METHOD AND DEVICE FOR CONTROLLING/REGULATING A ROTATORY DRIVE OF A WORKING UNIT OF A TRACK MAINTENANCE MACHINE

Abstract

A method for controlling/regulating a rotatory drive of a working unit of a track maintenance machine includes using a sensor to record a measurement parameter with an approximately periodic history function derived from a rotation of the drive. A frequency or period duration of the history function is determined with an evaluation device, and the frequency or period duration is compared with a target parameter for stipulating a control signal. A number of time-discrete measurement values is formed for the measurement parameter and an auto-correlation of these measurement values is carried out by a calculation unit for determining the frequency or the period duration. As compared to a conventional method with zero stage recording, a precise recording of frequency changes is therefore also possible between two zero stages. A device for carrying out the method is also provided.

Claims

1-15. (canceled)

16. A method for controlling/regulating a rotatory drive of a working unit of a track maintenance machine, the method comprising: using a sensor to record a measurement parameter (X) with an approximately periodic history function derived from a rotation of the drive; determining a frequency (f) or a period duration (T) of the history function by using an evaluation device; comparing the frequency (f) or the period duration (T) with a target parameter for stipulating a control signal; forming a number of time-discrete measurement values (xi) for the measurement parameter (X); and carrying out an auto-correlation of the measurement values (xi) by using a calculation unit for determining the frequency (f) or the period duration (T).

17. The method according to claim 16, which further comprises: using a further sensor to record a further measurement parameter (Y) with an approximately periodic history function derived from a rotation of a further drive; forming a further number of time-discrete measurement values (yi) for the further measurement parameter (Y); and using the calculation unit to carry out a cross-correlation of the measurement values (xi, yi) of both measurement parameters (X, Y) for determining a phase offset (φ).

18. The method according to claim 16, which further comprises stipulating a cycle time for forming the time-discrete measurement values (xi), and using the cycle time to determine an evaluation period.

19. The method according to claim 16, which further comprises providing an iterative calculation of function values (Ψi) of a correlation function by using a summation of a consistently identical number of measurement value products with every new measurement value (xi).

20. The method according to claim 19, which further comprises subtracting a measurement value product with an oldest measurement value from a sum of measurement value products of a preceding calculation for a current calculation of a function value (Ψi), and adding a new measurement value product to the current measurement value.

21. The method according to claim 16, which further comprises calculating interpolated measurement values prior to a calculation of function values (Ψi) of a correlation function.

22. The method according to claim 16, which further comprises filtering the measurement values (xi) prior to a calculation of function values (Ψi) of a correlation function.

23. The method according to claim 16, which further comprises calculating interpolated function values (Ψi) prior to a determination of a maximum of a correlation function.

24. The method according to claim 16, which further comprises recording a distance from a sensitive element of the sensor to an eccentric shaft driven by the drive, as a measurement parameter (X).

25. The method according to claim 16, which further comprises recording a field strength of a magnetic field rotated with the drive and acting on the sensor, as a measurement parameter (X).

26. A device for carrying out a method for controlling/regulating a rotatory drive of a working unit of a track maintenance machine according to claim 16, the device comprising: a sensor for recording a measurement parameter (X) with an approximately periodic history function derived from a rotation of a drive; an evaluation device for determining a frequency (f) or a period duration (T) of the history function; a controller for regulated control of the drive; a calculation unit supplied with time-discrete measurement values (xi) of the measurement parameter (X), said calculation unit containing an algorithm for carrying out an auto-correlation of the measurement values (xi) for determining the frequency (f) or the period duration (T).

27. The device according to claim 26, which further comprises: a further sensor for recording a further measurement parameter (Y) derived from a rotation of a further drive; said calculation unit being supplied with time-discrete measurement values (yi) of the further measurement parameter (Y); and said calculation unit containing an algorithm for carrying out a cross-correlation of the measurement values (xi, yi) of both measurement parameters for determining a phase offset (φ).

28. The device according to claim 26, wherein said calculation unit is a microprocessor.

29. The device according to claim 26, which further comprises: a first assembly including said calculation unit and a communication interface; and a dedicated assembly for the drive, said dedicated assembly coupled with said first assembly, and said dedicated assembly including power electronics, a regulator unit and a communication interface.

30. The device according to claim 26, wherein said sensor includes a sensitive element for a capacitive or inductive or magnetic recording of the measurement parameter (X).

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The invention will now be explained by way of example with reference to the enclosed Figures. These show in schematic illustrations:

[0025] FIG. 1 track maintenance machine with tamping unit

[0026] FIG. 2 track maintenance machine with stabilisation unit

[0027] FIG. 3 tamping unit in a side view

[0028] FIG. 4 cross-section through track with tamping machine

[0029] FIG. 5 sensor with distance measurement

[0030] FIG. 6 sensor with field strength measurement

[0031] FIG. 7 measurement values

[0032] FIG. 8 correlation function

[0033] FIG. 9 forming of function terms

[0034] FIG. 10 system illustration

[0035] FIG. 11 system illustration with assemblies

[0036] FIG. 12 signal processing diagram

DESCRIPTION OF THE EMBODIMENTS

[0037] The track maintenance machine 1 illustrated in FIG. 1 is a tamping machine and comprises a machine frame 3 that is displaceable on on-track undercarriages 2. A tamping unit is arranged on the machine frame 3 as a working unit 4. The tamping machine serves for processing a track 5, where rails 7 fitted to sleepers 6 are positioned in a ballast bed 8. During the tamping process the track panel made of sleepers 6 and rails 7 is lifted into a target position by means of a lifting/lining unit 9 and a measuring system 10 and displaced sideways if required. Vibrating tamping tools 11 of the working unit 4 are submersed into the ballast bed 8 for fixing this position. The submersed tamping tools 11 are squeezed towards each other and compress the ballast below the lifted sleepers 6 here.

[0038] The tamping tools 11 are coupled with a vibration generator 12 for generating vibration. The vibration generator 12 comprises a rotatory drive 13, which drives an eccentric shaft 14. Squeezing drives 15 are mounted on the eccentric shaft 14. When the eccentric shaft 14 rotates, its eccentricity generates the desired vibration amplitude.

[0039] After a tamping process, the track 5 is normally stabilised to pre-empt settlement. The track maintenance machine 1 illustrated in FIG. 2 serves for this. This is a dynamic track stabiliser (DGS) with two stabilisation units as working units 4 here. Each stabilisation unit comprises a vibration generator 12 with rotating imbalances, which are driven by means of a rotatory drive 13. The active vibration generator 12 starts a vibration of the vibration unit transversely to the longitudinal track direction. The stabilisation unit straddles the rails 7 of the track 5 with roller tongs 16 here, which transmits the vibration to the track panel. This effects a shaking of the track panel into the ballast bed 8.

[0040] For tamping as well as for stabilisation as well as for further vibration generators 12 used in track construction, the vibration generated should fulfil different requirements. A vibration frequency of 35 Hz is for example stipulated for optimal ballast compression. A higher frequency f of approx. 45 Hz is desired for a submersion process of the tamping tools 11 in order to reduce submersion resistance. Outside of the ballast bed 8, the frequency f should be lower to reduce noise pollution.

[0041] Further requirements are explained with reference to the working unit 4 in FIGS. 3 and 4. The working unit 4 comprises four tamping devices 17 which can be lowered into the ballast bed 8 separately from each other by means of a height adjustment drive. Opposing tamping tools 11 are coupled with their own vibration generator 12 via squeezing drives 15 on each tamping device 17. The vibration generators 12 are controlled via a common controller 18. Phase offsets φ are stipulated for the generated vibration in relation to each other in addition to the frequency f. A mirror-inverted synchronisation of the tamping devices is normally desired to minimise the reactive vibration acting on the machine frame 3 as well as noise development.

[0042] It is necessary for controlling/regulating a rotatory drive 13 in a vibration generator 12 to continuously record the frequencies f or the period duration T. High requirements also exist during track maintenance with regard to the robustness of a sensor arrangement. The evaluation of zero stages is known to have the disadvantage that frequency changes may be recognised late with the same. It is therefore envisaged according to the invention to determine the frequency f or the period duration T by means of auto-correlation. A measurement parameter X, which represents an approximately periodic function development of the generated vibration, remains the basis for this.

[0043] To that end, a sensor 19 that is magnetically, inductively or capacitively coupled with the vibration generator 12 is arranged. The distance sensor 19 illustrated in FIG. 5 for example comprises a sensitive element, which measures a distance to an eccentric sleeve surface of an eccentric shaft 14 via inductive coupling. This distance, which changes when the eccentric shaft 14 rotates, is the measurement parameter X, the approximately periodic development of which is subsequently evaluated.

[0044] Alternatively to this a magnetic component 20 of a vibration generator 12 is arranged in FIG. 6, which is rotated by means of the rotatory drive 13. This creates a rotating magnetic field, which is recorded by means of a fixed sensor 19. With this variant the field strength changing during a rotation is recorded and evaluated as measurement parameter X.

[0045] FIG. 7 shows an exemplary processing of the measurement parameter X (or Y for a further vibration generator 12). Measurement parameter development over time t is illustrated in the upper diagram. This is an approximately periodic history function, wherein interference may occur as a consequence of external influences. The aim of the following method steps is the determination of the period duration T or frequency f=1/T. Measurement values x.sub.i (or y.sub.i for a further vibration generator 12) are recorded in the second diagram for measurement parameter X, wherein the time interval between the measurement values x.sub.i is stipulated by a cycle time. For this, information is either requested from the sensor 19 in cycles, or a time-discrete number of measurement values x.sub.i is formed from an analogue sensor signal by means of an analogue digital transducer (serial values with index i).

[0046] Incorrect measurement values x.sub.f are advantageously recognised and eliminated by means of a digital filter. It is also expedient to improve the sensor signal by means of an IIR filter, for example of the fourth order. A high pass of the second order removes steady components here, whilst a low pass of the second order dampens high-frequency interference of the signal.

[0047] An interpolation of the measurement value development takes place during a next step to obtain an improved data basis for forming the auto-correlation function Ψ.sub.xx(i). An additional value is for example interpolated between each recorded measurement value x.sub.i (3.sup.rd diagram in FIG. 7).

[0048] An example of the history function values Ψ.sub.i of the auto-correlation function Ψ.sub.xx(i) is illustrated in FIG. 8. The values Ψ.sub.i of the auto-correlation function Ψ.sub.xx(i) result from a totalling of measurement value products x.sub.n-x.sub.n-i as follows:

[00001]${\Psi }_{\mathrm{xx}}\left(i\right)=\underset{n}{.Math.}{x}_n.Math.x_{n-i}$

[0049] It is expedient for the present invention if a consistently identical number of measurement value products is totalled for each function value

[0050] An evaluation of frequency f or period duration T by means of a new calculation of the auto-correlation function Ψ.sub.xx(i) conveniently takes place with every cycle stipulated by the cycle time. For this the measurement value product with the oldest measurement value is subtracted from the sum of measurement value products of the preceding calculation during a current calculation of a function value, and a new measurement value product with the current measurement value is added:

Σ.sub.(i,t+1)=Σ.sub.(i,t)−x.sub.(2n-1,t).Math.x.sub.(2n,t)+x.sub.(0,t).Math.x.sub.(i,t)

[0051] The corresponding summation is demonstrated in FIG. 9. This simplified iterative approach allows the continuous calculation of the auto-correlation function Ψ.sub.xx(i) with a limited calculation effort almost in real time.

[0052] Further function values are interpolated from the development of the function values Ψ.sub.i during an optimisation step illustrated in the upper diagram of FIG. 8. The result is clear from the lower diagram. As the period duration T to be calculated is determined by the position of the first maximum (with the exception of zero), an interpolation is expediently carried out only for this range. This expected range is generally known from the frequencies stipulated in track construction.

[0053] A calculation unit 22 for carrying out the method steps stated is arranged in an evaluation means 21. The measurement values x.sub.i of several vibration generators 12 of a controlled working unit 4 are supplied to the calculation unit 22 in the system illustration in FIG. 10 here. The individual drives 13 of the working unit 4 are controlled via associated performance levels 23 on the output side.

[0054] This system construction is illustrated in more detail in FIG. 11. A first assembly 24 comprises the calculation unit 22, an analogue-digital converter 25, a pre-processing unit 26 and communication interfaces 27. Measurement signals of the sensors 19 are prepared for the calculation unit 22 via the pre-processing unit 26 and the analogue-digital converter 25. Time-discrete measurement value series are in fact formed, which represent the respective periodic measurement value development.

[0055] The calculation unit 22 is coupled with a configuration and diagnosis unit 28 and a controller 29 via the communication interfaces 27 for the stipulation of general control commands. Dedicated assemblies 30 are also envisaged for each vibration generator 12, which are coupled with the calculation unit 22 via communication interfaces 27. Each one of these assemblies 30 here comprises a regulator unit 31 and power electronics 32 for controlling the associated drive 13.

[0056] An exemplary processing of the measurement signals or measurement values x.sub.i is illustrated for four controlled/regulated vibration generators 12 of a working unit 4 in FIG. 12. A sensor 19 for generating a periodic development for the corresponding measurement parameter X is allocated to each vibration generator 12. Measurement value series are formed from this with respective filtering 33 and a scanning rate conversion 34.

[0057] The auto-correlation function Ψ.sub.xx(i) is formed continuously from the measurement value series of a vibration generator 12 for determining the respective current frequency f. A frequency determination 36 by means of an extreme value determination 35 results from this.

[0058] A determination of the associated phase offset φ takes place three times parallel to this for two vibration generators 12 each. For this a cross-correlation is first formed from the two measurement value series. The following cross-correlation function results from the measurement values x.sub.i of the one vibration generator 12 and the measurement values y.sub.i of the other vibration generator 12:

[00002]${\Psi }_{xy}\left(i\right)=\underset{n}{.Math.}{x}_n.Math.y_{n-i}$

[0059] The corresponding phase offset determination 37 is realised by means of an extreme value determination 35 from the respective continuously formed development of the cross-correlation Ψ.sub.xy(i).