US-12618746-B2 - Method of extending the lifetime of a rolling element bearing

US12618746B2US 12618746 B2US12618746 B2US 12618746B2US-12618746-B2

Abstract

A method of extending the lifetime of a rolling-element bearing of an electrical machine, the method including: a) defining a first set of control parameter values, b) estimating a first remaining useful life, RUL, of the rolling-element bearing, based on a signal that provides a measure of vibrations of the rolling-element bearing obtained when the electrical machine is controlled by the first set of control parameter values, c) changing at least one control parameter value in the first set of control parameter values to thereby obtain a second set of control parameter values, d) estimating a second RUL of the rolling-element bearing based on the signal, obtained when the electrical machine is controlled by the second set of control parameter values, e) comparing the first RUL with the second RUL, f) in case the second RUL is longer than the first RUL, replacing the control parameter values in the first set of control parameter values with those of the second set of control parameter values, and replacing the value of the first RUL with that of the second RUL, and g) repeating steps c)-f) over and over during operation of the electrical machine.

Inventors

Panagiotis Kakosimos
Ingo Stroka
Dmitry Svechkarenko

Assignees

ABB SCHWEIZ AG

Dates

Publication Date: 20260505
Application Date: 20211020
Priority Date: 20201022

Claims (15)

1 . A method of extending the lifetime of a rolling-element bearing of an electrical machine, the method comprising: a) defining a first set of control parameter values, b) estimating a first value of remaining useful life, RUL, of the rolling-element bearing, based on a signal that provides a measure of vibrations of the rolling-element bearing obtained when the electrical machine is controlled by the first set of control parameter values, c) changing at least one control parameter value in the first set of control parameter values to thereby obtain a second set of control parameter values, d) estimating a second value of RUL of the rolling-element bearing based on the signal, obtained when the electrical machine is controlled by the second set of control parameter values, e) comparing the first value of RUL with the second value of RUL, f) in case the second value of RUL is longer than the first value of RUL, replacing the control parameter values in the first set of control parameter values with those of the second set of control parameter values, and replacing the first value of RUL with the second value of RUL, and g) repeating steps c)-f) during operation of the electrical machine.
2 . The method as claimed in claim 1 , wherein the estimating in steps b) and d) involves using a degradation model of the rolling-element bearing, using an Extended Kalman Filter or using neural networks.
3 . The method as claimed in claim 1 , wherein step f) involves replacing the control parameter values in the first set of control parameter values only in case the second RUL deviates with more than a threshold value from the first RUL.
4 . The method as claimed in claim 1 , comprising, prior to step c): A) determining a harmonic content of the signal, B) changing at least one control parameter value in a third set of control parameter values in case an amplitude higher than a threshold value is present in the harmonic content, and C) repeating steps A)-B) until a set of third control parameter values that causes the amplitude to decrease to a value below the threshold value has been found; and D) performing steps c-g), wherein the at least one control parameter value that is changed in step c) is of that at least one control parameter that is associated with the at least one control parameter value used in step B), and wherein the change of the at least one control parameter value in step B) is smaller than the change in step c).
5 . The method as claimed in claim 4 , wherein the changing of the at least one control parameter value in step B) is performed stepwise in each iteration.
6 . The method as claimed in claim 4 , wherein the changing in step c) is based on the at least one control parameter value selected in step B).
7 . The method as claimed in claim 4 , wherein in initial iterations of step B) the control parameter value of an individual control parameter is changed, and in later iterations when changes to all individual control parameters have been performed, combinations of control parameter values of control parameters are changed.
8 . The method as claimed in claim 4 , wherein steps A)-C) form an inner control loop and steps c)-g) form an outer control loop.
9 . A non-transitory storage medium encoded with computer code which when executed by processing circuitry of a control system causes the control system to perform the following steps: a) defining a first set of control parameter values, b) estimating a first value of remaining useful life, RUL, of the rolling-element bearing, based on a signal that provides a measure of vibrations of the rolling-element bearing obtained when the electrical machine is controlled by the first set of control parameter values, c) changing at least one control parameter value in the first set of control parameter values to thereby obtain a second set of control parameter values, d) estimating a second value of RUL of the rolling-element bearing based on the signal, obtained when the electrical machine is controlled by the second set of control parameter values, e) comparing the first value of RUL with the second value of RUL, f) in case the second value of RUL is longer than the first value of RUL, replacing the control parameter values in the first set of control parameter values with those of the second set of control parameter values, and replacing the first value of RUL with the second value of RUL, and g) repeating steps c)-f) during operation of the electrical machine.
10 . A control system for extending the lifetime of a rolling-element bearing of an electrical machine, the control system comprising: a storage medium comprising a computer code, and processing circuitry, wherein when the processing circuitry executes the computer code, the control system is configured to: a) define a first set of control parameter values, b) estimate a first value of remaining useful life, RUL, of the rolling-element bearing, based on a signal that provides a measure of vibrations of the rolling-element bearing obtained when the electrical machine is controlled by the first set of control parameter values, c) change at least one control parameter value in the first set of control parameter values to thereby obtain a second set of control parameter values, d) estimate a second value of RUL of the rolling-element bearing based on the signal, obtained when the electrical machine is controlled by the second set of control parameter values, e) compare the first value of RUL with the second value of RUL, f) in case the second value of RUL is longer than the first value of RUL, replace the control parameter values in the first set of control parameter values with those of the second set of control parameter values, and replace the first value of RUL with the second value of RUL, and g) repeat steps c)-f) during operation of the electrical machine.
11 . The control system as claimed in claim 10 , wherein the processing circuitry is configured to estimate the first and second RUL using a degradation model of the rolling-element bearing, using an Extended Kalman Filter or using neural networks.
12 . The control system as claimed in claim 10 , wherein the processing circuitry is configured to, prior to step c): A) determine a harmonic content of the signal, B) change at least one control parameter value in a third set of control parameter values in case an amplitude higher than a threshold value is present in the harmonic content, and C) repeat steps A)-B) until a set of third control parameter values that causes the amplitude to decrease to a value below the threshold value has been found; and D) perform steps c-g), wherein the at least one control parameter value that is changed in step c) is of that at least one control parameter that is associated with the at least one control parameter value used in step B), and wherein the change of the at least one control parameter value in step B) is smaller than the change in step c).
13 . The control system as claimed in claim 12 , wherein the processing circuitry is configured to change the at least one control parameter value in step B) stepwise in each iteration.
14 . The control system as claimed in claim 12 , wherein the processing circuitry is configured to perform the change in step c) based on the at least one control parameter value selected in step B).
15 . The control system as claimed in claim 12 , wherein the processing circuitry is configured to change the control parameter value of an individual control parameter in initial iterations of step B), and in later iterations when changes to all individual control parameters have been performed, to change combinations of control parameter values of control parameters.

Description

TECHNICAL FIELD The present disclosure generally relates to rolling-element bearings used in electrical machines. BACKGROUND Electric motors are a key component that enables modern, efficient production. For a company to globally stay competitive its motors should be running 24/7. More than half of the breakdowns of electrical motors are due to failures of the bearings. The improper handling of the bearing during its life is the main reason for its early failure. In academia and industry, many researchers focus on how to detect the incipient bearing faults. It is understood that the failure types are numerous and very complicated to be identified due to their stochastic nature. Therefore, artificial intelligence algorithms are usually employed. However, detecting a bearing failure only brings the benefit of planning the bearing replacement in the next maintenance period. U.S. Pat. No. 6,326,758 discloses a model-based diagnostic approach along with qualitative/casual model information. The aim is to monitor the health condition and then act upon detecting an imminent failure. The goal is to change the control parameters and verify that the system can operate reliably until the next planned maintenance. The method involves a health assessment technique that evaluates the machine condition. A health indication is then given to the controller where a model of the overall system is determined. Finally, a multi-objective optimization method is involved to optimize the system performance. Different kinds of parameters are given as inputs to the optimization routine, e.g., control parameters, operating conditions, etc. The final goal is to act in the control system before imminent failures render it inoperable. SUMMARY A general object of the present disclosure is to provide a method that solves or at least mitigates the problems of the prior art. There is hence according to a first aspect of the present disclosure provided a method of extending the lifetime of a rolling-element bearing of an electrical machine, the method comprising: a) defining a first set of control parameter values, b) estimating a first remaining useful life, RUL, of the rolling-element bearing, based on a signal that provides a measure of vibrations of the rolling-element bearing obtained when the electrical machine is controlled by the first set of control parameter values, c) changing at least one control parameter value in the first set of control parameter values to thereby obtain a second set of control parameter values, d) estimating a second RUL of the rolling-element bearing based on the signal, obtained when the electrical machine is controlled by the second set of control parameter values, e) comparing the first RUL with the second RUL, f) in case the second RUL is longer than the first RUL, replacing the control parameter values in the first set of control parameter values with those of the second set of control parameter values, and replacing the value of the first RUL with that of the second RUL, and g) repeating steps c)-f) over and over during operation of the electrical machine. By this iterative method, the lifetime of the rolling-element bearing may be extended without the need of an optimisation routine. The optimisation routine disclosed in U.S. Pat. No. 6,326,758 requires the existence of a model. If not available a signal has to be injected, the system response must be analysed, and a transfer function (stimulus response) has to be calculated. Then, the transfer function is considered in the optimization function. For the present method, only the direction of the trend of the RUL estimation is needed to draw a conclusion about the impact of the change of the control parameter value. This also allows the implementation to be realized with minimum hardware resources. Further, according to U.S. Pat. No. 6,326,758 the optimality of the solution relies on the selected optimization function and its members. The application constraints are also put in the optimization routine, so the result is always optimal. Nevertheless, this requirement renders the method not easily implementable in practice. According to the present method, the power converter/drive may perform actions and make decisions similar to those of a human, so they have justifiable benefits, and safety issues are not raised. On the other hand, if any member of the optimization function in U.S. Pat. No. 6,326,758 is erroneous, the system will fail. The RUL is the amount of time that a group of apparently identical bearings will complete or exceed before the formation of a fatigue spall. The estimation of the first and second RUL may preferably be performed during steady-state conditions. This is because the estimation of the RUL is complex and takes time and may otherwise be affected by load or speed transients. The method may in step c) involve changing the at least one control parameter value stepwise in each iteration. The present method may thus be search based