control by the electric power steering device 1000 can continue. Similarly, in the second calculation process CP2, the first output value θ1 obtained based on the other first sensor 410 is also used, so even if an abnormality occurs in one of the two first sensors 410 and the second sensor 420, the second calculation process CP2 can be executed, and the assist control by the electric power steering device 1000 can continue.

[0117] At least some of the functions of each component of the control device 100 described above may be implemented by hardware including circuit sections such as LSI (Large Scale Integration), ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array), and GPU (Graphics Processing Unit), or by the cooperation of software and hardware. The storage unit that stores the program causing the processor 200 of the control device 100, which is a computer, to execute the control method described above, is implemented by a storage medium such as RAM (Random Access Memory), ROM (Read Only Memory), HDD (hard disk drive), and flash memory. The storage unit is not particularly limited as long as it can store the program causing the computer to execute the control method described above, and may be a microcomputer or a disk medium such as a CD-ROM. The storage unit may be provided separately from the control device 100. In this case, the control device 100 may communicate with the storage unit by wired communication or wireless communication and execute the program stored in the storage unit.

[0118] The present invention is not limited to the embodiments described above, and other configurations and methods may be adopted within the scope of the technical idea of ​​the present invention. In the first calculation process, the input value input to the inverse nominal model of the model following control unit may be calculated in any way, as long as it is calculated based on a first output value obtained based on a first sensor and indicating the output of the motor, and a second output value obtained based on a second sensor and indicating the output of the reduction mechanism. The cutoff frequency in the high-pass filter processing of the first calculation process may be different from the cutoff frequency in the low-pass filter processing of the first calculation process. The cutoff frequencies in the high-pass filter processing and the low-pass filter processing of the first calculation process are not particularly limited. The abnormality determination unit may determine abnormalities in the three sensors by comparing three values, including two first output values ​​and one second output value multiplied by the reduction ratio of the reduction mechanism. In other words, the abnormality determination process may include determining abnormalities in the three sensors by comparing three values, including two first output values ​​and one second output value multiplied by the reduction ratio of the reduction mechanism. The three values ​​compared in the anomaly detection unit may be values ​​that have not undergone low-pass filtering. The anomaly detection unit may not be provided.

[0119] Furthermore, this technology can be configured as follows: [1] A control device for controlling a portion of an electric power steering system mounted on a vehicle, which includes an input shaft to which a steering wheel operated by a driver is connected, an output shaft connected to the input shaft via a torsion bar, and a motor connected to the output shaft via a reduction mechanism, the control device comprising: a model following control unit that generates a correction torque to correct the input torque input to the control device based on a nominal model based on the configuration of the control device; and a calculation unit that calculates the input value input to the model following control unit, wherein the model following control unit calculates the inverse model of the nominal model A control device having an inverse nominal model to which the input value is input, and configured such that the transfer function of the controlled object is constrained to the transfer function of the nominal model in a frequency band where the complementary sensitivity gain, which is the gain in the gain characteristic of the complementary sensitivity function for the modeling error between the controlled object and the nominal model, is approximately 1, the calculation unit is capable of performing a first calculation process to calculate the input value, and in the first calculation process, the calculation unit calculates the input value based on a first output value obtained based on a first sensor and indicating the output of the motor, and a second output value obtained based on a second sensor and indicating the output of the reduction mechanism. [2] The control device according to [1], wherein in the first calculation process, the calculation unit calculates the input value by adding a value obtained by applying a high-pass filter process and a division process by dividing the first output value by the reduction ratio of the reduction mechanism, and a value obtained by applying a low-pass filter process to the second output value. [3] The control device according to [2], wherein the cutoff frequency in the high-pass filter processing is the same as the cutoff frequency in the low-pass filter processing. [4] The control device according to [2] or [3], wherein the cutoff frequency in the low-pass filter processing is 15 Hz or more and 50 Hz or less. [5] The control device according to any one of [1] to [4], further comprising an abnormality determination unit that determines whether or not an abnormality has occurred in the two first sensors and the one second sensor based on two first output values ​​obtained based on each of the two first sensors and one second output value obtained based on one second sensor, wherein the abnormality determination unit compares with each other three values ​​including two values ​​obtained by dividing the two first output values ​​by the reduction ratio of the reduction mechanism and one second output value, or three values ​​obtained by multiplying the two first output values ​​and one second output value by the reduction ratio, and determines that an abnormality has occurred in the sensor used to obtain the one value among the three sensors including the two first sensors and the one second sensor. [6] The control device according to [5], wherein each of the three values ​​has been subjected to a low-pass filter. [7] The control device according to [5] or [6], wherein the calculation unit is capable of performing a second calculation process for calculating the input value, and if the abnormality determination unit determines that no abnormality has occurred in at least one of the two first sensors and the second sensor, the calculation unit calculates the input value by the first calculation process, and if the abnormality determination unit determines that no abnormality has occurred in at least one of the two first sensors and that an abnormality has occurred in the second sensor, the calculation unit calculates the input value by the second calculation process, and in the second calculation process, the calculation unit performs a division process on the first output value by the reduction ratio of the reduction mechanism to obtain the input value. [8] If the abnormality determination unit determines that no abnormality has occurred in the two first sensors, the calculation unit uses the average value of the two first output values ​​obtained based on the two first sensors as the first output value used in the first calculation process, the control device according to any one of [5] to [7]. [9] The control device according to any one of [5] to [8], wherein the abnormality determination unit determines that an abnormality has occurred in one of the two first sensors and that no abnormality has occurred in the other first sensor, and the calculation unit uses the first output value obtained based on the other first sensor as the first output value used in the first calculation process. A motor device comprising a control device described in any one of items [1] to [9], and the motor. An electric power steering system comprising the motor device described in

[11]

[10] and a steering mechanism having the input shaft, the output shaft, and the torsion bar.

[12] A control method for controlling a portion of an electric power steering system mounted on a vehicle, which includes an input shaft to which a steering wheel operated by a driver is connected, an output shaft connected to the input shaft via a torsion bar, and a motor connected to the output shaft via a reduction mechanism, the control method comprising: performing model following control to generate a correction torque for correcting the input torque input to the control object based on a nominal model based on the configuration of the control object; and correcting the modeling error between the control object and the nominal model by the model following control. A control method comprising: constraining the transfer function of the controlled object to the transfer function of the nominal model in a frequency band where the complementary sensitivity gain, which is the gain in the gain characteristics of the complementary sensitivity function with respect to the difference, is approximately 1; and performing a first calculation process to calculate the input value to be input in the model-following control to an inverse nominal model, which is the inverse model of the nominal model, wherein the first calculation process includes calculating the input value based on a first output value obtained based on a first sensor and indicating the output of the motor, and a second output value obtained based on a second sensor and indicating the output of the reduction mechanism.

[13] The control method according to

[12] , wherein the first calculation process includes adding a value obtained by applying a high-pass filter process and a division process by dividing the first output value by the reduction ratio of the reduction mechanism, and a value obtained by applying a low-pass filter process to the second output value, to calculate the input value.

[14] The control method according to

[13] , wherein the cutoff frequency in the high-pass filter processing is the same as the cutoff frequency in the low-pass filter processing.

[15] The control method according to any one of

[12] to

[14] , comprising an abnormality determination process for determining whether an abnormality has occurred in the two first sensors and the one second sensor, based on two first output values ​​obtained based on each of the two first sensors and one second output value obtained based on the one second sensor, wherein the abnormality determination process compares with each other three values ​​including two values ​​obtained by dividing the two first output values ​​by the reduction ratio of the reduction mechanism and one second output value, or three values ​​obtained by multiplying the two first output values ​​and one second output value by the reduction ratio, and if one of the three values ​​differs from the other two values ​​by a predetermined threshold or more, it is determined that an abnormality has occurred in the sensor used to obtain the one of the three sensors including the two first sensors and the one second sensor.

[16] The control method described in

[15] , wherein the three values ​​are each subjected to low-pass filtering.

[17] The control method according to

[15] or

[16] , comprising: executing a second calculation process for calculating the input value; calculating the input value by the first calculation process if it is determined in the abnormality determination process that no abnormality has occurred in at least one of the two first sensors and the second sensor; and calculating the input value by the second calculation process if it is determined in the abnormality determination process that no abnormality has occurred in at least one of the two first sensors and that an abnormality has occurred in the second sensor, wherein the second calculation process includes performing a division operation on the first output value by the reduction ratio of the reduction mechanism to obtain the input value.

[18] In the abnormal determination process, when it is determined that no abnormality has occurred in the two first sensors, the average value of the two first output values respectively obtained based on the two first sensors is used as the first output value used in the first arithmetic processing, the control method according to any one of

[15] to

[17] .

[19] In the abnormal determination process, when it is determined that an abnormality has occurred in one of the two first sensors and no abnormality has occurred in the other first sensor, the first output value obtained based on the other first sensor is used as the first output value used in the first arithmetic processing, the control method according to any one of

[15] to

[18] .

[20] A program for causing a computer to execute the control method according to any one of

[12] to

[19] .

[0120] As described above, the configurations and methods described in this specification can be appropriately combined within a range that does not conflict with each other.

Explanation of symbols

[0121] 100... control device, 100a... motor device, 230... model following control unit, 231... inverse nominal model, 291... abnormal determination unit, 292... arithmetic unit, 410... first sensor, 420... second sensor, 521... steering wheel, 524a... input shaft, 524b... output shaft, 530... steering mechanism, 543... motor, 544... reduction mechanism, 546... torsion bar, 560... control target, 1000... electric power steering device, Cf3... cut-off frequency, CP1... first arithmetic processing, CP2... second arithmetic processing, GT... complementary sensitivity gain, N... reduction ratio, T(s)... complementary sensitivity function, T f ... correction torque, T r ... input torque, θ1,θ 1a ,θ 1b ,θ 1c ... first output value, θ2... second output value, θ 3a ,θ 3b ,θ 3c ... three values, θ i ... input value

Claims

1. A control device for controlling the portion of an electric power steering system mounted on a vehicle, which includes an input shaft to which a steering wheel operated by the driver is connected, an output shaft connected to the input shaft via a torsion bar, and a motor connected to the output shaft via a reduction mechanism, the portion including the motor and the reduction mechanism being controlled, A model-following control unit generates a correction torque for correcting the input torque input to the controlled object based on a nominal model based on the configuration of the controlled object, A calculation unit that calculates the input value input to the model following control unit, Equipped with, The model-following control unit has an inverse nominal model which is the inverse model of the nominal model and to which the input value is input, and is configured such that the transfer function of the controlled object is constrained to the transfer function of the nominal model in a frequency band where the complementary sensitivity gain, which is the gain in the gain characteristic of the complementary sensitivity function for the modeling error between the controlled object and the nominal model, is approximately 1. The calculation unit is capable of performing a first calculation process to calculate the input value, In the first calculation process, the calculation unit calculates the input value based on a first output value obtained based on a first sensor and indicating the output of the motor, and a second output value obtained based on a second sensor and indicating the output of the reduction mechanism.

2. The control device according to claim 1, wherein in the first calculation process, the calculation unit calculates the input value by adding a value obtained by applying a high-pass filter process and a division process by dividing the first output value by the reduction ratio of the reduction mechanism, and a value obtained by applying a low-pass filter process to the second output value.

3. The control device according to claim 2, wherein the cutoff frequency in the high-pass filter processing is the same as the cutoff frequency in the low-pass filter processing.

4. The control device according to claim 2, wherein the cutoff frequency in the low-pass filter processing is 15 Hz or more and 50 Hz or less.

5. The system includes an abnormality determination unit that determines whether or not an abnormality has occurred in the two first sensors and the one second sensor, based on two first output values ​​obtained based on each of the two first sensors and one second output value obtained based on one of the second sensors. The control device according to claim 1, wherein the abnormality determination unit compares three values, which include two values ​​obtained by dividing the two first output values ​​by the reduction ratio of the reduction mechanism and one second output value, or three values ​​obtained by multiplying the two first output values ​​and one second output value by the reduction ratio, and determines that an abnormality has occurred in the sensor used to acquire the one value among the three sensors, which include the two first sensors and one second sensor.

6. The control device according to claim 5, wherein each of the three values ​​is a value that has been subjected to low-pass filtering.

7. The calculation unit is capable of performing a second calculation process to calculate the input value, If the abnormality determination unit determines that no abnormality has occurred in at least one of the two first sensors and the second sensor, the calculation unit calculates the input value by the first calculation process, If the abnormality determination unit determines that no abnormality has occurred in at least one of the two first sensors, and that an abnormality has occurred in the second sensor, the calculation unit calculates the input value by the second calculation process, The control device according to claim 5, wherein in the second calculation process, the calculation unit performs a division operation on the first output value by the reduction ratio of the reduction mechanism to obtain the value obtained from the first output value, and the result is the input value.

8. The control device according to claim 5, wherein, when the abnormality determination unit determines that no abnormality has occurred in the two first sensors, the calculation unit uses the average value of the two first output values ​​obtained based on the two first sensors as the first output value used in the first calculation process.

9. The control device according to claim 5, wherein if the abnormality determination unit determines that an abnormality has occurred in one of the two first sensors and no abnormality has occurred in the other first sensor, the calculation unit uses the first output value obtained based on the other first sensor as the first output value used in the first calculation process.

10. A control device according to any one of claims 1 to 9, The motor and, A motor device equipped with the following features.

11. The motor device according to claim 10, A steering mechanism having the input shaft, the output shaft, and the torsion bar, An electric power steering system equipped with this system.

12. A control method for controlling the portion of an electric power steering system mounted on a vehicle, which includes an input shaft to which a steering wheel operated by the driver is connected, an output shaft connected to the input shaft via a torsion bar, and a motor connected to the output shaft via a reduction mechanism, wherein the control target is the portion including the motor and the reduction mechanism. This involves performing model-following control to generate a correction torque for correcting the input torque input to the controlled object, based on a nominal model derived from the configuration of the controlled object. The model-following control constrains the transfer function of the controlled object to the transfer function of the nominal model in a frequency band where the complementary sensitivity gain, which is the gain in the gain characteristic of the complementary sensitivity function with respect to the modeling error between the controlled object and the nominal model, is approximately 1. The first calculation process is performed to calculate the input values ​​that are input in the model-following control to the inverse nominal model, which is the inverse model of the nominal model. Includes, The first calculation process is a control method that includes calculating the input value based on a first output value obtained based on a first sensor and indicating the output of the motor, and a second output value obtained based on a second sensor and indicating the output of the reduction mechanism.

13. The control method according to claim 12, wherein the first calculation process includes calculating the input value by adding a value obtained by applying a high-pass filter process and a division process by dividing the first output value by the reduction ratio of the reduction mechanism, and a value obtained by applying a low-pass filter process to the second output value.

14. The control method according to claim 13, wherein the cutoff frequency in the high-pass filter processing is the same as the cutoff frequency in the low-pass filter processing.

15. The process includes an abnormality determination process that determines whether or not an abnormality has occurred in the two first sensors and the one second sensor, based on two first output values ​​obtained based on each of the two first sensors and one second output value obtained based on one second sensor. The control method according to claim 12, wherein the abnormality determination process includes comparing three values, which are obtained by dividing the two first output values ​​by the reduction ratio of the reduction mechanism and one second output value, or three values, which are obtained by multiplying the two first output values ​​and one second output value by the reduction ratio, and determining that an abnormality has occurred in the sensor used to acquire the one value among the three sensors, which include the two first sensors and one second sensor, if one of the three values ​​differs from the other two values ​​by a predetermined threshold or more.

16. The control method according to claim 15, wherein each of the three values ​​is a value that has been subjected to a low-pass filter.

17. The second calculation process is performed to calculate the aforementioned input value, If the abnormality determination process determines that no abnormality has occurred in at least one of the two first sensors and the second sensor, the input value is calculated by the first calculation process. If, in the abnormality determination process, it is determined that no abnormality has occurred in at least one of the two first sensors, and an abnormality has occurred in the second sensor, the input value is calculated by the second calculation process. Includes, The control method according to claim 15, wherein the second calculation process includes performing a division operation on the first output value by the reduction ratio of the reduction mechanism to obtain the input value.

18. The control method according to claim 15, wherein, in the abnormality determination process, it is determined that no abnormality has occurred in the two first sensors, the average value of the two first output values ​​obtained based on the two first sensors is used as the first output value used in the first calculation process.

19. The control method according to claim 15, wherein, in the abnormality determination process, it is determined that an abnormality has occurred in one of the two first sensors and that no abnormality has occurred in the other first sensor, the first output value obtained based on the other first sensor is used as the first output value used in the first calculation process.

20. A program that causes a computer to execute the control method described in any one of claims 12 to 19.

Images