WO2023208945A1

WO2023208945A1 - Method for identifying and compensating for a position-measurement error

Info

Publication number: WO2023208945A1
Application number: PCT/EP2023/060830
Authority: WO
Inventors: Armin Steinhauser; Joachim Weissbacher
Original assignee: B&R Industrial Automation GmbH
Priority date: 2022-04-28
Filing date: 2023-04-25
Publication date: 2023-11-02

Abstract

The aim of the invention is to provide a method for position measurement and position control for an electromagnetic transport system (1) comprising a stator (2), a rotor (3) and a position-control circuit (200), which method is improved in that it takes into account systematic measurement errors. To achieve this aim: at least one error vibration parameter (NxK) describing the vibration behavior of a position-measurement error (nx) at a vibration-component frequency (fK) is determined; a position-measurement signal (Xact) is corrected by means of the error vibration parameter (NxK) to obtain a position-correction signal (Xkorr); and the position-correction signal (Xkorr) is fed back in the position-control circuit (200) and used to control the position (Xreal) of the rotor (3).

Description

Method for identifying and compensating a position measurement error

The present invention relates to a method for operating an electromagnetic transport device with a stator, a rotor and a position control loop, wherein the rotor is movable along the stator along at least one direction of movement at a rotor speed, and wherein on the stator along the at least one direction of movement a plurality of position sensors is arranged stationary and spaced apart from one another. The position sensors generate a position measurement signal which describes a measurement position that deviates from the true position of the rotor by a position measurement error and which has a vibration component with a vibration component frequency.

Long stator linear motors (LLM) and planar motors (PM), their applications and their functionality are well known from the prior art. LLM generally consist of a stator (also referred to as a “long stator linear motor stator” or “LLM stator”) and at least one rotor (also referred to as a “shuttle” or “transport unit”). As described, for example, in US 6,876,107 B2, an LLM stator is usually composed of a large number of stator segments, with a large number of drive coils (also “LLM coils”) being arranged stationary next to one another on the stator or on the stator segments. The stator segments can have different geometries, such as straight lines, curves, switches, and can be assembled into a desired LLM stator by lining them up. The LLM stator thus forms a conveyor path along which one or more runners can be moved. The runners are held and guided along the conveyor track.

Planar motors (PM) are also generally known in the prior art. For example, US 9,202,719 B2 discloses the basic structure and functionality of a PM. A PM essentially also has a stator, which in a PM, however, forms a transport plane in which one or more rotors can be moved at least two-dimensionally. In PM, drive coils are usually arranged in the transport level, and in some versions also in several levels.

In order to be able to bring about, regulate and/or control the movement of a rotor, drive magnets (permanent or electromagnets) and position sensors (AMR sensors, Hall elements, etc.) are arranged on a rotor in addition to the drive coils arranged on the stator. By controlling the drive coils, which can be done in particular by applying a corresponding coil voltage to generate a drive current in the drive coils, a moving magnetic field, a so-called magnetic drive field, can be generated, which interacts with the drive magnets of the rotor to move a rotor. Drive coils that are controlled, i.e. energized, for the purpose of generating a magnetic drive field are referred to here and also in the following explanations as “active” drive coils. In order to move a rotor along the stator, some of the drive coils are always active drive coils, with normally de-energized drive coils becoming active drive coils and active drive coils becoming de-energized drive coils. A runner can thus be moved in the direction of the moving magnetic drive field.

For the targeted movement of a runner, a target position into which the position of the runner is to be transferred, or a target speed for the speed of the runner or a target acceleration for the acceleration of the runner can be specified. A target position or a target speed or a target acceleration or another target value (e.g. a magnetic flux) can be specified at a constant time, for example in the form of a fixed value in a fixed value control. However, a target position or a target speed or a target acceleration or another target value can also be specified in a time-varying manner.

Of particular practical relevance here is the specification of ramp-shaped courses for the target position. If a ramp with a constant ramp gradient is specified for the course of a target position, this results in constant target speeds (usually at least approximately) in the case of a cascaded control loop structure with a superimposed position controller and a subordinate speed controller, which is usual for LLM or PM. As is well known, in a cascaded control loop structure, a higher-level controller (e.g. position controller) determines further target specifications as manipulated variables, which are to be implemented by lower-level controllers (e.g. speed controller). In such a cascaded control loop structure, a target speed that is constant over time (at least approximately) can be generated for a target position by means of a ramp-shaped course. This procedure is particularly relevant in cases where an existing position controller cannot be deactivated, but operation at a constant speed is still desired.

As mentioned, target values such as target position, target speed or target acceleration can be implemented in LLM or PM by suitable control of the drive coils arranged on the stator. In order to determine the coil voltages required for such control, as is usual in control technology, a comparison is often carried out in each control magazine between a predetermined setpoint and a measurement signal, usually referred to as an “actual value”, of the time course of the variable to be controlled. Such a comparison can be carried out by forming the difference between the setpoint and actual value, the result of which is known to be referred to as a control error. In the case of an LLM or a PM, such a control error can be fed into a controller provided for controlling the variable to be controlled, in which the coil voltages required to control the drive coils can be determined directly based on the control error. On the other hand, if there is a cascaded control loop, a higher-level controller (e.g. a position controller) can initially also specify a setpoint for a lower-level controller (e.g. speed controller). In the case of a control circuit constructed with a cascade of controllers, the coil voltages required to control drive coils are usually determined by the innermost, i.e. the last, controller in the cascade. In LLM or PM, the innermost controller of such a cascade is often a current controller.

In the manner described, a large number of rotors can also be moved independently of one another along a conveyor path formed by a stator. More detailed information can be found in WO 2013/143783 A1, WO 98/50760 A2, US 6,876,107 B2, US 2013/0074724 A1 or EP 1 270 311 B1.

Regardless of the specific implementation of a control loop, the quality and accuracy of the measurement signals used in the control loop, for example actual values of the control, sometimes have a significant influence on the operating quality that can be achieved, for example on the accuracy of a position control. If there are incorrect position measurement signals in an LLM or PM, the impression that a runner is already at a specified target position could arise due to the equality of the position measurement signal and the target position, although this is not (yet) the case. Dirt effects such as measurement noise and/or quantization can also seriously affect the control dynamics and control accuracy that can be achieved.

To control the movement of the rotor of an LLM or PM, it is therefore essential to know exactly its current position relative to the stator in order to be able to correctly control the drive coils to generate the moving magnetic field. Determining the position of the runner therefore plays a special role in the operation of the LLM or PM. When measuring position, particular attention must be paid to systematic measurement errors caused by the measuring principle of the position sensors used. When measuring position with sin/cos encoders, principle-related offset errors, phase errors and amplitude errors sometimes occur, with all of the disadvantageous consequences mentioned above. Such problems are also known in the prior art.

EP 1 195 880 A1 describes the use of additional sensors to increase the positioning accuracy. The state of the art also includes the use of complex, external linear scales or laser trackers to determine systematic position measurement errors.

US 10,914,620 B2, in contrast, describes a data-based approach in which measured signal curves are compared with stored target values. The saved target values must be known in advance, which can prove problematic in many practical situations. Possible changes (wear and aging) of a transport system during operation, which can also affect position measurement technology, are not taken into account in US 10,914,620 B2.

For the cited publications, the approaches described therein either require the use of additional components, such as additional sensors or additional measuring equipment, or that previously known data sets are required. Both are disadvantageous in practical use.

It is therefore an object of the present invention to provide an improved method for position measurement and position control for electromagnetic transport systems, particularly with regard to taking systematic measurement errors into account.

This task is solved for an electromagnetic transport system mentioned at the beginning by the features of the characteristics of the independent claims. The independent claims describe a method for operating an electromagnetic transport device and an electromagnetic transport device.

According to the invention, during operation of the electromagnetic transport device, which can be, for example, an LLM or a PM, at least one position vibration parameter is determined, which describes the vibration behavior of the at least one vibration component of the position measurement signal, and it is determined by means of a predetermined description of a relationship between the position measurement signal and the position measurement error, at least one error vibration parameter is determined from the at least one position vibration parameter, which describes the vibration behavior of the position measurement error at the vibration component frequency. Based on this, the position measurement signal is corrected using the error vibration parameter to form a position correction signal, which is fed back in the position control loop and used to regulate the position of the rotor. With this procedure, systematic position measurement errors are described in a very simple way, in the simplest case only by an error vibration parameter. This approach allows systematic position measurement errors to be determined without much effort and without any problems Integration of the error vibration parameter into the control of the electromagnetic transport device. According to the invention, the control is taken into account by correcting the position measurement signal using the error vibration parameter, whereby the control accuracy in the position control loop can be improved to a remarkable degree.

In a preferred manner, the oscillation component of the position measurement signal corresponds to a fundamental wave of the position measurement signal, the fundamental wave frequency of the fundamental wave being determined by the rotor speed and the arrangement of the position sensors. Furthermore, a position oscillation parameter preferably corresponds to a position fundamental wave parameter, which describes the oscillation behavior of the fundamental wave of the position measurement signal, and an error oscillation parameter preferably corresponds to an error fundamental wave parameter, which describes the oscillation behavior of the position measurement error at the fundamental wave frequency. This procedure is based on the knowledge that the fundamental wave of the position measurement error in particular is crucial for the systematic position measurement errors. The invention basically allows any vibration components to be taken into account and thus fundamental and harmonic waves to be taken into account. However, since the fundamental wave is often the dominant vibration component, taking the fundamental wave into account often makes a noticeable improvement in the aforementioned control accuracy in the position control loop possible.

Accordingly, in a further preferred embodiment of the invention, an oscillation component of the position measurement signal corresponds to a harmonic of the position measurement signal, the harmonic frequency of which corresponds to an integer multiple of a fundamental wave frequency, which is determined by the rotor speed and the arrangement of the position sensors, and a position oscillation parameter Position harmonic parameter that describes the oscillation behavior of the harmonic of the position measurement signal, and an error oscillation parameter an error harmonic parameter that describes the oscillation behavior of the position measurement error at the harmonic frequency. By taking additional vibration components into account, the operating behavior of the transport device can be noticeably further improved. In some cases it can be advantageous to take a harmonic into account instead of the fundamental wave mentioned at the beginning, especially if the behavior of the systematic measurement errors that occur is strongly influenced by harmonics.

In a further advantageous embodiment of the invention, a position compensation signal is determined from the at least one error vibration parameter, which is added to the position measurement signal to convert the position measurement signal to the position correction signal to correct. The position correction signal is subsequently used for improved position control. It should be mentioned that other approaches are also conceivable at this point; position compensation signals can also be determined which are subtracted from the position measurement signal, or by which the position measurement signal is divided, or by which the position measurement signal is multiplied. The invention is not restrictive at this point, but rather allows individual adaptation to a given application in order to always achieve the greatest possible improvement.

To implement the method according to the invention in an electromagnetic transport device, such as an LLM or a PM, the position correction signal in the position control loop for controlling the position of the rotor is compared with a predetermined target position, and a manipulated variable is determined in the position control loop from the result of this comparison, based on which Active drive coils of the stator are energized in order to generate a propulsive force acting on the rotor and to transfer the rotor to the specified target position.

In a further advantageous embodiment of the present invention, the stator is divided into a plurality of stator sections, with the stator sections each being assigned at least one error vibration parameter related to the respective stator section. In this way, a finely resolved consideration of the position measurement error along the stator is possible, and depending on the application, due to the great flexibility of the invention, the majority of the stator sections can be adapted in order to always achieve the best coordination of numerical complexity, which is achieved by an increase of Vibration parameters result from a possibly increased number of stator sections, and achievable improvement in control accuracy.

With a corresponding subdivision of the stator into stator sections, when correcting the position measurement signal to the position correction signal, those error vibration parameters can always be used that are assigned to that stator section of the stator over which the rotor is located at the time of correction of the position measurement signal. In this way, the position correction signal is specifically tailored to the respective conditions along the stator, which often allows a further improvement of the method according to the invention.

The at least one error vibration parameter can be identified within the scope of the invention by means of an identification journey of the rotor along the stator, but there is also the possibility of determining error vibration parameters computationally, for example from construction data of the transport device or also by specification by an operator or by another specification. However, the determination through an identification run offers the advantage that the error vibration parameter is determined specifically for the stator where it is also used for control.

Preferably, the runner is moved at a constant runner speed during the identification journey and/or the identification journey is carried out in a constant speed phase of the runner. The possibility of carrying out the aforementioned identification during constant-speed phases during operation allows data recorded during constant-speed phases, which can occur, for example, randomly during operation, to be used directly to further improve the operating quality of the transport unit.

A particularly advantageous variant of the method according to the invention results if, in addition to the position measurement signal, a target position specified for the position of the rotor is included in the determination of the at least one error vibration parameter. In the context of this variant, the position measurement signal can be subtracted from the target position in order to determine an uncorrected position control error, and the uncorrected position control error can be used to determine the at least one error vibration parameter. This approach is based on the knowledge that the information about the behavior of a position measurement error is transported by the position measurement signal, and the position measurement signal is usually part of a position control error used in the control. This is particularly the case if, as is common in control engineering, the position control error is formed by subtracting the position measurement signal from the target position. In certain cases, the position control error is easier or more available for calculations than a position measurement signal itself. The fact that the method according to the invention can also be carried out on the basis of a position control error often results in a significant expansion of the practical applicability of the invention.

Furthermore, the specified description of the relationship between the position measurement signal and the position measurement error can be given in the form of a linear and freely parameterizable transmission system. Specifically, this can be a linear and time-invariant transfer function, which is advantageous in many cases since the handling of linear and time-invariant transfer functions is well known from control engineering and a number of tools are available to handle them. It should be noted that instead of linear and time-invariant transfer functions, other forms of implementing the description of the relationship between the position measurement signal and the position Measurement errors can be used, for example tables or lookup tables or other forms of implementation.

With the help of the present invention, it is possible to determine the systematic position measurement error, which is due to the encoder principle (e.g. Hall encoder, AMR encoder), without the use of a reference system (e.g. measuring ruler, laser tracker, reference data sets, ...). and subsequently also to compensate.

The present invention also simplifies the commissioning of transport systems in the form of an LLM or PM. The elimination of an otherwise frequently required reference system in the form of a measuring ruler or a laser tracker allows the costs and complexity of the transport system to be reduced.

The invention also achieves increased positioning accuracy, both statically and dynamically. Increased positioning accuracy in turn brings with it a number of other advantages, such as smoother position progressions during operation. In the course of position control, smoother current curves result due to smoother position curves, which is usually accompanied by a lower current load. A lower current load can in turn achieve higher energy efficiency, less heating, etc. Further advantages in this context include improved smooth running, improved synchronization properties and improved acoustics.

The present invention can be used in a wide variety of transport or drive systems, especially if a magnetic field is used in these transport or drive systems both to generate force and to determine a position/position measurement values. This includes, among other things, long stator linear motors and planar motors, but also short stator linear motors.

The present invention is explained in more detail below with reference to FIGS. 1 to 9, which show advantageous embodiments of the invention by way of example, schematically and non-restrictively. This shows

Fig. 1 shows the functional principle of a long stator linear motor,

2 shows a typical magnetic field of an arrangement of drive magnets on the rotor,

3 shows a typical sensor signal of a position sensor,

4 shows a typical sensor response of a position sensor obtained from the sensor signal,

5 shows a course of an actual position of a runner and an error-prone measurement signal, 6 shows a course of a measurement error depending on the position of a runner,

7 shows a position control loop for a long stator linear motor,

8 shows an assignment of error vibration parameters to stator sections,

9 shows a comparison of measurement curves achieved with and without the method according to the invention.

In Fig. 1, a transport device 1 in the form of a long stator linear motor (LLM) is shown as an example. Although the invention is described below using the example of a long stator linear motor, the statements also apply in an analogous manner to planar motors.

In this embodiment, the LLM 1 consists of a plurality of separate stator segments Z ₁ , ..., Z _p , which are subsequently referenced using Z _m (with m≥1 as the running index), and which are assembled into a stationary stator 2 of the LLM 1 are. For this purpose, the stator segments Z ₁ , ..., Z _p can be arranged on a stationary support structure (not shown in FIG. 1). Furthermore, the stator segments Z ₁ , ... , Z _p , as also shown in FIG. 1, can be designed in different geometric shapes, for example as straight line segments or curve segments, in order to be able to realize different transport paths. The geometry of the stator 2 can basically be arbitrary and the stator 2 can be arranged in a plane, but can also be shaped three-dimensionally. In a planar motor (PM), stator segments Z ₁ , ... , Z _p form a plane.

A plurality of electrical drive coils L _m1 , ..., L _mn are arranged in a known manner along the stator 2 (shown in FIG. 1 only for the stator segment Z ₁ ), which are connected to drive magnets Y ₁ , ..., Y _L of the rotor 3 Interaction (a single drive magnet is referenced below with Y _k ). In the case of stator segments Z _m , each stator segment Z _m carries a plurality of drive coils L _m1 , ..., L _mn . In an equally known manner, using coil control units 101, 102, a drive magnetic field and thus a propulsive force F for the rotor 3 are created by regulating/setting coil voltages U _L1 ,..., _{U Ln} _that drop across the drive coils L ₁₁ , ..., L 1n _x generated in order to move the rotor 3 along the stator 2 with a rotor speed v _x along a direction of movement X.

If a stator 2 of an LLM 1 is composed of stator segments Z ₁ , ... , Z _p , which are designed as different geometric shapes, for example as straight line segments and/or as curve segments, a rotor 3 can under certain circumstances also move along other directions of movement. The movement of the rotor 3 is limited to the shape of the stator 2 in the sense of a constraint (e.g. holonomic, scleronomic, rheonometric constraints). For these cases, the following are available: Description of the position of a rotor 3 used position variables, such as x _real , x _mess , x _act , are to be understood as path parameters that assign a unique value to each point on the stator 2 and thus the position of a rotor 3 on the stator despite any curves that may be present 2 clearly describe. In the case of a PM, movement along at least one further direction of movement Y is usually possible in the transport plane.

As a rule, several so-called active drive coils L ₁₁ , ..., L _1q act on a rotor 3 at the same time, the number of which is noted below with the integer variable q. An active drive coil is a drive coil in the area of a rotor 3, or in the area of the drive magnets Y ₁ , ... , Y _L of a rotor 3, which is energized in order to generate a drive magnetic field that is connected to the drive magnets Y ₁ , .. . , Y _L of the runner 3 works together to move the runner 3.

For reasons of clarity, only two coil control units 101, 102 are shown in FIG. 1. Of course, each coil voltage U _L1 , ..., U _Ln of each drive coil L _m1 , ..., L _mn in each stator segment Z _m is regulated with a coil control unit 101, 102, whereby several coil control units 101, 102 can also be combined to form one control unit . Possible implementations of a coil control unit 101, 102 include microprocessor-based hardware, such as microcontrollers, and integrated circuits (ASIC, FPGA). Microprocessor-based hardware or integrated circuits can be clocked with a sampling time T _s .

In the embodiment of an LLM 1 shown in FIG. 1, additional runners 3 can of course be provided in addition to the runner 3 shown. By means of a transport control 100 superordinate to the coil control units 101, 102, all given runners 3 can be moved individually (speed, acceleration, path, direction) and independently (except for the avoidance of possible collisions) by other runners 3. The coil control units 101, 102 receive target variables SG1, ..., SG _n for the control from the transport controller 100. Since a single runner 3 is already sufficient for the principle of the present invention, Fig. 1 shows only one runner 3. The extension of the principle according to the invention to a large number of runners represents a purely routine activity for a person skilled in the field of LLM and therefore no difficulty.

In the embodiment shown in FIG. 1, a plurality of position sensors S _mj are arranged on the stator 2 in order to measure the position x _real of a rotor 3 on the stator 2 and to generate a corresponding position measurement signal x _act . Here, m references the stator segment Z _m and j the position sensors arranged therein. In general cases without subdividing a stator 2 into stator segments Z _m, referencing can only be done with an index j, i.e. Sj, may be sufficient. In a known manner, position measurement signals can be supplied to the transport control 100 and/or the coil control units 101, 102 in order to be used to control the movement of a rotor 3 or to control the entire LLM 1 or a PM. The specific structure of such a control, which can be implemented in a coil control unit 101, 102 or in a transport control 100 or partially in a coil control unit 101, 102 and/or partially in the transport controller 100 and which takes into account position measurement signals generated by position sensors S _mj , is discussed addressed separately at a later point.

In the area of LLM and PM, magnetic, magnetoresistive or magenostrictive position sensors S _mj (such as an anisotropic magneto-resistive sensor, a tunnel magneto-resistance sensor or giant magneto-resistance sensor) and Hall sensors have been established for position determination. Such position sensors S _mj measure a magnetic field M generated by the rotor 3, for example a magnetic field M generated by the drive magnets Y ₁ , ... , Y _L of a rotor 3 or by its own position magnets on the rotor 3, or a property of this magnetic field M, for example the Magnetic field intensity, i.e. the amount of the magnetic field M (e.g. a Hall sensor) or the direction of the magnetic field M (e.g. a magnetoresistive sensor) in the area of the respective position sensor S _mj . In addition, other types of sensors that are able to detect a magnetic field M can also be used.

With regard to the arrangement of position sensors S _mj, their distance from one another is particularly important. Position sensors S _mj are usually arranged at a constant distance from one another. However, the distance can also change, especially in curve segments. In many practically relevant cases, position sensors S _mj and drive coils L _m1 , L _mn arranged on the stator 2 are arranged alternately from one another, for example a position sensor S _mj is arranged after each of the drive coils L _m1 , ..., L _mn . In many cases, the arrangements of drive coils L _m1 , ..., L _mn and position sensors S _mj are designed depending on one another, for example by placing a position sensor S on the stator 2 for each drive coil L _m1 , ..., L _mn at an always the same distance _mj is arranged.

As is known, the distance between two drive magnets Y ₁ , ... , Y _L of a rotor 3 (from pole center to pole center) is referred to as pole pitch τ _p , the distance between two drive coils L _m1 , ... , L _mn (from pole center to pole center ) on the stator 2, in contrast, as a slot pitch τ _n . In the above-described options for arranging position sensors S _mj, the distance between two position sensors S _mj corresponds to the slot pitch τ _n . This is also assumed in the following explanations, i.e. a spacing of position sensors S _mj on the stator 2 corresponding to the slot pitch τ _n . However, this assumption does not represent a limitation of the present invention. Within the scope of the present invention, the distance between the position sensors S _mj mentioned could also be designed to be variable, i.e. change along the stator 2, and in particular be different from the slot pitch τ _n .

In Fig. 2, a magnetic field M of an arrangement of drive magnets Y ₁ , ... , Y _L on a rotor 3 is shown as an example in the form of magnetic field lines. From this magnetic field M acting on the position sensor S _mj , a position sensor S _mj first generates a sensor response in a known manner. A sensor response can, for example, include a sine trace and/or a cosine trace (in the form of an electrical voltage measurement value u), as shown in FIG. 3. As is known, from such a sensor response (sine and cosine trace), for example in the form of a voltage u, depending on the angle y of the magnetic field M (the angle at which the magnetic field M strikes the position sensor S _mj ), the position x _k of an individual can be determined in a linear approximation movable drive magnet Y _k proportional to the magnetic field angle γ of the magnetic field M can be determined in an evaluation unit, for example as x _k = r * arctan (sin (2γ) / cos (2γ)), with r as a known constant factor, e.g. r = τ _p / (2π )). In order to generate a magnetic field M for position measurement, position magnets arranged specifically for this purpose on a rotor 3 can also be provided in a known manner.

A position sensor S _mj can also directly provide a position x _k (of a drive magnet Y _k or a rotor 3) as a sensor response, or a sine and/or cosine trace, which is then evaluated, or angle information y. Likewise, the position sensor S _mj (depending on the type of sensor) can have an absolute value |A| for the magnitude of the magnetic field M (amplitude of the sine and cosine trace), for example as . The sensor response can therefore also have several sizes

contain, for example angle information y and absolute value |A|.

If a drive magnet Y _k or a position magnet is moved past a position sensor S _mj , a sawtooth-like curve of the angle information y between zero and 2π (value depending on the sensor) as well as the absolute value |A| typically results depending on the position x _k of the drive magnet Y _k relative to the position sensor S _mj , as shown in Fig. 4. Each voltage of such a sawtooth is assigned to a unique position x _k of a drive magnet Y _k relative to the position sensor S _mj . After the installation position of the position sensor S _mj and the structural and geometric structure of the rotor 3 are known, a position measurement signal x _act can be determined, which describes a measurement position x _mess . However, as will be explained below, a measuring position x _mess can deviate from the true position x _real of a runner 3. Further details about the position sensors mentioned, their measuring principle, the subsequent signal processing, or the consideration of the exact installation locations of the drive magnets Y _k on the rotor 3 when measuring the position can be found in the relevant literature, such as WO 2021/105387 A1 or EP 3 376 166 B1.

As explained at the beginning, however, in the practical implementation of LLM 1 and PM there are sensor errors when measuring the position x _real of a rotor 3. The main causes of sensor errors are manufacturing tolerances, material errors or electrical and/or non-electrical dirt effects or interference with the position sensors used S _mj .

How such sensor errors can affect the situations shown in Figures 2, 3 and 4 is shown by the dashed lines in Figures 3 and 4

and

shown. The sine and cosine traces described can deviate from an ideal sine or cosine shape, as shown in the line

shown in Fig. 3. As a direct consequence, a measurement angle signal identified from this can also deviate from a correct course of the angle signal, as shown in FIG. 4 based on

shown.

5 further shows how the errors described can affect a position measurement signal x _act that is ultimately determined from error-containing sensor responses. Specifically, FIG. 5 shows an actual position x _real of a runner 3 over time t. The course of a measuring position x _meas is also shown. The measuring position x _mess corresponds to the position of the rotor 3 described by an error-containing position measuring signal x _act . Due to the sensor errors described, the measuring position x _mess can also lie alternately above or below the true position x _real of the rotor 3.

In order to illustrate the nature of the sensor errors mentioned and the effects in more detail, Fig. 6 shows several measured courses of position measurement errors n _x depending on a true position x _real of a runner 3. Fig. 6a shows measured courses of position measurement errors n _x over a stator segment Z _m , FIG. 6b shows a section of the curves from FIG. 6a in the corresponding stator segment Z _m . The length of the entire _stator segment Z _m is denoted by L _seg,m _in FIG. It can be seen from the enlarged detail in Fig. 6b that in Fig. 6b a large number of measurements with different runners 3 were superimposed in order to validate the courses shown.

_Based on _Figures 6a _and 6b it can be seen that the actual position measurement _errors n Of this periodicity are In particular, the fundamental wave GW shown and the first harmonic OW1 are decisive. As can be seen from the curves shown, in the present case the fundamental wave GW of the position measurement error n _x is dominant.

A fundamental wave GW and harmonics OW (such as a first harmonic OW1, and/or a second harmonic OW2, etc.) are also generally referred to below as oscillation components. It should be noted that in other cases or in other embodiments, other vibration components SK can also dominate the temporal behavior of the position measurement error n _x . Other oscillation components SK can, for example, have frequencies that have a non-integer relationship to the fundamental wave frequency f ₀ of the fundamental wave. As will be shown below, the present invention also allows such other, general vibration components SK to be taken into account.

The present invention takes up these circumstances to improve position measurement and position control in an LLM 1 or a PM. The invention is based, among other things, on the knowledge that the shape of the curves shown schematically in FIG. 6 usually only changes slightly or not at all from operating point to operating point. An operating point can be defined by a specific rotor speed v _x and/or by a specific controller parameterization and/or by a specific temperature and/or by a specific electrical load and/or by a specific mechanical load.

If a rotor 3 moves with a variable rotor speed v _x along a corresponding stator 2, the frequencies of the frequency components of a position measurement error n _x , such as the frequency f ₀ of the fundamental wave GW or the frequency f ₁ of the first harmonic OW1, change or the frequency f _K of a general oscillation component SK, but its amplitudes and phase positions remain (at least approximately) unchanged. Based on this knowledge, the present invention proposes to identify the position-dependent amplitude curve of the position measurement error n _x , as shown schematically in FIG. 6, once, and then the result of the identification also for other operating points with, for example, other rotor speeds _v of a rotor 3 to correct the measured position measurement signal x _act .

In the course of the present invention, it was also recognized that the connection between a measured position measurement signal x _act and a position measurement error n _x is known (sometimes even very precisely) in many cases relevant to practice. This knowledge allows the above-described identification of the position measurement error n _x based on a measured position measurement signal x _act in a simple manner to carry out. For this purpose, the position measurement error _n _x can be calculated from the position measurement signal x act based on a (usually model and mathematical) description of the relationship between the measured position measurement signal x _act and the position measurement error n _x . It should be noted here that the position measurement error n _x can also be determined in another way for a subsequent compensation of the position measurement error n _x , which will be described in detail later. For example, additional reference measuring systems can be used for this purpose, or a measured position measurement signal x _act can be compared with previously known reference data or comparison data.

The invention makes it possible to determine and describe the described position measurement error n _x in a very general form, ie independently of influencing factors such as rotor speed v _x and/or load case and/or controller parameterization and/or temperature. Based on this general form of description, the uniquely identified position measurement error n _x can be used for compensation at any operating points (different rotor speed v _x , different load case, different controller parameterization or temperature).

The following shows how the steps described above can be carried out using the example of a specific position control loop 200 for the position control of a runner 3. 7 shows in a possible embodiment of the control a position control circuit 200 constructed in the form of a cascade from a position controller F ₁ and a speed controller F ₂ for controlling the position x _real of a rotor 3. However, the following statements are by no means intended to be viewed as restrictive. A position control loop 200 for applying the method according to the invention could also contain further regulators, such as a current or voltage regulator, or be constructed in a non-cascaded form, or have one or more feedforward control paths, or a plurality of filters for filtering those in the position control loop 200 Have signals or other components.

In the position control loop 200 shown in FIG. 7, the blocks F _1; F ₂ , F ₃ represent generic, linear and freely parameterizable transmission systems. As mentioned, the transmission system F ₁ takes on the role of the position controller, the transmission system F ₂ takes on the role of the speed controller. _The position controller F ₁ determines from a target position x _specified _for the position of the rotor ₃ a target speed _v Propulsive force f is determined as a manipulated variable. The target propulsive force f must be set on the LLM 1, in particular by energizing the active drive coils involved. The ultimately up The force _F

The transmission system G also represents a model of an LLM 1 with a rotor 3, the position x of which is _actually controlled. In block G, a single stator segment Z _m of the controlled LLM 1 is shown schematically, with a rotor 3 and with position sensors S _i-1 , S _mi , S _mi+1 .

The block F3 is used in the position control loop 200 shown to determine the rotor speed v _xact from the measured position measurement signal x _act . For this purpose, block F ₃ can be designed, for example, as a differentiator with a downstream low-pass filter. To implement the force f determined by the speed controller F ₂ , additional controllers can be present in a known manner, such as a force controller or a current controller. However, these controllers are not necessary to illustrate the principle of the invention and are therefore not shown. It is also known that power electronics (not shown) are also provided on the LLM 1 in order to generate the required coil voltages and apply them to the active drive coils involved.

The input signal d also represents an input disturbance, which can optionally be taken into account. A typical input disturbance d for LLM 1 is the so-called cogging torque (cogging effect). The position measurement error n _x under consideration is also shown as an input variable which acts as an output disturbance on the position control loop 200 and influences the position measurement signal x _act .

The following statements are based on the assumption that the elements F _1; F ₂ , F ₃ , G can be described as linear, time-invariant SISO transfer functions. However, this assumption in no way limits the present invention. Other linear and nonlinear, time-invariant and time-variant, SISO ("single-input, single-output") and MIMO ("multiple-input, multiple-output") transmission systems are also available for the elements F _1; F ₂ , F ₃ , G conceivable. Only the concrete mathematical execution would change in such cases.

For control loops such as the position control loop 200 shown, integral transformations are known to be common for analysis (for discrete-time signals, the z-transformation, for continuous signals, for example, the Laplace or Fourier transformation). Using an integral transformation for the signals given in the position control loop 200 and the usual introduction of capital letters, for example Z{x _act } = X _act for the z-transform of the position measurement signal x _act , the measured position measurement _signal are described in three parts:

The transfer function from the output disturbance n _x (position measurement error) to the output is called a sensitivity transfer function (“sensitivity” for short).

denotes that from the input disturbance d to the output , as

Input sensitivity transfer function.

At this point it should be noted that the input interference d (cogging effect) occurs at the typical frequencies f ₀ of a fundamental wave GW of a position measurement error n _x , as well as at the frequencies of the harmonics OW above (e.g. harmonic frequencies f ₁ , f ₂ , the first two harmonics OW1, OW2 etc.) of a position measurement error n _x can be neglected. If, moreover, the course _of the target position x _{should is} chosen appropriately, for example in the form of a ramp to ensure a constant rotor speed _v

_It was _recognized that in the exemplary embodiment shown in FIG certain frequencies of interest (fundamental wave frequency f ₀ or harmonic frequency f ₁ or vibration component frequency f _K ) to determine the estimated position measurement error n _x . The sensitivity transfer function Q represents a possible description Q of the relationship between the position measurement signal x _act and the position measurement error n _x . However, a description Q of the relationship between the position measurement signal x _act and the position measurement error n _x can also be given elsewhere be, such as through a table, or a Bode diagram, or a locus, or a state space model.

_In a next _step _, the transformed _position measurement signal ). _To _determine transformed signals, such as a transformed position measurement signal The _results For the purpose of identifying the position measurement error n _x according to the invention _, an oscillation parameter

N _x (ω ) = [Q(ω )] ^-1 X _act (ω ).

In this way, oscillation parameters (ω) (typically complex numbers) can be determined that describe the oscillation behavior (amplitude and phase) of the position measurement error n _x at a frequency ω.

Below, the vibration parameters described are also noted in the well-known sine-cosine form, for example using a vibration parameter for the real part a _x = Re{N _x ( ω )} and using a vibration parameter for the imaginary part b _x = lm{N _x ( ω )} of the original vibration parameter N _x (ω ).

As described, the position measurement error _n The vibration parameter belonging to the fundamental wave GW of the position measurement signal x _act is subsequently noted _as Xo _and referred to as the position fundamental wave parameter X ₀ ; vibration parameters belonging to harmonics are subsequently noted as

If vibration components SK with a vibration component frequency f _K are generally considered, the associated parameters are generally referred to as position vibration parameters X _K. In a corresponding manner, oscillation parameters of the position measurement error n _x are referred to as error fundamental wave parameters N _x0 and as error harmonic parameters N _x1 , or generally as error oscillation parameters N _xK .

If the fundamental wave frequency f ₀ and the frequency of the first harmonic f ₁ are selected for the calculation described above, i.e. ω ₀ = 2π f ₀ and ω ₁ = 2πf ₁ , where f ₁ = 2f ₀ usually applies, an error fundamental wave parameter N _x0 and an error harmonic parameter N _x1 according to the regulations

or.

be determined. If the position sensors S _mj are spaced along the stator 2 by a constant sensor distance d _s , the frequency f ₀ for evaluating the fundamental wave GW results according to the expression used for the evaluation

The angular frequency then depends on the rotor speed v _x and the

Sensor distance d _s . The frequencies of the harmonics f ₁ , f ₂ , etc. can result in a known manner as integer multiples of the fundamental wave frequency f ₀ . However, other harmonic frequencies f ₁ , f ₂ can also be selected which have no integral relationship to the fundamental wave frequency f ₀ .

If the position sensors S _mj are spaced along the stator 2 by a constant slot pitch τ _n , the frequency f ₀ for evaluating the fundamental wave GW results according to the expression. The circular frequency used for the evaluation depends

then depends on the rotor speed v _x and the slot pitch τ _n .

With these oscillation parameters N _x0 , N _x1 or a general error oscillation parameter N _xK , the behavior of the position measurement error n _x along a stator 2 can be described with great precision. Since, as described, in many cases the fundamental wave GW represents the dominant signal component of the position measurement error n _x , it is often sufficient to determine the error fundamental wave parameter N _x0 alone to describe the position measurement error n _x .

Based on the knowledge that with changing rotor speeds v _x , the frequencies of the fundamental wave GW and harmonics OW of the position measurement error _n the determination of the error fundamental wave parameter N _x0 and the error harmonic parameter N _x1 with only one rotor speed v _x a comprehensive and exact description of the position measurement error n _x can be achieved for any rotor speed v _x . The amplitude of a fundamental wave GW or a harmonic OW represents in a known manner the maximum deflection of the fundamental wave GW or harmonic OW from the position of its arithmetic mean value. The phase position of a fundamental wave GW or a harmonic OW, in contrast, represents the temporal position in a known manner Position of the fundamental wave GW or harmonic OW, for example the time shift of its maximum at the beginning of a measurement interval.

In a particularly advantageous manner, the identification of the error vibration parameters according to the invention can take place at a constant rotor speed v _x or at a constant rotor target speed v _{x, should} . The frequency f ₀ for evaluating the fundamental wave GW then results from the rotor speed v _x and the arrangement or installation positions of the position sensors S _mj . However, if a possibly changing rotor speed v _x is taken into account, changing rotor speeds _v With regard to the choice of a constant runner speed v _x for an identification run, it was recognized that the constant runner speed v _x should not be chosen smaller than v _min = f ₀ τ _n . If this minimum speed v _min is not reached, it can happen that the frequencies of interest (fundamental wave frequency f ₀ , harmonic frequencies f ₁ , f ₂ , etc.) are not sufficiently excited.

The constant rotor speed v _x should not exceed the maximum speed depending on the sampling time T _s of the intended position measuring system of the LLM 1. If this maximum speed is exceeded

v _max it can happen that not enough measurement samples can be recorded and as a result a sufficiently fine frequency resolution can no longer be ensured. In this case, it may happen that the vibration parameters belonging to the frequencies of interest (fundamental wave frequency f ₀ , harmonic frequencies f ₁ , f ₂ , etc.) can no longer be calculated.

M _min represents the minimum number of samples of the measured position measurement signal x _act that are required to identify the desired harmonics of the position measurement error n _x . Specifically, the minimum number of samples M _min required results from the desired number of harmonics n _harm to be identified as M _min = 2 (n _harm + 1), so that, for example, M _min = 2 for the sole identification of the fundamental wave. In practice, however, this theoretical maximum speed is usually far higher than realistic operation, so that there are usually always enough measured samples available.

In an advantageous manner, when calculating the position measurement error n _x, the control error or tracking error e _x can also be used. Since the course of the target position x _should in the usual cases not provide any contributions to the frequencies required for the characterization of the position measurement error n _x , and since the control error or tracking error e _x in the control system is often more easily available for calculations, the The modification described below can be used. Specifically, the expression can be used for the sensitivity Q

be applied so that the position measurement error n _x depends on the control error e _x as

can be calculated. E _x (ω ) stands for the corresponding oscillation parameters of the control error e _x . This means that the position measurement error n _x can be determined based on variables that are required and available anyway for the control.

In practical implementation, it often proves to be advantageous to determine separate or individual vibration parameters N _x for given stator sections of the stator 2. Stator sections can in particular correspond to the stator segments Z _m of the stator 2. However, stator sections can also be chosen differently and include several stator segments Z _m , or only include parts of stator segments Z _m .

The determination of individual vibration parameters N _x is often advantageous since vibration parameters N _x do not remain constant along the stator 2 in many cases. With a suitable choice of the predetermined stator sections, a compromise tailored to the respective application can be achieved between complexity, due to the large number of oscillation parameters N _x to be stored, and a sufficiently precise description of the position measurement error n _x along the stator 2.

In the case of position sensors S _mj spaced apart with a constant slot pitch τ _n , it often proves to be advantageous to divide the stator 2 or the given stator segments Z _m into stator sections of length τ _n (slot pitch). With a total length L _stat of the stator 2, stator sections result, so that for the index i the

Stator sections i ∈ [0, i _max - 1] applies.

During an identification run, at least one is then created for each stator section

Vibration parameter N _x calculated. For this purpose, a position measurement signal x _act measured and buffered in this stator section is advantageously used. Depending on how many harmonics OW can be identified in addition to the fundamental wave GW, a large number of oscillation parameters N _x can also be determined and assigned to the stator sections mentioned.

The result of such an identification run can be stored, for example, as a table, as shown in FIG. 8, which establishes an association between a stator section and the vibration parameters of the position measurement error n _x . In the embodiment shown in FIG. 8, the table is constructed using parameters a ₀ , b ₀ , etc., which correspond to the sine-cosine form described above. However, other parameters that provide the information about the position measurement error n _x transported by the vibration parameters N _x or a ₀ , b ₀ , etc. are also conceivable at this point.

In order to subsequently use the determined error vibration parameters N _x to correct an incorrect position measurement signal x _act , the determined error parameters can be used Vibration parameters N _x for each predetermined stator section of the stator 2, a compensation signal x _comp can be determined. The error oscillation parameters N _x can be read from a table shown in FIG. 8. This compensation signal x _comp can then be added to the measured position measurement signal x _act in order to generate the measured position measurement signal x _act into a corrected position correction signal x _korr . A corrected position correction signal x _korr can, as shown in FIG. 7, be fed back in the position control loop 200 and used to control the position x _real of the rotor 3.

To determine the compensation signal x _comp, a compensation unit 201 is provided in the embodiment shown in FIG. 7. A table, as shown in FIG. 8, can be stored in the compensation unit 201 in an advantageous manner. Furthermore, the measured position measurement signal x _act can also be supplied to a compensation unit 201 in order to inform the compensation unit 201 of the current stator section on the stator 2 for which the compensation is to be calculated. Possible implementations of a compensation unit 201 include microprocessor-based hardware, such as microcontrollers, and integrated circuits (ASIC, FPGA).

How the correction described above can be carried out in concrete terms is shown below as an example using the correction of the fundamental wave GW of a position measurement error n _x . As part of the compensation, the current stator section above which the rotor 3 is located is first calculated from the measured position x _act . For this purpose, the index of the stator section can be determined according to. Can build on that

The corresponding sine-cosine coefficients are read from a table, as shown in FIG. 8, and the parameters a ₀ and b ₀ for the fundamental wave GW. Using the parameters a ₀ and b ₀ , for example, according to the regulation

a compensation signal x _comp dependent on the measured position x _act can be determined. Instead of the position signal x _act , the above calculation can also be done with a relative position x _rel (e.g. relative to the slot pitch τ _n ), i.e. x _rei = mod(x _act , τ _n ). The runner speed v _x at which a runner 3 is allowed to move as part of the compensation is not restricted. Depending on the stator section, the corresponding parameters a ₀ , b ₀ ,... can be obtained depending on i _act and the above regulation can be parameterized with the related parameters a ₀ , b ₀ ,... .

Since error vibration parameters N _x or associated parameters a _x , b _x in sine-cosine form usually change only slightly or not at all from operating point to operating point change, error vibration parameters N _x can be identified at a first rotor speed v _x , but can subsequently also be used for compensation at other rotor speeds v _x during operation. For example, error vibration parameters N _x can be identified at a first rotor speed v _x1 , but can subsequently be used to correct the position measurement signal x _act at a second rotor speed v _x2 during operation.

What is crucial here is that, for example, error vibration parameters N _x or associated parameters a _x , b _x can be determined at a first identification frequency, but can also be used for compensation at frequencies that are completely different from the identification frequency.

The compensation signal x _comp can be added to the measured position x _act depending on the measured position x _act . Fig. 9 shows the effect of the compensation _described on the control error _e Fundamental wave GW is clearly visible and almost completely disappears in the compensated case.

As part of the compensation, a so-called “fading out” can also be activated, which linearly scales the activation of the compensation signal x _comp in the edge areas of the stator segments Z _m (where the segment transition in particular dominates). If the width of the edge areas is denoted by x _fade , scaled compensation can be carried out for these areas using the rule

calculated and specified.

As described at the beginning, the present invention can also be applied to PM in the same way as to LLM 1. The basic principle remains the same when used in PM, only the complexity of implementing the invention increases. Due to the six degrees of freedom for the movement of a rotor 3 that are usually given in PM, six control loops are usually used to regulate the movement of the rotor in PM operation. In this case, the identification of position measurement errors n _x according to the invention is usually extended to six control loops.

Claims

Patent claims

1. Method for operating an electromagnetic transport device (1) with a stator (2), a rotor (3) and a position control loop (200), the rotor (3) moving along the stator (2) along at least one direction of movement (X). _a rotor speed ₍ _v , whereby the position measurement signal (x _act ) describes a measurement position (x _mess ) which deviates from the true position (x _real ) of the rotor (3) by a position measurement error (n _x ), and where the position measurement signal (x _act ) has at least one vibration component (SK) with a vibration component frequency (f _K ), characterized in that at least one position vibration parameter (X _K ) is determined, which determines the vibration behavior of the at least one vibration component (SK) of the position measurement signal (x _act ) describes that at least one error vibration parameter (N xK) is determined from the at least one position vibration parameter (X _K ) by means of _a predetermined description (Q) of a relationship between the position measurement signal (x _act ) and the position measurement error (n _x ). , which describes the oscillation behavior of the position measurement error (n _x ) at the oscillation component frequency (f _K ), and that the position measurement signal (x _act ) is corrected to a position correction signal (x _corr ) using the error oscillation parameter (N _{x K} ), which is fed back in the position control loop (200) and used to regulate the position (x _real ) of the rotor (3).

2. The method according to claim 1, characterized in that an oscillation component (SK) of the position measurement signal (x _act ) corresponds to a fundamental wave (GW) of the position measurement signal (x _act ), the fundamental wave frequency (f ₀ ) of which is determined by the rotor speed (v _x ) and the arrangement of the position sensors (S _j ) is determined so that a position oscillation parameter (X _K ) corresponds to a position fundamental wave parameter (X ₀ ), which describes the oscillation behavior of the fundamental wave (GW) of the position measurement signal (x _act ), and that an error oscillation parameter (N _xK ) corresponds to an error fundamental wave parameter (N _x0 ), which describes the oscillation behavior of the position measurement error (n _x ) at the fundamental wave frequency (f ₀ ).

3. The method according to any one of claims 1 or 2, characterized in that a vibration component (SK) of the position measurement signal (x _act ) corresponds to a harmonic (OW) of the position measurement signal (x _act ), the harmonic frequency (f ₁ ) of which is an integer multiple a fundamental wave frequency (f ₀ ), which is determined by the rotor speed (v _x ) and the arrangement of the position sensors (S _j ), corresponds to that a position Oscillation parameter (X _K ) corresponds to a position harmonic parameter (X ₁ ), which describes the oscillation behavior of the harmonic (OW) of the position measurement signal (x _act ), and that an error oscillation parameter (N _xK ) corresponds to an error harmonic parameter (N _x1 ). , which describes the oscillation behavior of the position measurement error (n _x ) at the harmonic frequency (f ₁ ).

4. The method according to any one of the preceding claims, characterized in that a position compensation signal (x _comp ) is determined from the at least one error vibration parameter (N _xK ), which is added to the position measurement signal (x _act ) to produce the position measurement signal (x _act ) to correct the position correction signal (x _corr ).

5. Method according to one of the preceding claims, characterized in that the position correction signal (x _korr ) in the position control loop (200) for controlling the position (x _real ) of the rotor (3) is compared with a predetermined target position (x _target ), and that in the position control loop (200) a manipulated variable (f) is determined from the result of this comparison, based on which active drive coils (L _m1 ,...,L _mq ) of the stator (2) are energized in order to act on the rotor (3 ) acting propulsive force (F _x ) and to transfer the rotor (3) to the specified target position (x _should ).

6. The method according to any one of the preceding claims, characterized in that the stator (2) is divided into a plurality of stator sections, the stator sections each having at least one error vibration parameter (N _{x K} ) related to the respective stator section. is assigned.

7. The method according to claim 6, characterized in that when correcting the position measurement signal (x _act ) to the position correction signal (x _korr ), those error vibration parameters (N _xK ) are used which are assigned to that stator section of the stator (2), above which the rotor (3) is located at the time the position measurement signal (x _act ) is corrected.

8. Method according to one of the preceding claims, characterized in that the at least one error oscillation parameter (N _xK ) is identified by means of an identification journey of the rotor (3) along the stator (2).

9. The method according to claim 8, characterized in that the runner (3) is moved during the identification run at a constant runner speed (v _x ) and / or the identification run is carried out in a constant speed phase of the runner (3). .

10. The method according to claim 8 or 9, characterized in that an identification journey is carried out several times during the operation of the electromagnetic transport device (1).

11. Method according to one of the preceding claims, characterized in that when determining the at least one error vibration parameter (N _xK ), in addition to the position measurement signal (x _act ), a target position (x _target ) specified for the position of the rotor (3) is taken into account .

12. The method according to claim 11, characterized in that the position measurement signal (x _act ) is subtracted from the target position (x _target ) in order to determine an uncorrected position control error (e _x ), and that the uncorrected position control error (e _x ) is used is used to determine the at least one error vibration parameter (N _xK ).

13. The method according to any one of the preceding claims, characterized in that the predetermined description (Q) of a relationship between the position measurement signal (x _act ) and the position measurement error (n _x ) is given in the form of a linear transmission system.

14. The method according to claim 14, characterized in that the predetermined description (Q) corresponds to a linear and time-invariant transfer function.

15. Electromagnetic transport device in the form of a long stator linear motor (1) or planar motor, comprising a stator (2) on which a plurality of electrical drive coils (L _m1 ,...,L _mn ) each with a coil control unit (101, 102,... ., 10n) is arranged, with at least one rotor (3) movable along the stator (2), on which a plurality of excitation magnets (Y ₁ , ... , Y _L ) is arranged, and with a supply unit, which is designed , into a number q ≤ k of active drive coils (L _m1 ,... ,L _mq ), which are involved in the movement of the at least one rotor (3), each an electrical coil current (i _Lm1 ,... ,i _Lmn ) to generate a magnetic drive field, which interacts with the drive magnets (Y _n1 ,...,Y _nk ) of the at least one rotor (3) to move the rotor (3), and with a transport control (100), in which has a position control loop (200) for controlling the position (x _real ) of the at least one rotor (3), and with a plurality of position sensors (S _j ) arranged in a fixed position and spaced apart from one another, which are designed to generate a position measurement signal (x _act ), whereby the position measurement signal (x _act ) describes a measurement position (x _mess ) which deviates from the true position (x _real ) of the rotor (3) by a position measurement error (n _x ), and where the position measurement signal (x _act ) has at least one vibration component (SK) with a vibration component frequency (f _K ), characterized in that the transport control (100) is designed to determine at least one position vibration parameter (X _K ) which determines the vibration behavior of the at least a vibration component (SK) of the position measurement signal (x _act ) describes that the transport control (100) is designed by means of a predetermined Description (Q) of a relationship between the position measurement signal (x _act ) and the position measurement error (n _x ) from the at least one position oscillation parameter (X _K ) to determine at least one error oscillation parameter (N _{x K} ), which determines the oscillation behavior of the position measurement error (n _x ) at the vibration component frequency (f _K ), and that a compensation unit (201) is provided, _which is designed to convert the position measurement signal (x _act ) into a position correction signal ( x _corr ) and to feed back in the position control loop (200) in order to use the position correction signal (x _corr ) in the position control loop (200) to regulate the position (x _real ) of the rotor (3).