EP0731051A1 - Device and method for damping vibrations on an elevator cage - Google Patents

Abstract

The device includes actuators equipped with linear motors (7). The stationary part (16) of the motor is attached to the frame of the lift car, and the moving part of the motor (17) is attached to guides (21). The moving part of the motor is a magnet which is attached by a tension-compression component to a roller lever serving as the guide.

Description

The invention relates to a device and a method for vibration damping on an elevator car guided on rails, which has movably connected guide elements between two end positions, vibrations occurring transversely to the direction of travel being measured by a plurality of inertial sensors attached to the car and for controlling at least one between the car and Guide elements arranged actuator are used, which works to the vibrations occurring and opposite to the direction of the vibrations.

Cross vibrations act on the cab due to unevenness in the guide rails, as well as due to the wind, as a result of lateral tensile forces that are transmitted by the traction cables or when the load changes position during travel. A method for damping such vibrations in an elevator car or in a part thereof is known from US Pat. No. 5,027,925; After determination of the undesired transverse accelerations, corresponding counterforces are exerted on the cabin by a vibration damper which is arranged between the cabin and the frame. However, this method requires a complex floating storage of the cabin in a cabin frame, which in addition to the high expenditure on equipment requires a very large amount of additional space. In addition, the force acts on the frame, which can cause it to jerk back and forth between the guides at low frequencies. Such a system is hardly controllable in terms of control technology.

In contrast, the invention has for its object to simplify the vibration damping device and the method and to achieve a satisfactory damping of the various vibrations acting on the cabin at any time. This object is achieved by the teaching specified in the characterizing part of claim 1.

The use of one linear motor per actuator is particularly advantageous because these motors can generate large dynamic and static forces and have low energy consumption. In addition, they are light in weight and have small moving masses and are relatively easy to regulate. With the invention, the transverse accelerations exerted on the guide elements and transverse forces acting directly on the cabin are reduced to such an extent that they can no longer be felt in the cabin. The device for vibration damping remains functional even under asymmetrical stress; it adjusts itself automatically when the cabin is tilted relative to the guide rails, so that a sufficient damping path is available on both sides.

The equipment required to carry out the method is low and the masses which are moved quickly are very small. This is also achieved in that all measurement signals are fed to a common control and this acts on only one actuator per guide element. Structural resonances can also be suppressed by adjusting the frequency response of the controller.

The measures listed in the dependent claims allow advantageous developments and improvements of the invention.

The position feedback for returning the guide elements to the central position is particularly advantageous, the position feedback is only active in the low frequency range.

Fig.1

eine schematische Darstellung einer an Schienen geführten Aufzugskabine,

Fig.2

einen als Linearmotor ausgebildeten Aktuator,

Fig.3

eine Frontansicht einer Rollenführung,

Fig.4

eine Seitenansicht einer Rollenführung,

Fig.5a,b,c

drei Varianten eines Drehantriebs für den Aktuator,

Fig.6a

schematisch eine Aufzugskabine mit Aktuatoren und Sensoren in x_k-Richtung,

Fig.6b

schematisch eine Aufzugskabine mit Aktuatoren und Sensoren in y_k-Richtung,

Fig.7

den Reglerteil des aktiven Systems, und

Fig.8

das Blockdiagramm für das ganze System.

The invention is explained in more detail below with reference to an embodiment shown in the drawing. Show it:

Fig. 1

1 shows a schematic representation of an elevator car guided on rails,

Fig. 2

an actuator designed as a linear motor,

Fig. 3

a front view of a roller guide,

Fig. 4

a side view of a roller guide,

Fig.5a, b, c

three variants of a rotary drive for the actuator,

Fig.6a

schematically an elevator car with actuators and sensors in the x _k direction,

Fig. 6b

schematically an elevator car with actuators and sensors in the y _k direction,

Fig. 7

the controller part of the active system, and

Fig. 8

the block diagram for the whole system.

1 shows a schematic representation of the device according to the invention. A cabin 1 is guided by means of roller guides 2 on rails 3, which are attached in a shaft, not shown. The cabin 1 is elastically mounted in a cabin frame 4 for passive vibration damping. Rubber springs 4.1 are used for this purpose, which are designed to be relatively stiff in order to suppress the occurrence of low-frequency torsional vibrations about the y-axis. The Roller guides 2 are attached to the cabin frame 4 laterally below and above. They consist of a stand 5, actuators 6 and guide elements in the form of two lateral rollers 8 and a central roller 9 arranged at 90 ° to it.

Unevenness in the rails 3, lateral tensile forces caused by the traction ropes or changes in the position of the load while driving cause vibrations of the cabin frame 4 and the cabin 1 and thus impair driving comfort. Such vibrations of the cabin 1 are to be reduced. Two position sensors 10 per roller guide 2 each measure the distance between the car 1 and the rail 3. At least three or five inertial sensors 11 measure the vibrations or accelerations occurring transversely to the car 1. The inertial sensors 11 are preferably arranged in the center of gravity of the cabin frame 4 and in pairs far from each other in order to also be able to detect rotations around the z-axis. In addition, vibrations generated by wind and cable forces are well recorded.

By processing the measured values, the actuators 6 arranged on the roller guides 2 are regulated, which operate simultaneously with the vibrations occurring and opposite to the direction of the vibrations. Damping of the vibrations acting on the cabin 1 is thus achieved. Vibrations are reduced in such a way that they no longer have an effect on the cabin 1 in a way that is perceptible to the passenger. Each roller guide 2 is equipped with two actuators 6. This allows five degrees of freedom or axes of cabin 1 to be controlled: displacement in the x and y directions, and rotation about the x, y and z axes.

It is also possible to equip only the lower two roller guides 2 with two actuators 6 each. This allows the three degrees of freedom of a plane or three axes to be controlled: displacement in the x and y directions, and Rotation around the z axis according to the coordinate system in Fig. 1.

2 shows a linear motor 7 of an actuator 6 according to the device according to the invention. The linear motor 7 is based on the principle of the moving magnet. It consists of a laminated stator 16 provided with windings 15 and a movable motor part 17 designed as a magnet. A magnet 18 is attached to the movable motor part 17. Advantages of the linear motor 7 are its simple controllability, as well as low weight and small moving masses and a large dynamic and static force with low energy consumption.

3 and 4 show a roller guide 2 according to the device according to the invention. The stand 5 is fastened to the cabin frame 4 by means of fastening elements 19. Each roller guide 2 is equipped with two actuators 6, each of which is provided with a linear motor 7. One shifts the middle roller 9, the other linear motor 7 the two side rollers 8. The rollers 8, 9 are fastened to roller levers 21 by means of axle bolts 20. The roller levers 21 of the two lateral rollers 8 are connected to one another via a pull rod 22. To transmit the movements emanating from the actuators 6, the roller levers 21 are connected in an articulated and low-friction manner to the stand 5 by means of axle bolts 23, and the roller levers 21 of the two lateral rollers 8 are connected in an articulated and low-friction manner to the tie rod 22 by means of axle bolts 24. On the stands 5 guide rods 25 with pressure springs 26 are attached. The pressure springs 26 are each fixed to the outer end 27 of the guide rods 25. The guide rods 25 run through a passage 28 in the roller levers 21, so that the pressure springs 26 rest on the outer sides 29 of the roller levers 21 and press the rollers 8, 9 against the guide rail 3.

A fastening plate 30 is attached to the stand 5 by means of fastening elements 31 such as screws. The stators 16 of the actuators 6 are screwed onto the fastening plate 30 with fastening elements 32. The movable motor part 17 is connected by means of screws 33 to the roller lever 21 and thus to the rollers 8, 9. In order for the air gap 34 of the linear motor 7 to be retained, lateral guidance is still required. This consists of ball-bearing rollers 35 and is almost frictionless. Two brackets 36 enable the mounting of the ball-bearing rollers 35 and form the lateral boundaries of the movable motor part 17. A low-friction bearing is necessary in order to be able to precisely control the force to be generated by the actuator 6. Starting from a central position 37, the length of the stator 16 of the linear motor 7 determines the maximum possible inner and outer end positions. The travel is limited by elastic stops 38 and 39.

One variant consists in connecting the movable motor part 17 to the roller lever 21 by means of a push-pull member. The movable motor part 17 is then stored independently of the roller lever 21.

Due to the parallel connection of the pressure spring 26 with the actuator 6, the roller guide 2 remains functional even in the event of a partial or complete failure of the active vibration damping, since the pressure springs 26 press the rollers 8, 9 against the guide rail 3 independently of the actuator 6.

5a, 5b and 5c show variants of using a rotary drive 43 instead of the linear motor 7. This drive has a swivel angle of approximately 90 degrees and drives the roller lever 21 with a crank 44 and a pull-push member 45 (FIG. 5a) or a flexible pulling means 46 (FIG. 5b) or with a cam plate 47 (FIG .5c).

6a and 6b show an elevator car 1 with actuators and sensors in the x _k direction and in the y _k direction according to the inventive device. To simplify the illustration, the x _k and y _k directions are each shown separately.

• Eine Verschiebung x_k in x_k-Richtung
• Eine Drehung φ_ky um die y_k-Achse
• Eine Verschiebung y_k in y_k-Richtung
• Eine Drehung φ_kx um die x_k-Achse
• Eine Drehung φ_kz um die z_k-Achse

The regulation for suppressing the cabin vibrations and for correcting the setting of the cabin 1 with respect to the two guide rails 3 is based on a dynamic model of the system. This model is a mathematical description that summarizes all existing practical and theoretical experience with the system. The cabin vibrations to be dampened by this device occur in the following degrees of freedom:

• A displacement x _k in the x _k direction
• A rotation φ _ky around the y _k axis
• A displacement y _k in the y _k direction
• A rotation φ _kx around the x _k axis
• A rotation φ _kz around the z _k axis

The system model describes the dynamics of the elevator system in all degrees of freedom mentioned above. This model also takes into account all relevant structural resonances that arise due to the elasticities between the different masses and within the cabin frame 4.

Based on the system model, a controller is used which treats all degrees of freedom described by the model at the same time. For this purpose, the methods of robust multi-size control are used (Multi-Input Multi-Output or MIMO Robust Design). These methods use the existing system model to design an observer-based controller. The observer is a dynamic part of the controller and has the task, based on the existing measurements (eg accelerations different measuring points) to estimate all movement states that cannot be measured directly (e.g. speeds and positions of the different masses) in real time. Thus, the controller will have a maximum of information about the system. Based on all movement states and not only on their measurable part, the controller provides the best answer for every degree of freedom, which significantly increases the quality of the control. Since the model and the observer based on it take into account all relevant structure resonances, the controller does not excite any of these resonances. The model-based control ensures the necessary stability of the system. This would not be the case if the system dynamics were not taken into account in the controller design.

The robust controller is designed in such a way that it only becomes effective in a certain frequency range so that it does not react to undesired frequency-dependent system dynamics and interference. This is done without having to connect additional filters to the controller.

Additional filters can limit the effectiveness of the controller and easily lead to instabilities. They also significantly increase the computational complexity of the control algorithm. Another advantage of the robust design method is the consideration of the model error during the design. This is done by quantifying the inaccuracies of the model as frequency-dependent variables and taking them into account in the controller design. The resulting controller thus has sufficient robustness against any malfunctions and modeling errors.

The first goal of the controller is the suppression of cabin vibrations in the high frequency range (between 0.9 and 15 Hz) without the controlled elevator outside this range becoming worse than the uncontrolled one. On the other hand, the controller must ensure that the setting of the cabin frame 4 with respect to the two guide rails 3 is so it is regulated that there is a sufficient damping path on each roller 8, 9. This is particularly important when the cabin 1 is loaded asymmetrically. For the first control purpose, an acceleration or a speed feedback with inertial sensors 11 should be sufficient, a position feedback being necessary for the second control target. If the absolute position of the cabin 1 could be measured and returned for the control, the second return would have no conflict with the first. However, since only measurements of the relative positions between the rollers 9 and the car frame 4 are available, the absolute position of the car 1 cannot be measured, but only the position of the car frame 4 relative to the guide rails 3 4 and keep roller lever 21 constant, which is nothing other than following the unevenness of the rails. Therefore, the two returns have two contradicting goals. To avoid the conflict between acceleration (or speed) and position feedback, the following strategy is followed:

Two controllers are used to generate a common output signal. The first controller has the measurements from the inertial sensors 11 alone and is therefore responsible for the suppression of the vibrations. The second controller has the position measurements alone and is responsible for the guiding games of cabin 1. The setpoints of the forces that the first controller demands from the actuators 6 are added to the corresponding quantities of the second controller. The solution to avoid the conflict between the two controllers is based on the fact that the forces responsible for the skew of the cabin 1 (a non-symmetrical loading of the cabin, a large lateral rope force, etc.) change much more slowly than the other sources of interference, which cause the cabin vibrations (mainly Rail unevenness and air disturbances). For this reason, the position control operating in the low frequency range, which is rather harmful for the suppression of the vibrations, is limited to 0 to 0.7 Hz. This means that there is no adverse influence on the suppression of the vibrations, because this only works from 0.9 Hz. The feedback of the signals from the inertial sensors 11 must not be effective in the frequency range below 0.9 Hz, so that the zero sensor error and, in the case of an acceleration sensor, the measured part of gravity, which is not constant due to the tilting movement, has no influence on the position control. This also avoids the risk of the actuators 6 being overdriven. Limiting the bandwidth of each feedback loop using the robust design process becomes particularly important for this purpose.

Another advantage of this solution is that the controller does not contain any non-linearity. Nonlinearity makes stability analysis very difficult, if at all possible. Since the two feedback loops are designed at the same time, the method takes both control loops into account in the stability analysis.

The mounting of the inertial sensors 11 on the cabin frame 4 instead of on the cabin body 1 or on the roller guides 2 is particularly advantageous for an efficient control. If the sensors were mounted on the cabin body 1, the measurements would show considerable phase losses due to the elastic suspension of the Go back cabin 1. Far higher vibration amplitudes occur on the roller guides and the influence of gravity would have to be compensated for.

The controllers are designed for the system in the cabin coordinate system. The measurements from the coordinate system of each sensor to the car body coordinate system are mapped using various linear transformations. Another transformation from the cabin coordinate system too the actuator coordinate systems is necessary for the output of the force setpoints.

• Acht Linearmotoren 7 oder Drehantriebe 43
• Acht Verstärker und Kraftregler 50 für die Linearmotoren 7 oder Drehantriebe 43
• Fünf Trägheitssensoren 11 (Beschleunigungs- oder Geschwindigkeitsaufnehmer)
• Fünf Spannungs/Stromwandler 51 für die Ausgänge der Trägheitssensoren 11
• Acht Positionssensoren 10

The active system for damping the cabin vibrations and for adjusting the cabin 1 in five degrees of freedom (x _k , φ _ky , y _k , φ _kx , φ _kz ) consists of the following elements:

• Eight linear motors 7 or rotary drives 43
• Eight amplifiers and force controllers 50 for the linear motors 7 or rotary drives 43
• Five inertial sensors 11 (acceleration or speed sensors)
• Five voltage / current converters 51 for the outputs of the inertial sensors 11
• Eight position sensors 10

In the case of a cheaper version of the active system, only three degrees of freedom of the cabin are regulated (x _k , y _k , φ z). Therefore, linear motors 7 and sensors 10, 11 are only mounted below. In this case, the computational effort is significantly lower, which enables the use of a slow and inexpensive real-time computer.

7 shows the controller part of the active system according to the device according to the invention. Since the distances between the sensors and an analog / digital converter unit 55 are relatively long, the measurement signals must be transmitted as current signals and not as voltage signals. The position sensors 10 already deliver their output signals as current. In contrast, the inertial sensors 11 deliver their outputs in the form of voltage signals. In this case, a voltage / current converter 51 is necessary for the output of each inertial sensor 11. Since the analog / digital converter 55 can only sample voltage signals, an analog signal processing unit 56 is needed on the part of the real-time computer 57, which has one channel for each measurement signal. A channel consists of a current / voltage converter 58, one Anti-aliasing low-pass filter 59, which is necessary for the scanning, and an ordinary voltage amplification 60 for adapting the signal range.

The core of the real-time computer 57 is the digital signal processor 61, which is responsible for all mathematical calculations. In order to be able to record the necessary measurements from the hardware, a multi-channel analog / digital converter unit 55 is required. A multichannel digital / analog converter unit 63 is used to output the force setpoints to the linear motors 7. The entire controller algorithm with all required programs is stored in an EEPROM 64. This program is supplied by a host computer 65 during the commissioning of the active system and is adapted to the cabin 1 to be controlled. After commissioning, the host computer 65 is uncoupled, the program stored on the EEPROM 64 remaining there until it is modified or replaced by the host computer 65 during the next calibration. A RAM 66 is used by the digital signal processor 61 as a memory for the intermediate values of the calculations. A data bus 67 is used for communication between the digital signal processor 61 and all of these components. A module in the form of a communication port 68, which is responsible for the connection to the host computer, is also connected to this data bus 67.

It is possible to divide the computing task between two digital signal processors 61 that are connected to the same data bus 67 if the task cannot be solved quickly enough by a single signal processor 61.

8 shows the block diagram for the entire system according to the device according to the invention. The real-time computer 57 is programmed in this application in such a way that it uses the controller algorithm at a specific frequency in real time calculated.

• Die Verarbeitung der Messungen aus den fünf Trägheitssensoren 11 auf dem Kabinenrahmen 4 in x_k- sowie in y_k-Richtung. Die gemessenen Signale werden in Spannungs/Strom-Wandler 51 umgewandelt und durch die analoge Signalverarbeitungseinheit 56 übertragen und von Analog/Digital-Wandlerkanälen 55 abgetastet.
• Die oben erwähnten Messungen liegen in den Koordinatensystemen der Trägheitssensoren 11 vor. Da die Regelung im Kabinenkoordinatensystem geschieht, müssen sie in dieses Koordinatensystem transformiert werden. Zu diesem Zweck benutzt der Algorithmus die lineare Transformation T_KT. Die Ausgänge dieser Transformation sind:
  • Die Translationsbeschleunigung (bzw. die Translationsgeschwindigkeit) der Kabine 1 in x_k-Richtung (ẍ_k, bzw. ẋ_k).
  • Die Drehbeschleunigung (bzw. die Drehgeschwindigkeit) der Kabine 1 um die y_k-Achse (φ̈_ky, bzw. φ̇_ky).
  • Die Translationsbeschleunigung (bzw. die Translationsgeschwindigkeit) der Kabine in y_k-Richtung (ÿ_k, bzw. ẏ_k).
  • Die Drehbeschleunigung (bzw. die Drehgeschwindigkeit) der Kabine 1 um die x_k-Achse (φ̈_kx, bzw. φ̇_kx).
  • Die Drehbeschleunigung (bzw. die Drehgeschwindigkeit) der Kabine 1 um die z_k-Achse (φ̈_kz, bzw. φ̇_kz).

The algorithm consists of the following steps, which do not necessarily have to be carried out in the order given:

• The processing of the measurements from the five inertial sensors 11 on the cabin frame 4 in the x _k and y _k directions. The measured signals are converted into voltage / current converters 51 and transmitted by the analog signal processing unit 56 and sampled by analog / digital converter channels 55.
• The above-mentioned measurements are in the coordinate systems of the inertial sensors 11. Since the regulation takes place in the cabin coordinate system, it must be transformed into this coordinate system. For this purpose the algorithm uses the linear transformation T _KT . The outputs of this transformation are:
  • The translation acceleration (or the translation speed) of the cabin 1 in the x _k direction (ẍ _k , or ẋ _k ).
  • The rotational acceleration (or the rotational speed) of the cabin 1 around the y _k axis (φ̈ _ky , or φ̇ _ky ).
  • The translation acceleration (or the translation speed) of the cabin in the y _k direction (ÿ _k , or ẏ _k ).
  • The rotational acceleration (or the rotational speed) of the cabin 1 around the x _k axis (φ̈ _kx , or φ̇ _kx ).
  • The rotational acceleration (or the rotational speed) of the cabin 1 around the z _k axis (φ̈ _kz , or φ̇ _kz ).

• Einen Kraftsollwert F^T _xs in x_k-Richtung.
• Einen Drehmomentsollwert M^T _ys um die y_k-Achse.
• Einen Kraftsollwert F^T _ys in y_k-Richtung.
• Einen Drehmomentsollwert M^T _xs um die x_k-Achse.
• Einen Drehmomentsollwert M^T _zs um die z_k-Achse.

The target value for each of these variables is zero. Therefore, the five signals are subtracted from zero before they are passed on to the robust multi-size controller I. This controller I reacts to the five signals simultaneously according to the concept described above and delivers the following signals at the output:

• A force setpoint F ^T _xs in the x _k direction.
• A torque setpoint M ^T _ys around the y _k axis.
• A force setpoint F ^T _ys in the y _k direction.
• A torque setpoint M ^T _xs around the x _k axis.
• A torque setpoint M ^T _zs around the z _k axis.

• Die Messungen aus den Positionssensoren 10 in x_k-Richtung und in y_k-Richtung. Die gemessenen Signale werden durch die analoge Signalverarbeitungseinheit 56 übertragen und von Analog/Digital-Wandlerkanälen 55 abgetastet. Da diese Messungen im Positionssensorkoordinatensystem vorliegen, müssen sie ins Kabinenkoordinatensystem transformiert werden. Dafür wird eine lineare Transformation T_KP benutzt. Diese Transformation liefert fünf Positionssignale als Ausgang. Um die Positionsfehlersignale zu erhalten, wird jedes dieser Signale von Null subtrahiert. Somit werden zwei translatorische (x^E _K und y^E _K) und drei rotatorische Positionsfehlersignale (φ^E _Kx, φ^E _Xy und φ^E _Kz) erhalten.
• Auf die fünf Positionsfehler reagiert ein robuster Mehrgrössenregler II gemäss dem oben erwähnten Konzept und liefert als Ausgang die folgenden Sollwerte zur Korrektur der Aufzugseinstellung:
  • Den Kraftsollwert F^P _xs für die Verschiebung in x_k-Richtung.
  • Den Drehmomentsollwert M^P _ys für die Drehung um die Yk-Achse.
  • Den Kraftsollwert F^P _ys für die Verschiebung in y_k-Richtung.
  • Den Drehmomentsollwert M^P _xs für die Drehung um die x_k-Achse.
  • Den Drehmomentsollwert M^P _zs für die Drehung um die z_k-Achse.
• Anhand der linearen Transformation T^P _AK werden die Sollwerte aus dem Regler II in das Aktuatorkoordinatensystem transformiert. Der Unterschied zwischen den beiden linearen Transformationen T^T _AK und T^P _AK liegt darin, dass die von der zweiten Transformation resultierenden Kraftsollwerte der Linearmotoren 6 in x_k-Richtung nur Druckkräfte auf die Schienen 3 verursachen. Das wird erreicht, indem ein einziger Aktuator in x_k-Richtung unten und ein einziger Aktuator in x_k-Richtung oben gleichzeitig vom Regler II betätigt werden. Somit wird sichergestellt, dass keine der vier Rollen 9 in x_k-Richtung den Kontakt mit den Führungsschienen 3 verliert. Im Fall der ersten Transformation ist dies nicht möglich, weil sie wesentlich weniger Kräfte verlangt als die zweite Transformation.
• Die entsprechenden Ausgänge aus den zwei Transformationen T^T _AK und T^P _AK werden zusammen addiert, um die Kraftsollwerte für jeden der acht Linearmotoren 7 zu berechnen.
• Die Kraftsollwerte werden von den Digital/Analog-Wandlerkanälen 63 zu analogen Signalen umgewandelt. Die resultierenden Signale treiben die entsprechenden Leistungsverstärker und Kraftregler 50, welche die Ströme der Linearmotoren 7 durch analoge Rückführungen regeln. Die Leistungsverstärker 50 sind pulsbreitenmoduliert. Der Kabinenrahmen 4 wird nun von den resultierenden Kräften so beeinflusst, dass die zwei Regelziele erreicht werden. Sollte der jeweilige Kraftsollwert (bei störungsfreier Fahrt) den Wert Null annehmen, so übt der zugeordnete Aktuator keine Kräfte aus.

Using a linear transformation T ^T _AK , the setpoints from controller I are transformed into the actuator coordinate systems.

• The measurements from the position sensors 10 in the x _k direction and in the y _k direction. The measured signals are transmitted by the analog signal processing unit 56 and sampled by analog / digital converter channels 55. Since these measurements are available in the position sensor coordinate system, they must be transformed into the cabin coordinate system. A linear transformation T _{KP is} used for this. This transformation provides five position signals as an output. To obtain the position error signals, each of these signals is subtracted from zero. Two translatory (x ^E _K and y ^E _K ) and three rotatory position _{error signals} (φ ^E _Kx , φ ^E _Xy and φ ^E _Kz ) are _thus obtained.
• A robust multi-size controller II reacts to the five position errors in accordance with the above-mentioned concept and provides the following setpoints as an output for correcting the elevator setting:
  • The force setpoint F ^P _xs for the displacement in the x _k direction.
  • The torque setpoint M ^P _ys for rotation about the Yk axis.
  • The force setpoint F ^P _ys for the displacement in the y _k direction.
  • The torque setpoint M ^P _xs for the rotation around the x _k axis.
  • The torque setpoint M ^P _zs for the rotation around the z _k axis.
• Using the linear transformation T ^P _AK , the setpoints from the controller II are transformed into the actuator coordinate system. The difference between the two linear transformations T ^T _AK and T ^P _AK is that the force setpoints of the linear motors 6 resulting from the second transformation in the x _k direction only cause pressure forces on the rails 3. This is achieved by simultaneously actuating a single actuator in the x _k direction at the bottom and a single actuator in the x _k direction at the top by the controller II. This ensures that none of the four rollers 9 loses contact with the guide rails 3 in the x _k direction. In the case of the first transformation, this is not possible because it requires far less force than the second transformation.
• The corresponding outputs from the two transformations T ^T _AK and T ^P _AK are added together in order to calculate the force setpoints for each of the eight linear motors 7.
• The force setpoints are converted by the digital / analog converter channels 63 to analog signals. The resulting signals drive the corresponding power amplifiers and force regulators 50, which regulate the currents of the linear motors 7 by means of analog feedback. The power amplifiers 50 are pulse width modulated. The cabin frame 4 is now influenced by the resulting forces so that the two control goals are achieved. If the respective force setpoint (with trouble-free travel) assumes the value zero, the assigned actuator does not exert any forces.

The execution of all linear transformations as well as the calculation of the controller algorithm is carried out by the digital signal processor 61 in every sampling period.

Claims (10)

Device for reducing vibrations of an elevator car (1) guided on rails (3), which has guide elements (21) movably connected to it between stops (38, 39), vibrations occurring at right angles to the direction of travel of inertia sensors (4) attached to the car frame (4) 11) measured and used to regulate at least one actuator (6) which is arranged between the cabin (1) and guide elements (21) and which operates in relation to the vibrations occurring and in the opposite direction to the vibrations
characterized,
that the actuator or actuators (6) are equipped with linear motors (7), the fixed motor part (16) being fixed to the frame of the cabin (1) and the movable motor part (17) being fastened to the guide elements (21). Device according to claim 1,
characterized,
that the movable motor part (17) is a magnet. Device according to claim 1 or 2,
characterized,
that a roller lever, to which the movable motor part (17) is attached, serves as the guide element (21). Device according to claim 1 or 2,
characterized,
that a roller lever is used as the guide element (21), which is connected to the movable motor part (17) via a push-pull member. Device according to one of claims 1 to 4,
characterized,
that a low-friction guide (35) maintains an air gap (34) between the fixed motor part (16) and the movable motor part (17). Device for reducing vibrations of an elevator car (1) guided on rails (3), which has guide elements (21) movably connected to it between two stops (38, 39), vibrations occurring at right angles to the direction of travel of inertia sensors attached to the car frame (4) (11) measured and used to regulate at least one actuator (6) which is arranged between the cabin (1) and guide elements (21) and which works in relation to the vibrations occurring and in the opposite direction to the vibrations,
characterized,
that the actuators (6) are equipped with rotary drives (43), the movable motor part being fastened to the guide elements (21) by means of a crank (44) and a tension-compression member (45). Device according to the preamble of claim 6,
characterized,
that the actuators (6) are equipped with rotary drives (43), the movable motor part being connected to the guide elements (21) via a cam disc (47) or via a flexible traction means (46). Method for reducing vibrations of an elevator car (1) guided on rails (3), which has guide elements (21) movably connected to it between two stops (38, 39), with vibrations of inertia sensors attached to the car frame (4) occurring transversely to the direction of travel (11) measured and used to regulate at least one actuator (6) which is arranged between the cabin (1) and guide elements (21) and which works in relation to the vibrations occurring and in the opposite direction to the vibrations,
characterized,
that the actuators (6) are equipped with regulated linear motors (7) or regulated rotary drives (43), the regulation being carried out by the outputs of two controllers combined into a force setpoint, an acceleration feedback system operating in the high frequency range and a position feedback system operating in the low frequency range. A method according to claim 8,
marked by,
a feedback to the guide elements (21), which reacts to the measured vibration paths or their time derivatives, in order to minimize the actual vibrations of the cabin (1) and is characterized in that for the guide elements (21) relative to their end positions (38, 39) Middle position (37) is defined and, in the event of noticeable deviations therefrom, they are returned to this position by the actuators (6) in the low frequency range. Method according to one of the preceding claims,
characterized,
that the acceleration feedback working in the high frequency range and the position feedback working in the low frequency range are realized by two control loops with one controller (I) and one controller (II) belong to a control part implemented as a computer program in a preferably digital signal processor (61).

Images