EA043332B1 - METHOD AND DEVICE FOR CONTROL/REGULATION OF A ROTATING DRIVE OF A WORKING UNIT OF A TRACK MACHINE - Google Patents

Description

Field of technology

The invention relates to a method for controlling/regulating a rotating drive of a working unit of a track machine, while using a sensor, the measured parameter created by the rotation of the drive is recorded together with an almost periodic characteristic function, while using a processing device, the frequency or duration of the period of the characteristic function is determined, and at the same time the frequency or the duration of the period is compared with a given parameter to generate a control signal. The invention also concerns a device for performing the method.

State of the art

From the publication by Auer F. et al.: High - Tech-Stopfaggregate fur nachhaltige Gleislageverbesserung, ElEisenbahningenieur, November 2015, pp. 18-22, a method for regulating the rotation speed of a rotating drive of a sleeper tamping unit is known. An eccentric shaft is driven to produce vibrations, which are transmitted via an auxiliary drive to the sleeper tamper.

This makes it possible to purposefully change the vibration frequency during the sleeper tamping cycle. A high frequency is set (42-59 hertz) during the process of immersing the tamping sleeper into the crushed stone bed. During the auxiliary movement process of the tamper, the optimal frequency of 35 hertz is used. In the lifting state, a reduced idle speed (approximately 28 hertz) is established, at which the unit rotates in the idle position.

In a track machine with speed control, phase stabilization is also used. In this case, during idling, the speed of all vibration generators is synchronized and the corresponding phase shift of the rotating drives is established in such a way that superimposition of vibrations is minimized.

Rotating vibration drives are also used in other working units of track machines. For example, WO 2008/009314 A1 describes a so-called track stabilizer, in which stabilizing units with rotating balancers are subjected to vibration. In this case, two synchronized stabilizing units with an adjustable vibration frequency are driven.

From Hauke R. et. Al: Bettungsreinigungsmaschinen - ein ϋberblick, El-Spezial Gleisbaumaschinen und - gerate, May 2016, pp. 30-35 machines for cleaning crushed stone beds with various screening systems are known. This also uses rotary drives with adjustable vibration frequency.

Brief description of the invention

The basis of the claimed invention is to improve a method of the type indicated above so that it is possible to perform an accurate determination of the frequency or duration of a period when changes are quickly recorded. It is further an object of the claimed invention to propose a suitable device for carrying out the improved method.

In accordance with the claimed invention, these problems are solved thanks to the features of independent claims 1 and 11 of the claims. Preferred embodiments of the invention are described in the dependent claims.

In this case, a series of time-discrete measured values is formed for the measured parameter, and with the help of a computing device, autocorrelation of these measured values is performed to determine the frequency or duration of the period. Compared to the known zero-crossing method, it is therefore possible to accurately record frequency changes between two zero-crossings. With recorded time-discrete measured values, the values of the autocorrelation function can be determined at any time. The result of the corresponding calculation of the function is the magnitude of the function on the time axis. On the time axis, the time interval between zero and the first maximum of the duration of the period of the function of the characteristic of the measured parameter is obtained. In this way, a new frequency determination can be carried out immediately with each new measured value recorded.

In another embodiment of the invention, another measured parameter is recorded using another sensor, resulting from the rotation of another drive, with an approximately periodic characteristic function, while for another measured parameter another series of time-discrete measured values is formed, and at the same time, using a computing device, cross-correlation of the measured values of both measured parameters to determine the phase displacement. As a result of the cross-correlation carried out at any time, the specified phase shifts can be determined immediately. This ensures precise synchronization of multiple rotating drives (phase stabilization).

Preferably, the duration of the cycle is set to form discrete measured values, wherein the duration of the cycle determines the processing period. Thus, with each new recorded measured value, the frequency or period duration and in this case the phase displacement are processed. The accuracy of the method increases with a reduced cycle time.

- 1 043332

In another improved embodiment of the invention, it is provided that with each new measured value the value of the correlation function is recalculated by forming a sum of a conserved number of products of the measured values. Thus, the number of computational operations is limited and the opportunity for subsequent simplifications is created.

In this case, it turns out to be especially appropriate for the actual calculation of the value of the function to subtract from the sum of the products of the measured values of the previous calculation the product of the measured values with the oldest measured value and add the new product of the measured values with the actual measured value. Thus, only minor computational operations are necessary to update the values of the correlation function. Thanks to the low computational requirements, calculations can be carried out almost in real time using a cost-effective and small computing device.

In another improved embodiment of the invention, it is provided that interpolated measured values are determined before calculating the value of the correlation function. By interpolating additional measured values between specified repeat values, the position of the maximum of the correlation function can be more accurately determined. In this way, a more precise determination of the frequency or duration of the period can be made.

The quality of frequency determination is improved when the measured values are filtered before calculating the value of the correlation function. For example, the measurement signal is prepared using a so-called pulse response filter (4th order IIR bandpass filter). In this case, the high-pass filter limits the DC component of the signal, while the low-pass filter dampens and removes high-frequency noise in the signal. The IIR filter also has the advantage that, in contrast to other digital types of filters (for example, the FIR filter), it has significantly fewer computational operations. This property of the filter is very useful in this case, because the requirements for the amount of computational work of the computing device remain limited.

An improvement of the method is also achieved due to the fact that before determining the maximum of the correlation function, the interpolated values of the function are determined. It is more appropriate to interpolate only in the area of extreme values in order to be able to more accurately determine the position of extreme values. Increased accuracy can be achieved as a result with less computational effort.

In a preferred embodiment of the invention, the distance of the sensing element of the sensor to the eccentric shaft driven by the drive is recorded as a measured parameter. Due to the margin of error in frequency determination using the correlation function, there is no need to perform precise sensor mounting or calibration procedures. If the drive is driven by an eccentric shaft made of ferromagnetic material, then an inductive distance sensor is useful without additional adjustment of the rotating elements.

In an alternative embodiment of the invention, it is provided that the force of the magnetic field acting on the sensor, moving during the rotation of the drive, is recorded as a measured parameter. This magnetizes the drive shaft or adds a magnet to produce a rotating magnetic field. The need for additional space when placing a magnet turns out to be insignificant. Using an appropriate sensor located close to the shaft, the changing strength of the magnetic field is recorded as the shaft rotates.

In the case of the claimed device for performing the described method, this device includes a sensor for recording a measured parameter occurring during rotation of the drive with an approximately periodic characteristic function, a processing device for determining the frequency or duration of the period of the characteristic function, and a design group for controlled activation of the drive. In this case, time-discrete measured values of the measured parameters are sent to the computing device, and an algorithm is installed in the computing device to perform autocorrelation of these measured values to determine the frequency or duration of the period. Thus, with each new measured value recorded, an accurate frequency determination can be carried out immediately.

In an improved embodiment of the device, another sensor is installed to register another measured parameter resulting from the rotation of another drive, while discrete time measured values of other measured parameters are sent to the computing device, and at the same time an algorithm is installed in the computing device to perform cross-correlation of the measured values both measured parameters to determine the phase displacement. This results in a simple design for phase stabilization.

To implement a device with fewer and more compact components, it is advisable if the computing device is a microprocessor. Wherein

- 2 043332 optimal algorithms and an efficient signal processing method are used to match the volume of calculations with the properties of the microprocessor along with the limited working storage.

The computing device with a communication interface is advantageously located in a first structure group, which includes power electronics, a control unit and a communication interface that is connected to the first structure group. These groups allow you to linearly change the device. Thus, a computing device can be used to determine the frequency and period duration of multiple drives.

For effective recording of the measured parameters, it is preferable if the sensor includes a sensing element for capacitive or inductive or magnetic recording of the measured parameters. There are no special requirements for mounting the sensor, because neither the offset nor the exact amplitude of the sensor signal is significant for further processing.

Brief description of drawings

The claimed invention is explained below in more detail using examples of its implementation with reference to the attached figures. The figures schematically depict:

in fig. 1 shows a track machine with a sleeper tamping unit;

in fig. 2 shows a track machine with a stabilizing unit;

in fig. Figure 3 shows a sleeper tamping unit in a side view;

in fig. 4 shows a cross section along a rail track with a track machine;

in fig. 5 shows a sensor with distance measurement;

in fig. 6 shows a sensor with field strength measurement;

in fig. 7 shows the measured values;

in fig. 8 shows the correlation function;

in fig. Figure 9 shows the formation of functional terms;

in fig. 10 shows the system;

in fig. 11 shows a system with structural groups;

in fig. Figure 12 shows a signal processing diagram.

Description of embodiments of the invention

In fig. 1 shows a track machine 1, which is a sleeper tamping machine and includes a machine frame 3 moving on rail running gears 2. On the machine frame 3 a sleeper tamping unit 4 is located as a working unit. The sleeper tamping machine is designed for processing a rail track 5, in which rails 7 fixed to sleepers 6 are located on a crushed stone bed 8. During the tamping process, the railway grid formed by sleepers 6 and rails 7 is raised to a given position using a lifting unit 9 and a measuring system 10 and in this case it moves to the side. To fix this position, vibrating sleeper tamping tools 11 of the working unit 4 are immersed in the crushed stone bed 8. The immersed sleeper tamping tools 11 move towards each other and at the same time compact the crushed stone under the raised sleepers 6.

To create vibrations, the tamping tools 11 are connected to a vibration generator 12. The vibration generator 12 includes a rotating drive 13, which drives an eccentric shaft 14. Auxiliary drives 15 are located on the eccentric shaft 14. When the eccentric shaft 14 rotates, the desired vibration amplitudes are produced due to their eccentricity.

After the tamping process, the rail track 5 is generally stabilized to eliminate subsidence. For this purpose, the one shown in Fig. 2nd track machine 1. This is a dynamic track stabilizer (DGS) with two stabilizing units as working units. Each stabilizing unit includes a vibration generator 12 with rotating imbalances, which are driven by a rotating drive 13. The active vibration generator 12 vibrates the stabilizing unit in a transverse direction relative to the longitudinal direction of the rail track. At the same time, the stabilizing unit is covered by roller grips 16 on rails 7 of the rail track 5, as a result of which vibrations are transmitted to the railway grid. This causes the railway grid to sink into the crushed stone bed 8.

Both during tamping and during stabilization, as well as during subsequent applications of vibration generators 12 during construction work on the rail track, the vibration generated must correspond to the specified characteristics. For example, for the optimal process of crushed stone compaction, the vibration frequency is set to 35 hertz. For the driving process of the tamping tools 11, a higher frequency f, approximately 45 hertz, is desirable in order to reduce the driving resistance. Outside the crushed stone bed 8 the frequency f should be lower to reduce the noise load.

Other conditions regarding the work unit 4 are explained with reference to FIG. 3 and 4. The working unit 4 includes four sleeper tamping blocks 17, which are lowered separately from each other into the crushed stone bed 8 using a height-adjustable drive. In each sleeper tamping block 17, the sleeper tamping tools 11 located opposite each other are connected through auxiliary drives 15 with their own vibration generator 12. Vibration generators 12 are controlled using a common control device 18. For the vibrations produced, the displacements occurring among themselves are set along with the frequency f phases φ. As a rule, it is desirable to have the same synchronization of the tamping blocks in order to reduce the backlash of vibrations affecting the track machine 1, as well as the noise load.

To control/regulate the rotating drive 13 in the vibration generator 12, it turns out to be necessary to constantly record the frequency f or the duration of the period T. Moreover, when carrying out construction work on a rail track, there are high requirements for the reliability of the design of the sensor device. Processing when passing zero points is known, with the disadvantage that frequency changes are thereby detected with a delay. Therefore, according to the claimed invention, it is possible to determine the frequency f or the duration of the period T using autocorrelation. The basis for this is created by the measured parameter X, which is an almost periodic function of the characteristic of the vibration produced.

To do this, a sensor 19 is placed, which is connected to the vibration generator 12 magnetically, inductively or capacitively. For example, includes the one shown in FIG. 5 distance sensor 19 is a sensitive element which, thanks to inductive coupling, measures the distance to the eccentric surface of the body of the eccentric shaft 14. This distance, which changes when the eccentric shaft 14 rotates, represents the parameter X and its approximately periodic characteristic is processed subsequently.

Alternatively, according to FIG. 6, a magnetic component 20 of the vibration generator 12 is installed, which rotates using a rotating drive 13. This creates a rotating magnetic field, which is recorded using a stationary sensor 19. With this option, the field strength changing during rotation is recorded as a measured parameter X and processed .

In fig. 7 shows, as an example, the processing of the measured parameter X (for example, Y for another vibration generator 12). The top diagram shows the characteristic curve of the measured parameter over time t. In this case, we are talking about an approximately periodic function of the characteristic, and interference may occur due to external influences. The purpose of the subsequent stage of the method is to determine the duration of the period T or the frequency f = 1/T. For the measured parameter X, the measured values xi (or yi for another vibration generator 12) are recorded on the second diagram, and the time interval between the measured values x _i is set through the duration of the cycle. For this purpose, either in accordance with the clock cycle, the request of the sensor 19 or, based on the analog signal of the sensor, a discrete time series of measured values xi (series values with index i) is formed using an analog-to-digital converter.

Preferably, erroneous measured values xi are identified using a digital filter and removed. In this case, it turns out to be advisable to improve the sensor signal using a filter - IIR, for example, of the fourth order. In this case, the second-order high-pass filter removes DC components and the second-order low-pass filter dampens high-frequency signal noise.

In the next step, the characteristics of the measured quantities are interpolated to obtain an improved database for the formation of the autocorrelation function ψ _xx (i). For example, between each recorded measured value xi an additional value is interpolated (third diagram in Fig. 7).

An exemplary characteristic curve of the magnitudes of the function ψ _i of the autocorrelation function ψ _xx (i) is shown in FIG. 8. The values ψ _i of the autocorrelation function ψ _xx (i) are obtained as arising from the resulting sum of products of the measured quantities

HP ^x HP-1:

Ψ _χχ (ΐ) = Σ _η Χ _η χΧη-1

For the claimed invention, it turns out to be appropriate if for each value of the function ψ _i the remaining equal number of products of the measured values is summed up.

Advantageously, at each clock time defined by the clock duration, the processing of the frequency f or period duration T is carried out by a new calculation of the autocorrelation function ψ _xx (i). In this case, during the current calculation of the value of the function, the product of the measured values with the oldest measured value is subtracted from the sum of the product of the measured values of the previous calculation and the new product of the measured values with the current measured value is added:

Σ (i, t + 1) - Σ (i, _t ) - X (2n - i, t) ^X X (2n,t) + X (0,t) ^x X(i,x)

The resulting corresponding amount is shown in FIG. 9. Thanks to a simple iterative procedure, continuous computation of the autocorrelation function ψ _xx (i) can be performed with a limited amount of computation in almost real time.

Based on those shown in the top diagram in FIG. 8, characteristics of quantities

-

Claims (4)

functions ψ _i are interpolated at the optimization stage by other values of the function. The result is shown in the bottom diagram. Since the resulting duration of period T is determined by the position of the first maximum (except at zero), it is more appropriate to interpolate only in this range. This expected range is known, as a rule, due to the specified frequencies during construction work on the railway track. To carry out the specified method steps, a computing device 22 is located in the processing device 21. In this case, in the device shown in FIG. 10, the system is supplied to the computing device 22 by the measured values xi of several vibration generators 12 of the controlled working unit 4. At the output, the individual drives 13 of the working unit 4 are controlled through the corresponding power stages 23. This more detailed system design is shown in FIG. 11. The first structural group 24 includes a computing device 22, an analog-to-digital converter 25, a pre-processing unit 26 and a communication interface 27. The measurement signals of the sensor 19 are processed using a pre-processing unit 26 and an analog-to-digital converter 25 for the computing device 22. Specifically, discrete time series of measured quantities are created, which represent the corresponding periodic characteristics of the measured quantities. Through a communication interface 27, the computing device 22 is connected to the configuration and diagnostic device 28 and the control unit 29 to provide general control orders. Each vibration generator 12 is provided with its own construction group 30, which is connected via a communication interface 27 to the computing device 22. Each construction group 30 includes a control unit 31 and power electronics 32 for controlling the corresponding drive 13. For controlled/adjustable vibration generators 12 of the working unit 4 is shown in FIG. 12 is an example of processing a measuring signal or measuring variable xi. For each vibration generator 12, a sensor 19 is designed to create a periodic characteristic for the corresponding measured parameter X. Based on this, series of measured values are formed using appropriate filtering 33 and converting the sampling frequency 34. To determine the corresponding actual frequency f, a constant autocorrelation function ψ _χχ (ι) is formed from the series of measured values of the vibration generator 12. By determining the maximum value 35, a frequency determination 36 is then obtained from this. For the corresponding two vibration generators 12, the corresponding phase shift φ should be determined in parallel three times. To do this, a cross-correlation is first formed based on both series of measured values. Using the measured values x _i of the vibration generator 12 and the measured values yi of the other vibration generator 12, the following cross-correlation function is obtained: Based on the continuously corresponding curve of the cross-correlation function ψ _xy (i), the corresponding definition of the phase shift 37 is obtained by defining the extreme value 35. CLAIM 1. A method for controlling/regulating the rotating drive (13) of the working unit (4) of the track machine (1), while the measured parameter (X) created as a result of rotation of the drive is recorded using a sensor (19), together with an approximately periodic function of the characteristic, while using a processing device (21) determine the function (f) or duration of the period (T) of the characteristic function and compare the frequency (f) or duration of the period (T) with a given parameter to supply a control signal, characterized in that it is created for the measured parameter (X) a series of time-discrete measured values (x _i ) and that, using a computing device (22), autocorrelation of these measured values (x _i ) is performed to determine the frequency (f) or the duration of the period (T). 2. The method according to claim 1, characterized in that with the help of another sensor (19) another measured parameter (U) obtained as a result of rotation of another drive (13) is recorded with an approximately periodic function of the characteristic that is created for the other measured parameter (U) another discrete time series of measured values (yi) and that, using a computing device (22), a cross-correlation of the measured values (xi, yi) of both measured parameters (X, Y) is performed to determine the phase displacement (φ). 3. The method according to claim 1 or 2, characterized in that for the formation of discrete time measured values (xi) a takt duration is set, and that the takt duration determines the processing period. 4. The method according to one of claims 1-3, characterized in that with each new measured value (xi) an iterative calculation of the values of the correlation function (ψ _i ) is performed by forming a preserving -