Method for the detection of errors in pump units

• US 8,401,806 B2
• Filed: 02/23/2008
• Issued: 03/19/2013
• Est. Priority Date: 03/23/2007
• Status: Active Grant

• Current Assignee: Grundfos Management A/S (POUL DUE Jensen's FOND)
• Original Assignee: Grundfos Management A/S (POUL DUE Jensen's FOND)
• Inventors: Borsting, Hakon, Munk, Flemming
• Primary Examiner(s): Raymond, Edward
• Application Number:
  US12/532,284
• Publication Number:
  US 20100082275A1
• Time in Patent Office: 1,851 Days
• Field of Search: 702/56, 702/35, 702/34, 702/39, 702/185, 702/145
• US Class Current: 702/56
• CPC Class Codes: F04D 15/0066 by changing the speed, e.g. of the driving engine
• CPC Class Codes: F04D 15/0088 Testing machines
• CPC Class Codes: F04D 29/669 especially adapted for liquid pumps F04D29/18 takes precedence
• CPC Class Codes: F05D 2270/709 with neural networks

Abstract

A method is provided for detecting faults in a pump assembly having an electric motor, or faults in an electric motor, having at least one rotating shaft. A vibration signal is detected, the detected vibration signal is processed in a manner such that the influence of the current rotational speed of the shaft is eliminated, periodic signals are filtered out of the processed vibration signal, and the vibrational operating condition, in particular possible faults, is recognized by way of the periodic signals. Further, a pump assembly having an electric motor or an electric motor is provided in which this method is applied.

Claims

1. A method for detecting faults in an electric motor, wherein the electric motor has at least one rotating shaft, the method comprising:

detecting a vibration signal in the electric motor, processing the detected vibration signal in a manner such that influence of a current rotational speed of the at least one rotating shaft is eliminated by multiplying a sampling frequency value by a current rotational speed value and dividing the result by a constant rotational speed value,

filtering out periodic signals of the processed vibration signal, and

recognizing a vibrational operating condition including any possible faults by way of the periodic signals.

2. The method according to claim 1, wherein the processed vibration signal is subjected to a Cepstral analysis and the vibrational operating condition is recognized by way of a Cepstral diagram produced by the Cepstral analysis.

3. The method according to claim 2, wherein a pattern recognition is carried out for recognizing the vibrational operating condition from the Cepstral diagram, wherein the pattern recognition particularly takes into account at which locations swings occur in the Cepstral diagram.

4. The method according to claim 2, wherein the vibrational operating condition is recognized by way of previously known patterns in the Cepstral diagram.

5. The method according to claim 2, wherein the recognition of the vibrational operating condition is effected in predefined sections of the Cepstral diagram.

6. The method according to claim 2, wherein the recognition of the vibrational operating condition is effected by way of the Cepstral diagram and by way of at least one of a neural network and fuzzy logic.

7. The method according to claim 2, wherein after the Cepstral analysis and before the recognition of the vibrational operating condition, a step of normalizing the Cepstral diagram takes place in a manner such that swings in the diagram are set in relation to background noises.

8. An electric motor, wherein the electric motor has at least one rotating shaft and a fault detection system, the fault detection system comprising:

at least one vibration sensor,

an evaluation device connected to the at least one vibration sensor, wherein the evaluation device is provided with a processing module configured to eliminate influence of a current rotational speed of the at least one rotating shaft from a vibration signal captured by the at least one vibration sensor by multiplying a sampling frequency value by a current rotational speed value and dividing the result by a constant rotational speed value,

a filter module configured for filtering periodic signals out of the vibration signal processed by the processing module, and

a recognition module configured for recognizing a vibrational operating condition by way of the periodic signals.

9. The electric motor according to claim 8, wherein:

the filter module comprises a Cepstral analysis module that is configured for carrying out a Cepstral analysis on the vibration signal processed by the processing module, and

the recognition module is configured for recognition of the vibrational operating condition by way of a Cepstral diagram produced by the Cepstral analysis module.

10. The electric motor according to claim 8, wherein the recognition module comprises at least one of a fuzzy logic and a neural network.

11. The electric motor according to claim 9, wherein:

the evaluation device comprises a memory module in which characteristic patterns of a previously stored Cepstral diagram are stored for certain operating conditions, and

the recognition module is configured for recognizing the vibrational operating condition on the Cepstral diagram by way of the stored patterns.

12. The electric motor according to claim 8, wherein:

the at least one vibration sensor is arranged on a mechanical structure of the electric motor, in a terminal box, within at least one of an arrangement of electronic components and in a fluid conduit for a fluid to be delivered by a pump assembly.

13. The electric motor according to claim 8, further comprising a wireless signal transmission device for transmitting signals between the at least one vibration sensor and the evaluation device.

14. The electric motor according to claim 9, wherein the evaluation device comprises a normalization module configured for normalizing the Cepstral diagram produced by the Cepstral analysis module in such a manner that swings in the Cepstral diagram are set in relation to background noises.

15. A method for detecting faults in a pump assembly having an electric motor, wherein the pump assembly has at least one rotating shaft, the method comprising:

detecting a vibration signal in the pump or electric motor,

processing the detected vibration signal in a manner such that influence of a current rotational speed of the at least one rotating shaft is eliminated by multiplying a sampling frequency value by a current rotational speed value and dividing the result by a constant rotational speed value,

filtering out periodic signals of the processed vibration signal, and

recognizing a vibrational operating condition including any possible faults by way of the periodic signals.

16. A pump assembly having an electric motor, wherein the pump assembly has at least one rotating shaft and a fault detection system, the fault detection system comprising:

at least one vibration sensor,

an evaluation device connected to the at least one vibration sensor, wherein the evaluation device is provided with a processing module configured to eliminate influence of a current rotational speed of the at least one rotating shaft from a vibration signal captured by the at least one vibration sensor by multiplying a sampling frequency value by a current rotational speed value and dividing the result by a constant rotational speed value,

a filter module configured for filtering periodic signals out of the vibration signal processed by the processing module, and

a recognition module configured for recognizing a vibrational operating condition by way of the periodic signals.