JP2011087402A - Motor controller and electric power steering device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a motor controller for controlling a motor precisely by estimating an angular velocity accurately regardless of a temperature change or the like of a motor with a brush. <P>SOLUTION: An angular velocity estimation value ωe is used for motor control of an electric power steering device, and is obtained as follows. A resistance value Rc is calculated from a current detection value Im and a voltage detection value Vm obtained for each control period in a steering holding state. When the number of times for calculating the resistance value Rc is not less than the prescribed Ca times during one steering holding period, an average of the resistance value Rc for the nearest Ca times is calculated as a new resistance value Rnew (S54). When the number of times for calculating the resistance value Rc is C times, which is smaller than Ca times, a weighted mean of an average R1 of the resistance value Rc for the nearest C times and an average Rold of the resistance value Rc for the Ca times in the last steering holding period is calculated as the new resistance value Rnew (S62). The angular velocity estimation value ωe is calculated by the new resistance value Rnew (S58). <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a motor control device that drives an electric motor such as a motor with a brush, and an electric power steering device using the same.

In the motor control device in the electric power steering apparatus, in determining the target value of the current to be passed through the motor, the angular speed of the motor corresponding to the steering speed is considered in addition to the steering torque and the vehicle speed. On the other hand, when controlling a brushless motor, an angle detection sensor for detecting the rotational position of the motor such as a resolver is usually used. Therefore, the motor angular velocity is obtained by differentiating the output signal of the angle detection sensor with respect to time. Can be obtained. However, in motor control that does not use an angle sensor, such as control of a motor with a brush, the motor current and the voltage between the motor terminals are detected, and the motor angular velocity ω is calculated based on the following equation.
ω = (V−I × R) / k (1)
Here, V is a voltage between motor terminals, I is a motor current, R is a motor resistance (resistance between motor terminals), and k is a counter electromotive force constant.

  The following conventional techniques are known in relation to the present invention. That is, in Patent Literature 1, in a power steering apparatus including a steering assist motor driven in accordance with a steering torque applied to a steered wheel, a detection result of a motor drive current and a terminal voltage detection corresponding thereto. The internal resistance of the motor is obtained from a plurality of data with the average value of the result, the back electromotive force is obtained by subtracting the product of the internal resistance and the drive current from the terminal voltage, and the rotational angular velocity of the motor is obtained from this back electromotive force. The structure of requesting is disclosed. In Patent Document 2 and Patent Document 3, in an electric power steering apparatus including a control unit that drives a motor in accordance with the magnitude of steering torque, the control unit detects the detected inter-terminal voltage and motor current. The resistance value between the motor terminals is obtained based on the value, the winding temperature of the motor is estimated based on the resistance value between the motor terminals, the back electromotive force constant is temperature-corrected based on the estimated winding temperature, A configuration is disclosed in which the motor angular velocity is estimated using the temperature-corrected counter electromotive force constant.

Japanese Patent Publication No. 7-25313 Patent No. 4258502 Patent No. 3767035

  By the way, the motor resistance R varies depending on temperature change, manufacturing variation, contact state between the brush and the commutator of the motor with brush, and the like. Further, the back electromotive force constant k is not necessarily constant, and fluctuates due to temperature changes, manufacturing variations, and the like. The estimation accuracy of the motor angular velocity ω according to the above equation (1) is lowered by such a change in the motor resistance R or the counter electromotive force constant k. As a result, for example, in the electric power steering apparatus using the motor control apparatus as described above, the accuracy of control using the motor angular velocity is lowered, and there is a possibility that steering assist cannot be performed satisfactorily.

  Therefore, an object of the present invention is to accurately estimate the motor angular velocity even if there is a temperature change or manufacturing variation in a motor driven without using a high-precision motor angle detection sensor such as a motor with a brush. It is to provide a motor control device capable of controlling the motor with high accuracy using an estimated value. Another object of the present invention is to provide an electric power steering apparatus provided with such a motor control device.

The first invention is a motor control device that calculates an estimated value of the rotational speed of the electric motor based on a counter electromotive force generated in an armature winding of the electric motor and drives the electric motor using the estimated value. And
Current detecting means for detecting a current flowing through the electric motor;
Voltage detecting means for detecting a voltage applied to the electric motor;
Based on a plurality of detection values sequentially acquired for a predetermined physical quantity when the electric motor is in a predetermined state, at least one of the resistance value of the electric motor and the back electromotive force constant, which are parameters in the circuit equation of the electric motor. Parameter value update means for calculating the value of one parameter as an update parameter value;
Using the updated parameter value calculated by the parameter value updating unit, the current detection value acquired by the current detection unit, and the voltage detection value acquired by the voltage detection unit, the electric motor based on the circuit equation An estimated value calculating means for calculating a speed estimated value indicating the rotational speed of the motor,
The parameter value updating means includes
When the number of detection values acquired for the physical quantity for a new calculation of the update parameter value is equal to or greater than a predetermined number, a new update parameter value is obtained by the first predetermined method using the acquired detection value. Calculate
When the number of detection values acquired for the physical quantity for a new calculation of the update parameter value is smaller than the predetermined number, the acquired detection value and the update parameter value already obtained are used. A new update parameter value is calculated by the second predetermined method,
When the new update parameter value is calculated by the parameter value update unit, the estimated value calculation unit calculates the speed estimated value using the new update parameter value.

According to a second invention, in the first invention,
A determination means for determining whether or not the rotation speed of the electric motor is equal to or less than a predetermined value near 0;
The parameter value updating means is sequentially acquired as the physical quantity detection value by the current detection means and voltage detection means when the determination means determines that the rotation speed of the electric motor is equal to or less than the predetermined value. Calculate a resistance value of the electric motor based on a plurality of current detection values and voltage detection values,
The first predetermined method obtains a resistance value of the electric motor at the detection time as a calculated resistance value from the current detection value and the voltage detection value at each detection time of the plurality of current detection values and voltage detection values, It is a method of calculating a moving average value for a predetermined number of the calculated resistance values as a new update parameter value,
In the second predetermined method, a resistance value of the electric motor at the detection time is obtained as a calculated resistance value from the current detection value and the voltage detection value at each detection time of the plurality of current detection values and voltage detection values. The weighted average value of the average value of the calculated resistance value and the update parameter value already obtained is calculated as a new update parameter value.

According to a third invention, in the first invention,
A simple speed detecting means for detecting the rotational speed of the electric motor with an accuracy lower than that required for drive control of the electric motor;
The parameter value update means includes a plurality of current detection values sequentially acquired by the current detection means and the voltage detection means as detection values of the physical quantity when the rotation speed detected by the simple speed detection means is a predetermined value or more. And a resistance value and a back electromotive force constant of the electric motor based on the voltage detection value and a plurality of rotation speed detection values sequentially obtained by the simple speed detection means,
In the first predetermined method, the current detection value, the voltage detection value, and the rotation speed detection value at each detection point of the plurality of current detection values and voltage detection values satisfy at least approximately the circuit equation. , Calculating the resistance value and back electromotive force constant of the electric motor as a new update parameter value,
In the second predetermined method, the current detection value, the voltage detection value, and the rotation speed detection value at each detection time of the plurality of current detection values and voltage detection values satisfy at least approximately the circuit equation. A method of calculating a resistance value and a back electromotive force constant of the electric motor as a temporary parameter value, and calculating a weighted average value of the temporary parameter value and the update parameter value already obtained as a new update parameter value It is characterized by being.

According to a fourth invention, in the first invention,
A simple angle detecting means for detecting the rotation angle of the electric motor with a resolution such that the rotation speed of the electric motor is obtained with an accuracy lower than the accuracy required for drive control of the electric motor;
The parameter value updating means changes the simple angle from the first time point to the first time point when the rotation angle detected by the simple angle detection means is changed and a new rotation angle is detected as the first rotation angle detection value. Until the second time point when the rotation angle detected by the detection unit changes and a new rotation angle is detected as the second rotation angle detection value, the current detection unit and the voltage detection unit sequentially acquire the detection value of the physical quantity. A resistance value of the electric motor is calculated based on a plurality of detected current values and detected voltage values,
The first predetermined method is based on an equation obtained by integrating the circuit equation from the first time point to the second time point, using the plurality of current detection values and voltage detection values and the first and second rotation angles. It is a method of calculating the resistance value of the electric motor as a new update parameter value,
The second predetermined method is based on an equation obtained by integrating the circuit equation from the first time point to the second time point using the plurality of current detection values and voltage detection values and the first and second rotation angles. It is a method of calculating a resistance value of the electric motor as a temporary parameter value, and calculating a weighted average value of the temporary parameter value and the already obtained update parameter value as a new update parameter value. To do.

According to a fifth invention, in the first invention,
A simple speed detecting means for detecting the rotational speed of the electric motor with an accuracy lower than that required for drive control of the electric motor;
The parameter value update means includes a plurality of current detection values sequentially acquired by the current detection means and the voltage detection means as detection values of the physical quantity when the rotation speed detected by the simple speed detection means is a predetermined value or more. And a voltage detection value and a plurality of rotation speed detection values sequentially obtained by the simple speed detection means, to calculate a counter electromotive force constant of the electric motor,
The first predetermined method uses the current detection value, the voltage detection value, and the rotation speed detection value at each detection time of the plurality of current detection values and voltage detection values to determine the detection time at the detection time from the circuit equation. A method of obtaining a counter electromotive force constant of an electric motor as a calculated counter electromotive force constant, and calculating a moving average value for the predetermined number of the calculated counter electromotive force constants as a new update parameter value,
The second predetermined method uses the current detection value, the voltage detection value, and the rotation speed detection value at each detection time of the plurality of current detection values and voltage detection values to determine the detection time at the detection time from the circuit equation. A back electromotive force constant of the electric motor is obtained as a calculated back electromotive force constant, and a weighted average value of the average value of the plurality of calculated back electromotive force constants and the update parameter value already obtained is used as a new update parameter value. It is a method of calculating.

According to a sixth invention, in the first invention,
Temperature detecting means for detecting the temperature of the component of the device corresponding to each of the plurality of resistance components constituting the resistance of the electric motor as the physical quantity;
The parameter value updating means includes
Including temperature characteristic storage means for individually holding the temperature characteristics of the plurality of resistance components,
Calculate a resistance value of the electric motor based on the temperature characteristics using a plurality of temperature detection values sequentially acquired for each component by the temperature detection means,
The first predetermined method uses the temperature detection value at each detection time of the plurality of temperature detection values sequentially acquired for each component, and detects the detection based on the temperature characteristic of the resistance component corresponding to the component. The value of the resistance component at the time is acquired, the sum of the values for the plurality of resistance components is obtained as the calculated resistance value of the electric motor, and the moving average value for the predetermined number of calculated resistance values is newly updated. It is a method of calculating as a parameter value,
The second predetermined method uses the temperature detection value at each detection time of the plurality of temperature detection values sequentially acquired for each component, and detects the detection based on a temperature characteristic of a resistance component corresponding to the component. The value of the resistance component at the time is acquired, the sum of the values for the plurality of resistance components is obtained as the calculated resistance value of the electric motor, and the average value for the plurality of calculated resistance values is already obtained. The method is characterized in that a weighted average value with the update parameter value is calculated as a new update parameter value.

A seventh aspect of the invention is an electric power steering device that applies a steering assist force by an electric motor to a steering mechanism of a vehicle,
A motor control device according to any one of the first to sixth inventions;
The motor control device drives an electric motor that applies the steering assist force to the steering mechanism.

  In the motor control device according to the first aspect of the invention, the number of acquired detection values of the predetermined physical quantity required for new calculation of the update parameter value used for calculating the speed estimation value indicating the rotation speed of the electric motor is predetermined. When the number is greater than or equal to the number, a new update parameter value is calculated by the first predetermined method using those detected values. As a result, a stable update parameter value with less noise can be obtained, so that the speed estimation value can be calculated with high accuracy. Further, when the number of detection values of the physical quantity required for new calculation of the update parameter value is smaller than a predetermined number, the update value parameter value already obtained is used in addition to the acquired detection value. Thus, a new update parameter value is calculated by the second predetermined method. Therefore, it is possible to calculate a stable update parameter value with less noise even when the number of acquired physical quantity detection values is small or when the period of a predetermined state in which the physical quantity can be detected is short. Can do. Therefore, according to the first invention, even if there is an individual difference in the electric motor, the temperature changes or the electric motor is a brushed motor, and the contact state between the brush and the commutator changes. However, the estimated speed value of the electric motor is calculated with high accuracy, thereby improving the control performance of the motor control device.

  In the motor control device according to the second aspect of the invention, when the rotation speed of the electric motor is equal to or less than a predetermined value near 0, a current detection value and a voltage detection are detected as detection values of the physical quantity necessary for calculating a new update parameter value. When a value is acquired and the number of detection values acquired is equal to or greater than a predetermined number, the resistance value of the electric motor at the detection time is calculated based on the current detection value and the voltage detection value at each detection time of the detection values. The moving average value for the predetermined number of calculated resistance values is calculated as a new update parameter value. Further, even when the number of acquired current detection values and voltage detection values for the new calculation of the resistance value of the electric motor as the update parameter value is smaller than the predetermined number, the acquired current detection The resistance value of the electric motor at the detection time is calculated as a calculated resistance value based on the current detection and voltage detection at each detection time of the value and the voltage detection value, and the average value for the calculated resistance value of the acquired number A weighted average value with the already obtained update parameter value is calculated as a new update parameter value. Therefore, regardless of the number of acquired current detection values and voltage detection values for calculating a new update parameter value, the period during which the rotation speed of the electric motor is equal to or less than a predetermined value near 0 (acquisition of calculated resistance value). Regardless of the length of the possible period, a stable update parameter value (resistance value of the electric motor) with less noise can be calculated with high frequency. According to the second invention as described above, even if there is an individual difference in the electric motor, the temperature changes, or the electric motor is a brushed motor, and the contact state between the brush and the commutator changes. However, since the update parameter value follows the change and the speed estimation value of the electric motor is calculated with high accuracy, the control performance of the motor control device is improved.

  In the motor control device according to the third aspect of the present invention, when the rotational speed detected by the simple speed detecting means is equal to or higher than a predetermined value, a plurality of current detections are detected as detection values of the physical quantity necessary for calculating a new update parameter value. Value and voltage detection value are acquired sequentially, and when the number of detection values acquired is equal to or greater than a predetermined number, the current detection value and voltage detection value at each detection time of those detection values and the rotation speed detection value by the simple speed detection means And the resistance value of the electric motor and the back electromotive force constant are calculated as new update parameter values so that the circuit equation of the electric motor at least approximately satisfies the circuit equation. As a result, it is possible to obtain not only the resistance value of the electric motor but also the counter electromotive force constant as a stable update parameter value with less noise by using a low-cost simple speed detection means. Further, when the number of acquired current detection values and voltage detection values for the new calculation of the updated parameter value is less than a predetermined number, the resistance value and back electromotive force of the electric motor obtained in the same manner as described above The constant is set as a temporary parameter value, and a weighted average value between the temporary parameter value and an already obtained update parameter value is calculated as a new update parameter value. Therefore, regardless of the number of acquired current detection values and voltage detection values for calculating a new update parameter value, and regardless of the length of the period when the rotation speed is equal to or higher than the predetermined value, the noise is stable and stable. As the update parameter value, not only the resistance value of the electric motor but also the counter electromotive force constant can be calculated with high frequency. According to the third aspect of the invention, even if there is an individual difference in the electric motor, the temperature changes or the electric motor is a brushed motor, and the contact state between the brush and the commutator changes. However, since the update parameter value follows the change and the estimated speed value of the electric motor is calculated with higher accuracy, the control performance of the motor control device is further improved.

  In the motor control device according to the fourth aspect of the invention, the physical quantity necessary for calculating a new update parameter value between two time points when the rotation angle changes every time the rotation angle detected by the simple angle detection means changes. The resistance value of the electric motor is calculated from an equation obtained by integrating the circuit equation of the electric motor with time based on a plurality of current detection values and voltage detection values sequentially obtained as the detection values. The calculated resistance value of the electric motor becomes a new update parameter value when the number of acquired current detection values and voltage detection values is a predetermined number or more. Thereby, the resistance value of the electric motor can be obtained as a stable update parameter value with less noise by using a simple angle detection means with low cost and low resolution. In addition, when the number of current detection values and voltage detection values obtained for new calculation of the updated parameter value is less than a predetermined number, the resistance value of the electric motor obtained in the same manner as described above is the temporary parameter value. Then, a weighted average value between the temporary parameter value and the already obtained update parameter value is calculated as a new update parameter value. Therefore, regardless of the number of current detection values and voltage detection values obtained for calculating a new update parameter value, and regardless of the length of the period when the rotation speed is equal to or higher than the predetermined value, stable update with less noise The parameter value (electric motor resistance value) can be calculated with high frequency. According to the fourth aspect of the invention, even if there is an individual difference in the electric motor, the temperature changes, the electric motor is a brushed motor, and the contact state between the brush and the commutator changes. However, since the update parameter value follows the change and the speed estimation value of the electric motor is calculated with high accuracy, the control performance of the motor control device is improved.

  In the motor control device according to the fifth aspect of the invention, when the rotational speed detected by the simple speed detecting means is equal to or higher than a predetermined value, a plurality of current detections are detected as the physical quantity detection values necessary for calculating a new update parameter value. Value and voltage detection value are sequentially acquired, and when the number of detection values acquired is equal to or greater than a predetermined number, the current detection value and voltage detection value at each detection time of these detection values and the rotation speed detection value by the simple speed detection means And the counter electromotive force constant at the detection time is obtained as a calculated counter electromotive force constant from the circuit equation of the electric motor, and the moving average value for the predetermined number of calculated counter electromotive force constants is a new update parameter value. Is calculated as Thereby, the counter electromotive force constant of the electric motor can be obtained as a stable updated parameter value with less noise by using a low-cost simple speed detecting means. Further, even when the number of acquired current detection values and voltage detection values for the new calculation of the update parameter value is smaller than the predetermined number, the calculation back-up occurs for each detection time point in the same manner as described above. A power constant is calculated, and a weighted average value between the average value for the calculated counter electromotive force constant of the acquired number and the update parameter value already obtained is calculated as a new update parameter value. Therefore, regardless of the number of current detection values and voltage detection values obtained for calculating a new update parameter value, and regardless of the length of the period when the rotation speed is equal to or higher than the predetermined value, stable update with less noise The parameter value (back electromotive force constant of the electric motor) can be calculated with high frequency. According to the fifth aspect of the invention, even if there is an individual difference in the electric motor, the temperature changes, or the electric motor is a brushed motor, and the contact state between the brush and the commutator changes. However, since the update parameter value follows the change and the speed estimation value of the electric motor is calculated with high accuracy, the control performance of the motor control device is improved.

  In the motor control device according to the sixth aspect of the invention, when the electric motor is in a predetermined state, a plurality of temperature detection values are sequentially acquired as detection values of the physical quantity necessary for calculating a new update parameter value. When the obtained number is equal to or greater than the predetermined number, the temperature detection values at the detection times of these detection values are used to determine the respective values at the detection time based on the temperature characteristics of the plurality of resistance components constituting the resistance of the electric motor. The value of the resistance component is acquired, the sum of the values for the plurality of resistance components is obtained as the calculated resistance value of the electric motor, and the moving average value for the predetermined number of calculated resistance values is used as the new update parameter value. Calculated. This makes it possible to obtain a calculated resistance value that accurately reflects the temperature characteristics of the motor resistance by considering the temperature characteristics of each resistance component individually, and the moving average value for the calculated resistance value is a stable update parameter with less noise. Calculated as a value. In addition, even when the number of detected temperature values for new calculation of the updated parameter value is less than the predetermined number, the calculated resistance value is calculated for each detection time in the same manner as described above. A weighted average value of the average value for the calculated resistance value of the acquired number and the update parameter value already obtained is calculated as a new update parameter value. Therefore, a stable update parameter value (electric motor resistance value) with less noise can be calculated with high frequency regardless of the number of acquired temperature detection values for calculating a new update parameter value. According to the sixth aspect of the invention, even if the temperature changes, the updated parameter value accurately follows the resistance value change of the electric motor due to the temperature change, and the estimated speed value of the electric motor is accurately calculated. Therefore, the control performance of the motor control device is improved.

  According to the seventh aspect of the invention, the number of acquired detection values of the predetermined physical quantity necessary for new calculation of the update parameter value used for calculating the speed estimation value indicating the rotation speed of the electric motor in the electric power steering apparatus. Even when there is a small amount of noise, even when the period of the predetermined state in which the physical quantity can be detected is short, a stable update parameter value with little noise can be calculated. Accordingly, the estimated speed value of the electric motor is calculated with high accuracy, thereby improving the control performance of the motor control device and obtaining a good steering feeling in the electric power steering device.

It is the schematic which shows the structure of the electric power steering apparatus using the motor control apparatus which concerns on this invention with the structure of the vehicle relevant to it. It is a block diagram which shows the structure of the motor control apparatus which concerns on the 1st Embodiment of this invention. It is a flowchart which shows the motor control process for implement | achieving the control part in the said 1st Embodiment. It is a wave form diagram for demonstrating the basic angular velocity estimation process used as the foundation of the angular velocity estimation process in the said 1st Embodiment. It is a block diagram which shows an example of a functional structure of the angular velocity estimation part in the said 1st Embodiment. It is a flowchart which shows the said basic angular velocity estimation process about the said 1st Embodiment. It is a figure for demonstrating the update method of the resistance map in the said 1st Embodiment. It is a flowchart which shows the 1st angular velocity estimation process for implement | achieving the angular velocity estimation part in the said 1st Embodiment. It is a flowchart which shows the 2nd angular velocity estimation process for implement | achieving the angular velocity estimation part in the said 1st Embodiment. It is a block diagram which shows one structural example of the angular velocity estimation part in the 2nd Embodiment of this invention. It is a flowchart which shows the 1st angular velocity estimation process for implement | achieving the angular velocity estimation part in the said 2nd Embodiment. It is a flowchart which shows the 2nd angular velocity estimation process for implement | achieving the angular velocity estimation part in the said 2nd Embodiment. It is a flowchart which shows the 3rd angular velocity estimation process for implement | achieving the angular velocity estimation part in the said 2nd Embodiment. It is a block diagram which shows one structural example of the angular velocity estimation part in the 3rd Embodiment of this invention. It is a flowchart which shows the 1st angular velocity estimation process for implement | achieving the angular velocity estimation part in the said 3rd Embodiment. It is a flowchart which shows the 2nd angular velocity estimation process for implement | achieving the angular velocity estimation part in the said 3rd Embodiment. It is a block diagram which shows one structural example of the angular velocity estimation part in the 4th Embodiment of this invention. It is a flowchart which shows an example of the angular velocity estimation process for implement | achieving the angular velocity estimation part in the said 4th Embodiment. It is a block diagram which shows one structural example of the angular velocity estimation part in the 5th Embodiment of this invention. It is a characteristic view which shows the temperature characteristic of each resistance value used with the resistance value update part of the angular velocity estimation part in the said 5th Embodiment. It is a flowchart which shows an example of the angular velocity estimation process for implement | achieving the angular velocity estimation part in the said 5th Embodiment. It is a flowchart which shows the resistance value calculation process in the angular velocity estimation process for implement | achieving the angular velocity estimation part in the said 5th Embodiment.

<1. Electric power steering device>
FIG. 1 is a schematic diagram showing the configuration of an electric power steering device using a motor control device according to the present invention, together with the configuration of a vehicle related thereto. This electric power steering apparatus is a column assist type electric power steering apparatus including a brushed motor 1, a speed reducer 2, a torque sensor 3, a vehicle speed sensor 4, and an electronic control unit (ECU) 5 as a motor control apparatus. is there. The motor control device according to the present invention preferably includes a sensor (hereinafter referred to as “rotation sensor”) 7 for detecting the presence or absence of rotation of the motor 1, and the electric power steering device shown in FIG. A hall sensor is provided as the rotation sensor 7. The rotation sensor 7 is not limited to a hall sensor, and any sensor that generates a pulse signal according to the rotation of the motor 1 may be used. Further, since it is not necessary to detect the rotation direction of the motor 1, a low-cost sensor can be used as the rotation sensor 7. Furthermore, this electric power steering apparatus includes a steering angle sensor (or an operation angular velocity sensor) 7h. However, one of the rotation sensor 7 and the steering angle sensor 7h can be omitted (details will be described later).

  As shown in FIG. 1, a steering wheel (steering wheel) 101 is fixed to one end of the steering shaft 102, and the other end of the steering shaft 102 is connected to a rack shaft 104 via a rack and pinion mechanism 103. Both ends of the rack shaft 104 are connected to a wheel 106 via a connecting member 105 composed of a tie rod and a knuckle arm. When the driver rotates the handle 101, the steering shaft 102 rotates, and the rack shaft 104 reciprocates accordingly. As the rack shaft 104 reciprocates, the direction of the wheels 106 changes.

  The electric power steering device performs the following steering assistance in order to reduce the driver's load. The torque sensor 3 detects the steering torque Ts applied to the steering shaft 102 by the operation of the handle 101, the steering angle sensor 7h detects the steering angle θh indicating the operation amount of the handle 101, and the vehicle speed sensor 4 detects the vehicle speed S. To detect. The rotation sensor 7 generates pulses at a frequency corresponding to the rotation speed of the rotor of the motor 1 and outputs a pulse signal P including these pulses. When the rotor of the motor 1 is stopped, the rotation sensor 7 does not generate a pulse.

  The ECU 5 receives power supplied from the in-vehicle battery 8 and drives the brushed motor 1 based on the steering torque Ts, the vehicle speed S, and the pulse signal P (or the steering angle θh). The brushed motor 1 generates a steering assist force when driven by the ECU 5. The speed reducer 2 is provided between the brushed motor 1 and the steering shaft 102. The steering assist force generated by the brushed motor 1 acts to rotate the steering shaft 102 via the speed reducer 2.

  As a result, the steering shaft 102 rotates by both the steering torque applied to the handle 101 and the steering assist force generated by the brushed motor 1. As described above, the electric power steering apparatus performs steering assist by applying the steering assist force generated by the brushed motor 1 to the steering mechanism of the vehicle.

<2. First Embodiment>
<2.1 Motor controller configuration>
FIG. 2 is a block diagram showing the configuration of the motor control device according to the first embodiment of the present invention. This motor control device is used in the electric power steering device and is configured using the ECU 5 to drive the brushed motor 1. The ECU 5 includes a control unit 10 and a drive unit 20. The control unit 10 is realized using a microcomputer (hereinafter abbreviated as “microcomputer”). The drive unit 20 includes a motor drive circuit 30, a current detection resistor 35, a PWM signal generation circuit 21, a voltage detection circuit 24, and a current detection circuit 25. In the present embodiment, it is assumed that the rotation sensor 7 is used and the rudder angle sensor 7h is not used. In addition, the temperature sensors 41, 42, and 43 indicated by dotted lines in FIG. 2 are not used in this embodiment (the temperature sensors 41, 42, and 43 are used in a fifth embodiment described later).

  The ECU 5 receives the steering torque Ts output from the torque sensor 3 and the vehicle speed S output from the vehicle speed sensor 4, and these are given as input data to the microcomputer 10. In the drive unit 20, as will be described later, the voltage applied to the motor 1, that is, the voltage across the terminals of the motor 1 (hereinafter referred to as “motor voltage”) is detected by the voltage detection circuit 24, and the motor 1 (internal electric machine) A current (hereinafter referred to as “motor current”) flowing through the child winding is detected by the current detection circuit 25, and a voltage detection value Vm and a current detection value Im as these detection results are also given to the microcomputer 10.

  The microcomputer 10 executes a program stored in a memory (not shown) built in the ECU 5, thereby setting a target current setting unit 12, a subtractor 14, a control calculation unit as functional units for controlling the motor 1. 16 and the angular velocity estimation unit 50 are realized by software. Based on the voltage detection value Vm, the current detection value Im, and the pulse signal P output from the rotation sensor 7, the angular velocity estimation unit 50 calculates an angular velocity estimation value ωe indicating the rotation speed of the motor 1 (the rotor). To do. The target current setting unit 12 determines a target value It of current to be passed through the motor 1 based on the steering torque Ts and the vehicle speed S and the estimated angular velocity value ωe. The subtractor 14 calculates a deviation (It−Im) between the current target value It and the current detection value Im. The control calculation unit 16 calculates a command value D indicating a voltage to be applied to the motor 1 in order to cancel the deviation (It-Im) by a proportional-integral control calculation based on the deviation (It-Im). The command value D is output from the microcomputer 10 and given to the PWM signal generation circuit 21 of the drive unit 20.

  The PWM signal generation circuit 21 generates first and second rightward PWM signals SRd1 and SRd2 and first and second leftward PWM signals SLd1 and SLd2 according to a command value D given from the microcomputer 10. . Here, the sign of the command value D indicates that the torque to be generated by the motor 1 is a torque in a direction assisting rightward steering (hereinafter referred to as “rightward torque”) or a torque in a direction assisting leftward steering (hereinafter “left”). Direction torque). When generating the right direction torque, the right direction PWM signals SRd1 and SRd2 are generated as PWM signals having a duty ratio corresponding to the command value D, and the left direction PWM signals SLd1 and SLd2 are generated as inactive signals. When generating the left direction torque, the left direction PWM signals SLd1 and SLd2 are generated as PWM signals having a duty ratio corresponding to the command value D, and the right direction PWM signals SRd1 and SRd2 are generated as inactive signals.

  The motor drive circuit 30 is a bridge circuit composed of power field-effect transistors (hereinafter referred to as “FETs”) 31 to 34 that are four switching elements, and this bridge circuit is connected to the power line of the battery 8 and the ground. The motor 1 is connected as a load. In other words, this bridge circuit includes power supply line side FETs 31 and 32 connected to the power supply line, and ground line side FETs 33 and 34 connected to the ground line via the current detection resistor 35. The positive terminal of the motor 1 is connected to the connection point N1 with the ground line side FET 33, and the negative terminal of the motor 1 is connected to the connection point N2 between the power line side FET 32 and the ground line side FET 34. The right direction PWM signals SRd1 and SRd2 are applied to the gate terminals of the power supply line side FET 31 and the ground line side FET 34, respectively, and the left direction PWM signal is applied to the gate terminals of the power supply line side FET 32 and the ground line side FET 33. SLd1 and SLd2 are applied, respectively. As a result, the FETs 31 and 34 or the FETs 32 and 33 are turned on and off at a duty ratio corresponding to the command value D, whereby a voltage having a polarity and a magnitude corresponding to the sign of the command value D is transferred from the motor drive circuit 30 to the motor 1. Applied. As a result, a current is supplied from the motor drive circuit 30 to the motor 1, and the motor 1 generates a steering assist force according to the steering torque Ts, the vehicle speed S, and the estimated angular velocity value ωe.

  While the motor 1 is driven as described above, the voltage applied to the motor 1, that is, the voltage across the terminals of the motor 1 is detected by the voltage detection circuit 24, and the current flowing through the motor 1 is between both ends of the current detection resistor 35. Is detected by the current detection circuit 25 based on the voltage of. The detected voltage detection value Vm and detected current value Im are input to the microcomputer 10 and, as described above, calculation of the angular velocity estimation value ωe and the above deviation (It−Im) for the feedback control of the motor 1. Is done.

<2.2 Motor control processing>
FIG. 3 is a flowchart showing a motor control process for realizing a control unit (ECU) 5 for controlling the driving of the motor 1 in software in the present embodiment. In this motor control process, first, information necessary for motor control such as vehicle information is acquired (step S2). The information obtained here includes parameter values (for example, the initial value of the back electromotive force constant k and the motor resistance value R, which are necessary for calculation in the control calculation unit 16 and calculation of the angular velocity estimation value ωe in the angular velocity estimation unit 50, Etc.), and flags and variables used in an angular velocity estimation process (see FIG. 6, etc.) described later are also initialized in step S2 (for example, Fh = 0, C = 0, Cn = 50, etc.).

  After the execution of step S2, the cycle of calculating the target value It of the current to be supplied to the motor 1 based on the output signals from the sensors 3, 4, 7 and the detection circuits 24, 25 in step S3. Steering torque Ts, vehicle speed S, pulse signal P, voltage detection value Vm, and current detection value Im are sequentially acquired for each control cycle.

  In the next step S4, an angular velocity estimation process for realizing the angular velocity estimation unit 50 is executed, and thereby an angular velocity estimation value ωe of the motor 1 is calculated.

  Thereafter, in step S6, processing for current control of the motor 1 and the like is performed using the estimated angular velocity value ωe. That is, in step S6, the current target value It is determined based on the angular velocity estimated value ωe, the processing steering torque Ts, and the vehicle speed S, and the deviation (It−Im) between the current target value It and the current detection value Im is determined. A command value D indicating a voltage to be applied to the motor 1 is calculated by the proportional integral control calculation based on the calculation. That is, the target current setting unit 12, the subtracter 14, and the control calculation unit 16 shown in FIG.

  After such processing as current control, the process returns to step S3. Thereafter, while the motor control unit (ECU) 5 in the electric power steering apparatus is operating, steps S3, S4, and S6 are repeatedly executed in the control cycle described above.

<2.3 Configuration and operation of angular velocity estimation unit>
FIG. 4 is a waveform diagram for explaining the basic angular velocity estimation processing that is the basis of the angular velocity estimation processing for realizing the angular velocity estimation unit 50 in the present embodiment. FIG. 4A shows a pulse output from the rotation sensor 7 when the steering angle θh changes as shown in FIG. 4B in the electric power steering apparatus using the motor control apparatus according to the present embodiment as the ECU 5. FIG. 4C shows the waveform of the signal P. FIG. 4C shows the change in the count value C of the counter 535 in the average value calculation unit 53 described later when the steering angle θh changes as shown in FIG. 4B. Show.

  FIG. 5 is a block diagram illustrating an example of a functional configuration of the angular velocity estimation unit 50 according to the present embodiment. As illustrated in FIG. 5, the angular velocity estimation unit 50 includes a steering determination unit 51, a resistance calculation unit 52, An average value calculating unit 53 and an angular velocity estimated value calculating unit (hereinafter abbreviated as “estimated value calculating unit”) 56 are provided, and the steering determination unit 51, the resistance calculating unit 52, and the average value calculating unit 53 are resistant to each other. The value update unit 200 is configured. The voltage detection value Vm and the current detection value Im obtained by the voltage detection circuit 24 and the current detection circuit 25 are input to the resistance calculation unit 52 and the estimated value calculation unit 56.

  In the basic angular velocity estimation process, the pulse signal P from the rotation sensor 7 is input to the steering determination unit 51, and the steering determination unit 51 does not output a pulse for a predetermined time ΔT1. Assuming that the rotation speed of the motor 1 is equal to or less than a predetermined value near 0, it is determined that the vehicle is in the steered state. If the rotation sensor 7 outputs a pulse during the time ΔT1, it is determined that the vehicle is in the steering state (FIG. 4 (a)). This determination result is input to the resistance calculation unit 52 and the average value calculation unit 53 as the rotation state signal St. Further, in the basic angular velocity estimation process, the steering is detected when the rotation sensor 7 does not output a pulse during the time ΔT1, but when a pulse is output from the rotation sensor 7 after the detection, at that time The rotation state signal St changes to indicate the steering state.

In general, the angular velocity ω of the rotor of the brushed motor 1 is given by the following equation as described above.
ω = (V−I × R) / k (2)
Here, V is a motor voltage (terminal voltage), I is a motor current, R is a motor resistance (terminal resistance), and k is a counter electromotive force constant. Therefore, if the value of the motor resistance R is known in addition to the voltage detection value Vm and the current detection value Im, the angular velocity ω can be calculated.

When the electric power steering device is in the steering-holding state and the rotor of the motor 1 is stopped, the angular velocity ω = 0, so the motor resistance R is calculated by the following equation (3) based on the above equation (2). Can do.
R = V / I (3)
Therefore, when the rotation state signal St indicates the steered state, the resistance calculation unit 52 regards the angular velocity ω = 0 and substitutes V = Vm and I = Im into the above equation (3) to set the value of the motor resistance R. (Hereinafter, this value is referred to as “calculated resistance value” and is indicated by the symbol “Rc”). The calculated resistance value Rc (= Vm / Im) thus obtained is input to the average value calculation unit 53.

  Based on the rotation state signal St, the average value calculation unit 53 is a calculation that is input from the resistance calculation unit 52 for each control cycle during a period in which the electric power steering device is in the steering state (hereinafter referred to as “steering period”). The average value of the resistance value Rc is obtained and output as the resistance average value Rnew. A specific method for calculating the resistance average value Rnew will be described later.

  The estimated value calculation unit 56 calculates the angular velocity estimated value ωe by the above equation (2) using the resistance average value Rnew together with the voltage detection value Vm and the current detection value Im. That is, the estimated value calculation unit 56 calculates the angular velocity estimated value ωe by substituting V = Vm, I = Im, and R = Rnew into the above equation (2). In the present embodiment, it is assumed that the back electromotive force constant k is set to a known value in step S2 of FIG.

  In the above configuration, the estimated angular velocity value ωe is calculated using the above resistance average value Rnew as the motor resistance value R in the above formula (2). Instead, refer to a resistance map as shown in FIG. The estimated angular velocity value ωe may be calculated using the resistance value Rm thus obtained.

  In the case of a configuration in which the resistance value Rm is obtained with reference to the resistance map, the angular velocity estimation unit 50 includes a map update unit 54 and a map holding unit 55 in addition to the above-described components 51, 52, 53, and 56. . In this case, the current detection value Im used for the calculation of the calculated resistance value Rc in the resistance calculation unit 52 is also input to the average value calculation unit 53 for each control cycle as the calculated current value Ic, and the average value calculation unit 53 During the steering period, the average value of the calculated current value Ic is calculated as the current average value Inew, and the current average value Inew is also output for each control cycle together with the resistance average value Rnew. These resistance average value Rnew and current average value Inew are input to the map update unit 54.

  The map holding unit 55 holds, as a resistance map, a table that associates motor current with motor resistance based on the current-resistance characteristics of the motor 1. The map update unit 54 updates the resistance map based on the resistance average value Rnew and the current average value Inew. That is, as shown in FIG. 7, the characteristic curve representing the current-resistance characteristic passes through the point A indicating the resistance average value Rnew and the current average value Inew (the characteristic curve is a solid line in FIG. 7). Change the resistance map so that it changes from a curve to a dotted curve. In the second angular velocity estimation process (FIG. 9) to be described later, the resistance map is updated when the steering state is switched to the steering state (step S72). However, the present invention is not limited to this configuration. For example, each time the resistance value Rc is calculated a predetermined number of times in the rudder state, the resistance map may be updated based on the resistance average value Rnew and the current average value Inew.

  The estimated value calculation unit 56 refers to the resistance map in the map holding unit 55 to obtain a motor resistance value Rm corresponding to the current detection value Im, and uses the resistance value Rm as the voltage detection value Vm and the current detection value. The angular velocity estimated value ωe is calculated by the above formula (2) using together with Im. That is, the estimated value calculation unit 56 calculates the angular velocity estimated value ωe by substituting V = Vm, I = Im, and R = Rm into the above equation (2). In the second angular velocity estimation process (FIG. 9) described later, a resistance map is used to obtain the resistance value Rm used for calculating the angular velocity estimated value ωe in the steering state (step S76). Is not limited to this configuration, and a resistance map may be used to determine the resistance value Rm used for calculating the estimated angular velocity value ωe even in the steering state.

<2.4 Specific example of angular velocity estimation processing>
Hereinafter, some specific examples of angular velocity estimation processing (step S4 in FIG. 3) for realizing the angular velocity estimation unit 50 in the present embodiment in software will be described.

<2.4.1 Basic angular velocity estimation process>
FIG. 6 is a flowchart showing details of the basic angular velocity estimation processing that is the basis of the angular velocity estimation processing S4 in the present embodiment. The angular velocity estimation unit 50 realized by this basic angular velocity estimation process is configured not to include the map update unit 54 and the map holding unit 55 (see FIG. 5). Hereinafter, the basic angular velocity estimation process will be described in detail with reference to FIGS. 5 and 6. The basic angular velocity estimation process is executed every control cycle (see FIG. 3), and the microcomputer 10 operates as follows in this basic angular velocity estimation process.

  First, it is determined based on the pulse signal P from the rotation sensor 7 whether or not the electric power steering apparatus is in a steering-holding state (step S10). Note that the determination method for determining whether or not the vehicle is in the steered state is not limited to this. For example, instead of the rotation sensor 7, the steerable state is based on an output signal from the rudder angle sensor (or steering angular velocity sensor) 7 h. You may make it determine whether it is in a state. Further, for example, it may be determined whether or not the vehicle is in the steering holding state based on the estimated angular velocity value ωe calculated in step S58 described later in the basic angular velocity estimating process in the previous control cycle. These points are the same in the first angular velocity estimation process described later.

  As a result of the determination in step S10, if the steering state is maintained (Yes), the process proceeds to step S11. If the steering state is not established, that is, if the steering state is (No), the process proceeds to step S30. The flag Fh indicating whether or not the steering is maintained is reset (Fh = 0), and then the process proceeds to step S24.

  In step S11, it is determined based on the flag Fh whether or not the steering was maintained in the previous control cycle. As described above, the flag Fh is reset in the initial state (Fh = 0), and is set in step S12 (Fh = 1) described later when the steering holding state is reached. Therefore, in step S11, it is determined whether or not the steering state is maintained even in the previous control cycle based on whether or not the flag Fh is set (for example, if Fh = 1, the steering state is maintained in the previous control cycle). Determined).

  If the result of determination in step S11 is that the steering state is also maintained in the previous control cycle (Fh = 1), the process proceeds to step S12, and if the steering state is not maintained in the previous control cycle (Fh = 0). In the case), the above described count value C, variable SR for integrating the calculated resistance value Rc, and variable value SI for integrating the calculated current value Ic are initialized to 0 (step S32), and thereafter The process proceeds to step S12.

  In step S12, the flag Fh is set (Fh = 1), and in the next step S14, the count value C is incremented by 1, and then V = Vm (voltage detection value) and I = Im (current detection) in equation (3). The value of the motor resistance R is obtained as a calculated resistance value Rc by substituting (value) (step S16). Next, in order to integrate the calculated resistance value Rc, a value obtained by adding the calculated resistance value Rc to the variable SR is set as a new variable SR, and the current detection value Ic = Im at the time of calculation of the calculated resistance value Rc is to be integrated. A value obtained by adding the calculated current value Ic to the variable SI is set as a new variable SI (step S18).

  Thereafter, it is determined whether or not the count value C is equal to or greater than a predetermined end value Cn (for example, 50) (step S20). As a result of the determination, if the count value C has reached the end value Cn (Yes in S20), the process proceeds to step S34. If the count value C has not reached the end value Cn (No in S20), Proceed to step S24.

  In step S34, the variable SR is divided by the count value C = Cn, and the variable SI is divided by the count value C = Cn, thereby calculating the respective average values of the calculated resistance value Rc and the calculated current value Ic at the latest Cn times. Are obtained as the resistance average value Rnew and the current average value Inew, respectively. After the resistance average value Rnew and the current average value Inew are obtained in this way, the count value C and the variables SR and SI are initialized to 0 (step S38), and the process proceeds to step S24.

  In step S24, the angular velocity estimated value ωe is calculated by the equation (2) using the resistance average value Rnew together with the current detection value Im and the voltage detection value Vm. That is, the estimated angular velocity value ωe is calculated by substituting I = Im, V = Vm, and R = Rm into the equation (2). Thereby, the angular velocity estimation process in one control cycle is terminated, and the process returns to the motor control process routine shown in FIG.

  By the basic angular velocity estimation process shown in FIG. 6, the angular velocity estimation unit 50 having the configuration shown in FIGS. 4 and 5 is realized by software. According to such basic angular velocity estimation processing, every time the resistance value Rc and the current value Ic are calculated Cn times in the steering state (every time the count value C reaches the end value Cn), the angular velocity estimation value ωe is calculated. Since the resistance average value Rnew used for the calculation is updated (steps S34 and S24), even if there is an individual difference in the motor 1, the contact state between the brush and the commutator of the motor 1 and the temperature change. However, the resistance average value Rnew with less error compared to the actual motor resistance value is obtained, and the estimated angular velocity value ωe can be accurately calculated using the resistance average value Rnew. As a result, the control performance of the motor control device is improved, and a good steering feeling can be obtained in an electric power steering device using the motor control device.

<2.4.2 First angular velocity estimation process>
In the above basic angular velocity estimation process, the resistance average value Rnew used for calculating the angular velocity estimated value ωe is updated every time the count value C reaches the final value Cn in the steering state. The resistance average value Rnew is not updated when the count value does not reach the final value Cn in the (holding state period). For example, when the final value Cn is set to 100, the resistance average value Rnew is not updated even if C = 90 is counted during the steering period (even if the calculated resistance value Rc is calculated 90 times). Further, when the count value (the number of times of calculation of the calculated resistance value Rc in one steering period) often does not reach the final value Cn (for example, 100) due to external influences, etc., the resistance average value Rnew is updated. The cycle becomes longer. When this update cycle becomes excessively long, a difference occurs between the resistance average value Rnew used for calculating the angular velocity estimated value ωe and the actual motor resistance value R, and the control performance as a motor control device is deteriorated.

  Therefore, in the first angular velocity estimation process for realizing the angular velocity estimation unit 50 in the present embodiment by software, the count value C is a predetermined value (final value) of the resistance average value Rnew used for calculating the angular velocity estimated value ωe. Instead of updating every time Cn) is reached, the resistance average value Rnew is configured to be updated sequentially (every control cycle) in the steering period. FIG. 8 is a flowchart showing the first angular velocity estimation process. The first angular velocity estimation process is executed for each control cycle (see FIG. 3), and in this first angular velocity estimation process, the microcomputer 10 operates as follows.

  First, in step S10, it is determined whether or not the electric power steering device is in a steering holding state. Since the contents of step S10 and steps S11 to S16, S30, and S32 that can be executed thereafter are the same as the contents of the steps having the same number in the basic angular velocity estimation process (FIG. 6), description thereof is omitted.

  In the first angular velocity estimation process, a FIFO memory is realized in software using the memory in the microcomputer 10, and when the calculated resistance value Rc is newly calculated in step S16, the resistance value Rc is stored in the FIFO memory. Stored (step S50). The FIFO memory is a first-in first-out memory having a predetermined value Ca (for example, 50), and when one new calculated resistance is stored in a state where Ca calculated resistance values Rc are stored. The calculated resistance value Rc in the FIFO memory stored first is discarded. For this reason, when the calculated resistance value Rc is stored in the FIFO memory more than Ca times, only the Ca calculated resistance values Rc stored in the most recent Ca time exist in the FIFO memory.

  After the newly calculated resistance value Rc is stored in the FIFO memory, it is determined whether or not the count value C is equal to or greater than a predetermined value Ca (step S52). As a result of this determination, if the count value C has reached the predetermined value Ca (Yes in S52), the process proceeds to step S54. If the count value C has not reached the predetermined value Ca (No in S52), Proceed to step S60. Since C = 1 at the time of the first execution of step S52, the process proceeds to step S60, but the angular velocity estimation process (first angular velocity estimation process) in FIG. 8 is repeatedly executed to calculate the resistance value Rc. When the number of times becomes Ca or more, the process proceeds to step S54 (C ≧ Ca).

  When the process proceeds to step S54, Ca resistance values Rc calculated in the most recent Ca times at that time exist in the FIFO memory, and an average value Rnew of these Ca resistance values Rc is obtained. This means that the resistance average value Rc is obtained by subjecting the resistance value Rc sequentially obtained for each control cycle to a moving average process. Thereafter, in step S56, the resistance average value Rnew is substituted into Rold, and the process proceeds to step S58.

When the process proceeds to step S60, C resistance values Rc calculated in the most recent C times at that time are present in the FIFO memory, and an average value R1 of these C resistance values Rc is obtained. Here, C is an integer greater than or equal to 1 and smaller than Ca. When C is small, the average value R1 is not a stable value but varies, that is, the noise included in the average value R1 is relatively large. It is not preferable to use it for calculation of the estimated angular velocity value ωe as it is. Therefore, the resistance average value Rnew is obtained by the following equation (4) indicating the weighted average of the resistance values Rold and R1 (step S62).
Rnew = {(1−f (C)) · Rold + f (C) · R1} (4)
Here, f (C) is a monotonically increasing function of the count value C, which is 0 when C = 0 and 1 when C = Ca. For example, the function f (C) = C / Ca or f (C) = C ² / Ca ² can be used. The above equation (4) corresponds to the number of acquired current detection values Im and voltage detection values Vm for obtaining the resistance average value Rnew as a new update parameter value (this is equal to the number of acquisitions of the calculated resistance value Rc). The weighted average of the resistance value Rold and the resistance value R1 is shown. When the resistance average value Rnew is obtained by the equation (4), the process proceeds to step S58.

  In step S58, the angular velocity estimated value ωe is calculated by the equation (2) using the resistance value R = Rnew together with the current detection value I = Im and the voltage detection value V = Vm. For the calculation of the estimated angular velocity value ωe, the resistance value Rnew calculated in step S54 or S62 is used in the steering period. In the steering period, the resistance value Rnew set in the previous steering period is used. used. When the angular velocity estimation value ωe is calculated in step S58, the angular velocity estimation process in one control cycle is terminated, and the process returns to the motor control process routine shown in FIG.

  According to such a first angular velocity estimation process, the resistance average value Rnew for calculating the angular velocity estimation value ωe is sequentially obtained in each steering holding period and updated every control cycle. For this reason, the resistance average value Rnew for calculating the estimated angular velocity value ωe can follow the change in the actual motor resistance value R, and the deviation from the actual motor resistance value R is suppressed. As a result, even if the contact state or temperature between the brush and the commutator of the motor 1 changes, the estimated angular velocity value ωe is accurately calculated using the resistance average value Rnew with less error than the actual motor resistance value. be able to. As a result, the control performance of the motor control device is improved, and a good steering feeling can be obtained in an electric power steering device using the motor control device.

<2.4.3 Second angular velocity estimation process>
In the first angular velocity estimation process, the resistance value Rnew used for calculating the angular velocity estimated value ωe in the steering holding period is updated every control cycle (steps S52 to S56, S60, and S62 in FIG. 8). However, in the steering period, the resistance average value Rnew calculated last in the immediately preceding steering period is used to calculate the angular velocity estimated value ωe, and is not updated during the steering state. Therefore, in the second angular velocity estimation process for realizing the angular velocity estimation unit 50 in the present embodiment in software, as shown in FIG. 9, when it is determined that the vehicle is in the steering state (No in step S10). The resistance map is updated based on the resistance average value Rnew and current average value Inew last calculated in the steering period immediately before (step S72), and the resistance used for calculating the estimated angular velocity value ωe in the steering period. The value Rm is determined with reference to the resistance map for each control period (step S76), and the angular velocity estimated value ωe is calculated using the resistance value Rm (step S78).

  The contents of the other steps S10 to S16, S32, S50 to S62 in the second angular velocity estimation process are the same as the contents of the corresponding steps S10 to S16, S32, S50 to S62 in the first angular velocity estimation process (FIG. 8). Basically the same. However, in the second angular velocity estimation process, as described above, in order to update the resistance map, not only the resistance average value Rnew but also the current average value Inew is necessary, so that steps S50, S54, The contents of S56, S60, and S62 are slightly different from the contents of the corresponding steps S50, S54, S56, S60, and S62 in the first angular velocity estimation process (see FIGS. 8 and 9).

  According to such second angular velocity estimation processing, in addition to the same effects as in the case of the first angular velocity estimation processing, the calculation accuracy of the angular velocity estimation value ωe during the steering period is improved. As a result, the control performance of the motor control device is further improved, and a better steering feeling can be obtained in an electric power steering device using the motor control device.

<3. Second Embodiment>
<3.1 Motor controller configuration>
Similarly to the first embodiment, the motor control device according to the second embodiment of the present invention is used in the electric power steering device as shown in FIG. 1 is driven. The configuration of the motor control device is basically the same as that of the first embodiment except that the steering angle sensor 7h is provided instead of the rotation sensor 7 and the configuration of the angular velocity estimation unit 50. As shown in FIG. However, the temperature sensors 41, 42, and 43 of the respective parts indicated by dotted lines in FIG. 2 are not used in this embodiment. In this embodiment, in calculating the angular velocity estimated value ωe, the steering angle θh indicated by the output signal of the steering angle sensor 7h is not used as it is, but is obtained by differentiating the steering angle θh as described later. A rough value ωs of the motor angular velocity corresponding to the steering angular velocity ωh (hereinafter referred to as “angular velocity approximate detection value ωs” or simply “detection value ωs”) is used. Therefore, instead of the rudder angle sensor 7h, a steering angular velocity sensor or a low-cost simple angular velocity sensor that easily detects the angular velocity of the motor may be used. With the configuration in which the rotation speed of the motor 1 is detected using the steering angle sensor, the steering angular speed sensor, or the simple angular speed sensor in this way, the detection accuracy of the rotation speed is lower than the accuracy necessary for the drive control of the electric motor, but it is markedly different. In addition, the speed detection means (simple speed detection means) of the motor 1 can be realized at low cost.

  Hereinafter, the configuration of the angular velocity estimation unit 50 in the present embodiment will be described, and the same or corresponding components will be denoted by the same reference numerals for the other portions, and the description thereof will be omitted (also in other embodiments described later). The same). However, there is a slight difference in the motor control processing shown in FIG. 3 with respect to the first embodiment, and this difference will be appropriately referred to in the description of the configuration of the angular velocity estimation unit 50 (described later). The same applies to other embodiments).

  FIG. 10 is a block diagram showing a functional configuration of the angular velocity estimation unit 50 in the present embodiment. The angular velocity estimating unit 50 includes a velocity determining unit 61, a parameter value calculating unit 62, and an angular velocity estimated value calculating unit 66. The velocity determining unit 61 and the parameter value calculating unit 62 constitute a parameter value updating unit 400. . In the angular velocity estimation unit 50, an angular velocity value obtained based on an output signal from the steering angle sensor (or steering angular velocity sensor) 7h is input as the angular velocity approximate detection value ωs. The detection accuracy of the steering angle sensor (or steering angular velocity sensor) 7h is higher as the rotational speed of the motor 1 is higher. For this reason, the speed determination unit 61 determines whether or not the angular velocity approximate detection value ωs is equal to or greater than a predetermined minimum value ωmin, and outputs a signal indicating the determination result as the rotation state signal Sp. Based on the rotation state signal Sp, the parameter value calculation unit 62 performs the angular velocity approximate detection value ωs, the current detection value Im, and the voltage detection value when the motor 1 is rotating at a predetermined speed or higher corresponding to the minimum value ωmin. A resistance value Rnew and a counter electromotive force constant knew are obtained based on a plurality of sets of detected values ωs, Im, and Vm with Vm as one set (details will be described later). The angular velocity estimated value calculation unit 66 calculates the angular velocity estimated value ωe using the resistance value Rnew and the back electromotive force constant knew. Further, the parameter value calculation unit 62 determines the resistance value Rnew and the reverse value when the motor 1 is rotating at a speed smaller than a predetermined speed corresponding to the minimum value ωmin or stopped based on the rotation state signal Sp. The electromotive force constant knee is not newly calculated (not updated). In this case, the angular velocity estimated value calculation unit 66 uses the already obtained resistance value Rnew and the counter electromotive force constant knew as it is, or uses the resistance value Rnew obtained by another method or the like to calculate the angular velocity estimated value ωe. Calculate (details will be described later).

<3.2 Specific example of angular velocity estimation processing>
Hereinafter, some specific examples of angular velocity estimation processing (step S4 in FIG. 3) for realizing the angular velocity estimation unit 50 in the present embodiment in software will be described. In the present embodiment, in step S2 in FIG. 3, the number of detection values (Vm, Im, etc.) necessary for the new calculation of the minimum value ωmin, the time limit value Tlim, and parameter values (Rnew, knew) described later. An appropriate value based on an experiment or the like is set for the predetermined value ja or the like that defines the value, and an appropriate value based on an experiment or the like is set as an initial value for the resistance values Rold, Rnew and the back electromotive force constants kold, knew, and the parameters A variable j indicating the number of detection values (Vm, Im, ωs) used for calculation of the values (Rnew, knew) is initialized to zero. In step S3 of FIG. 3, the steering torque Ts, the vehicle speed S, the angular velocity approximate detection value ωs, the voltage detection value Vm, for each control cycle based on the output signals from the sensors 3, 4, 7h and the detection circuits 24, 25. A current detection value Im is acquired.

<3.2.1 First angular velocity estimation process>
FIG. 11 is a flowchart showing details of a first angular velocity estimation process which is an example of the angular velocity estimation process S4 in the present embodiment. The first angular velocity estimation process in the present embodiment is also executed for each control cycle (see FIG. 3), and in this first angular velocity estimation process, the microcomputer 10 operates as follows.

  First, it is determined whether or not the angular velocity approximate detection value ωs obtained based on the output signal from the steering angle sensor (or steering angular velocity sensor) 7h is equal to or greater than a predetermined minimum value ωmin (step S100). As a result of this determination, if the angular velocity approximate detection value ωs is equal to or greater than the minimum value ωmin (in the case of Yes), the process proceeds to step S102, and if the angular velocity approximate detection value ωs is smaller than the minimum value ωmin (in the case of No). The process proceeds to step S120 without updating the parameter values (resistance value Rnew, counter electromotive force constant knew) used to calculate the estimated angular velocity value ωe.

  In step S102, it is determined whether the variable j is 0 or not. If the variable j is 0 as a result of the determination, the process proceeds to step S103, and if the variable j is not 0, the process proceeds to step S104. The microcomputer 10 has a built-in timer, and in step S103, the timer is started. As a result, the elapsed time Telp from the start of the timer is initialized to 0, and then the process proceeds to step S104. When the process proceeds to step S104, the value of the variable j is increased by 1, and then the process proceeds to step S106.

  Also in the first angular velocity estimation process in the present embodiment, a FIFO memory is realized in software using the memory in the microcomputer 10. This FIFO memory is a first-in first-out type memory whose size is a predetermined value ja (for example, 30), and a set of four values consisting of the detected values ωs, Im, Vm of the angular velocity, current, and voltage and the elapsed time Telp. When one set of detection data is stored in the state where the ja set of detection data is stored, one set of detection data in the FIFO memory stored first (the detection value ωs constituting the first set) , Im, Vm and the elapsed time Telp) are discarded.

  In step S106, the current detected values ωs, Im, Vm and the elapsed time Telp are stored in such a FIFO memory as a set of detected data.

  In the next step S108, a difference ΔTel between the elapsed time Telp in the first group of the FIFO memory and the elapsed time Telp indicated by the timer is obtained. This difference ΔTelp indicates the elapsed time from the detection time point of the detected values ωs, Im, Vm in the first set to the present time. In the next step S110, whether the difference ΔTelp is equal to or smaller than a predetermined time limit value Tlim. Determine whether. At this time, the elapsed time Telp in the first group is substantially 0 and ΔTelp is substantially equal to the elapsed time Telp indicated by the timer. However, when the first group is taken out or discarded from the FIFO memory as described later, the second group Is the first group, and the elapsed time Telp in the first group is a positive value (non-zero).

  As a result of the determination in step S110, when the difference ΔTelp is equal to or smaller than the time limit value Tlim, the process proceeds to step S112, and it is determined whether or not the variable j is equal to the predetermined value ja. If it is determined that the variable j is equal to the predetermined value ja, the process proceeds to step S114. If the variable j is not equal to the predetermined value ja (that is, smaller than the predetermined value ja), the process proceeds to step S128. Since j = 1 at the time when step S112 is first executed, the process proceeds to step S128. However, the angular velocity estimation process (first angular velocity estimation process) in FIG. When ωs, Im, Vm, and Telp are stored in the FIFO memory, the process proceeds to step S114 (j = ja).

  In step S114, the resistance value Rnew and the counter electromotive force constant knew are determined as follows from the detected values ωs, Im, Vm of the ja set among the above-described values ωs, Im, Vm, Telp of the ja set stored in the FIFO memory. And calculate.

First, the detected values ωs, Im, and Vm of the ja set in the FIFO memory are written as ωs (i), Im (i), and Vm (i) (i = 1, 2,..., Ja), and the ja row 2 A matrix X of columns, a ja-dimensional vector Y, and a two-dimensional vector φ are defined as follows (the vector Y can also be referred to as a ja × 1 matrix, and the vector φ can be referred to as a 2 × 1 matrix).
X = (Ω II) (5)
Y = (Vm (1) Vm (2)... Vm (ja)) ^T (6)
φ = (k R) ^T (7)
Here, Ω is Ω = (ωs (1) ωs (2)... Ωs (ja)) ^T (8)
And II is II = (Im (1) Im (2)... Im (ja)) ^T (9)
The superscript T is a symbol indicating transposition (for example, A ^T represents a transposed matrix of the matrix A). From Equation (2) k · ωs (i) + R · Im (i) = Vm (i) (10)
(I = 1, 2,..., Ja), which is expressed by X, Y, and φ defined as above,
X · φ = Y (11)
It becomes.

Therefore, in step S114, k and R calculated by the following equations are set as the above-described counter electromotive force constant knee and resistance value Rnew (this is the counter electromotive force used for calculating the angular velocity estimated value ωe in step S120). It means updating the constant knew and the resistance value Rnew).
φ = X ⁺ Y (12)
Here, X ⁺ is a pseudo inverse matrix of the matrix X, and can be calculated by the following equation.
X ⁺ = (X ^T · X) ⁻¹ · X ^T (13)
Further, φ = (k R) ^T calculated by the expression (12) using the pseudo inverse matrix X ^T as described above minimizes the following Ψ.
Ψ = {V (1) −k · ωs (1) −R · Im (1)} ² + {V (2) −k · ωs (2) −R · Im (2)} ² ...
… + {V (j) −k · ωs (j) −R · Im (j)} ²
= (Y−X · φ) ^T · (Y−X · φ) (14)
Therefore, if the by phi = (k R) and ^T φ = (knew Rnew) ^T calculated by the equation (12), the counter electromotive force constant k = knew and resistance R = Rnew is the FIFO memory j sets of A value for approximating the detected values ωs (i), Im (i), Vm (i) (i = 1, 2,..., Ja) using the linear form Vm = k · ωs + R · Im by the least square method; It has become. That is, the calculation of φ by the equation (12) means that all of the j sets of detected values ωs (i), Im (i), Vm (i) (i = 1, 2,..., Ja) are motors 1. This means obtaining φ = (k R) ^T that at least approximately satisfies the circuit equation Vm = k · ωs + R · Im.

  When the new resistance value Rnew and the back electromotive force constant knew are obtained as described above, the resistance value Rnew and the back electromotive force constant Knew are then set as Rold and kold, respectively (step S116), and the first group is set in the FIFO memory. In order to discard the detected data (ωs, Im, Vm, Telp) and store the next new set of detected data (ωs, Im, Vm, Telp), the variable j is reduced by 1 ( Step S118). Then, it progresses to step S120.

  When the process proceeds to step S128, it is determined whether or not the variable j is 2 or more. As a result of the determination, if the variable j is smaller than 2 (in the case of No), since the resistance value R1 and the counter electromotive force constant k1 cannot be calculated in step S130 described later, the resistance used for calculating the estimated angular velocity value ωe. The process proceeds to step S120 without updating the value Rnew and the back electromotive force constant knew. On the other hand, if it is determined that the variable j is 2 or more (Yes), the process proceeds to step S130.

At the time of proceeding to step S130, the FIFO memory stores j sets of detected values ωs, Im, Vm and the elapsed time Telp, and among these sets of j detected values ωs, Im, Vm, the same as step S114 Thus, the resistance value R1 and the counter electromotive force constant k1 are calculated. That, niv = j as by the phi = (k R) and ^T phi = (k1 R1) ^T is calculated by the equation (12) obtains the resistance value R1 and the counter electromotive force constant k1 in step S130. The variable j at this time is an integer greater than or equal to 2 and smaller than ja. When the variable j is small, the resistance value R1 and the counter electromotive force constant k1 vary and are not stably calculated. Since the contained noise is relatively large, it is not preferable to use the resistance value R1 and the counter electromotive force constant k1 as they are for the calculation of the estimated angular velocity value ωe.

Therefore, the resistance value Rnew and the counter electromotive force constant knew to be used for the calculation of the estimated angular velocity value ωe, the weighted average of the resistance values Rold and R1, and the weighted average of the counter electromotive force constants kold and k1 are expressed by the following equation (15). And (16) (step S132).
Rnew = {(1−f (j)) · Rold + f (j) · R1} (15)
knee = {(1−f (j)) · kold + f (j) · k1} (16)
Here, f (j) is a monotonically increasing function of variable j, which is 0 when j = 0 and 1 when j = ja. For example, function f (j) = j / ja or f (j) = j ² / ja ² can be used. The above equations (15) and (16) are obtained as the number of acquired current detection value Im and voltage detection value Vm and angular velocity approximate detection value ωs for obtaining the resistance value Rnew and the back electromotive force constant knew as new update parameter values. The weighted average of the resistance values Rold and R1 and the weighted average of the back electromotive force constants knew and k1 are shown. When the resistance value Rnew and the back electromotive force constant knew are obtained from these equations (15) and (16), the process proceeds to step S120.

  As a result of the determination in step S110, when the difference ΔTel indicating the elapsed time from the detection point of the detected values ωs, Im, Vm in the first group of the FIFO memory exceeds the time limit value Tlim (in the case of No) Advances to step S122, and the detected value ωs, Im, Vm of the first set is too old to be used for calculating the current resistance value Rnew and the like, and the detected data (ωs, Im, Vm of the first set is read from the FIFO memory. , Telp) is taken out (discarding the first set of detection data).

  Next, the variable j is decremented by 1 (step S124) in accordance with the extraction of the first set of detection data, and then it is determined whether or not the variable j is 0 (step S126). If the result of this determination is that the variable is not 0, processing returns to step S108.

  Thereafter, steps S110 → S122 → S124 → S126 → S108 are repeatedly executed until the difference ΔTelp in the first group in the FIFO memory becomes equal to or less than the time limit value Tlim. However, if the variable j becomes 0 during this period, since there is no detection data (ωs, Im, Vm, Telp) stored in the FIFO memory, the resistance value Rnew and the back electromotive force used for the calculation of the estimated angular velocity value ωe. The process proceeds to step S120 without updating the constant knee. If the difference ΔTelp is equal to or smaller than the time limit value Tlim while repeatedly executing Steps S110 → S122 → S124 → S126 → S108, the process proceeds to Step S112, and after executing Steps S128 to S132 described above, the process proceeds to Step S120. move on.

  In step S120, the angular velocity estimated value ωe is calculated by the equation (2) using the detected current value I = Im and the detected voltage value V = Vm together with the resistance value R = Rnew and the counter electromotive force constant k = knew. For the calculation of the estimated angular velocity value ωe, when the above step S114 or S132 is executed, the resistance value Rnew and the back electromotive force constant knew calculated there are used. However, both of the above steps S114 and S132 are used. Is not executed, the resistance value Rnew and the counter electromotive force constant knew set before the current control cycle are used. When the angular velocity estimation value ωe is calculated in step S120, the angular velocity estimation process in one control cycle is terminated, and the process returns to the motor control process routine shown in FIG.

  According to such first angular velocity estimation processing, when the detected values Vm, Im, and ωs are obtained ja times or more when the angular velocity approximate detection value ωs is equal to or greater than the predetermined minimum value ωmin, it is obtained in the latest ja times. Using the detected values Vm, Im, ωs, that is, the detected values Vm, Im, ωs existing in the FIFO memory, the resistance value Rnew and the counter electromotive force constant knew are calculated by the least square method (step S114), and these resistance values are calculated. Rnew and the counter electromotive force constant knew are used to calculate the angular velocity estimated value ωe (step S120). When the number of times the detected values Vm, Im, and ωs are acquired when the angular velocity approximate detected value ωs is equal to or greater than the predetermined minimum value ωmin, the detected value Vm obtained at the latest j times is obtained when the number of times the detected values Vm, Im, ωs are acquired is less than the predetermined value ja , Im, ωs, the resistance value R1 obtained by the least square method and the back electromotive force constant k1 are not used as they are, but the set of detected values Vm, Im, ωs based on obtaining the detected values more than ja times before that. The resistance value Rnew and the counter electromotive force constant knew to be used for calculating the angular velocity estimated value ωe are also updated in consideration of the resistance value Rnew and the counter electromotive force constant knew calculated from (Step S132). Further, for the calculation of the resistance values Rnew, R1 and the counter electromotive force constants knew, k1, the detected values Vm, Im, ωs that have passed the time limit value Tlim from the detection time are not used. (Steps S108, S110, S122 to S126).

  Therefore, according to the first angular velocity estimation process, when the angular velocity approximate detection value ωs is equal to or larger than the predetermined minimum value ωmin, the resistance value Rnew and the counter electromotive force constant knew used for calculating the angular velocity estimation value ωe are used. A stable and highly accurate value can be calculated with high frequency. For this reason, not only the resistance value Rnew for calculating the estimated angular velocity value ωe but also the counter electromotive force constant knew can follow changes in the actual motor resistance value R and the counter electromotive force constant k. Deviation from the value R and the back electromotive force constant k is suppressed. As a result, the angular velocity estimated value ωe can be calculated with higher accuracy. Thereby, the control performance of the motor control device is improved, and a better steering feeling can be obtained in an electric power steering device using the motor control device.

<3.2.2 Second angular velocity estimation process>
In the first angular velocity estimation process, the resistance average value Rnew and the counter electromotive force constant knew used for calculating the angular velocity estimation value ωe are updated at a high frequency when the angular velocity approximate detection value ωs is equal to or greater than the predetermined value ωmin. However, when the angular velocity approximate detection value ωs is smaller than the predetermined value ωmin, that is, when the motor 1 is rotating at a speed lower than the predetermined value or is stopped (typically, the electric power steering device is in the steered state). Is not updated). Therefore, in order to improve this point, when in the steered state, resistance value update processing (steps S10 to S16, S30, S32, S50) in the angular velocity estimation processing (for example, angular velocity estimation processing shown in FIG. 8) in the first embodiment. ~ S56, S60, S60), the resistance value Rnew is updated, and when in the steering state, the process (hereinafter, "steps S100-S118, S122-S132" in the first angular velocity estimation process (Fig. 11) in the present embodiment). A configuration may be considered in which the resistance value Rnew and the counter electromotive force constant knew are updated in S250 (referred to as parameter value update processing).

  FIG. 12 is a flowchart showing an angular velocity estimation process for realizing such a configuration. Hereinafter, this angular velocity estimation processing will be described as second angular velocity estimation processing for realizing the angular velocity estimation unit 50 in the present embodiment in software. This second angular velocity estimation process is also executed for each control cycle (see FIG. 3), and in this second angular velocity estimation process, the microcomputer 10 operates as shown in the flowchart of FIG.

  As can be seen by comparing FIG. 12 with FIG. 8, the resistance value update processing (steps S10 to S16, Steps S10 to S16 in the angular velocity estimation processing of FIG. 8 is performed in the steering state in the second angular velocity estimation processing of the present embodiment. S30, S32, and S50 to S62), the corresponding steps are denoted by the same reference numerals and description thereof is omitted.

  In the angular velocity estimation process of FIG. 8, if it is determined in step S10 that the steering state is maintained (in the case of Yes), after resetting the flag Fh in step S30, step S58 (step of calculating the angular velocity estimated value ωe is performed as it is. However, in the angular velocity estimation process of FIG. 12 in the present embodiment, if it is determined in step S10 that the steering is not maintained, the flag Fh is reset in step S30 and then the parameter value steering update process S250 is executed. Thereafter, the process proceeds to step S120 (step of calculating the angular velocity estimated value ωe). As this parameter value steering update process S250, a parameter value update process (a process surrounded by a dotted line in FIG. 11) S250 for updating the resistance value Rnew and the back electromotive force constant knew in the angular velocity estimation process of FIG. 11 is adopted. Has been. In addition, also when it determines with it being a steering holding state, if resistance value Rnew is updated (step S54, S62), it will progress to step S120.

  In step S120, the angular velocity estimation value ωe is calculated by the equation (2) using the resistance value R = Rnew, the counter electromotive force constant k = knew, the current detection value I = Im, and the voltage detection value V = Vm. As the resistance value Rnew and the counter electromotive force constant knew in the equation (2), the resistance value Rnew updated in step S54 or S62 is used when the steering state is maintained, and when the steering state is not maintained, Usually, the resistance value Rnew and the back electromotive force constant knew updated in the parameter value steering update process S250 are used (see steps S114 and S132 in FIG. 11).

  According to the angular velocity estimation process (second angular velocity estimation process) shown in FIG. 12, the resistance value Rnew used for calculating the angular velocity estimated value ωe is stable regardless of whether the steering state is maintained or the steering state. High values can be calculated with high frequency. In the steering state, a stable and highly accurate value can be frequently calculated as the back electromotive force constant knew used for calculating the angular velocity estimated value ωe. Therefore, the update of the resistance value Rnew for calculating the estimated angular velocity value ωe can better follow the change in the actual motor resistance value R, and the difference between the actual motor resistance value R and the back electromotive force constant k. Is further suppressed. As a result, the angular velocity estimated value ωe can be calculated with high accuracy. Thereby, the control performance of the motor control device is further improved, and a better steering feeling can be obtained in the electric power steering device using the motor control device.

<3.2.3 Third angular velocity estimation process>
In the first angular velocity estimation process (FIG. 11), even when the number of detected values Vm, Im, and ωs is small, the resistance value Rnew and the counter electromotive force constant knee that are used to calculate the angular velocity estimated value ωe are calculated. Although updating can be performed at a high frequency (steps S128 to S132), the angular velocity estimation process can be simplified by reducing the updating frequency to some extent. FIG. 13 is a flowchart showing such angular velocity estimation processing (hereinafter referred to as “third angular velocity estimation processing”).

  As can be seen by comparing FIG. 13 with FIG. 11, among the steps included in the angular velocity estimation process (third angular velocity estimation process) of FIG. 13, the same steps as those included in the angular velocity estimation process of FIG. The steps S106 to S118 and S122 to S132 in the angular velocity estimation process of FIG. 11 are changed to steps S140 to S148 in the angular velocity estimation process of FIG.

  According to such an angular velocity estimation process of FIG. 13, every time the detection values Vm, Im, and ωs are acquired ja times, the acquired detection values Vm (i), Im (i), ω ( i) The resistance value Rnew and the back electromotive force constant knee are updated by setting φ calculated by the equation (12) using (i = 1 to ja) to φ = (Rnew knew) (steps S140 to S140). S148). However, if the elapsed time from the detection time point of the first set of detection values Vm (1), Im (1), and ω (1) out of the ja set exceeds the time limit value Tlim, the set of Vm (I), Im (i), ω (i) (i = 1 to ja) are not used for updating the resistance value Rnew and the counter electromotive force constant knew (steps S103 and S146), and are acquired thereafter. The resistance value Rnew and the back electromotive force constant knew are updated using the detection values Vm (i), Im (i), and ω (i) (i = 1 to ja) of the set ja (step S148). Even with such angular velocity estimation processing, when the angular velocity approximate detection value ωs is equal to or greater than the predetermined minimum value ωmin, the resistance value Rnew and the counter electromotive force constant knew used for calculating the angular velocity estimated value ωe are updated at a predetermined frequency. Therefore, even if the actual motor resistance value R and the counter electromotive force constant k change, the estimated angular velocity value ωe can be calculated with high accuracy. As a result, the control performance of the motor control device is improved, and a good steering feeling can be obtained in an electric power steering device using the motor control device.

<4. Third Embodiment>
<4.1 Configuration of motor control device>
As in the first embodiment, the motor control device according to the third embodiment of the present invention is also used in the electric power steering device as shown in FIG. 1 is driven. The configuration of the motor control device is basically the same as that of the first embodiment except that the steering angle sensor 7h is provided instead of the rotation sensor 7 and the configuration of the angular velocity estimation unit 50. As shown in FIG. However, the temperature sensors 41, 42, and 43 of the respective parts indicated by dotted lines in FIG. 2 are not used in this embodiment.

  Hereinafter, the configuration of the angular velocity estimation unit 50 in the present embodiment will be described. FIG. 14 is a block diagram showing a functional configuration of the angular velocity estimation unit 50 in the present embodiment. The angular velocity estimation unit 50 includes an angle change detection unit 71, a resistance calculation unit 72, and an angular velocity estimation value calculation unit 66. The angular velocity estimation unit 50 is acquired based on an output signal from the steering angle sensor 7h. The rotation angle θs of the motor 1 is input. The angle change detection unit 71 and the resistance calculation unit 72 constitute a resistance value update unit 200. According to the configuration in which the rotation angle of the motor 1 is detected using the steering angle sensor 7h (or the simple angle sensor that detects the angular velocity of the motor 1 with low resolution) as in the present embodiment, the cost is sufficiently low. An angle detection means (simple angle detection means) of the motor 1 can be realized. In this case, the resolution of the simple angular velocity detection means does not have to be such a resolution that the angular velocity can be obtained with the accuracy required for the drive control of the motor 1, so a sensor for detecting the angular velocity with the accuracy necessary for the drive control is provided. It is possible to calculate an accurate angular velocity estimation value at a lower cost than when it is used.

  The (rotor) rotation angle θs of the motor 1 can be calculated based on the steering angle θh obtained by the steering angle sensor 7h. The steering angle sensor 7h has a resolution higher than that of a sensor that directly detects the rotation angle of the motor 1. Is low. However, if the output signal from the rudder angle sensor 7h changes, the rotation angle θs obtained based on the output signal can be regarded as corresponding to the actual rotation angle. Therefore, in the angular velocity estimation unit 50 of the present embodiment, the angle change detection unit 71 determines whether or not the rotation angle θs acquired based on the output signal of the steering angle sensor 7h has changed, and a signal indicating the determination result. Is output as the angle change detection signal Sc.

  The resistance calculation unit 72 sets the rotation angle θs as the first rotation angle θ1 when the rotation angle θs changes based on the angle change detection signal Sc, and then the current and voltage obtained for each control cycle. The detection values Vm and Im are accumulated. Then, the resistance calculation unit 72 sets the rotation angle θs as the second rotation angle θ2 when the rotation angle θs subsequently changes based on the angle change detection signal Sc, and sets the first and second rotation angles θ1, A resistance value Rnew of the motor 1 is calculated based on θ2 and current and voltage detection values Vm and Im accumulated from the detection time point of the first rotation angle θ1 to the detection time point of the second rotation angle θ2. Thereafter, each time a change in the rotation angle θs is detected, the first rotation angle θ1 set before that time, the second rotation angle θ2 that is the rotation angle θs at that time, and the first and second rotation angles. The resistance value Rnew of the motor 1 is calculated based on the detected values Vm and Im of the current and voltage between the detection times of θ1 and θ2.

  The angular velocity estimated value calculation unit 66 calculates the angular velocity estimated value ωe by using the detected current values Im and Vm at each time point together with the resistance value Rnew thus calculated. In the present embodiment, the rotation angle θs of the motor 1 acquired based on the output signal of the steering angle sensor 7h is input to the angular velocity estimation unit 50 (FIG. 14), but the output signal of the steering angle sensor 7h is The angular velocity estimation unit 50 may receive the indicated steering angle θh, and the angular velocity estimation value calculation unit 66 may calculate the angular velocity estimation value ωe using the steering angle θh and the reduction ratio of the speed reducer 2.

<4.2 Specific example of angular velocity estimation processing>
Hereinafter, a specific example will be described for angular velocity estimation processing (step S4 in FIG. 3) for realizing the angular velocity estimation unit 50 in the present embodiment in software. In the present embodiment, in step S2 of FIG. 3, appropriate values based on experiments and the like are set for the back electromotive force constant k of the motor 1 and predetermined values ja and jmax described later, and the resistance values Roll and Rnew of the motor 1 are set. Also, an appropriate value based on an experiment or the like is set as an initial value, a variable j indicating the number of acquired voltage detection values Vm and current detection values Im used to newly calculate the resistance value Rnew, and a rotation angle θs. A flag Fs that is set when the first change is initialized to 0. In step S3 in FIG. 3, the steering torque Ts, the vehicle speed S, the steering angle θh, the voltage detection value Vm, and the current detection are performed for each control period based on the output signals from the sensors 3, 4, 7h and the detection circuits 24, 25. The value Im is obtained. In this embodiment, the steering angle sensor 7h is required, but the cost of the steering angle sensor is significantly lower than that of the rotation sensor that directly detects the rotation angle of the motor 1.

<4.2.1 First Angular Velocity Estimation Process>
FIG. 15 is a flowchart showing details of a first angular velocity estimation process which is an example of the angular velocity estimation process S4 in the present embodiment. The first angular velocity estimation process in the present embodiment is also executed for each control cycle (see FIG. 3), and in this first angular velocity estimation process, the microcomputer 10 operates as follows.

  First, it is determined based on the output signal from the rudder angle sensor 7h whether or not the rotation angle θs of the motor 1 has changed (step S150). If the result of this determination is that the rotation angle θs has changed, the process proceeds to step S152 assuming that a new rotation angle θs has been detected, and the process proceeds to step S172 if the rotation angle θs has not changed.

  In step S152, it is determined whether or not a variable j indicating the number of acquired current detection values and voltage detection values Vm and Im used for calculating a new resistance value Rnew is zero. If the variable j is 0 as a result of this determination, the current rotation angle (the newly detected rotation angle) θs is set as the first rotation angle θ1 in step S153 (θ1 = θs) and the flag Fs is set. After setting (Fs = 1), the process proceeds to step S154. If the variable j is not 0, the process proceeds to step S154.

  In step S154, the variable j is incremented by 1, and then the current detection value Im and the voltage detection value Vm are set to I (j) and V (j), respectively (step S156), and the current rotation angle θs is set to the first value. The rotation angle θ2 is set (step S158).

  Thereafter, in step S160, it is determined whether or not the variable j is greater than or equal to a predetermined value ja (ja is a positive integer, for example 50). Here, the predetermined value ja is the number of times the current and voltage are detected to newly calculate the resistance value Rnew, and an appropriate value based on an experiment or the like is set in advance in step S2 of FIG. If it is determined in step S160 that the variable j is equal to or greater than the predetermined value ja, the process proceeds to step S162. If the variable j is smaller than the predetermined value ja, the process proceeds to step S178. Since j = 1 at the time when step S160 is first executed, the process proceeds to step S178. However, the angular velocity estimation process (first angular velocity estimation process) in FIG. When the detected value I (i) and the detected voltage value V (i) (i = 1 to j, j ≧ ja) are acquired and the rotation angle θs changes, the process proceeds to step S162.

  In step S162, the resistance value Rnew is calculated from the first and second rotation angles θ1 and θ2, the current detection value I (i), and the voltage detection value V (i) (i = 1 to j) as follows. .

In the equation (2), when both sides are time integrated with V = V (t) and I = I (t), the following equation is obtained (the resistance value R is considered not to change during this time integration period).
θ (t) = (1 / k) ∫V (t) dt− (R / k) ∫I (t) dt (17)
Here, θ (t) = ∫ωdt. From the above equation (17), the motor resistance value R can be obtained by the following equation.
R = {∫V (t) dt−θ (t) · k} / ∫I (t) dt (18)
Now, assuming that the detection times of the first and second rotation angles θ1 and θ2 are t1 and t2, respectively, a new resistance value Rnew used to calculate the angular velocity estimated value ωe can be obtained from the following equation using the equation (18). it can.
Rnew = {Tc · Σ (i = 1, j) V (i) − (θ2−θ1) · k} / {Tc · Σ (i = 1, j) I (i)} (19)
Here, Tc is a control cycle (detection interval of current Im and voltage Vm),
Σ (i = 1, j) V (i) = V (1) + V (2) + ... + V (j),
Σ (i = 1, j) I (i) = I (1) + I (2) +... + I (j)
It is.

  Therefore, in step S162, using the first and second rotation angles θ1, θ2, the current detection value I (i), and the voltage detection value V (i) (i = 1 to j), The resistance value Rnew is calculated.

  When a new resistance value Rnew is obtained in this way, the resistance value Rnew is then set to Rold (step S164), and it is determined whether or not the variable j exceeds a predetermined value jmax (step S166). Here, the predetermined value jmax is larger than the predetermined value ja, and is introduced in order to avoid using the detection values Im and Vm that have passed for a long time from the detection time point in calculating the resistance value Rnew. It is. As a result of the determination in step S166, when the variable j exceeds the predetermined value jmax (in the case of Yes), the next new value is detected using the current detection value Im and the voltage detection value Vm detected after the present time. In order to calculate the resistance value Rnew, the current rotation angle θs is reset as the first rotation angle θ1, and the current detection value Im and the voltage detection value Vm are set to I (1) and V (1), respectively ( Step S168), and then the process proceeds to Step S170. On the other hand, as a result of the determination, if the variable j is equal to or less than the predetermined value jmax (in the case of No), the process proceeds to step S170 as it is.

  As a result of the determination in step S160, when the process proceeds to step S178 (No in step S160), it is determined whether or not the variable j is 2 or more. As a result of the determination, if the variable j is smaller than 2, the resistance value R1 cannot be calculated in step S180, which will be described later. Therefore, the process proceeds to step S170 without updating the resistance value Rnew used for calculating the estimated angular velocity value ωe. . On the other hand, as a result of the determination, if the variable j is 2 or more, the process proceeds to step S180.

In step S180, the following equation similar to equation (19) is obtained from the first and second rotation angles θ1 and θ2, and the current detection value I (i) and the voltage detection value V (i) (i = 1 to j). The resistance value R1 is calculated from (20).
R1 = {Tc · Σ (i = 1, j) V (i) − (θ2−θ1) · k} / {Tc · Σ (i = 1, j) I (i)} (20)
The variable j at this time is an integer greater than or equal to 2 and smaller than ja (see step S160). When the variable j is small, the resistance value R1 varies and is not calculated stably, that is, included in the resistance value R1. Since the generated noise is relatively large, it is not preferable to use the resistance value R1 as it is for the calculation of the estimated angular velocity value ωe. Therefore, the resistance value Rnew to be used for calculation of the estimated angular velocity value ωe is calculated by the following equation (21) indicating a weighted average of the resistance values Rold and R1 (step S182).
Rnew = {(1−f (j)) · Rold + f (j) · R1} (21)
Here, f (j) is a monotonically increasing function of variable j, which is 0 when j = 0 and 1 when j = ja. For example, function f (j) = j / ja or f (j) = j ² / ja ² can be used. The above equation (21) represents a weighted average of the resistance values Rold and R1 according to the number of acquired current detection values Im and voltage detection values Vm for obtaining the resistance value Rnew as a new update parameter value. . When the resistance value Rnew is obtained by this equation (21), the process proceeds to step S170.

  As a result of the determination in step S150, if the rotation angle θs has not changed (No in step S150), the process proceeds to step S172, and it is determined whether or not the flag Fs is set. As a result of this determination, if the flag Fs is not set (if Fs = 0), the change in the rotation angle θs has not been detected yet, and the process directly proceeds to step S170. On the other hand, if the determination result shows that the flag Fs is set (if Fs = 1), the process proceeds to step S174, where the variable j is incremented by 1, and the current detected value Im and voltage detected value Vm are set to the current detected value Im. These are set as I (j) and V (j), respectively (step S176). Thereby, after the change of the rotation angle θs is detected and the rotation angle θs is set as the first rotation angle θ1, the current detection value Im and the voltage detection value Vm are I even in the control cycle in which the rotation angle θs does not change. Accumulated as (j) and V (j). After execution of step S176, the process proceeds to step S170.

  In step S170, the angular velocity estimated value ωe is calculated by Equation (2) using the resistance value R = Rnew, the detected current value I = Im, and the detected voltage value V = Vm. For the calculation of the estimated angular velocity value ωe, when the above step S162 or S182 is executed, the resistance value Rnew calculated there is used, but when neither of the above steps S162 and S182 is executed. The resistance value Rnew set before the current control cycle is used. When the angular velocity estimation value ωe is calculated in step S170, the angular velocity estimation process in one control cycle is terminated, and the process returns to the motor control process routine shown in FIG.

  According to such first angular velocity estimation processing, each time a change in the rotation angle θs is detected based on the output signal from the steering angle sensor 7h, the current detection value I (i) and the voltage detection value V (i). On the basis of the integrated value, the resistance value Rnew is newly calculated by the equations (19) or (20) and (21) (steps S150, S162, S180, S182). At this time, even when the number of acquired current detection values Im and voltage detection values Vm after the acquisition time of the first rotation angle θ1 is small, the resistance value Rnew obtained before that is used to obtain the equation (21). Thus, a new resistance value Rnew is calculated (step S182). Therefore, a stable angular velocity estimated value ωe with little variation can be acquired with high frequency. For this reason, the resistance value Rnew used for calculation of the estimated angular velocity value ωe can follow the actual change in the motor resistance value R, and the calculation of the estimated angular velocity value ωe is not affected by a temperature change or the like. As a result, even if the contact state between the brush of the motor 1 and the commutator, the temperature, or the like changes, the estimated angular velocity value ωe can be accurately calculated using the resistance Rnew with less error than the actual motor resistance value. it can. As a result, the control performance of the motor control device is improved, and a good steering feeling can be obtained in an electric power steering device using the motor control device. The present embodiment including the angular velocity estimation process as described above is advantageous in terms of cost because it does not require a high resolution sensor compared to the conventional technique for differentiating the output signal from the steering angle sensor to obtain the angular velocity. It is.

<4.2.2 Second angular velocity estimation process>
In the first angular velocity estimation process (FIG. 15), the angular velocity estimation value ωe is calculated even when the number of acquired current detection values Im and voltage detection values Vm after the acquisition of the first rotation angle θ1 is small. It is possible to update the resistance value Rnew to be used at a high frequency (steps S178 to S182), but the angular velocity estimation process can be simplified by reducing the update frequency to some extent. FIG. 16 is a flowchart showing such angular velocity estimation processing (hereinafter referred to as “second angular velocity estimation processing”).

  As can be seen by comparing FIG. 16 with FIG. 15, among the steps included in the angular velocity estimation process of FIG. 16, the same steps as those included in the angular velocity estimation process of FIG. Then, steps S164, S166, S178 to S182 in the angular velocity estimation process of FIG. 15 are deleted in the angular velocity estimation process of FIG. 16, and in the angular velocity estimation process of FIG. 16, the variable j is smaller than the predetermined value ja in step S160. If it is determined (No), the process proceeds to step S170 without updating the resistance value Rnew used to calculate the estimated angular velocity value ωe.

  According to such an angular velocity estimation process of FIG. 16, every time the detected values Vm and Im are acquired ja times or more and a change in the rotation angle θs is detected, the acquired detected values I (i) and V (i ) (I = 1 to j) is used to calculate the resistance value Rnew by the equation (19) (step S162), thereby updating the resistance value Rnew used for calculating the angular velocity estimated value ωe. Thereby, a stable angular velocity estimated value ωe with little variation, that is, an angular velocity estimated value ωe in which the influence of noise included in the detected value is suppressed can be acquired with high frequency. As a result, even if the contact state between the brush of the motor 1 and the commutator, the temperature, or the like changes, the estimated angular velocity value ωe can be accurately calculated using the resistance Rnew with less error than the actual motor resistance value. it can.

<5. Fourth Embodiment>
<5.1 Configuration of motor control device>
Similarly to the first embodiment, the motor control device according to the fourth embodiment of the present invention is used in the electric power steering device as shown in FIG. 1 is driven. The configuration of the motor control device is basically the same as that of the first embodiment except that the steering angle sensor 7h is provided instead of the rotation sensor 7 and the configuration of the angular velocity estimation unit 50. As shown in FIG. However, the temperature sensors 41, 42, and 43 of the respective parts indicated by dotted lines in FIG. 2 are not used in this embodiment. In this embodiment, in calculating the angular velocity estimated value ωe, the steering angle θh indicated by the output signal of the steering angle sensor 7h is not used as it is, but the steering angular velocity ωh obtained by differentiating the steering angle θh is used. used. Therefore, instead of the steering angle sensor 7h, a low-cost angular velocity sensor that simply detects the angular velocity of the steering angular velocity sensor or the motor may be used (see FIGS. 1 and 10).

  FIG. 17 is a block diagram showing a functional configuration of the angular velocity estimation unit 50 in the present embodiment. The angular velocity estimation unit 50 includes a resistance value update unit 200, a back electromotive force constant update unit 300, and an angular velocity estimation value calculation unit 66.

  The resistance value update unit 200 updates the resistance value Rnew used for calculating the angular velocity estimated value ωe in the angular velocity estimation processing (first angular velocity estimation processing in the first embodiment) in FIG. 8 (FIG. 8). In this case, the resistance value updating process corresponding to the portion surrounded by the dotted line in FIG. Therefore, the resistance value update unit 200 is functionally configured by a steering determination unit 51, a resistance calculation unit 52, and an average value calculation unit 53 included in the angular velocity estimation unit 50 shown in FIG. However, in the present embodiment, the determination (step S10) as to whether or not the steered state is the steered state by the steerage determining unit 51 is not the pulse signal P from the rotation sensor 7, but the output signal from the steer angle sensor (or steering angular velocity sensor) 7h. This is performed using the steering angular velocity ωh obtained based on the above. Instead of such a configuration, the resistance value updating unit 200 is changed by a process for updating the resistance value Rnew used for calculating the angular velocity estimation value ωe among the other angular velocity estimation processes in each embodiment of the present invention. It may be realized.

  As shown in FIG. 17, the back electromotive force constant update unit 300 includes a speed determination unit 81, a back electromotive force constant calculation unit 82, and an average value calculation unit 83.

  In the angular velocity estimation unit 50, the steering angular velocity ωh obtained based on the output signal from the steering angle sensor (or steering angular velocity sensor) 7h is input. Since the detection accuracy of the steering angle sensor 7h is higher as the rotational speed of the motor 1 is higher, the speed determination unit 81 determines whether or not the steering angular speed ωh is equal to or higher than a predetermined minimum value ωhmin, and shows the determination result. The signal is output as the rotation state signal Sp. Based on the rotation state signal Sp, the back electromotive force constant calculation unit 82 determines the steering angular velocity ωh, the current detection value Im, and the voltage detection value when the motor 1 is rotating at a predetermined speed or higher corresponding to the minimum value ωhmin. The counter electromotive force constant k is calculated for each control period using Vm. The average value calculation unit 83 obtains the counter electromotive force constant average value knew by performing an averaging process on the counter electromotive force constant k sequentially obtained for each control period in this way. The angular velocity estimated value calculation unit 66 calculates the angular velocity estimated value ωe using the resistance value Rnew, the counter electromotive force constant average value knew, the detected current value Im, and the detected voltage value Vm.

<5.2 Specific example of angular velocity estimation processing>
Hereinafter, a specific example will be described for angular velocity estimation processing (step S4 in FIG. 3) for realizing the angular velocity estimation unit 50 in the present embodiment in software.

  FIG. 18 is a flowchart showing an example of the angular velocity estimation process S4 in the present embodiment. In the angular velocity estimation processing of FIG. 18, first, in order to realize the resistance value updating unit 200 in software, the angular velocity estimation value ωe is calculated using the current detection value Im and the voltage detection value Vm acquired up to the present time. In order to newly calculate the resistance value Rnew, the resistance value update process S200 is executed. In the present embodiment, as the resistance value update process S200, as described above, the resistance value update process described above included in the angular velocity estimation process of FIG. 8 (a process consisting of steps surrounded by a dotted line in FIG. 8). ) Is used. After the resistance value update process S200 is executed, the counter electromotive force constant update process S300 is executed in order to newly calculate the counter electromotive force constant knee for calculating the angular velocity estimated value ωe.

  In the angular velocity estimation process (FIG. 18) in this embodiment, the resistance value update process S200 and the counter electromotive force constant update process S300 are reversed in order, and the resistance value update is performed after executing the counter electromotive force constant update process S300. The process S200 may be executed. Since the details of the resistance value update process S200 are clear from the above description, details of the back electromotive force constant update process S300 will be described below. In the back electromotive force constant update process S300, a predetermined minimum value ωhmin and a reduction ratio Cg of the speed reducer 2 (see FIG. 1) regarding the steering angular velocity are used, which are appropriate based on experiments and the like. It is assumed that the value is set in step S2 of FIG. In addition, it is assumed that the variable Ck used in the counter electromotive force constant update process S300 is also initialized in step S2 of FIG.

  In the back electromotive force constant update processing S300, first, it is determined whether or not the steering angular velocity ωh obtained based on the output signal from the steering angle sensor 7h is equal to or greater than a predetermined minimum value ωhmin (step S202). As a result of this determination, if the steering angular velocity ωh is equal to or greater than the minimum value ωhmin (Yes), the process proceeds to step S204. If the steering angular velocity ωh is smaller than the minimum value ωhmin (No), the angular velocity is It progresses to step S220, without performing the process for updating the counter electromotive force constant knew used for calculation of estimated value (omega) e.

  In step S204, it is determined whether or not a variable Ck indicating the number of acquired voltage detection values Vm and current detection values Im used to newly calculate the counter electromotive force constant knew is zero. If the variable Ck is 0 as a result of the determination, the process proceeds to step S222, and if the variable Ck is not 0, the process proceeds to step S206. The microcomputer 10 has a built-in timer, and in step S222, the timer is started. As a result, the elapsed time Tkelp from the start of the timer is initialized to zero. Thereafter, the process proceeds to step S206.

  In step S206, it is determined whether or not the elapsed time Tkelp is less than or equal to the time limit value Tlim. As a result of this determination, when the elapsed time Tkelp is equal to or shorter than the time limit value Tlim (in the case of Yes), the variable Ck is increased by 1 (step S208), and the process proceeds to step S210. On the other hand, when the elapsed time Tkelp exceeds the time limit value Tlim (in the case of No), the variable Ck is initialized to 0 (step S224), and the process without updating the counter electromotive force constant knee is performed. Proceed to S220. As a result, the old detection value Im. The use of Vm is avoided.

In step S210, the steering angular velocity ωh, the update resistance value Rnew obtained in the resistance value update process S200, the reduction ratio Cg of the reduction gear 2, the current detection value Im, and the voltage detection value Vm are used. The counter electromotive force constant k is calculated by the following equation (22) (hereinafter, this counter electromotive force constant k is also referred to as “calculated counter electromotive force constant k”). The reduction ratio Cg is a ratio (for example, 1/15) of the rotational speed of the slave shaft to the rotational speed of the master shaft in the speed reducer 2. Further, in the case where the back electromotive force constant update process S300 is executed prior to the resistance value update process S200, the resistance value Rnew calculated when the angular velocity estimation process was previously executed or the step of FIG. The resistance value Rnew set in S2 is used as the resistance value Rnew of the following formula (22).
k = (Vm−Im · Rnew) / (ωh / Cg) (22)

  In the present embodiment, a FIFO memory is realized by software using the memory in the microcomputer 10, and when the counter electromotive force constant k is newly calculated in step S210 as described above, the counter electromotive force constant is calculated. k is stored in the FIFO memory (step S212). This FIFO memory is a first-in first-out memory having a predetermined value Cka (for example, 50), and in the state where Cka back electromotive force constants k are stored, another new back electromotive force constant k is further stored. When stored, the back electromotive force constant k in the FIFO memory stored first is discarded. For this reason, when the counter electromotive force constant k is stored Cka times or more in this FIFO memory, only the Cka number of counter electromotive force constants k stored in the latest Cka times exist in the FIFO memory.

  After the newly calculated counter electromotive force constant k is stored in the FIFO memory, it is determined whether or not the variable Ck is greater than or equal to a predetermined value Cka (step S214). As a result of this determination, if the variable Ck has reached the predetermined Cka (in the case of Yes), the process proceeds to step S216. If the variable Ck has not reached the predetermined Cka (in the case of No), the process proceeds to step S226. Since Ck = 1 at the time of the first execution of step S214, the process proceeds to step S226. However, the angular velocity estimation process of FIG. Then, the process proceeds to step S216.

  When the process proceeds to step S216, Cka back electromotive force constants k calculated in the most recent Cka times at that time are present in the FIFO memory, and the average value of these Cka back electromotive force constants k is newly inverted. Obtained as electromotive force constant knew. This means that a new counter electromotive force constant knew is obtained by performing a moving averaging process on the counter electromotive force constant k sequentially obtained for each control period. Thereafter, the new counter electromotive force constant knee is set to kold (step S218), and the process proceeds to step S220.

When the process proceeds to step S226, Ck counter electromotive force constants k calculated in the most recent Ck times at that time exist in the FIFO memory, and an average value k1 of these Ck counter electromotive force constants k is obtained. Here, Ck is an integer greater than or equal to 1 and smaller than Cka, and when Ck is small, the average value k1 is not a stable value but varies, that is, the noise included in the average value k1 is relatively large. Therefore, it is not preferable to use it for calculation of the angular velocity estimated value ωe as it is. Therefore, a new counter electromotive force constant knee is obtained by the following equation (23) indicating the weighted average of the counter electromotive force constants kold and k1 (step S228).
knee = {(1−f (Ck)) · kold + f (Ck) · k1} (23)
Here, f (Ck) is a monotonically increasing function of the variable Ck, which is 0 when Ck = 0 and 1 when Ck = Cka. For example, f (Ck) = Ck / Cka or f (Ck) = Ck ² / Cka ² can be used. The above equation (23) is the number of acquired current detection values Im and voltage detection values Vm for obtaining the counter electromotive force constant knew as a new update parameter value (this is equal to the number of acquisition of the calculated counter electromotive force constant k). The weighted average of the back electromotive force constants kold and k1 is shown. When a new counter electromotive force constant knew is obtained from the equation (23), the process proceeds to step S220.

  In step S220, the angular velocity estimated value ωe is calculated by the equation (2) using the resistance value R = Rnew and the counter electromotive force constant k = knew together with the current detection value I = Im and the voltage detection value V = Vm. In the calculation of the angular velocity estimation value ωe, when the above step S216 or S228 is executed, the back electromotive force constant knew calculated there is used, but neither of the above steps S216 or S228 is executed. The counter electromotive force constant knew set before the current control cycle is used. When the angular velocity estimation value ωe is calculated in step S220, the angular velocity estimation process in one control cycle is terminated, and the process returns to the motor control process routine shown in FIG.

<5.3 Effects>
According to the present embodiment as described above, not only the resistance value Rnew but also the counter electromotive force constant knew among the parameter values used for calculating the angular velocity estimated value ωe is updated, and the steering angular velocity ωh is a predetermined minimum value. When it is equal to or larger than ωhmin, the counter electromotive force constant knee is updated every control cycle. Further, the counter electromotive force constant knew includes a plurality of detection values Im. Since it is obtained by averaging the counter electromotive force constant k calculated from Vm according to the equation (22) (steps S216, S226, S228), the counter electromotive force constant k calculated according to the equation (22) is obtained. The angular velocity estimated value ωe is calculated using a stable counter electromotive force constant knew that is stable and has no variation. For this reason, even if the actual back electromotive force constant in the motor 1 changes, the angular velocity estimated value ωe can be calculated with high accuracy. As a result, the control performance of the motor control device is improved, and a good steering feeling can be obtained in an electric power steering device using the motor control device. In the angular velocity estimation process of FIG. 18, every time the variable Ck reaches a predetermined value Cka, a moving average process is performed on the calculated value of the counter electromotive force constant k (step S216). Instead, the counter electromotive force constant is calculated. A simple averaging process may be performed for each calculated value of k by a predetermined number.

<6. Fifth Embodiment>
<6.1 Motor controller configuration>
Similarly to the first embodiment, the motor control device according to the fifth embodiment of the present invention is used in the electric power steering device as shown in FIG. 1 is driven. In this motor control device, unlike the first embodiment, temperature sensors 41, 42, and 43 are provided for each part. Specifically, these temperature sensors include a sensor 41 provided for detecting the coil temperature Tw of the motor 1, a sensor 42 provided for detecting the brush temperature Tb of the motor 1, and the motor. This is a sensor 43 provided for detecting the substrate temperature Td of the drive circuit 30. Moreover, in this embodiment, the structure of the angular velocity estimation part 50 differs from other embodiment. The other configuration of the motor control device according to the present embodiment is basically the same as that of the first embodiment, as shown in FIG.

  FIG. 19 is a block diagram showing a functional configuration of the angular velocity estimation unit 50 in the present embodiment. The angular velocity estimation unit 50 includes a temperature characteristic storage unit 91, a resistance value calculation unit 92, an average value calculation unit 93, and an angular velocity estimation value calculation unit 66. The temperature characteristic storage unit 91, the resistance value calculation unit 92, and the average The value calculation unit 93 constitutes the resistance value update unit 200.

  As shown in FIG. 19, the temperature characteristic storage unit 91 holds a winding resistance map Mw, a brush resistance map Mb, a contact resistance map Mc, and a circuit resistance map Md. Here, the winding resistance map Mw is a table associating the coil temperature with the winding resistance based on the temperature characteristics of the winding resistance of the motor 1 as shown in FIG. 20A, and the brush resistance map Mb is shown in FIG. FIG. 20B is a table for associating the brush temperature with the brush resistance based on the temperature characteristic of the brush resistance of the motor 1 as shown in FIG. The circuit resistance map Md is a table for associating the brush temperature with the contact resistance based on the temperature characteristics of the motor drive circuit 30 based on the temperature characteristics of the substrate temperature of the motor drive circuit 30 as shown in FIG. It is a table corresponding to the resistance (hereinafter referred to as “circuit resistance”).

  The temperature characteristic storage unit 91 receives the coil temperature detection value Tw, the brush temperature detection value Tb, and the substrate temperature detection value Td obtained by the above-described temperature sensors 41, 42, and 43 for each control cycle (see FIG. 3). In step S3), the winding resistance value Rw corresponding to the detected coil temperature value Tw is obtained with reference to the winding resistance map Mw, and the brush resistance value Rb corresponding to the detected brush temperature value Tb is referred to the brush resistance map Mb. The contact resistance value Rc corresponding to the brush temperature detection value Tb is obtained with reference to the contact resistance map Mc, and the circuit resistance value Rd corresponding to the substrate temperature detection value Td is obtained with reference to the circuit resistance map Md. Ask. The winding resistance value Rw, brush resistance value Rb, contact resistance value Rc, and circuit resistance value Rd thus obtained are input to the resistance value calculation unit 92.

  The resistance value calculation unit 92 obtains a motor resistance value Rs for each control cycle from the winding resistance value Rw, the brush resistance value Rb, the contact resistance value Rc, and the circuit resistance value Rd. The average value calculation unit 93 obtains the resistance average value Rnew by performing an averaging process on the motor resistance value Rs obtained sequentially for each control period in this way. The angular velocity estimated value calculation unit 66 detects the resistance average value Rnew, a preset counter electromotive force constant k (step S2 in FIG. 3), the voltage detection value Vm from the voltage detection circuit 24, and the current detection from the current detection circuit 25. An angular velocity estimation value ωe is calculated using the value Im.

<6.2 Specific example of angular velocity estimation processing>
Hereinafter, a specific example will be described for angular velocity estimation processing (step S4 in FIG. 3) for realizing the angular velocity estimation unit 50 in the present embodiment in software. The temperature characteristic storage unit 91 is realized by the angular velocity estimation process and a memory in the microcomputer 10, and constitutes the winding resistance map Mw, the brush resistance map Mb, the contact resistance map Mc, and the circuit resistance map Md. It is assumed that data to be stored is stored in advance in the memory.

  FIG. 21 is a flowchart showing an example of the angular velocity estimation process S4 in the present embodiment. Also in this example, the angular velocity estimation process is executed for each control cycle (see FIG. 3), and the microcomputer 10 operates as shown in FIGS. 21 and 22 in the execution of the angular velocity estimation process.

  The angular velocity estimation process in FIG. 21 is different from the angular velocity estimation process in FIG. 8 in that steps S10 to S12, S30, and S32 are deleted and the processing content for calculating the resistance value Rs (step 16 in FIG. 8). Except for the difference in step S17) in FIG. 21, this is substantially the same as the angular velocity estimation processing in FIG. 8, and the corresponding steps are denoted by the same reference numerals and description thereof is omitted.

  FIG. 22 is a flowchart showing a detailed procedure of the process in step S17 in this embodiment, that is, the resistance value calculation process. In this resistance value calculation process, first, the coil temperature detection value Tw, the brush temperature detection value Tb, and the substrate temperature detection value Td obtained by the above-described temperature sensors 41, 42, and 43 are acquired (step S230). Next, a winding resistance value Rw corresponding to the detected coil temperature value Tw is obtained by referring to the winding resistance map Mw indicating the temperature characteristics of the winding resistance (step S232). Similarly, the brush resistance value Rb corresponding to the brush temperature detection value Tb is obtained by referring to the brush resistance map Mb indicating the temperature characteristic of the brush resistance (step S234), and the contact resistance map Mc indicating the temperature characteristic of the contact resistance is obtained. , The contact resistance value Rc corresponding to the brush temperature detection value Tb is obtained (step S236), and the circuit resistance map Md indicating the temperature characteristic of the circuit resistance is referred to, thereby corresponding to the substrate temperature detection value Td. A circuit resistance value Rd is obtained (step S238).

When the winding resistance value Rw, the brush resistance value Rb, the contact resistance value Rc, and the circuit resistance value Rd are obtained as described above, the motor resistance is calculated as the sum of these resistance values Rw, Rb, Rc, and Rd. A value (hereinafter referred to as “calculated resistance value”) Rs is obtained (step S240). That is,
Rs = Rw + Rb + Rc + Rd (24)
It is.

  When the microcomputer 10 executes the resistance value calculation process (S17) including steps S230 to S240 as described above, the resistance value calculation unit 92 illustrated in FIG. 19 is realized.

  When the calculated resistance value Rs is obtained as described above, the resistance value calculation process (S17) is terminated, the process returns to the angular velocity estimation process routine shown in FIG. 21, and the processes after step S50 are executed. In this process, a plurality of calculated resistance values Rs obtained for each control cycle up to the present time are used, and in the same manner as in the first angular velocity estimation process (angular velocity estimation process in FIG. 8) in the first embodiment, the resistance An average value Rnew (steps S54 and S62) is calculated, and an estimated angular velocity value ωe is calculated using the resistance average value Rnew (step S58). When the angular velocity estimation value ωe is calculated in step S58, the angular velocity estimation process in one control cycle is terminated, and the process returns to the motor control process routine shown in FIG.

<6.3 Effect>
According to the present embodiment as described above, the motor resistance value Rnew necessary for calculating the angular velocity estimated value ωe is determined by considering the temperature characteristics of the resistance value for each resistance (resistance component) of each part constituting the motor resistance R. It is calculated (see steps S230 to S240, FIG. 20). As a result, the estimated angular velocity value ωe is calculated using the resistance value Rnew that accurately reflects the temperature characteristic of the motor resistance, so that the accuracy of the estimated angular velocity value ωe can be improved. Further, since the resistance average value Rnew for calculating the estimated angular velocity value ωe is updated every control cycle in a state where the motor control device is operating (steps S54 and S62), the estimated angular velocity value ωe is calculated. This resistance average value Rnew can follow the change in the actual motor resistance value R, and the deviation from the actual motor resistance value R is suppressed. For this reason, even if the temperature changes, the estimated angular velocity value ωe can be accurately calculated using the resistance average value Rnew with less error than the actual motor resistance value. As described above, the control performance of the motor control device is improved, and a good steering feeling can be obtained in the electric power steering device using the motor control device.

<7. Variations>
In the first to fifth embodiments, the angular velocity estimation unit 50 is realized by software by the microcomputer 10 executing a predetermined program, but part or all of the angular velocity estimation unit 50 is realized by hardware. May be.

  The motor control devices according to the first to fifth embodiments are devices for driving the brushed motor 1, but the present invention is not limited to this. The present invention is based on the back electromotive force generated in the armature winding of the electric motor, based on the resistance value and back electromotive force constant of the motor, the applied voltage to the motor, and the detected value of the current flowing to the motor. As long as the estimated value of the rotational speed (angular speed) can be calculated, the present invention can be applied to other types of motors.

  The present invention can be applied not only to the above-described column assist type electric power steering apparatus but also to a pinion assist type or rack assist type electric power steering apparatus. The present invention can also be applied to a motor control device used in addition to the electric power steering device.

  DESCRIPTION OF SYMBOLS 1 ... Motor with a brush (electric motor), 5 ... ECU (motor control apparatus), 7 ... Rotation sensor, 7h ... Steering angle sensor (simple angle detection means), Steering angular velocity sensor (simple speed detection means), 51 ... Steering Determination unit (determination unit) 56 ... Angular velocity estimation value calculation unit (estimation value calculation unit) 91 ... Temperature characteristic storage unit 200 ... Resistance value update unit (parameter value update unit) 300 ... Back electromotive force constant update unit ( Parameter value updating means), 400... Parameter value updating section (parameter value updating means).

Claims (7)

A motor control device that calculates an estimated value of the rotational speed of the electric motor based on a back electromotive force generated in the armature winding of the electric motor, and drives the electric motor using the estimated value,
Current detecting means for detecting a current flowing through the electric motor;
Voltage detecting means for detecting a voltage applied to the electric motor;
Based on a plurality of detection values sequentially acquired for a predetermined physical quantity when the electric motor is in a predetermined state, at least one of the resistance value of the electric motor and the back electromotive force constant, which are parameters in the circuit equation of the electric motor. Parameter value update means for calculating the value of one parameter as an update parameter value;
Using the updated parameter value calculated by the parameter value updating unit, the current detection value acquired by the current detection unit, and the voltage detection value acquired by the voltage detection unit, the electric motor based on the circuit equation An estimated value calculating means for calculating a speed estimated value indicating the rotational speed of the motor,
The parameter value updating means includes
When the number of detection values acquired for the physical quantity for a new calculation of the update parameter value is equal to or greater than a predetermined number, a new update parameter value is obtained by the first predetermined method using the acquired detection value. Calculate
When the number of detection values acquired for the physical quantity for a new calculation of the update parameter value is smaller than the predetermined number, the acquired detection value and the update parameter value already obtained are used. A new update parameter value is calculated by the second predetermined method,
When the new update parameter value is calculated by the parameter value update unit, the estimated value calculation unit calculates the speed estimated value using the new update parameter value. A determination means for determining whether or not the rotation speed of the electric motor is equal to or less than a predetermined value near 0;
The parameter value updating means is sequentially acquired as the physical quantity detection value by the current detection means and voltage detection means when the determination means determines that the rotation speed of the electric motor is equal to or less than the predetermined value. Calculate a resistance value of the electric motor based on a plurality of current detection values and voltage detection values,
The first predetermined method obtains a resistance value of the electric motor at the detection time as a calculated resistance value from the current detection value and the voltage detection value at each detection time of the plurality of current detection values and voltage detection values, It is a method of calculating a moving average value for a predetermined number of the calculated resistance values as a new update parameter value,
In the second predetermined method, a resistance value of the electric motor at the detection time is obtained as a calculated resistance value from the current detection value and the voltage detection value at each detection time of the plurality of current detection values and voltage detection values. 2. The motor according to claim 1, wherein a weighted average value of an average value of the calculated resistance value and the update parameter value already obtained is calculated as a new update parameter value. Control device. A simple speed detecting means for detecting the rotational speed of the electric motor with an accuracy lower than that required for drive control of the electric motor;
The parameter value update means includes a plurality of current detection values sequentially acquired by the current detection means and the voltage detection means as detection values of the physical quantity when the rotation speed detected by the simple speed detection means is a predetermined value or more. And a resistance value and a back electromotive force constant of the electric motor based on the voltage detection value and a plurality of rotation speed detection values sequentially obtained by the simple speed detection means,
In the first predetermined method, the current detection value, the voltage detection value, and the rotation speed detection value at each detection point of the plurality of current detection values and voltage detection values satisfy at least approximately the circuit equation. , Calculating the resistance value and back electromotive force constant of the electric motor as a new update parameter value,
In the second predetermined method, the current detection value, the voltage detection value, and the rotation speed detection value at each detection time of the plurality of current detection values and voltage detection values satisfy at least approximately the circuit equation. A method of calculating a resistance value and a back electromotive force constant of the electric motor as a temporary parameter value, and calculating a weighted average value of the temporary parameter value and the update parameter value already obtained as a new update parameter value The motor control device according to claim 1, wherein: A simple angle detecting means for detecting the rotation angle of the electric motor with a resolution such that the rotation speed of the electric motor is obtained with an accuracy lower than the accuracy required for drive control of the electric motor;
The parameter value updating means changes the simple angle from the first time point to the first time point when the rotation angle detected by the simple angle detection means is changed and a new rotation angle is detected as the first rotation angle detection value. Until the second time point when the rotation angle detected by the detection unit changes and a new rotation angle is detected as the second rotation angle detection value, the current detection unit and the voltage detection unit sequentially acquire the detection value of the physical quantity. A resistance value of the electric motor is calculated based on a plurality of detected current values and detected voltage values,
The first predetermined method is based on an equation obtained by integrating the circuit equation from the first time point to the second time point, using the plurality of current detection values and voltage detection values and the first and second rotation angles. It is a method of calculating the resistance value of the electric motor as a new update parameter value,
The second predetermined method is based on an equation obtained by integrating the circuit equation from the first time point to the second time point using the plurality of current detection values and voltage detection values and the first and second rotation angles. It is a method of calculating a resistance value of the electric motor as a temporary parameter value, and calculating a weighted average value of the temporary parameter value and the already obtained update parameter value as a new update parameter value. The motor control device according to claim 1. A simple speed detecting means for detecting the rotational speed of the electric motor with an accuracy lower than that required for drive control of the electric motor;
The parameter value update means includes a plurality of current detection values sequentially acquired by the current detection means and the voltage detection means as detection values of the physical quantity when the rotation speed detected by the simple speed detection means is a predetermined value or more. And a voltage detection value and a plurality of rotation speed detection values sequentially obtained by the simple speed detection means, to calculate a counter electromotive force constant of the electric motor,
The first predetermined method uses the current detection value, the voltage detection value, and the rotation speed detection value at each detection time of the plurality of current detection values and voltage detection values to determine the detection time at the detection time from the circuit equation. A method of obtaining a counter electromotive force constant of an electric motor as a calculated counter electromotive force constant, and calculating a moving average value for the predetermined number of the calculated counter electromotive force constants as a new update parameter value,
The second predetermined method uses the current detection value, the voltage detection value, and the rotation speed detection value at each detection time of the plurality of current detection values and voltage detection values to determine the detection time at the detection time from the circuit equation. A back electromotive force constant of the electric motor is obtained as a calculated back electromotive force constant, and a weighted average value of the average value of the plurality of calculated back electromotive force constants and the update parameter value already obtained is used as a new update parameter value. The motor control device according to claim 1, wherein the motor control device is a method of calculating. Temperature detecting means for detecting the temperature of the component of the device corresponding to each of the plurality of resistance components constituting the resistance of the electric motor as the physical quantity;
The parameter value updating means includes
Including temperature characteristic storage means for individually holding the temperature characteristics of the plurality of resistance components,
Calculate a resistance value of the electric motor based on the temperature characteristics using a plurality of temperature detection values sequentially acquired for each component by the temperature detection means,
The first predetermined method uses the temperature detection value at each detection time of the plurality of temperature detection values sequentially acquired for each component, and detects the detection based on the temperature characteristic of the resistance component corresponding to the component. The value of the resistance component at the time is acquired, the sum of the values for the plurality of resistance components is obtained as the calculated resistance value of the electric motor, and the moving average value for the predetermined number of calculated resistance values is newly updated. It is a method of calculating as a parameter value,
The second predetermined method uses the temperature detection value at each detection time of the plurality of temperature detection values sequentially acquired for each component, and detects the detection based on a temperature characteristic of a resistance component corresponding to the component. The value of the resistance component at the time is acquired, the sum of the values for the plurality of resistance components is obtained as the calculated resistance value of the electric motor, and the average value for the plurality of calculated resistance values is already obtained. The motor control device according to claim 1, wherein the weighted average value with the update parameter value is calculated as a new update parameter value. An electric power steering device for applying a steering assist force to a steering mechanism of a vehicle by an electric motor,
The motor control device according to any one of claims 1 to 6, comprising:
The motor control device drives an electric motor that gives the steering assist force to the steering mechanism.

Images