Method for operating a linear motor

Abstract

For adjusting the movement of the transmission unit of the linear motor, a mass term JT is specified as a function of the adjustment variable (SG) of the active drive coil (4). k The quality function J(SG) is used, and the quality function J(SG) used to regulate the movement of the transmission unit (3) along the stator (2) is optimized with respect to the adjustment variable (SG) to determine the optimal adjustment variable (SG) for the corresponding time step of the movement regulation. opt ), and based on the determined optimal adjustment variable (SG) opt The active drive coil (9) is energized, and at least two movement stages are provided during the movement of the transmission unit (3) along the stator (2), wherein different mass functions J(SG) are used in the at least two movement stages to determine the optimal adjustment variable (SG). opt ), where different mass functions J(SG) are expressed through the mass term JT used. k The number k of (SG) and / or the mass term JT k (SG) and / or weighting factor k k To distinguish them.

Description

[0001] The present invention relates to a method for operating a linear motor having a stator on which a plurality of drive coils are arranged, and a transmission unit having a plurality of drive magnets arranged thereon that move along the stator, wherein an electromagnetic field is generated by energizing the drive coils in the region of the drive magnets of the transmission unit to interact with the drive magnets of the transmission unit to move the transmission unit.

[0002] A linear motor comprises a primary component and a secondary component (rotor) movably arranged relative to the primary component. A drive coil is arranged on the primary component, and a drive magnet is arranged on the secondary component, and vice versa. The drive magnet can be a permanent magnet, an electric coil, or a short-circuit winding. The drive coil is an electric coil energized to generate an electromagnetic field. Through the interaction of the (electro)magnetic fields of the drive magnet and the drive coil, a force acts on the secondary component, causing it to move relative to the primary component. For example, a linear motor can be implemented as a synchronous or asynchronous machine. A linear motor can also be implemented as a planar linear motor (also called a plane motor). The drive coil of a linear motor is arranged along one direction of movement, or in the case of a planar motor, along two directions of movement, i.e., within a plane of movement. The secondary component can move along one direction of movement or move freely within a plane of movement in two directions. Short-stator linear motors and long-stator linear motors can also be distinguished, wherein in a long-stator linear motor, the secondary component is shorter than the primary component, and in a short-stator linear motor, the primary component is shorter than the secondary component. In the case of a planar motor, the primary component is typically larger than the secondary component.

[0003] Linear motors are used in electromagnetic transmission systems, where the transmission unit moves to perform the transmission task. The transmission unit can be implemented as a secondary or primary component. Typically, such electromagnetic transmission systems are implemented in the form of long-stator linear motors or planar motors.

[0004] In the case of a long-stator linear motor, multiple electric drive coils are arranged side-by-side fixedly along the stator. Multiple drive magnets, acting as permanent magnets or as electric coils or short-circuited windings, are arranged on the transmission unit. These drive magnets are separated from the drive coils by an air gap and interact with them. Through the interaction of the (electro)magnetic fields of the drive magnets and the drive coils, a force is applied to the transmission unit, causing it to move along the stator. In the case of a planar motor, the drive coils on the stator are arranged in a plane. Similarly, drive magnets arranged in the plane are present on the transmission unit.

[0005] Typically, a certain number of drive coils are arranged on each stator segment. These stator segments may also have different geometries, such as straight lines, curves, turnouts, etc. Subsequently, the stator segments can be assembled into the desired stator by connecting them in series. However, for the linear motor according to the present invention, the use of such stator segments and the number of such stator segments are irrelevant, as are the number and arrangement of the drive coils on the stator segments.

[0006] By energizing the drive coils in the region of the drive magnet of the transmission unit, a magnetic field interacting with the magnetic field of the drive magnet can be generated, producing a propulsive force (in the direction of movement along the stator or, in the case of a planar motor, in the direction of movement along the plane of movement of the stator) and / or a normal force (in the direction transverse to the direction of movement) acting on the transmission unit. The generated force can be influenced by adjusting the generated magnetic flux through controlling the individual drive coils. Thus, by controlling the drive coils to generate a magnetic field moving in the direction of movement, the transmission unit can move along the transmission path (which includes the transmission plane in the case of a planar motor) in a desired manner. Alternatively, multiple transmission units can be arranged along the transmission path, and their movement can be controlled individually and independently by typically applying a voltage to energize the respective drive coils interacting with a transmission unit.

[0007] Examples of long-stator linear motors can be found in WO 2013 / 143783 A1, US 6,876,107 B2, US 2013 / 0074724 A1, or WO 2004 / 103792 A1. For example, US 9,202,719 B2 reveals the basic structure and function of a planar motor.

[0008] To regulate the movement of the transmission unit, a dq coordinate system that moves along with the transmission unit is typically used, similar to that used in rotary electric motors. In the case of planar electric motors, this is also separated into two directions of movement. The drive current required for the desired movement is then calculated in the dq coordinate system, which has a current component in the direction of movement (often referred to as the q component) and possibly a current component in the normal direction (i.e., transverse to the direction of movement, often referred to as the d component). Here, the q component is responsible for generating the propulsive force, while the d component is responsible for generating the normal force transverse to the propulsive force. However, for normal movement along the stator, an active normal force is usually not required (i.e., by energizing the drive coils). Subsequently, similar to rotary electric motors, the drive current in the dq coordinate system is converted using an inverse Park transformation into the coil current of the drive coils involved in force generation, which is then generated by power electronics by applying a corresponding voltage to the drive coils. This is repeated for a predetermined time step (typically in the range of 1 / 10 millisecond). Various drive coils are also involved during the movement of the transmission unit. This regulation applies whether forces are generated on both sides of the transmission unit (as seen in the direction of movement of the transmission unit) or only on one side.

[0009] For example, in Khong, PC et al., "Magnetic Guidance of the Mover in a Long- "Magnetic Guidance of the Motor in a Primary Linear Motor", IEEE Transactions on Industrial Applications, Vol. 47, Issue 3, May / June 2011, pp. 1319-1327 This adjustment is described in the dq model based on a long stator linear motor. In Khong, a long stator linear motor with drive coils arranged on both sides visible in the direction of movement is described, and the normal force on both sides is used to center the transmission unit so as to guide it in the middle.

[0010] EP 3 385 110 A1 describes a long stator linear motor in which the thrust and normal forces are adjusted independently of each other so that the normal force can adapt to the conditions of the transmission path. In EP 3 109 998 B1, the normal force or torque is used to control the movement of the transmission unit in a turnout.

[0011] To regulate the force and / or torque acting on the transmission unit, regulators are used, typically cascaded regulators consisting of multiple consecutive regulators, to compensate for deviations between a pre-given setpoint variable (e.g., set position or set speed) and the actual variable (e.g., actual position or actual speed). Often, multiple regulators are also present along the stator, such as one regulator per stator segment (as described in EP 3 422 558 A1). For this purpose, the regulator determines the regulated variable (e.g., the voltage applied to the drive coil) according to the implemented regulation rule (typically a PI or PID regulator), which is then implemented on the linear motor. For example, the regulated variable is implemented by power electronics for the drive coil; that is, it is generated and applied to the drive coil. Conventional regulators contain regulator parameters that must be adapted to the corresponding regulation path, in this case, the stator with the drive coil and the transmission unit with the drive magnet and its magnetic coupling, in order to adjust the desired regulation behavior, such as regulation error, overshoot, stability, response, etc. Therefore, the regulator must be parameterized, which means that appropriate values ​​must be assigned to the regulator parameters. These regulator parameters are usually fixed and do not change during operation.

[0012] EP 3 251 986 A1 describes the parameterization of regulator parameters in a long stator linear motor. Here, it can also be specified that different sets of regulator parameters are used for different transmission units or for different locations along the transmission path, so as to optimally adapt the regulation behavior to different transmission units, for example, due to different loads, wear conditions, designs, etc. For parameterization, the regulation loop is excited using an excitation signal superimposed on the regulation variable, and the response of the regulation loop to the excitation is evaluated in the form of frequency response, thereby deriving the regulator parameters. On the one hand, this is complex; on the other hand, it can only be performed at specific locations along the transmission path and for specific transmission units. While this method can improve regulation, it still lacks flexibility.

[0013] The objective of this invention is to improve the adjustment of the movement of the transmission unit of a linear motor, and in particular to make it more flexible.

[0014] To regulate the movement as precisely as possible, the quality function includes a quality term that evaluates the deviation between the pre-given setpoint variable and the actual variable of the regulation, which depends on the regulation variable. This allows for high regulation quality in the movement of the transmission unit. By using different quality functions in different movement stages of the transmission unit, the regulation can be flexibly adapted to the corresponding movement, and different regulation objectives can be achieved in each movement stage. The movement stages can be flexibly configured along the stator, and the quality functions in the movement stages can also be flexibly configured. This does not change the basic flow of regulating the movement of the transmission unit, because only the different quality functions are optimized.

[0015] Optimization-based adjustment of the movement of the transmission unit allows for both online and offline optimization of the quality function. This further increases the flexibility of movement adjustment.

[0016] In offline optimization, it is particularly advantageous to pre-create movement characteristic curves for various quality functions, mapping movement setpoint variables, especially the driving force, and the position of the transmission unit relative to the active drive coil, to optimal adjustment variables. Therefore, the optimal adjustment variable for regulating the movement of the transmission unit must be read online only from the correct characteristic curves.

[0017] Preferably, the quality function includes a quality term that evaluates the electrical power required for the movement of the transmission unit. This makes it possible to include a movement phase in which the transmission unit moves with the least possible electrical power loss. This allows for energy-efficient operation of the linear motor.

[0018] If the mass function includes a mass term that evaluates the sum of the control variables, the heat load on the components of the power electronic device used to implement the control variables and the heating of the stator can be reduced.

[0019] Advantageously, a mathematical model of the linear motor can be used to determine the variables of the mass term or the constraints that the model can be used for optimization. This model allows for the calculation of specific variables or parameters, rather than measuring them within the linear motor. It also allows the determination of the linear motor's state variables and all necessary output variables. When constraints are used, the physical laws of the linear motor are ensured to be considered, which can generally improve the regulating quality.

[0020] A particular advantage is the use of magnetoresistive networks (i.e., system equations derived from magnetoresistive networks that reflect physical laws) to mathematically model linear motors. Using magnetoresistive models allows for higher system quality (accuracy) relative to model complexity.

[0021] The following will refer to Figures 1 to 9 To explain the invention in more detail, Figures 1 to 4 Advantageous design features of the invention are illustrated, by way of example and not limitation. The accompanying drawings show:

[0022] Figure 1 and Figure 2 A possible structure for a long-stator linear motor is shown, which has a stator with teeth containing drive coils and a transmission unit with drive magnets.

[0023] Figure 3 and 4 The stator's magnetoresistive network is shown.

[0024] Figure 5 The rotor's reluctance network and its coupling with the stator are shown.

[0025] Figure 6 This demonstrates how to adjust the movement of the transmission unit by optimizing the quality function.

[0026] Figure 7 and Figure 8 Possible designs for adjusting the movement of the transmission unit are shown, and

[0027] Figure 9 An example of a motion characteristic curve is shown.

[0028] The present invention has been described using a long stator linear motor as an example of a linear motor embodiment, but it can also be applied in a similar manner to other embodiments of linear motors, such as planar motors or short stator linear motors.

[0029] exist Figure 1 The diagram shows a simplified example section of a long-stator linear motor 1. The long-stator linear motor 1 consists of a stator 2 and at least one transmission unit 3, which is movable along the stator 2 in the direction of movement x. In the illustrated embodiment, the stator 2 is the primary component of the linear motor, while the transmission unit 3 is a secondary component. Typically, multiple transmission units 3 (which are also different transmission units) move simultaneously and independently along the stator 2. The structural design and geometry of the stator 2 and the transmission unit 3 are not important to the present invention, but they will certainly affect the structure of the linear motor 1 model described below. How the transmission unit 3 is guided and held along the stator 2 is also not particularly important to the present invention.

[0030] In the direction of movement x, multiple drive coils 4 are arranged side-by-side along the stator 2. The drive coils 4 are typically arranged on stator teeth 5 made of a high-permeability material (e.g., iron), which are connected to each other by a stator yoke 6. A secondary tooth 7 without a drive coil 4 may also be provided between two stator teeth 5 with drive coils 4. However, the secondary tooth 7 may also be removed. A drive magnet assembly 8 with multiple drive magnets 9 is arranged on the transmission unit 3. The drive magnets 9 arranged side-by-side in the direction of movement x of the drive magnet assembly 8 are typically of opposite polarities. The drive magnet assembly 8 faces the drive coils 4 and is separated from the drive coils 4 by an air gap 14. The maintenance of the air gap 14 is typically ensured by mechanical and / or magnetic guidance (not shown) on the stator 2 of the transmission unit 3.

[0031] As previously mentioned, in the case of a planar motor, the drive coil 4 is arranged in the moving plane of the stator 2. Similarly, the drive magnet 9 is arranged on a plane. In the case of a planar motor, the transmission unit 3 typically has no mechanical guidance, but it is usually magnetically levitated.

[0032] Multiple drive coils 4 in the region of the drive magnet assembly 8 of the transmission unit 3 interact with the drive magnet 9 of the drive magnet assembly 8 to move the transmission unit 3. The number of drive coils 4 used for the movement of the transmission unit 3 can be assumed to be known and can be determined, but it can also depend on the position of the transmission unit 3 relative to the stator 2 and can also be determined by the optimization described below. Therefore, the number of drive coils 4 can also be known as a result of optimization, which determines how many drive coils 4 should be energized and how in the environment of the transmission unit 3. These energized drive coils 4 (hereinafter also referred to as active drive coils) are adjusted by the adjustment unit 10. During the movement of the transmission unit 3 relative to the stator 2, which drive coils among the drive coils 4 are active drive coils naturally changes.

[0033] However, it should be noted that not all drive coils 4 in the region of the drive magnet assembly 8 are necessarily active drive coils. It is conceivable that only some drive coils 4 in the region of the drive magnet assembly 8 may be used as active drive coils. For example, a drive coil 4 may have failed and therefore cannot be used for the movement of the transmission unit 3. Nevertheless, the transmission unit 3 can be moved using other drive coils 4 in the region of the drive magnet assembly 8 that are active drive coils.

[0034] To prevent the long stator linear motor 1 from malfunctioning when drive coil 4 fails, the adjustment unit 10 can be configured to simply ignore the faulty drive coil 4. However, this may cause some interference to the movement of the transmission unit 3, but in some cases, such as in moving sections where there are no precision requirements for the movement of the transmission unit 3, this interference can be ignored. However, the adjustment unit 10 can also compensate for the fault of drive coil 4 with the remaining active drive coils to avoid malfunction of the long stator linear motor 1. For this purpose, it is only necessary to notify the adjustment unit 10 which drive coil 4 has failed. This can also be easily implemented within the framework of the present invention, as further implemented below.

[0035] To determine if there is a fault in drive coil 4, a coil current or coil voltage can be applied to it. Subsequently, by measuring the coil voltage or coil current, it can be determined whether drive coil 4 is short-circuited or unloaded, which is equivalent to a fault occurring because the drive coil is no longer inoperable. This can be done continuously during the operation of the long stator linear motor 1, or during the adjustment of the movement of the transmission unit 3.

[0036] The drive coil 4 can also be arranged on both sides of the transmission unit 3 when viewed along the movement direction x. In this case, the transmission unit 3 can also have drive magnet assemblies 8 on both sides when viewed along the movement direction x. Therefore, in such a structure, a force acting on the transmission unit 3 can be generated on both sides of the transmission unit 3. However, this does not change the adjustment according to the invention described below.

[0037] The adjustment unit 10 uses, for example, the device control unit 11 (e.g., via such as...) Figure 1 The data communication bus shown has a pre-defined movement setting variable BS. The device control unit 11, for example, controls and monitors the simultaneous movement of multiple transmission units 3 on the stator 2. In each time step of the adjustment, typically within a 1 / 10 millisecond range, the adjustment unit 10 determines, based on the movement setting variable BS, an adjustment variable SG for the multiple active drive coils 4, or each active drive coil 4, involved in the movement of the transmission units 3, using adjustments implemented thereon (e.g., in the form of adjustment software running on processor-based hardware). However, an adjustment unit 10 may also be provided for each drive coil 4. The adjustment variables SG for the active drive coils 4 are typically not equal.

[0038] The adjustment variable SG can be the coil current i used to energize the drive coil 4. C Or coil voltage v C Coil current i C Or coil voltage v C This can be generated by, for example, a power electronic device (not shown) described in EP 3 249 803 A1, for each involved drive coil 4, and can be applied to the drive coil 4. Typically, the adjustment variable SG is the coil voltage v applied to the active drive coil 4. C .

[0039] In the coil current i C In the case where the regulating variable SG is used, as further implemented as follows, a voltage v can be set to the coil voltage. C The conversion, or the setting of the adjusting coil current i C The current regulator. This can be implemented in the regulation unit 10 or in the power electronic device.

[0040] A movement setting variable BS can be pre-defined for each time step of the adjustment, such as a set position, speed, or force of the transmission unit 3 along the stator 2. However, the movement setting variable BS can also be determined within the adjustment unit 10 itself (e.g., based on a predetermined movement along the stator 2 (e.g., as the position or speed of the transmission unit 3 in time or along a path)). The predetermined movement can be pre-defined to the adjustment unit 10, but it can also be calculated within the adjustment unit 10, for example, to reach a predetermined target position along the stator 2.

[0041] To adjust the movement of the transmission unit 3, it can also be specified, for example, that a known position sensor 13 is arranged along the stator 2 ( Figure 1 (as shown in the figure) to determine the actual variable (e.g., actual position) of the movement of the transmission unit 3 along the stator 2.

[0042] Typically, at each time step of the adjustment, the driving force and / or driving torque that must be applied to the transmission unit 3 is determined from the movement setting variable BS in the upstream movement regulator for the transmission unit 3, thereby enabling the transmission unit 3 to perform the desired and predetermined movement. The movement regulator can be implemented in the adjustment unit 10, or it can be a separate unit. The adjustment variable SG for the active drive coil 4 is then determined from the driving force and / or driving torque to be applied. Therefore, the active drive coil 4, used to perform the movement, is energized in a manner that generates the driving force and / or driving torque required to produce the predetermined movement.

[0043] Here, the adjustment unit 10 can also consider any faulty drive coil 4 and determine the adjustment variable SG for the active drive coil 4, so that the faulty drive coil 4 is compensated in the region of the drive magnet assembly 8 to implement the desired and predetermined movement. Thus, the electromagnetic field required for the movement of the transmission unit 3 is generated using the active drive coil 4 instead of the faulty coil.

[0044] This implements a method for operating a linear motor 1 having a stator 2 on which a plurality of drive coils 4 are arranged, and a transmission unit 3 moving along the stator 2 on which a plurality of drive magnets 9 are arranged, wherein an electromagnetic field is generated by energizing the active drive coils 4 in the region of the drive magnets 9 of the transmission unit 3 to interact with the drive magnets 9 of the transmission unit 3 to move the transmission unit 3. An adjustment unit 10 determines an adjustment variable SG for the active drive coils 4, preferably a coil current i. c Or coil voltage v c The active drive coil 4 is energized according to the determined adjustment variable SG to adjust the movement of the transmission unit 3. This can be done within a predetermined time step for adjusting the movement. The adjustment unit 10 has knowledge of the faulty drive coil 4, or detects the faulty drive coil 4 itself in the region of the drive magnet assembly 8. The action of the faulty drive coil 4 is compensated by the remaining active drive coil 4 in order to move the transmission unit 3. To this end, the adjustment unit 10 generates an adjustment variable SG for the remaining active drive coil 4, thereby compensating for the faulty drive coil 4 in order to move the transmission unit 3. This can be achieved through optimization as described below and is claimed in the patent claims.

[0045] The regulating unit 10 and / or the movement regulator can be microprocessor-based hardware on which regulating software is implemented. However, the regulating unit 10 and / or the movement regulator can also be implemented as computer software installed and implemented on available computer hardware. Furthermore, the regulating unit 10 and / or the movement regulator can be implemented as an integrated circuit, such as an application-specific integrated circuit (ASIC) or a field-programmable gate array (FPGA), on which a microprocessor can also be implemented. Alternatively, the regulating unit 10 can be implemented as an analog circuit, such as an analog computer. Combinations of these are also possible.

[0046] Therefore, in order to adjust the movement of the transmission unit 3, an adjustment variable SG, i.e., the coil current i, is determined for each active drive coil 4 in each time step of the adjustment. c Or preferably, the coil voltage v c This allows the active drive coil 4 to be energized to generate the magnetic flux required for movement.

[0047] To determine the adjustment variable SG of the active drive coil 4 (e.g., coil current i) c Or coil voltage v c The mass function J is used, which is a function of at least one control variable SG to be determined for the active drive coil 4, i.e., J = f(SG), where SG are vectors of control variables to be determined for each of the involved active drive coils 4. Specifically, the mass function J is a vector of k (k ≥ 1) weighted by k... k Weighted quality term JT k The sum of, i.e. At least one, preferably each mass item JT k It is a function of the adjustment variable SG of the active drive coil 4. A quadratic function is preferred because it has a global minimum and is therefore well-suited for optimization. However, the mass function J and therefore the mass term JT... k It can also depend on other variables of the linear motor 1, such as the position of the transmission unit 3 relative to the stator 2, or on magnetic variables (such as magnetic flux) in the stator 2.

[0048] If the regulating unit 10 knows whether the drive coil 4 has failed and which drive coil 4 has failed, this can be taken into account in the mass function J in the vector of the regulating variable SG. This will also automatically compensate for the failure of the drive coil 4.

[0049] The regulated mass function J(SG) is optimized (maximized or minimized) (usually minimized) to determine the regulation variable SG for the active drive coil 4 involved, which is generated in the corresponding time step to energize the drive coil 4. Optimization means finding those regulation variables SG that minimize or maximize the mass function J(SG). These regulation variables SG are also called optimal regulation variables SG. opt .

[0050] Optimization can be performed online, i.e., at each time step of the adjustment, or offline. In the case of offline optimization, for a pre-given movement scenario of the transmission unit 3 (e.g., the temporal trajectory of the transmission unit's position or velocity in time or path), the adjustment variable SG to be set during the execution of the movement scenario can be pre-calculated. To this end, the movement scenario can be discretized in certain time steps, and optimization can be performed in each of these time steps. Interpolation can also be performed between time steps. Thus, the adjustment variable SG exists at every necessary moment of the movement of the transmission unit 3, especially in each time step of the adjustment. In the transmission system, since the movement of the transmission unit 3 is usually pre-planned, the movement scenario is known. If the movement scenario is unknown or only partially known, the actual movement of the transmission unit 3 can be delayed so that sufficient information about the movement is known again.

[0051] In the case of offline optimization, a motion characteristic curve can also be generated. When the transmission unit 3 moves along the stator 2 at a specific speed corresponding to a specific driving force acting on the transmission unit 3 in the moving direction, the driving magnet 8 of the transmission unit 3 moves past the driving coil 4. However, this repeats the coil voltage v to be applied to the active driving coil 4. C or coil current i C This enables offline creation of coil voltage v for specific motion scenarios in the form of a combination of position and driving force. C or coil current i C The movement characteristic curve. Subsequently, during the movement of transmission unit 3, the required coil voltage v must be read out depending solely on the currently required driving force and current position shown in the movement characteristic curve. C and / or coil current i C Such movement characteristic curves can also be created for different mass functions J(SG).

[0052] In mathematics, optimization can often be written as... (As mentioned earlier, the cost function J can also depend on variables other than the adjustment variable SG).

[0053] Optimization can also be performed considering constraint g. Constraint g describes the physical laws of the linear motor 1. These laws are expressed in the form of a mathematical model of the linear motor 1, preferably as a function of the control variable SG, and, if necessary, as a function of other variables of the linear motor 1 (such as state variables). This mathematical model can be a description of the fundamental physical laws of the linear motor 1 (e.g., coil current i). c Or coil voltage u c A model that models the relationship between magnetic variables (such as magnetic flux). However, the mathematical model can also be a trained model, such as a local model network or a neural network. Typically, this model is used to map the response of the linear motor 1 to input variables (usually vectors) as output variables or state variables (usually vectors). The input variables are typically the regulation variable SG, for example, the coil current i. c Or coil voltage u c The output variable can be, for example, the driving force (propulsive force and / or normal force) or driving torque acting on the transmission unit 3, but it can also be the movement variable of the transmission unit 3, such as the position along the stator 2, the velocity in the movement direction x and / or normal y, or the acceleration in the movement direction x and / or normal y. In the case of a planar motor, the driving force can act on the movement plane, i.e., in the xz plane. The output variable can also be the total current (such as q or d current) of the involved drive coil 4. According to the fundamental laws of kinematics, the driving force / driving torque and the movement variable are equivalent, or in other words, the resulting movement variable is the result of the force / torque. The driving force / driving torque corresponds to the total current. The state variable describing the state of the long stator linear motor 1 is, for example, the magnetic variable (e.g., magnetic flux, magnetomotive force, or voltage) generated by the acting input variable in the long stator linear motor 1.

[0054] The faulty drive coil 4 can also be expressed as an additional constraint g (e.g., in the mathematical expression "the drive coil numbered xy has failed"). For example, this can be achieved by a vector for all drive coils 4 containing a binary value for each drive coil 4, where, for example, "0" can represent a fault and "1" can represent no fault.

[0055] In mathematics, optimization considering constraint g can usually be written in the following form (where, as mentioned earlier, the cost function J and / or constraint g may also depend on variables other than the adjustment variable SG).

[0056]

[0057] To solve such optimization problems, numerous known solution algorithms exist, to name just a few: such as the gradient method, Newton's method, evolutionary methods, or successive quadratic programming. The choice of solution algorithm is irrelevant to this invention, where, of course, the method advantageous in terms of computational cost and time is chosen (especially in the case of online optimization). A common feature of these solution methods is the search for (usually iterative) possible solutions to the optimization problem until a defined termination criterion is reached. For example, the termination criterion could be the number of iterations, or the difference between the solutions of two successive iterations of the optimization problem falling below a limit, or another termination criterion. The selection of the solution (i.e., the adjustment variable SG) is performed in each iteration step according to the rules of the solution method, where a suitable choice of solution can be pre-given as a starting value in the first iteration step. For example, in the gradient method, the gradient of the mass function (the derivative of the mass function with respect to the adjustment variable) is determined, and the adjustment variable is selected along this gradient for the next iteration step, where the step size from the current adjustment variable to the next adjustment variable is determined by pre-given rules of the solution method.

[0058] Since the movement of transmission unit 3 along stator 2 should be adjusted, the quality function J(SG) includes a quality term JTk(SG), which evaluates the deviation between a pre-given movement setting variable BS for adjusting the movement of transmission unit 3 and the actual variable IS, preferably the movement actual variable, which depends on the adjustment variable SG, as further detailed below. Here, as is typical in the case of adjustment, this relationship ensures that, at each time step of adjustment, the adjustment variable SG should be generated in such a way that the actual variable IS caused by the adjustment variable SG should correspond as closely as possible to the pre-given movement setting variable BS (usually represented by a specific maximum adjustment error). The actual variable IS is usually measured or determined based on other measured variables.

[0059] In the case of the linear motor 1 in the transmission system, the transmission unit 3 can typically move along the stator 2 on a single path or on multiple paths connected by switches. The path is planned before or during the movement of each transmission unit 3; this may also include planning the movement profile (movement variables, such as speed in time or along the path). Processing stations may also be provided along the path, where objects moving with the transmission unit 3 are processed. The transmission unit 3 can also transport different or varying objects, such as filled bottles, during its movement along the path. In short, the requirements for the movement of the transmission unit 3 can change during the movement of the linear motor 1 along the path.

[0060] To simplify the representation of the movement of the transmission unit 3, it is stipulated that there are at least two movement stages when the transmission unit 3 moves along the path, with a different mass function J(SG) used in each stage. The mass function J(SG) can be expressed by the mass term JT used. k The number k of (SG) and / or the mass term JT used k (SG) and / or via weighting factor k k This allows for differentiation. Therefore, the movement of the transmission unit 3 in different stages of movement can be flexibly adjusted to adapt to different requirements. This can even be done during the operation of the linear motor 1.

[0061] For example, in the first movement stage where adjusting the movement of the transmission unit 3 does not require particularly high precision, energy-optimized operation of the long stator linear motor 1 can be sought, while in the second movement stage where precise adjustment of movement is required, precise adjustment of the driving force can be expected. For example, in the turnout area, the adjustment of the movement of the transmission unit 3 can be carried out with a higher weighting of the adjustment of the normal force (in the y direction transverse to the movement direction x). However, outside the turnout, it can be done with higher energy efficiency.

[0062] Quality item JT k (SG) Evaluate the deviation between the setpoint variable BS and the actual variable IS, preferably as the square of the deviation, i.e., for example... or The weighting factor is k. e Such a quality item JT k (SG) is particularly useful for online optimization, where the current actual variable IS is available (e.g., by measurement). The actual variable IS can be measured, for example, using position sensor 13, but can also be determined based on existing known measurement variables (e.g., using an observer), where the mathematical model of linear motor 1 can also serve as the basis for the observer. In the case of offline optimization, the actual variable IS can also be determined based on the mathematical model of linear motor 1 (i.e., the model's output or state variables). Since the actual variable IS depends on the adjustment variable SG, the mass term JT... k (SG) also indirectly depends on the adjustment variable SG. Here, one can evaluate, for example, the deviation between the set speed and the actual speed, or the deviation between the set position and the actual position, or the deviation between the set driving force (or torque) and the actual driving force (or torque), or the deviation between the set magnetic flux and the actual magnetic flux. Therefore, the mass term JT k (SG) Evaluation of adjustment error. Since position and velocity are, of course, the result of the driving force and driving torque applied to transmission unit 3, and the driving force and driving torque depend on the magnetic flux, the above deviations are equivalent.

[0063] Here, the driving force or driving torque acting on the transmission unit 3 can be obtained from the mathematical model of constraint condition g(SG), where the driving force can be a propulsive force (also in the xz movement plane) and / or a normal force, but it can also be a force in the z direction. The driving torque can also typically be a torque about one of the axes x, y, and z. The driving force and driving torque can be vectors with components having different axes x, y, and z. The actual driving force or actual driving torque can then be calculated from the model. For example, the set driving force or set driving torque is determined based on the adjusted movement set variable BS, or it can be directly given as a set variable in advance. Preferably, a movement regulator is implemented that determines the required set driving force or required set driving torque from the movement set variable BS and the current actual movement variable IS (such as actual position or actual speed) according to the implemented regulation rule (e.g., PI or PID regulator).

[0064] Another quality item JT k (SG) can, for example, evaluate the electrical power required to move the transmission unit 3. Since the coil current i can be used... C Or coil voltage v C As the control variable SG, the electric power can therefore be easily determined from the control variable SG. Since the coil current i can be used... C Or, the coil voltage vC is used as the adjustment variable SG, therefore the square of the Euclidean norm of the adjustment variable SG (which can be combined into a vector) of the active drive coil 4 involved and energized is... The mass item JT provided for evaluating power k (SG), that is, for example The weighting factor is k. i As is well known, the Euclidean norm is the root of the sum of the squares of the components of the vector of the regulating variable SG. Therefore, its square is the sum of the squares of the vector components. Since electric power is proportional to the square of the current or voltage, the square of the Euclidean norm can be used... To assess the implemented electrical power, and thus also assess energy efficiency.

[0065] Another quality item JT k (SG) can be the sum of the values ​​of the adjustment variable SG of the drive coil 4 involved in the movement of the transmission unit 3. Σ It can also be the square of the sum, for example... or The weighting factor is k. ΣTherefore, the sum of the regulating variables can be evaluated. The smaller the sum of the regulating variables SG, the smaller the load on the power electronic devices used to generate the regulating variables SG. This may be advantageous for the long-term operation of the long stator linear motor 1, as the load on the power electronic devices is as uniform as possible. Consequently, the thermal load on the power electronic devices or the circuit components within them can also be reduced, especially due to lower losses. This also allows for less heating of the stator 2, which can simplify or eliminate the need for cooling the stator 2.

[0066] Therefore, the possible mass function J(SG) can be defined as follows.

[0067]

[0068] For example, if the thrust F is adjusted x and normal force F z Then the mass function J(SG) can be written as follows.

[0069]

[0070] In the case of a planar electric motor, the driving force F z It can also act in the z-direction, and a corresponding mass term JT is also included in the mass function J(SG). k , where F xS F yS F zS It is a pre-defined set value for the force, and F x (SG), F y (SG), F z (SG) is the actual value of the force determined using a mathematical model. Of course, only the propulsive forces Fx, Fz, or only the normal force Fy can be included in the mass function J(SG). In the same way, additional or other components of the driving force or driving torque can certainly be considered.

[0071] It is obvious that by changing the weighting factor k k (e.g., k) i k Σ k e k x k y k z ) and / or by adding or deleting certain quality items JT k (SG) (for example, This allows for simple changes to regulation goals and behaviors. Of course, in addition to the aforementioned quality items, other or additional quality items (JT) can also be included. k (SG) is also possible.

[0072] Therefore, by optimizing the mass function J(SG) based on the adjustment variable SG (online or offline), the optimal adjustment variable SG that minimizes (or maximizes) the mass function J(SG) can be determined. opt These optimal adjustment variables SG opt This is implemented in each time step of adjusting the movement of the transmission unit 3 on the linear motor 1, such as... Figure 6 This is illustrated schematically. For example, the moving set variable BS is for the desired thrust F. x F z and normal force F y Or the predetermined approach position or speed to be adjusted for transmission unit 3. For example, through mass items. Other or additional components of the driving force or driving torque are used to ensure compliance with the setpoint variable BS for this movement, wherein the accuracy of this compliance can be determined by the relevant weighting factor k. x k y k z k e Impact. In this context, other quality terms are unnecessary or can be weighted relatively less. If high precision is not required, higher weighting or the addition of quality terms characterizing electrical power, especially power loss, is possible. To achieve the most energy-efficient operation, the quality function J(SG) changes during different movement phases of the transmission unit 3. These different movement phases with different quality functions J(SG) can be pre-configured, or they can be determined only during movement (e.g., by the device control unit 11).

[0073] Of course, by setting further quality items JT k Other aspects of the operation of linear motor 1 can be considered.

[0074] refer to Figure 6 and Figure 7 It describes the implementation of the determined optimal control variable SG on the linear motor 1 or on the active drive coil 4 of the linear motor 1. opt Possible implementations.

[0075] According to Figure 7 In the example, the optimal adjustment variable SG was determined through optimization. opt (For example, the coil voltage v of the active drive coil 4) c The current is generated directly and applied to the drive coil 4 via a power electronic device (not shown). In this example, the movement setting variable BS is the setting position x of the transmission unit 3. set The setting position x set It can also be used for optimization, such as for cost items. To this end, the actual value of the movement variable IS can also be determined (e.g., measured or calculated), in this case, for example, the current actual position x. set And it is used in optimization. The driving force F is determined in the movement regulator RB (position regulator in this example) based on the movement set variable BS and the actual value IS of the movement variable. xS The driving force F must be adjusted. xS To minimize the adjustment error between the actual value IS and the moving setpoint variable BS within the current time step of adjustment. Driving force F xS In optimization, it can be used, for example, for cost items. However, the set value F of this driving force xS The movement setpoint variable BS can also be pre-defined. In this case, the movement regulator RB can also be implemented separately from the regulation unit 10 (e.g., as microprocessor-based hardware with software running on it, or as an integrated or analog circuit). In optimization, especially in the mathematical model of the linear motor 1, the coil current i of the active drive coil 4 can also be used for the optimization constraint g(SG). c The actual value can be measured or determined from other known values. This implementation of the adjustment unit 10 is particularly suitable for online optimization because the current actual value, such as the actual position x and / or coil current i, is detected on the linear motor 1 at each time step of the adjustment. c (or coil voltage v) c The actual value of ). This implementation can also be used for offline optimization.

[0076] According to Figure 8 In the embodiments, the determined optimal adjustment variable SG opt It is not directly generated and implemented, but rather adjusted on linear motor 1. The optimal adjustment variable SG, determined through optimization, is used. opt For example, the optimal coil current i for the active drive coil 4 c,opt These optimal coil currents i c,opt The control variable used in the control unit 10 is the control variable in the controller RS. Based on the optimal control variable SG... opt (Here, i is the coil current) c,opt ) and the current actual value of the adjustment variable SG (in this case, the coil current i of the active drive coil 4). c Specifically, based on its deviation, the regulator RS determines the control variable SG (here, the coil voltage v of the active drive coil 4) at each time step of the regulation according to the implemented regulation rule (e.g., a PI or PID regulator). c These adjustment variables SG (here, the coil voltage v) cThe voltage (v) is generated by a power electronic device (not shown) and applied to the drive coil 4. Of course, for each of the active drive coils 4, there exists such a variable regulator RS. The variable regulator RS can be implemented as software, microprocessor-based hardware, or integrated or analog circuitry. However, the variable regulator RS can also be part of the power electronic device. Furthermore, in optimization, the optimal coil voltage v... c,opt The optimal adjustment variable SG can also be determined. opt Subsequently, these optimal coil voltages v c,opt It can be used as a pre-tuning function for the regulator RS, such as Figure 8 The value is shown as a dashed line. Here, the optimal coil voltage v is... c,opt The applied coil voltage v is obtained by adding the control variable SG determined by the control variable regulator RS. c Similarly, in this embodiment, similar to according to Figure 7 In one implementation, a movement adjuster RB is provided to determine the driving force F based on a movement setting variable BS (e.g., a setting position). xS .

[0077] according to Figure 8 The implementation of the adjustment unit 10 is particularly suitable for offline optimization, but can also be used for online optimization.

[0078] In the case of offline optimization, a pre-created movement characteristic curve 12 can be used, which will drive force F x (and / or F) y F z The position x of transmission unit 3 relative to drive coil 4 is mapped to the optimal adjustment variable SG. opt For example, coil current i c,opt and / or coil voltage v c,opt As explained above. Figure 9 An exemplary representation of such a motion characteristic curve 12 is shown, which represents the position x of the transmission unit 3 relative to a plurality of active drive coils 4 and the driving force F. x The coil current i mapped to the active drive coil 4 c Of course, the movement characteristic curve 12 can be used with different defined mass functions J(SG) to determine the optimal adjustment variable SG in different movement phases. opt The movement characteristic curve 12 can be stored in the adjustment unit 10 or in an external storage unit. To adjust the movement of the transmission unit 3, simply read the optimal adjustment variable SG from the correct movement characteristic curve. opt Therefore, very little computing power is required.

[0079] The moving characteristic curve 12 can be stored in tabular form. Interpolation can then be performed between the table entries. However, it is also possible to match a mathematical function (e.g., a polynomial of a specific order) to a specific entry in the moving characteristic curve 12 using methods such as compensation calculations. In this case, only the specific mathematical function needs to be stored for the moving characteristic curve 12, and interpolation is no longer required.

[0080] For other directions of movement (e.g., the y or z direction), equivalent adjustments can certainly be achieved.

[0081] Quality item JT k (SG) variables (e.g.) F in x (SG), F y (SG) or F z (SG) can be determined using the mathematical model of linear motor 1. Here, the adjustment variable SG is used as the input variable of the model, and the mass term JT k The variables in (SG) are the output variables or state variables of the model.

[0082] Constraints g(SG) used for optimization and / or for determining the mass term JT based on physical conditions k Mathematical modeling of the linear motor 1 with variables of (SG) can be performed in different ways. The commonly used and well-known models of the linear motor 1 are the dq model or the finite element (FEM) model.

[0083] For example, the dq model mentioned at the beginning is composed of... Khong, PC, and others In the document described, or in Deng, Z. et al. The "Forces and Parameters of Permanent Magnet Linear Synchronous Machines "Forces and Parameters of Linear Synchronizers", IEEE Transactions on Magnetism, Vol. MAG-23, No. 1, January 1987 or Boldea, I. et al. "Linear Electric Actuators and Generators" by the University of Cambridge Publishing House, 1997 The dq model is described in Chapter 4. It describes the current, voltage, and magnetic flux in the dq coordinate system with the adjustment variable SG (coil current i). c and coil voltage v c The equations also include equations for the forces acting in the direction of movement x (driving force) and the lateral direction y (normal force), as well as other force or moment components.

[0084] Another advantageous possibility for modeling the behavior of the linear motor 1 is through a magnetoresistive network, which can be used to describe the coil current i. c Or coil voltage v cThis relates to the magnetic flux and the resulting driving force, propulsion force, and / or normal force, as well as the driving torque. The advantage of modeling using a magnetoresistive network is that it considers all forms of motor variables, not just the sinusoidal fundamental wave. Therefore, nonlinear effects (such as saturation) and cogging forces can be systematically and with sufficient abstraction to allow the model to be computed online.

[0085] Using a magnetoresistive network model of the linear motor 1 to regulate the movement of the transmission unit 3, even with a different regulation method than described above, is novel and inventive in itself. For example, the magnetoresistive network model can be used to create an observer that estimates the actual values ​​of the magnetic flux and / or force subsequently used for regulation.

[0086] A reluctance network describes the linear motor 1 as a network consisting of reluctance R (magnetoresistance), permeability G (magnetic conductance), and a magnetic voltage source. As is well known, reluctance describes the relationship between magnetic voltage and magnetic flux Φ. Permeability G is the reciprocal of reluctance R. The reluctance R of a magnetic conductor of length l and cross-section A is given by... Given, where µ0 is the free magnetic permeability, and µ r denoted as ρ, where ρ is the relative permeability of the magnetic conductor material.

[0087] Using reluctance network modeling as an example, taking a long stator linear motor as linear motor 1, according to... Figure 2 The structure of the long stator linear motor shown has been described. In other structural designs or with other linear motors 1 (such as in the case of a planar motor), corresponding changes may occur in the reluctance network and model.

[0088] Stator 2 is modeled as a magnetoresistive network RN with i = 1, …, n teeth (stator teeth 5 and auxiliary teeth 7). This is not necessarily all the teeth present on stator 2, but rather, for example, the teeth that interact with transmission unit 3. Typically, several drive coils 4 are selected, which are energized for the purpose of moving transmission unit 3. This yields the number of teeth n modeled using the magnetoresistive network RN. In j = 1, …, n c On each of the 5 stator teeth (where n) C ≤ n), with drive coils 4 arranged, and auxiliary teeth 7 arranged between these drive coils 4. Without the auxiliary teeth 7, the magnetoresistive network RN is correspondingly simplified. The dimensions used for the stator 2 and the transmission unit 3 are within Figure 2 The values ​​are given in the text, and for linear motor 1, they can be assumed to be known.

[0089] Each stator tooth 5 passes through differential reluctance (resistance per unit length). To model, the cross-sectional area A of stator tooth 5 is... c = w c ·bs By using the relative permeability µ, which depends on the magnetic flux Φ. r By performing nonlinear modeling of the material, the magnetic saturation of stator tooth 5 can be considered. Similarly, each auxiliary tooth 7 uses differential reluctance. To model, the cross-sectional area A of the secondary tooth 5 is... a = w a ·b s A width w is present between two adjacent teeth. ca Tooth coupling in the inter-tooth space is achieved through differential magnetic permeability. To model it, j driving coils 4 can be modeled as differential magnetic voltages. , where N C Indicates the number of turns of drive coil 4, l S It represents the length of the tooth in the y direction (i.e., transverse to the direction of movement x).

[0090] Then you can use, as follows Figure 3 The stator 2 is shown with a magnetoresistive network RN to model the differential segment dy in the y-direction. Each tooth forms a branch of the stator's magnetoresistive network RN, where the branches are connected by differential magnetic permeability G', thus forming a node of the magnetoresistive network RN. Where Φ i (y) represents the magnetic flux in the i-th tooth (stator tooth 5 or auxiliary tooth 7), while i (y) represents the magnetomotive force along the i-th tooth in the y-direction.

[0091] Therefore, at each of the n nodes of the stator's magnetoresistive network RN, a magnetic flux Φ along the tooth in the y direction can be established. i (y) and magnetic potential i The nodes of (y) are summed to obtain the following 2n equations as the system equations.

[0092]

[0093] The local differential description of flux and potential magnitude realizes the non-negligible stray flux, especially considering the interdental space.

[0094] By dividing these equations by dy and considering dy→0, we obtain the system equations:

[0095] .

[0096] In vector Φ(y) and In (y), n magnetic fluxes and magnetic potentials are summed, i.e. and The system matrices A1 and A2 are obtained.

[0097] and (Where diag represents a diagonal matrix), inputting matrix B yields:

[0098] .

[0099] To fully model stator 2, the ends at y = ls (i.e., stator yoke 6) and y = 0 (i.e., the transition to the air gap) must be modeled as boundary conditions, such as... Figure 4 The explanation given is that the distance between two adjacent teeth is at y = l. s The stator yoke at point 6 utilizes linear magnetic permeability G sy To model, among which , where µ ry It is the relative permeability of stator yoke 6, assumed to be constant, and A sy = b s ·l j This is the cross-sectional area of ​​stator yoke 6. When y=0, the tooth passes through the leakage magnetic permeability G. s0 Coupling, the leakage permeability G s0 For example, it can be used To model. Linear magnetic permeability G sy and leakage permeability G s0 The branches of the stator's magnetoresistive network RN are reconnected by creating more nodes within the network. Magnetic flux. This represents the magnetic flux flowing into the teeth of the stator 2 through the driving magnet 9 of the transmission unit 3, i.e., the magnetic flux in the air gap.

[0100] Using the same processing method as above, we again obtain the result for y = l s The system of equations with boundary conditions y = 0.

[0101] Where the matrix This supplements the system equations.

[0102] When the material of the teeth of stator 2 is assumed to be magnetically linear, the magnetic reluctance R' a and R' c The dependence on magnetic flux Φ disappears, and an analytical solution can be given for the above differential system equations.

[0103] For stator 2, j = 1, …, n c Chain flux in stator tooth 5 Subsequently obtained Having the i-th unit vector e i matrix The magnetic flux belonging to each stator tooth 5 in Φ(y) is selected by integration.

[0104] Conversely, if the material is nonlinear and magnetic saturation occurs, analytical solutions are impossible. In this case, approximation methods can be applied to solve the differential equations in the system equations.

[0105] For example, the i-th magnetic flux Φ in the tooth i and the i-th magnetic potential i We can use the approximate polynomial g j (y) is used to approximate the polynomial, which has a length l along the tooth. s The quantity (in the y-direction) is the coefficient at N support positions. and Its form is:

[0106]

[0107] By appropriately selecting the approximate polynomial g j (y) can transform the system of nonlinear differential equations into a solvable system of nonlinear algebraic equations. The approximate polynomial g j One possible choice for (y) is of the form: , where j = 0, …, N is the Lagrange interpolation polynomial. This transformation is achieved by approximating the polynomial g. j (y) is obtained by differentiating it to y and evaluating it at N support positions. Therefore, based on... or We obtain a constant differential matrix D. Therefore, matrix D has entries. The system of nonlinear algebraic equations, which serves as the system equations, then follows:

[0108]

[0109] in , as well as The transformed system matrix is ​​obtained ,in

[0110] and

[0111] and having an identity matrix I and having and Kronecker of .

[0112] The equations for the boundary can also be transformed in this way, thus obtaining the transformed equations.

[0113]

[0114] The selection matrix is ​​T. L And T0, which is at the boundary support points (y = 0 and l) s Select at ) and The entry.

[0115] Therefore, a chain flux that depends on the approximate magnetic flux can be given. An approximation can usually be given as The matrix W is derived from the approximate polynomial g. j (y) is obtained from the selection of N support points.

[0116] To determine the magnetic flux generated in the air gap by the driving magnet 9 of the transmission unit 3, the rotor's reluctance network RN is also utilized. L Modeling the air gap and transmission unit 3, such as Figure 5 What is being explained.

[0117] Transmission unit 3 uses air gap magnetic permeability G a,ij Coupled with stator 2, its indices i = 1, …, n represent the number of teeth, while indices j = 1, …, p represent the number of driving magnets 9 on transmission unit 3. Stator 2 is coupled by the magnetomotive force at the edge of stator 2 when y=0. Generated magnetic voltage Represented. Using the above approximation, for example, magnetic voltage is represented by a matrix. And the selection matrix T0 explained above get.

[0118] For example, the driving magnet 9 is composed of a constant magnetic voltage u msj = H c l m (with a thickness of permanent magnet l) m And the known magnetic coercivity H for the permanent magnet used c Voltage source and linear permeability The permanent magnet described, wherein A m is the known cross-sectional area of ​​the permanent magnet, and µ rm It is the known constant relative permeability of the permanent magnet. For other driving magnets 9, other models with non-constant magnetic voltages can be used. The yoke connecting the driving magnet 9 to the transmission unit 3 utilizes linear permeability. To describe it, it has a known cross-sectional area A. t and the known constant relative permeability µ rt and the distance w between the two driving magnets 9 t (For example, such as) Figure 2 (As shown in the center-center diagram). The leakage flux of the transmission unit 3 towards the stator 2 is transmitted through the leakage magnetic conduction. The description describes a structure having a gap width w between two driving magnets 9. ml Therefore, each driving magnet 9 forms a structure with linear magnetic permeability G. m and magnetic voltage u with magnetic voltage source msj The rotor's magnetoresistive network RN L The branches, these branches form the reluctance network RN of the rotor. L In the case of nodes, the leakage magnetic permeability G at the air gap ml The linear magnetic permeability at the yoke is connected at the end of each branch.

[0119] The rotor's reluctance network RN is connected at the node in the air gap. L Other branches, in which air gap magnetic permeability G is set a,ij Each of the p driving magnets 9 is connected via an air gap magnetic permeability G. a,ij It is connected to each of the n considered teeth in the stator's reluctance network RN. It has an air gap permeability G. a,ij The branches are respectively in the magnetoresistive network RN that constitutes the rotor. L In the case of other nodes, through the magnetic voltage u representing stator 2 s The voltage source is connected.

[0120] The air gap permeability G describes the magnetic coupling between stator 2 and transmission unit 3. a,ij (x s , y s It depends on the position of the transmission unit 3 relative to the stator 2 in the moving direction x and the lateral direction y. s y s Each driving magnet 9 is magnetically coupled to each considered tooth of the stator 2, or in other words, each driving magnet 9 is magnetically coupled to the magnetic flux Φ flowing into a tooth. Σ Make a contribution. Air gap permeability G a,ij (x s , y s The position x of the drive magnet 9 relative to the stator tooth 5 can be assumed to be known (e.g., determined in advance by measurement or simulation). Therefore, it is only necessary to determine the position x of the drive magnet 9 relative to the stator tooth 5. s y s air gap permeability G a,ij (x s , y s If auxiliary tooth 7 is used, it also depends on its position relative to auxiliary tooth 7. For example, the air gap permeability G a,ij (x s , y s It can then be stored in the adjustment unit 10 in tabular form, as a mathematical formula, or as a characteristic curve.

[0121] The rotor's magnetoresistive network RN can also be depicted.L The magnetic coupling between two adjacent transmission units 3. Therefore, this dependence will also be taken into account through this adjustment.

[0122] In order to obtain the magnetoresistive network RN from the air gap and transmission unit 3 L A system of independent equations can be derived in different ways. One possible approach is to use well-known graph theory, as illustrated below. The use of graph theory is advantageous because it provides a system for establishing and solving equations based on the independent variables.

[0123] Rotor reluctance network RN L Topology ( Figure 5 The network is divided into a tree and a Ko tree. Here, the tree network connects all nodes without forming a net (a closed loop on the branches), and the Ko tree includes all elements that are not part of the tree. The choice of tree is largely free, wherein preferably all magnetovoltage sources are located in the tree. For example, a suitable choice is to summarize the last air gap permeability in the Ko tree. and leakage permeability All air gap permeability except Subsequently, the magnetic flux Φ of the tree and the magnetic flux Φ of the Ko tree. c It can be summarized into a vector and The tree elements can be partitioned into the magnetic voltage source of stator 2 with index ts, the magnetic voltage source of driving magnet 9 with index tm, and the magnetic permeability in the tree with index tg. This yields a partial vector of the magnetic flux. , , In a similar manner, the magnetic voltage u of the tree is obtained. t And the magnetic voltage u of the Ko tree c The vector of magnetic voltage, where the vector of magnetic voltage is... and as well as The magnetic flux and voltage of the tree and Ko tree are transmitted via matrix V. and The connection is in the form of a magnetoresistive network RN, where matrix V is formed by the magnetoresistive network RN. L The topology and the partitioning of trees and Ko trees are obtained. Matrix V can be derived from the above partitioning. Divide again.

[0124] Matrix V depends on the number of stator tooth elements selected.

[0125] For n c 4 driving coils (and thus n = 2 n c +1 tooth), matrix It can be represented as , where element 0 describes the zero vector, and It has dimensions The lower triangular matrix, and all entries of the triangular matrix are 1.

[0126] For example, for five driving magnets 8 (p=5), the matrix get It has dimension Submatrices. For example, these submatrices yield... , , , , .

[0127] From the connection of magnetic flux and voltage between the tree and the Ko tree, matrix V is also obtained. g The matrix contains 0 and 1 again.

[0128] Furthermore, the magnetic flux Φ and magnetic voltage u are also transmitted via magnetic permeability. and The connection is in the form of , where the magnetic permeability matrix of the tree is . And the magnetic permeability matrix of the Ko tree is .

[0129] In order to obtain the desired result for the entire magnetoresistive network RN (RN S + RN L The system equations, and the magnetic flux Φ under the boundary conditions at y = 0. Σ Expressed by the magnetic flux in the air gap, thus obtaining .matrix and Select the air gap flux of each of the n teeth and sum them up. For example, we can obtain... .

[0130] The stator reluctance network RN and the rotor reluctance network RN L The resulting equations can be combined, and thus can be processed according to the system of equations. The system equations for the entire reluctance network RN are expressed in the form of SG=i. This set of equations represents the model of the linear motor 1 and can be used as a model with SG=i c The constraint condition g(SG,x) of the state vector x, or the mass term JT used to determine the mass function J(SG). k The variable (SG) can be given for any topology of the magnetoresistive network RN.

[0131] For according to Figure 3 , 4 The embodiment of 5 yields a result with state variables. State vector and the entire system matrix

[0132] The input vector follows .

[0133] To obtain a well-defined system of equations, you can add a line to the entire system equations. Among them, it is determined .

[0134] Therefore, this model can be used to calculate all system variables, especially the coil current i. c The chain flux ψ of the driving coil 4 is a function of the function. c The entire system of equations can also be easily transformed to determine the magnetic flux ψ. c The driving current i of the function c .

[0135] Typically, in the case of a linear motor, the coil voltage v is generated by means of a power electronic device. c And apply it to the drive coil to make the coil voltage v c Used as an input variable. When using the known law of electromagnetic induction... In this case, using the known ohmic resistance R of the drive coil 4 c Given the inductance L, we can determine the coil current i. c Determine the coil voltage v c ,vice versa.

[0136] Based on the system equations of the magnetoresistive network RN, for example, by applying the known principle of magnetic co-energy, the driving force and / or driving torque, such as the driving force F, can be derived. x and axial force F y Its form is

[0137] ,

[0138] Therefore, based on the relative position of transmission unit 3 with respect to stator 2 [x] s , y s The driving force and driving torque acting on transmission unit 3 can be determined. The driving force and driving torque can supplement the model of the long stator linear motor 1.

[0139] The reluctance model of the long stator linear motor described can of course be extended to the case of a planar motor as linear motor 1.

Claims

1. A method for operating a linear motor (1) having a stator (2) on which a plurality of drive coils (4) are arranged, and a transmission unit (3) having a plurality of drive magnets (9) arranged thereon, which moves along the stator (2), wherein an electromagnetic field is generated by energizing the active drive coils (4) in the region of the drive magnets (9) of the transmission unit (3) to move the transmission unit (3), wherein a mass function J(SG) is used as a function of the adjustment variable SG of the active drive coils (4), wherein the mass function J(SG) comprises a number of k ≥ 1 weighted by k k The weighted mass term JT depends on the moderating variable. k The sum of (SG), wherein the quality function J(SG) includes a quality term JT that evaluates the deviation of a pre-given movement setting variable BS for adjusting the movement of the transmission unit (3) from the actual variable IS of the adjustment, which depends on the adjustment variable SG. k (SG), wherein the quality function J(SG) used to adjust the movement of the transmission unit (3) along the stator (2) is optimized with respect to the adjustment variable SG to determine the optimal adjustment variable (SG) for the corresponding time step of the adjustment of the movement. opt ), and based on the determined optimal adjustment variable (SG) opt The active drive coil (4) is energized, characterized in that, During the movement of the transmission unit (3) along the stator (2), there are at least two movement stages, wherein different mass functions J(SG) are used in the at least two movement stages to determine the optimal adjustment variable (SG). opt The different mass functions J(SG) are derived from the mass term JT used. k The number k of (SG) and / or the mass term JT k (SG) and / or weighting factor k k To distinguish them.

2. The method as described in claim 1, characterized in that, The quality function J(SG) is optimized online at each time step of the adjustment during the movement of the transmission unit (3) to determine the optimal adjustment variable (SG) at the corresponding time step of the adjustment of the movement. opt ).

3. The method as described in claim 1, characterized in that, The quality function J(SG) is optimized offline for a pre-given movement of the transmission unit (3) along the stator (2) to pre-calculate the optimal adjustment variable (SG) to be set during the execution of the movement. opt ).

4. The method as described in claim 1, characterized in that, For different quality functions J(SG) in the at least two movement stages, movement characteristic curves (12) are pre-created through offline optimization. These movement characteristic curves map the movement setpoint variable (BS) and the position of the transmission unit (3) relative to the active drive coil (4) to the optimal adjustment variable (SG). opt During the movement of the transmission unit (3), the relevant optimal adjustment variable (SG) is read from the movement characteristic curve (12) based on the determined actual position of the transmission unit (3) relative to the active drive coil (4) and the pre-given set value of the movement setting variable (BS). opt ).

5. The method according to any one of claims 1 to 4, characterized in that, The optimal adjustment variable (SG) opt It is adjusted in the regulator.

6. The method according to any one of claims 1 to 4, characterized in that, The mass function J(SG) includes a mass term JT that evaluates the electrical power required to move the transmission unit (3). k (SG).

7. The method as described in claim 6, characterized in that, As a quality item JT k (SG) is the square of the Euclidean norm of the vector of the adjustment variable (SG) with the energized drive coil (4), i.e. .

8. The method according to any one of claims 1 to 4, characterized in that, The mass function J(SG) includes the form or Quality item JT k (SG).

9. The method according to any one of claims 1 to 4, characterized in that, The quality function J(SG) comprises the sum of the moderating variables (SG) used to evaluate the quality function. Σ Quality item JT k (SG), that is .

10. The method according to any one of claims 1 to 4, characterized in that, The quality function J(SG) comprises the sum of the moderating variables (SG) used to evaluate the quality function. Σ The square of the mass term JT k (SG), that is .

11. The method according to any one of claims 1 to 4, characterized in that, Quality item JT k (SG) includes variables determined using the mathematical model of the linear motor (1).

12. The method according to any one of claims 1 to 4, characterized in that, During optimization, the constraint condition g(SG) is considered in the form of a mathematical model of the linear motor (1).

13. The method as described in claim 11, characterized in that, The linear motor (1) is modeled using a magnetoresistive network (RN) as a branched network in which magnetoresistive R, magnetic permeability G or magnetic voltage sources are arranged and magnetic flux Φ flows in the magnetoresistive network.

14. The method as described in claim 13, characterized in that, The stator (2) of the linear motor (1) is modeled using n teeth connected by a stator yoke (6), and the differential segment dy of each of the n teeth, which are branches of the stator's reluctance network (RN), utilizes differential reluctance. , To model and, in the case of a driving coil, utilize a differential magnetic voltage on one of the said number of teeth. The model is based on a magnetic voltage source, where a magnetic potential is applied at the end of each branch. n (y) n (y+dy), and magnetic flux Φ n (y) flows into the branch, and two adjacent branches are connected by differential permeability G' in the case of nodes constituting the magnetoresistive network (RN) of the stator, wherein differential permeability G' models the magnetic coupling of two adjacent teeth through the inter-tooth space, and the stator yoke (6) is connected by linear permeability G. sy To model, the linear magnetic permeability G sy In the case of another node constituting the stator's magnetoresistive network (RN), the two adjacent branches of the stator's magnetoresistive network (RN) are connected, and the air gap (14) at the opposite end of the stator yoke (6) of the stator (2) is filled with leakage magnetic conductance G. s0 And the magnetic flux Φ through the air gap (14) Σn To model, the leakage magnetic permeability G s0 In the case of another node constituting the magnetoresistive network (RN) of the stator, connect the two adjacent branches of the magnetoresistive network (RN) of the stator.

15. The method as described in claim 14, characterized in that, At each node of the stator's magnetoresistive network (RN), establish the magnetic flux Φ in the branch. n (y) and magnetic potential on the branch n The nodes of (y+dy) are summed to obtain a set of equations as the system equations of the stator (2), which model the stator (2) of the linear motor (1).

16. The method according to any one of claims 13 to 15, characterized in that, The transmission unit (3) and the air gap (14) between the teeth of the stator (2) and the drive magnet (9) connected by a yoke on the transmission unit (3) utilize the rotor's reluctance network (RN) L The model is constructed using p driving magnets (9) as the magnetoresistive network (RN) of the rotor. L The branch utilizes linear magnetic permeability G. m And utilizing the magnetic voltage u of the magnetic voltage source msp To model, wherein the ends of the branches are in the magnetoresistive network (RN) constituting the rotor L In the case of the node, the leakage magnetic conductance G at the air gap (14) is... ml and the linear magnetic permeability G at the yoke b To connect, at the node at the air gap (14), the rotor's magnetoresistive network (RN) is connected. L Another branch of ), in which air gap magnetic permeability G is provided. a,ij Each of the p driving magnets (9) in the rotor's magnetoresistive network (RN) L ) respectively via air gap magnetic permeability G a,ij It is connected to each of the n teeth in the stator's magnetoresistive network (RN) and has an air gap permeability G. a,ij The branches are respectively in the magnetoresistive network (RN) constituting the rotor. L In the case of another node of the stator (2), the magnetic voltage u is represented by the stator (2). s Connect it to the voltage source.

17. The method as described in claim 16, characterized in that, According to the rotor's magnetoresistive network (RN) L The system equations are formed as the system equations of the rotor, which model the air gap (14) and the transmission unit (3).

18. The method as described in claim 15 or 17, characterized in that, From the magnetoresistive network (RN) S ), (RN L The system equations formed by the system are used to derive the driving force and / or driving torque acting on the transmission unit (3).

19. A linear motor (1) having a stator (2) on which a plurality of drive coils (4) are arranged, and a transmission unit (3) movable along the stator (2) having a plurality of drive magnets (9) arranged thereon, wherein an active drive coil (4) is energized in the region of the drive magnets (9) of the transmission unit (3) to generate an electromagnetic field, wherein the electromagnetic field interacts with the drive magnets (9) of the transmission unit (3) to move the transmission unit (3), wherein an adjustment unit (10) is provided, in which a mass function J(SG) is implemented as a function of the adjustment variable (SG) of the energized active drive coil (4), wherein the mass function J(SG) contains a number of k ≥ 1 weighted by k k The weighted mass term JT depends on the moderating variable. k The sum of (SG), wherein the quality function J(SG) includes a quality term JT that evaluates the deviation of a pre-given movement setting variable (BS) for adjusting the movement of the transmission unit (3) from the actual variable (IS) of the adjustment, which depends on the adjustment variable (SG). k (SG), wherein the adjustment unit (10) optimizes the quality function J(SG) for adjusting the movement of the transmission unit (3) along the stator (2) with respect to the adjustment variable (SG) to determine the optimal adjustment variable (SG) for the corresponding time step of the adjustment of the movement. opt ), and based on the determined optimal adjustment variable (SG) opt The driving coil (4) is energized, characterized in that, The adjustment unit (10) includes at least two movement phases during the movement of the transmission unit (3) along the stator (2), wherein the at least two movement phases are configured to determine the optimal adjustment variable (SG). opt Different mass functions J(SG), wherein the different mass functions J(SG) are determined by the mass term JT used. k The number k of (SG) and / or the mass term JT k (SG) and / or weighting factor k k To distinguish them.

Images