Patent Application
20220016983
The present invention relates to a method for operating a drive train and to a drive train.
In the case of converter-fed electric motors, it is generally known that the signal electronics determines the actual value of the torque based on the acquired motor current, the motor voltage, and/or the characteristic of the acquired angular position of an electric motor.
In the same way, it is generally known in converters to supply an electric motor with a motor voltage such that the electric motor is controlled to a torque setpoint value.
From the document DE 102007028914 A1 as the most proximate related art, a device for controlling and/or regulating the speed of an electric motor is known.
From the document DE 69120486 T2, a method for calculating the moment of inertia in a motor speed controller is known.
Therefore, the present invention is based on the object of achieving the best possible control quality in the control of an electric motor.
According to the present invention, the objective in the method is achieved by the features as recited in claim 1, and in the drive train it is achieved by the features as recited in Claim 9.
Important features of the present invention in the method for operating a drive train are that the drive train has an electric motor and a load which is driven by the electric motor or by a gear unit driven by the electric motor, and a sensor is provided, in particular is situated on the electric motor, for acquiring the angular position of a rotor of the electric motor,
a motor voltage is supplied to the electric motor with the aid of the converter, in particular with the aid of a motor control means of the converter, for achieving a torque setpoint value M_Soll,
and an angular velocity actual value ω_ist and an angular acceleration actual value α_ist are determined, in particular in a recurrent manner over time, from the values of the angular position of the rotor acquired with the aid of the sensor,
the torque setpoint value M_Soll is determined from a moment of inertia J_ges and an angular acceleration setpoint value α_Soll, which is determined as an actuation variable by an rpm control member to which the difference, in particular the system deviation, between angular velocity actual value ω_ist and angular velocity setpoint value ω_Soll, is supplied,
the moment of inertia J_ges is determined as the sum of the moment of inertia J_Mot of the drive train without a load and the moment of inertia J_Last of the load,
the moment of inertia J_Last of the load is determined from a torque actual value M_ist and from the angular acceleration actual value α_ist, in particular while taking moment of inertia J_Mot into account.
This has the advantage that a control of the rotational speed and of the torque of the load is able to be carried out with a high control quality. This is so because moment of inertia J_ges is able to be determined from the quotient of the torque actual value and the angular acceleration actual value. The determination of the torque actual value according to the related art is able to be realized based on the motor current and preferably another variable such as the motor voltage and/or the angular position of the rotor of the electric motor. The real value of the moment of inertia is therefore instantaneously determinable and can be utilized for the motor control, i.e. in particular the rpm control. The motor control is likewise able to be carried out according to the related art, in which case the parameter for the moment of inertia used there is determined according to the present invention. Thus, in a change in the drive train, e.g., a change in the moment of inertia of the load, the moment of inertia is able to be determined and the control can therefore be carried out using a current value of the moment of inertia in each case.
In one advantageous embodiment, torque actual value M_ist is determined from the motor current acquired with the aid of a current sensor and from the acquired values of angular position x of the rotor. This offers the advantage that a method known from the related art can be used for determining torque actual value M_ist.
In one advantageous embodiment, torque actual value M_ist is determined from the motor current acquired with the aid of a current sensor and from the acquired motor voltage U and/or from the acquired values of angular position x of the rotor. This offers the advantage that a method known from the related art is able to be used for determining torque actual value M_ist.
In one advantageous embodiment, moment of inertia J_Mot of the drive train without a load is predefined as a parameter. This is advantageous insofar as the moment of inertia of the rotor of the motor is able to be determined in a highly precise manner during the production of the motor, and J_Mot is therefore made known very precisely. Only the load-side moment of inertia or other moment of inertia thus has to be determined in order to obtain a reliable value for the moment of inertia.
In one advantageous embodiment, the torque setpoint value M_Soll is formed as a product of moment of inertia J_ges and angular acceleration setpoint value α_Soll. This has the advantage that the torque is easily determined.
In one advantageous embodiment, moment of inertia J_Last of the load is determined as a quotient, reduced by moment of inertia J_Mot, of torque actual value M_ist and angular acceleration actual value α_ist. This offers the advantage that a currently existing value of the moment of inertia of the load is able to be determined.
In one advantageous embodiment, at a first instant, moment of inertia J_Last of the load is determined as a first quotient, reduced by moment of inertia J_Mot, of torque actual value M_ist and angular acceleration actual value α_ist, and angular acceleration actual value α_ist lies in a first value range,
and at a second instant after the first instant, moment of inertia J_Last of the load is determined as a second quotient, reduced by moment of inertia J_Mot, of torque actual value M_ist and angular acceleration actual value α_ist, and angular acceleration actual value α_ist lies in a second value range, and an updated moment of inertia J_Last is determined by adding to the previously determined value of moment of inertia J_Last the difference between the second and the first quotient. This has the advantage that the moment of inertia is able to be adapted, i.e. updated.
In one advantageous embodiment, the first range is a value range of the angular acceleration, and the second range is another value range of the angular acceleration,
and the two value ranges do not overlap but are spaced apart from each other. This offers the advantage that the best possible signal-to-noise ratio is achievable given sufficient spacing of the value ranges, and a reliable determination of the value of the moment of inertia is therefore achievable.
In one advantageous embodiment, the respectively determined values, in particular of moment of inertia J_Last of the load, are filtered, in particular low-pass filtered. This has the advantage that the measuring noise is able to be reduced.
In one advantageous embodiment, the rpm control member is embodied as a PI controller,
it has a proportional element and an integral element, which is switched in parallel with the proportional element,
the proportionality constant of the proportional element is a function of moment of inertia J_Mot and total moment of inertia J_ges,
and the time constant of the integral element is a function of moment of inertia J_Last and the total moment of inertia J_ges. This offers the advantage that the function is able to be selected so that a drive train featuring backlash, that is to say, in particular a connection featuring play, is able to be operated with the same control behavior, in particular with the same time constant and the same transient characteristic, as the same drive train without the backlash, that is to say, in particular with a connection free of play instead of the connection featuring play. In addition, the P element reacts according to the part of the drive train that is connected to the rotor of the electric motor free of play so that a correspondingly satisfactory control behavior is achievable within the backlash.
In one alternative advantageous embodiment, the rpm control member is embodied as a PI controller,
it has a proportional element and an integral element, which is switched in parallel with the proportional element,
the proportionality constant of the proportional element is a function of moment of inertia J_Last and total moment of inertia J_ges,
and the time constant of the integral element is a function of moment of inertia J_Mot and of total moment of inertia J_ges. This offers the advantage that the function is able to be selected in such a way that a drive train featuring backlash, and thus in particular with a connection featuring play, is able to be operated with a similar control behavior as the same drive train without the backlash, i.e. in particular with a connection free of play instead of the connection exhibiting play. In addition, a control behavior that in essence is adapted to the load moment of inertia is also already activated within the backlash so that the transition from the range of the backlash to the range outside the backlash is carried out with barely a jerk. However, even outside the backlash, a low-jerk, greatly smoothed behavior is able to be achieved as a result of the small time constant.
In a further, alternative and advantageous embodiment, the rpm control member is embodied as a PI controller,
it has a proportional element and an integral element, which is switched in parallel with the proportional element,
the proportionality constant of the proportional element is a function of the product of moment of inertia J_Mot and of moment of inertia J_Last of the load,
and/or the time constant of the integral element is a function of the product of moment of inertia J_Mot and moment of inertia J_Last of the load. This is advantageous insofar as an optimized adjustment of the backlash is achievable because moment of inertia J_Mot and moment of inertia J_Last of the load are determined separately and multiplied by each other.
Important features in the drive train are that it is designed to carry out a previously mentioned method. This offers the advantage that a high-quality control of the rotation of the load is able to be carried out.
Additional advantages result from the dependent claims. The present invention is not restricted to the feature combination of the claims. One skilled in the art will discover further meaningful combination options of claims and/or individual claim features and/or features of the description and/or the figures, in particular from the posed objective and/or the objective posed by a comparison with the related art.
The present invention will now be described in greater detail with the aid of schematic illustrations.
According to the present invention,
In contrast to
As illustrated in the figures, in a method for a motor control according to the present invention, a converter supplies a motor voltage U, in particular a three-phase voltage, to an electric motor 33, which preferably is developed as a three-phase motor.
With the aid of a sensor, the angular position of a rotor shaft of electric motor 33 is acquired and forwarded as actual value x_ist to an rpm controller, which determines from the system deviation of actual value x_ist from a setpoint value x_Soll, which particularly is predefined as a parameter value, a value of an actuation variable.
As illustrated in
This actuation variable is transmitted to means 31 for determining torque setpoint value M_Soll, which determines torque setpoint value M_Soll while taking torque actual value M_ist into account.
It is forwarded to motor control means 32 as a setpoint value, and motor control means 32 supplies a motor voltage U to electric motor 33 as a function of this setpoint value.
A sensor acquires the angular position, i.e. angular position x, of the rotor of electric motor 33, rpm actual value ω_ist is determined therefrom, and the difference between rpm setpoint value ω_Soll and rpm actual value ω_ist is forwarded as a system deviation to rpm control member 30, which preferably is implemented as a linear controller such as a PI controller or a PID controller.
To determine torque setpoint value M_Soll, moment of inertia J_Last of the load driven by electric motor 33 is first determined in means 31 according to
J_Last=M_ist/(dω_ist/dt)−J_mot.
In this context, J_mot is the moment of inertia of the rotor of electric motor 33 known as a parameter of the electric motor, and dω_ist/dt is the time derivative of the angular velocity determined from the acquired angle values. In addition, torque actual value M_ist is determined by means 1 for determining the torque actual value or by means 20 for determining the torque actual value.
Based on determined moment of inertia J_Last and moment of inertia J_Mot available as a parameter, total moment of inertia J_ges, or in other words, the total mass inertia, is able to be determined by a summation J_ges=J_Mot+J_Last.
This is used for determining torque setpoint value M_Soll according to
M_Soll=J_ges*α_Soll
Thus, torque setpoint value M_Soll is determined by a simple multiplication of total moment of inertia J_ges by actuation variable α_Soll of rpm control member 30.
According to the present invention, the correct load mass inertia thus is available to the control at all times because it is able to be adjusted during the operation.
In another exemplary embodiment, if an additional torque such as a load torque, i.e. a torque introduced by the load into the drive train, or a friction torque occurs in addition to the torque for accelerating the drive, two ranges having different accelerations are utilized for determining the change ΔJ_Last of the moment of inertia, according to:
ΔJ_Last=M_ist_1/(dω_ist_1/dt)−M_ist_2/(dω_ist_2/dt)
where dω_ist_1/dt is the time derivative of the angular velocity, i.e. the angular acceleration, which lies in the first range, and M_ist_1 is the associated torque actual value,
where dω_ist_2/dt is the time derivative of the angular velocity, i.e. the angular acceleration, which lies in the second range, and M_ist_2 is the associated torque actual value.
Moment of inertia J_Mot of the rotor of the electric motor is constant here so that change ΔJ_ges of total moment of inertia J_ges corresponds to the change ΔJ_Last of moment of inertia of the load.
The first range is a value range of the angular acceleration, and the second range is another value range of the angular acceleration; the two value ranges do not overlap and are spaced apart from each other.
The determination of torque actual value M_ist_1 and of angular acceleration dω_ist_1/dt from the first range is carried out at a different time from the determination of torque actual value M_ist_2 and angular acceleration dω_ist_2/dt from the second range.
Moment of inertia J_Last is therefore able to be adapted to the change.
For example, the drive accelerates a load in a first time segment, and a first acceleration value occurs. In a subsequent time segment, the rotational speed of the motor is kept constant, and in a following, third time segment, the load is decelerated so that a second acceleration value occurs. The second acceleration value may be equal to the first acceleration value in its absolute amount.
Thus, the determination of moment of inertia J_Last of the load is able to be carried out by such a driving process, the influence of the friction being eliminated.
In this way, the moment of inertia is able to be determined, in particular adapted, during the ongoing operation of the system including the drive. Whenever the acceleration lies in the first range or in the second range, the respective quotient of the associated torque actual value and the angular acceleration is able to be updated.
Thus, even if there is a change in the load mass inertia during the operation, moment of inertia J_Last, and thus the total moment of inertia of the drive train including the load and the rotor, is able to be determined in an updating manner.
A drive train should be understood as electric motor 33 supplied by the converter together with the load driven by electric motor 33, possibly via a gear unit.
In further exemplary embodiments according to the present invention, the value characteristic of the load moment of inertia is additionally filtered in a suitable manner in order to reduce noise caused by noise of the measuring variables.
As illustrated in
As illustrated in
Kp2 is used here as the value of the proportionality constant of the proportional element, and Tn2 is used as a time constant of the integral element.
A stable control behavior is achievable in particular if the following applies:
Kp2=Kp1*J_Mot/(J_Mot+J_Last)
Tn2=Tn1*(J_Mot+J_Last)/J_Mot
Here, total moment of inertia J_ges is equal to the sum of J_Mot and J_Last, in which case moment of inertia J_Mot pertains to the particular rotatably mounted part of the drive train that describes the rotor of the electric motor of the drive train up to the connection featuring play, in particular up to the backlash, of the drive train, and J_Last describes the remaining part of the drive train.
According to the present invention, J_Last is determined, which therefore allows for an adaptation of the values.
In further exemplary embodiments according to the present invention, values according to:
Kp2=Kp1*f(J_Mot,J_ges)
Tn2=Tn1*g(J_Last,J_ges)
are used, with f and g being functions.
The separate determination of J_Last and J_Mot is therefore necessary. To this end, J_Mot is able to be determined in a first method step by determining J_Mot according to M_ist_1/(dω_ist_1/dt) when no load is present or when no load is coupled with the rotor shaft of the electric motor of the drive shaft. In a second method step, load moment of inertia J_Last is then determined according to
M_ist_1/(dωist_1/dt)−M_ist_2/(dωist_2/dt).
Later changes in the load moment of inertia J_Last are then correspondingly able to be determined in further method steps, or in other words, the value of load moment of inertia J_Last is adaptable.
In further exemplary embodiments according to the present invention, values are used according to
Kp2=Kp1*f(J_Last,J_ges)
Tn2=Tn1*g(J_Mot,J_ges)
and f and g are functions.
Thus, the separate determination of J_Last and J_Mot is required. For this purpose, J_Mot is able to be determined in a first method step in that J_Mot is determined according to M_ist_1/(dω_ist_1/dt) when no load is present or when no load is coupled with the rotor shaft of the electric motor of the drive train. In a second method step, load moment of inertia J_Last is then determined according to
M_ist_1/(dω_ist_1/dt)−M_ist_2/(dω_ist_2/dt)
Later changes in the load moment of inertia J_Last are then correspondingly able to be determined in further method steps, or in other words, the value of load moment of inertia J_Last is adaptable.
In further exemplary embodiments according to the present invention, values are used according to
Kp2=Kp1*f(J_Mot×J_Last)
Tn2=Tn1*g(J_Mot×J_Last)
and f and g are functions and therefore depend on the product of the two moments of inertia.
The separate determination of J_Last and J_Mot is therefore required. For this purpose, J_Mot is able to be determined in a first method step by determining J_Mot according to M_ist_1/(dω_ist_1/dt) while no load is present or while no load is coupled with the rotor shaft of the electric motor of the drive train. In a second method step, load moment of inertia J_Last is then determined according to
M_ist_1/(dω_ist_1/dt)−M_ist_2/(dω_ist_2/dt)
Later changes in the load moment of inertia J_Last are then correspondingly able to be determined in later method steps, or in other words, the value of load moment of inertia J_Last is adaptable.
| Number | Date | Country | Kind |
|---|---|---|---|
| 10 2018 009 518.6 | Dec 2018 | DE | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP2019/025422 | 11/28/2019 | WO | 00 |
