JPH08130893A - Method for controlling brushless motor - Google Patents

Abstract

PURPOSE: To maintain a stable control state even if noise is superposed on detection current. CONSTITUTION: When controlling the current (output torque) of a motor 10 using the position and speed of a rotor of the motor 10 estimated by a position/ speed estimator 18, the amount of change of these values is limited by a position/speed change limiter 24 when the a position estimation value and a speed estimation value greatly changes unnaturally. The amount of reference position change and the amount of reference change which becomes the reference of limitation are set based on a result obtained by estimating a load torque TL of the motor 10 by a load torque estimator 26.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control method for controlling a brushless motor without speed detection.

[0002]

2. Description of the Related Art "Proposal of Sensorless Control Using Current Control Error of Brushless DC Motor", Yasuda et al. In 6-57 ~ 6-58, brushless DC
There has been proposed a control method that does not need to detect the rotational angular velocity when controlling a motor, that is, a sensorless control.

In this control method, the voltage equation on the (d, q) coordinate system of the brushless motor is basically used.

[Equation 1] Is used. In the (d, q) coordinate system, the direction of the field flux is d
The coordinate system has an axis and a direction orthogonal to the axis as the q axis. When the equation (1) is linearly approximated at time n to enable discrete processing,

[Equation 2] Becomes Here, considering a (γ, δ) coordinate system that rotates following a field flux similar to the (d, q) coordinate system, it is considered that the difference Δθ (n) between the two angular positions at time n is sufficiently close to 0. After that, when the equation (2) is transformed from the (d, q) coordinate system to the (γ, δ) coordinate system,

(Equation 3) Becomes Rewrite equation (3) using vector expression,
Solving for the current vector i (n + 1) at time n + 1,

[Equation 4] Becomes

Furthermore, a model for modeling the equation (4) and estimating the current vector at time n + 1 from the current vector i (n) at time n

(Equation 5) think of. The difference between the actual current represented by equation (4) and the model current represented by equation (5) is the difference between the angular position of the (d, q) coordinate system and the angular position of the (γ, δ) coordinate system. This is due to the difference Δθ (n). Therefore, the difference Δi between equation (4) and equation (5)
Considering (n + 1) = i (n + 1) -i _M (n + 1), and assuming that the speed error of the model with respect to the actual one can be ignored at high speed operation,

(Equation 6) Is obtained. When this equation is transformed,

(Equation 7) Then, the formulas for estimating the angular position and speed of the rotor at time n + 1 are obtained. The actual angular position θ (n) of the rotor does not appear in Expression (7), and the measured value is the armature current (here, i _γ (n) and i _δ (n)) at the previous time n. An observer for estimating the angular position and speed of the rotor at time n + 1 can be obtained only by inputting, that is, without detecting the position and speed of the rotor.

[0005]

By constructing the control system using such an observer, the sensor for detecting the position and speed of the rotor can be eliminated, and the size and simplification can be realized. Here, in order to realize the above-mentioned observer, it is necessary to obtain the current error of the model with respect to the actual value, and therefore it is necessary to detect the actual current value. But,
If an error occurs in the output of the current sensor for that purpose, a calculation error occurs in the position estimation value and the like as shown in FIG. In addition, a calculation error may occur in the CPU that performs the position estimation calculation. If a position estimation value including such a calculation error or a calculation error is used for controlling the motor, the control becomes unstable. This is due to the nature of the algorithm that repeatedly executes the calculation to focus on the true value, even if the current detection value obtained by sampling for a relatively short time is used for the control, it makes the control unstable. There are many.

The present invention has been made to solve the above problems, and corrects the position or velocity estimated value when it can be considered that an error or abnormality has occurred in the position or velocity estimated value. In addition, it is an object of the present invention to realize a control method in which the control state is stable irrespective of noise superposition on the output of the current sensor and a rotational position or rotational speed sensor is not required.

[0007]

In order to achieve such an object, the present invention estimates the angular position and / or rotational angular velocity of the rotor of a brushless motor based on the detected value of the current flowing through the brushless motor. And a step of determining whether or not the estimated value of the angular position or the rotational angular velocity at the present time has changed by more than the reference change amount compared with the estimated value at the previous time, and when it is determined that the change is large. , Assuming that an error has occurred in the estimated value of the angular position or the rotational angular velocity at the present time, and replacing the estimated value at the present time with an alternative value so that the estimated value at the previous time does not change significantly, The step of controlling the current based on the angular position and / or the rotational angular velocity is repeatedly executed at predetermined intervals.

The present invention further includes a step of estimating the load torque of the brushless motor, a step of setting a reference change amount of an angular position and / or a rotational angular velocity based on the estimated load torque, and an estimated value at a previous time point. The step of setting the value obtained by adding the reference change amount to the above alternative value,
It is characterized in that it is repeatedly executed every cycle.

According to the present invention, when estimating the load torque of the brushless motor, first, the instantaneous load torque of the brushless motor is obtained based on the estimated values of the angular position or the rotational angular velocity at a plurality of recent times, and the obtained instantaneous load is calculated. It is characterized in that the torque is averaged over a period longer than the time constant of current detection to obtain an average load torque, and the obtained average load torque is used for setting the reference change amount.

[0010]

In the present invention, first, based on the detected value of the current flowing through the brushless motor, the angular position of the rotor and / or
Alternatively, the rotational angular velocity is estimated. Further, it is determined whether or not the estimated value of the angular position or the rotational angular velocity at the current time point is larger than the reference change amount as compared with the estimated value at the previous time point. If the change is large, it can be considered that an error has occurred in the estimated value of the angular position or the rotational angular velocity at the present time due to, for example, noise superimposed on the current detection. Therefore, in such a case, the estimated value at the present time is replaced with the alternative value so that the estimated value at the previous time does not significantly change. By controlling the current flowing through the brushless motor using the estimated value corrected in this way, the control state becomes stable irrespective of noise superposition on the current sensor output. Further, as in the conventional case, no rotational position or rotational speed sensor is required.

Further, in the present invention, the load torque of the brushless motor is estimated, the above-mentioned reference change amount is set based on the result, and the value obtained by adding the reference change amount to the estimated value at the previous time point is the alternative value. Is set to. As described above, by determining the reference change amount and the alternative value according to the load torque, the operation reflecting the load torque fluctuation is realized, so that the applicable range becomes wider.

In the present invention, when the load torque of the brushless motor is estimated, the instantaneous load torque of the brushless motor is first obtained based on the estimated values of the angular position or the rotational angular velocity at a plurality of recent times. Further, the obtained instantaneous load torque is averaged over a longer period than the time constant of current detection, and the average load torque is obtained. The obtained average load torque is used for setting the reference change amount. Therefore, it is possible to use the estimated value of the angular position or the rotational angular velocity that has been obtained before in estimating the load torque. Further, even when the noise of current detection has an influence on the estimated value of the angular position or the rotational angular velocity obtained previously, this influence can be eliminated by averaging.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT A preferred embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 shows the configuration of a motor control system according to an embodiment of the present invention. The motor 10 to be controlled in the system of this figure is a synchronous motor having a permanent magnet as a field, and the system of this figure is used as a drive system of an electric vehicle, for example.

When the system shown in this figure is used as a drive system for an electric vehicle, a higher-order control device (not shown) is
Determine the required output torque according to the operation of the accelerator pedal or brake pedal by the vehicle operator. The higher-order control device obtains a motor current corresponding to the obtained output torque (torque command), and uses this as a current command to control the current control unit 12
Supply to. The current control unit 12 generates and outputs a voltage command indicating the PWM threshold value in the PWM control unit 14 in the subsequent stage based on the input current command.

The PWM control unit 14 compares the voltage command with a carrier having a predetermined waveform to obtain pulse width modulation (PW).
M) generated switching signal. The inverter 16 provided in the subsequent stage of the PWM control unit 14 is composed of a predetermined number of switching elements, and these switching elements switch according to the switching signal supplied from the PWM control unit 14. Here, the voltage command is generated according to the current command, and the switching signal is generated according to the voltage command. Therefore, the inverter 1 responds to such a switching signal.
By executing the switching operation of 6, the DC power supplied to the inverter 16 from a power source (not shown), for example, a vehicle-mounted battery can be converted into an AC current according to the current command and supplied to the motor 10. As a result of such current control, the output torque of the motor 10 becomes a value corresponding to the torque command determined by the host control device.

In this embodiment, the position / speed estimator 18 is used as an observer for estimating the position and speed of the rotor of the motor 10. Position / velocity estimator 18
In estimating the position and velocity, first, the current and voltage of the motor 10 are detected by the current sensor 20 and the voltage sensor 22, and the detected current and detected voltage that are the results are input. However, a voltage command may be input from the current controller 12 instead of or in addition to the detected current. The position / speed estimator 18 regards the detected current (or voltage command) and the detected voltage as values in the (γ, δ) coordinate system, or changes from the (d, q) coordinate system to the (γ, δ) coordinate system. After the conversion, the algorithm shown in the above-mentioned paper is executed to estimate the position and speed of the rotor. The rotor position estimation value θ _M (n + 1) and the speed estimation value θ ^· _M (n + 1) obtained by the position / velocity estimator 18 are given by the position / velocity change limiter 24.
Is input to the current control unit 12 via. Current control unit 12
Is the estimated position value θ _M (n + 1) and the estimated speed value θ
^The current command is converted from the coordinate system fixed to the rotor to the stationary coordinate system by using _M (n + 1), and the detected current obtained by the current sensor 20 is compared with the current command input from the host controller. Thus, the change in the current command value is detected, and the voltage command is generated based on the result.

This embodiment is characterized by the position
A speed change limiter 24 and a load torque estimator 26.
The position / speed change limiter 24 calculates the position estimated value θ _M (n + 1) at the time n + 1 obtained by the position / speed estimator 18.
And the estimated speed value θ ^· _M (n + 1) are significantly different from the estimated position value θ _M (n) and the estimated speed value θ ^· _M (n) at time n, the change is referred to as the reference position. Limit to the amount of change and reference speed change. Load torque estimator 26
Is an observer for estimating the load torque _TL of the motor 10, and sets the reference position change amount and the reference speed change amount in the position / speed change limiter 24 based on the estimation result.

FIG. 2 shows the operation of the load torque estimator 26 in this embodiment. As shown in this figure, the load torque estimator 26 firstly estimates the speed values θ ^· _M (n−1) and θ ^· _M (n−1) obtained by the position / speed estimator 18 at times n−1 and n. n)

(Equation 8) It performs a computation of the instantaneous rotation angle acceleration θ ^·· at time n
_{Determine M} (n) (100). Load torque estimator 26
Next, after obtaining the output torque T _M of the motor 10 from the detected current and the current command, the instantaneous rotational angular acceleration θ
^{.. Based on} _M (n), output torque T _M, and known vehicle inertia J

[Equation 9] Of the instantaneous load torque T at time n
_L (n) is calculated (102). When obtaining the output torque T _M , it is preferable to use the detected current in order to improve the responsiveness. The load torque estimator 26 further includes

[Equation 10] The average load torque T ^- _L (n) at time n is obtained by the moving average calculation based on the equation (104). The averaging period k is a time period (e.g., 0.1 seconds to) that is sufficiently longer than the electrical time constant of the current sensor 20 (e.g., several microseconds to several milliseconds).
(Several seconds).

The load torque estimator 26 uses the average load torque T ^- _L (n) thus obtained, and formula (9)
The reference angular acceleration θ ^·· _ref is set by the calculation opposite to (106). That is, the next operation

[Equation 11] To execute. The load torque estimator 26, in addition to the obtained reference angular acceleration theta ^{· ·} _ref, the speed estimated value at time n theta
^· _M (n) using a,

(Equation 12) Then, the reference position change amount Δθ _ref and the reference speed change amount Δθ ^· _ref are set (108).

FIG. 3 shows the operation of the position / speed change limiter 24. As shown in this figure, the position / speed change limiter 24 uses the position estimation value θ at time n + 1.
_M (n + 1) and estimated speed value θ ^· _M (n + 1) and time n
Position estimation value θ _M (n) and velocity estimation value θ
^- the difference between _M (n), that is

(Equation 13) Δθ _M and Δθ ^· _M shown in
After inputting from the speed estimator 18, these and Δθ _ref
And the absolute value of the difference between Δθ ^· _ref and the small value ε ₁ or ε
₂ and 200 respectively (200, 202). ε ₁ and ε
Both ₂ are positive values, and are determined in consideration of the output torque T _M of the motor 10 and the influence on the efficiency. Step 200
The result of the comparison in

[Equation 14] If it is determined that the change amount Δθ _M is large, the position / speed change limiter 24 uses the position estimated value θ.
_To limit the amount of change in the position estimation value θ _M (n + 1) with respect to _M (n) to Δθ _ref

(Equation 15) The process of is executed (204). The result of the comparison in step 202,

[Equation 16] When the above condition holds, it can be considered that the amount of change Δθ ^· _M is large, and therefore, the estimated speed value θ with respect to the estimated speed value θ ^· _M (n) is
^{・ To} limit the amount of change in _M (n + 1) to Δθ ^・ _ref

[Equation 17] The process of is executed (206). In this way, the position estimation value θ _M (n + 1) whose amount of change is limited under the predetermined condition
And the estimated speed value θ ^· _M (n + 1) are supplied to the current control unit 12 and used for control, and the position / speed estimator 18 and the load torque estimator 2 are also used for processing at a later time.
6 is also supplied.

Therefore, according to the present embodiment, since the position / speed change limiter 24 limits the amount of change in the position estimation value θ _M (n + 1) and the speed estimation value θ ^· _M (n + 1), the current As a result of noise being superimposed on the detection current output from the sensor 20, the position estimation value θ _M (n + 1) and the speed estimation value θ output from the position / speed estimator 18
^{Even when} _M (n + 1) becomes the calculation error data as shown in FIG. 4, the position estimation value θ _M (n + 1) and the speed estimation value θ ^· _M (n + 1) supplied to the current control unit 12 are set. No significant change appears, so that the current control (torque control) of the motor 10 does not become unstable. Further, the reference position change amount Δθ _ref and the reference speed change amount Δθ ^· _ref used in the position / speed change limiter 24 are set based on the load torque T _L which is considered to have little variation in the electric vehicle traveling motor or the like. Therefore, Δθ _M
Also, the use of Δθ _ref and Δθ ^· _ref instead of Δθ ^· _M does not impair the control stability. Furthermore, Δθ
_In setting _ref and Δθ ^· _ref , not the instantaneous load torque T _L (n) but the average load torque T ⁻ _L (n)
Is used and the averaging period is set longer than the electrical time constant of the current sensor 20,
It is possible to preferably eliminate the influence of noise superimposed on the output of the.

In the above description, the load torque estimator 26 estimates the average load torque T ^- _L (n), and based on this, the reference position change amount Δθ _ref and the reference speed change amount Δ.
Although θ ^/ _ref is set, the average load torque T
^- it may be used _L instead of (n) instantaneous load torque _T L (n). With such a configuration, the instantaneous load torque T
_It can be used for applications where the fluctuation of _L (n) is small. Further, in an application in which the fluctuation of the instantaneous load torque T _L (n) is extremely small, the load torque estimator 26 is abolished because the load torque T _L is considered to be constant, and the reference position change amount Δθ _ref and the reference speed change amount Δθ are set. ^· _ref 2008 may be set to a constant.

Further, in the above embodiment, the instantaneous load torque T
By calculating the moving average of _L (n), the current sensor 20
The electric time constant of the noise, that is, the noise can be ignored. However, as another method of making the influence of the noise ignorable, there is another method that can be regarded as the time when the noise is not generated.
Using the instantaneous load torque T _L (n−i) estimated in step 1, the reference position change amount Δθ _ref and the reference speed change amount Δθ ^·
There is also a method of setting _ref . Whether or not noise is generated can be known from the position / speed change limiter 24 as the determination result of steps 200 and 202.

Although the drive system of an electric vehicle has been shown as an application target, the present invention can be applied to other uses.

[0026]

As described above, according to the present invention,
If the estimated value of the angular position or rotational angular velocity at the current time has changed more than the reference change amount compared to the estimated value at the previous time point, it is considered that there is an error in the estimated value of the angular position or rotational angular velocity at the current time. Since the estimated value at the present time is replaced with an alternative value so that the estimated value at the time does not change significantly, the control state is stable regardless of the superposition of noise on the output of the current sensor. Further, as in the conventional case, no rotational position or rotational speed sensor is required.

According to the present invention, since the load torque of the brushless motor is estimated and the reference change amount and the alternative value are set based on the result, the operation reflecting the load torque fluctuation is realized and the applicable range is increased. Get wider.

According to the present invention, when estimating the load torque, the instantaneous load torque is obtained on the basis of the estimated values of the angular position or the rotational angular velocity at a plurality of recent times, and this is calculated over a long period as compared with the time constant of the current detection. Since the average load torque obtained by averaging is used for setting the reference change amount, it is possible to use the estimated value of the angular position or rotational angular velocity previously obtained for load torque estimation, and the previously obtained angular position or rotational speed. Even if the noise of current detection influences the estimated value of the angular velocity, this influence can be eliminated by averaging.

[Brief description of drawings]

FIG. 1 is a block diagram illustrating a configuration of a system according to an embodiment of the present invention.

FIG. 2 is a flowchart showing a flow of a load torque estimation operation in this embodiment.

FIG. 3 is a flowchart showing a flow of a position / speed change limiting operation in this embodiment.

FIG. 4 is a waveform diagram showing a conventional problem.

[Explanation of symbols]

10 motor 12 current control unit 18 position / speed estimator 20 current sensor 24 position / speed change limiter 26 load torque estimator n time k average period J vehicle inertia T _M motor output torque T _L (n) at time n ( Instantaneous) Load torque T ^- _L (n) Average load torque at time n θ _M (n) Position estimation value at time n θ ^· _M (n) Speed estimation value at time n θ ^·· _M (n) Instant at time n angular acceleration theta ^{· ·} _{r ef} reference angular acceleration [Delta] [theta] _ref reference position change amount [Delta] [theta] ^· _ref variation epsilon ₁ of the reference speed variation [Delta] [theta] _M variation [Delta] [theta] ^· _M estimated speed of the position estimate, epsilon ₂ minute value

Claims (3)

[Claims] 1. A step of estimating an angular position and / or a rotational angular velocity of a rotor of a brushless motor based on a detected value of a current flowing in the brushless motor, and a step of estimating the current based on the estimated angular position and / or the rotational angular velocity. And a step of determining whether or not the estimated value of the angular position or the rotational angular velocity at the current time point is larger than the reference change amount compared to the estimated value at the previous time point in the control method of repeatedly executing the controlling step and the predetermined step. If it is determined that there is a large change, it is considered that there is an error in the estimated value of the angular position or rotational angular velocity at the current time, and the current value is estimated so that there is no significant change in the estimated value at the previous time. Replacing the estimated value with an alternative value and using it for controlling the current described above may be repeated for each cycle. And a control method characterized by: 2. The control method according to claim 1, wherein a step of estimating a load torque of the brushless motor, a step of setting a reference change amount of an angular position and / or a rotational angular velocity based on the estimated load torque, And a step of setting a value obtained by adding a reference change amount to an estimated value at a time point as the alternative value, repeatedly performing the cycle every cycle. 3. The control method according to claim 2, wherein when estimating the load torque of the brushless motor, first, the instantaneous load torque of the brushless motor is obtained based on the estimated values of the angular position or the rotational angular velocity at a plurality of recent times. A control method characterized in that the obtained instantaneous load torque is averaged over a period longer than the time constant of current detection to obtain an average load torque, and the obtained average load torque is used for setting a reference change amount.