Monitoring machine

By receiving accelerometer data and calculating elliptical motion parameters, the problem of difficulty in detecting the vibration changes of rotating components in the prior art is solved, and automated monitoring of machine status and fault prediction are realized, reducing maintenance costs and safety risks.

CN119013634APending Publication Date: 2024-11-22BUHLER AG
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202380033570.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2022-04-19
Filing Date
2023-04-19
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

When monitoring machines with rotating components, it is difficult to detect vibration pattern changes and failures in the early stage, resulting in high maintenance costs and safety risks. Most of the existing methods are manual or local measurements, and lack automated and efficient monitoring methods.

Method used

By receiving accelerometer data, the elliptical motion parameters of the machine in the time and frequency domains are determined, the rotation set is divided and the elliptical parameters are calculated, which are used to monitor the operating status of the machine and control instructions are generated in response to condition changes.

Benefits of technology

It realizes automatic, low-cost and reliable monitoring of machine status, early detection of faults and potential problems, reduces maintenance costs, and improves the efficiency and safety of predictive maintenance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119013634A_ABST
    Figure CN119013634A_ABST
Patent Text Reader

Abstract

A method and apparatus for monitoring a machine having at least one rotating component is disclosed. In the method, accelerometer data collected for the machine is received by a data processing device. An ellipse describing movement of the machine in the time and / or frequency domain is determined based on the accelerometer data. At least one ellipse parameter defining the ellipse is then used to monitor the machine.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to methods, apparatus, and computer program products for computer-assisted monitoring of a machine, and more particularly, to using accelerometer data in monitoring a machine having at least one rotating component. Background Art

[0002] For example, machines with rotating components may be monitored for maintenance and safety reasons. Predictive maintenance aims to keep machines in good operating order and prevent failures that would slow down operations or even shut down the machine completely, causing damage or even catastrophic failure, fire and / or health and safety risks.

[0003] According to some prior art arrangements, specific machine conditions have been measured locally. The measurements can be manual. For example, during the commissioning of a machine, the movements of the machine have been recorded on a piece of paper with a pencil mounted on the machine. The paper-pencil analysis is then used to perform regular checks on the tracked movements, for example once every two weeks. Another example is the monitoring of the bearing condition which can be tracked based on temperature sensors or noise sensors. Such sensors are usually provided for each bearing. Electricity and smoke are also used as a basis for monitoring. These techniques are also used in various combinations. Summary of the invention

[0004] According to one aspect, a method for monitoring a machine having at least one rotating component by a data processing device is provided, the method comprising: receiving accelerometer data collected for the machine; determining an ellipse describing the movement of the machine in a time domain and / or a frequency domain based on the accelerometer data; determining at least one ellipse parameter defining the ellipse; and using at least one ellipse parameter when monitoring the machine.

[0005] According to another aspect, a monitoring device is provided, the monitoring device comprising at least one processor and at least one memory including computer program code, the at least one memory and the computer program code being configured to, using the at least one processor, cause the device to at least: receive accelerometer data collected for a machine having at least one rotating component, determine an ellipse describing movement of the machine in a time domain and / or a frequency domain based on the accelerometer data, determine at least one ellipse parameter defining the ellipse, and use the at least one ellipse parameter when monitoring the machine.

[0006] According to a more specific aspect, at least one ellipse parameter is determined by rotating the ellipse and moving the zero center point of the ellipse.

[0007] The at least one ellipse parameter may include a radius of the ellipse in the x-coordinate and / or the y-coordinate, a center point of the ellipse in the x-coordinate and / or the y-coordinate, and / or a transformation angle of the ellipse.

[0008] Further parameters may be derived from the at least one ellipsoidal parameter. Further parameters may relate to weight and / or imbalance.

[0009] A change in at least one condition of the machine may be detected based on the monitoring. In response to the detection, a control instruction may be generated.

[0010] The accelerometer data may be divided into one or more rotational sets. An ellipse describing the corresponding rotational set may be calculated, and ellipse parameters may be determined, which define the ellipse describing the corresponding rotational set. The division may include finding a rotational cut point for the division based on low pass filtering, wherein every other crossing of a zero point is determined as a cut point. Alternatively, a search for a local maximum may be provided to identify a cut point. The accelerometer data may then be divided into sets at the determined cut points.

[0011] The accelerometer data for the machine may be generated by a single multi-axis accelerometer or at least two single-axis accelerometers.

[0012] Accelerometer data may be received from a plurality of machines included in a machine system. The accelerometer data from the plurality of machines may be analyzed, and based on the analysis, the operation of the system may be monitored.

[0013] Further analysis may be provided based on the data from the monitoring and the data from the data storage device. The data storage device may include a record of the data associated with the machine type. The further analysis may include mapping the data from the storage device to at least one condition and / or at least one ellipsoidal parameter determined for the machine based on the received accelerometer data.

[0014] Monitoring may include monitoring at least one of a drive, belt, bearings, gears, chain, blades and / or imbalance of the machine, weight distribution of the machine, changes in mass flow, tilt of the machine and / or a lever.

[0015] A computer software product embodying at least part of the functionality described herein may also be provided.According to one aspect, a computer program comprises instructions for performing the method described herein. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Some aspects will now be described in further detail, by way of example only, with reference to the following examples and drawings, in which:

[0017] Figure 1 A schematic example of a system including a machine and a data processing device is illustrated;

[0018] Figure 2 is an example of a data processing device.

[0019] Figure 3 and Figure 4An example of a rotating pattern is shown as a basis for the example;

[0020] Figure 5A and Figure 5B illustrates drawing an ellipse based on raw accelerometer data; and

[0021] Figure 6 and Figure 7 A flow chart of operations according to certain examples is shown. DETAILED DESCRIPTION

[0022] The following description gives an exemplary description of some possibilities for practicing the present invention. Although the specification may refer to "one", "a kind" or "some" examples or embodiments at several places in the text, this does not necessarily mean that each reference is made to the same example of the embodiment, or that a particular feature only applies to a single example or embodiment. Individual features of different examples and embodiments may also be combined to provide other embodiments.

[0023] Figure 1 A schematic example of a machine 1 having at least one rotating component is shown. The machine 1 comprises a drive 12 and a unit 14 driven by the drive. An internal component 15 of the unit is rotated by the drive via a shaft 13. The machine may be mounted on a flexible base. The machine may also be supported on a suspended support arrangement.

[0024] Figure 1 Also shown is a sensor element providing an accelerometer 10, which is attached to the machine 1 so that it can measure vibrations on the machine. The advantage of vibration-based monitoring is that vibrations caused by developing conditions often appear already at an early stage. Changes in vibration patterns can be detected and action taken earlier than based on information from, for example, heat, noise and smoke sensors. Modern accelerometers are sensitive enough to detect relatively small movements and changes, so that their use is a viable option.

[0025] The accelerometer unit 10 may include at least some processing capabilities 11 to generate accelerometer data. The data may be communicated to and used by a data processing device 20 that provides analysis and monitoring functions. The accelerometer 10 is connected to the data processing device 20 via a data connection 16. The data connection 16 may be any suitable wired or at least partially wireless data connection. The data may be communicated from the accelerometer 10 based on any suitable data communication protocol.

[0026] The accelerometer data may be a single multi-axis accelerometer.According to possibilities, the accelerometer unit 10 comprises at least two single-axis accelerometers.

[0027] It should be noted that although Figure 1A single machine is shown in , but the monitored system may comprise a large number of machines and corresponding sensors connected to and monitored by the data processing device. The data processing device may be configured for monitoring a large-scale production facility, such as a factory or a chemical and / or food processing plant.

[0028] The data processing device 20 may include any suitable computer equipment, such as a maintenance server, which is configured to perform the monitoring and control functions as described herein. The data processing device 20 may be connected to / be part of a wider data communication system 22 (such as an intranet or the Internet). The functionality of the data processor device 20 may also be provided at least in part in a cloud or virtualized environment. Appropriate gateways may be provided between different networks and / or environments.

[0029] The data processing apparatus may include any suitable computing arrangement including at least one processor, memory, software and internal circuitry and components necessary to perform the tasks configured to implement at least some of the features described herein. Figure 2 An example of internal components of a data processing device arranged on a board 50 and comprising processors 52 , 53 and one or more memories 51 is shown. Figure 2 Also shown are connections between elements of the device and an interface 54 for connecting the data processing device to the data communication system 22 and enabling data communications with the accelerometer 10. The at least one memory may include at least one ROM and / or at least one RAM. The device may include other possible components used in software and hardware that assist in performing the tasks it is designed to perform. At least one processor may be coupled to the at least one memory. The at least one processor may be configured to execute appropriate software code to implement one or more of the following aspects. The software code may be stored in the at least one memory.

[0030] The accelerometer 10 measures the vibration of the machine 1 and provides raw data to the analysis module of the data processing device 20. The accelerometer 10 can be any suitable sensor capable of providing the functionality of an accelerometer, generating accelerometer data and communicating the generated accelerometer data to the data processing device 20. Various suitable accelerometers are commercially available. The characteristic features of the accelerometer include the internal sampling rate and the sampling output, i.e., the reporting frequency. For example, an accelerometer with an internal sampling and output frequency in the range of more than 500 Hz can be used. More sensitive and capable accelerometers are also available.

[0031] Figure 3A schematic example of a rotating shaft 30 and its range of possible deviations d is shown. Due to the constant rotational movement, there is no circumferential acceleration. Ideally, the acceleration is to the center to keep the shaft on a circular path. However, any irregularity, imbalance, unnecessary weight or load, developed faults (e.g., worn or damaged bearings, bent shafts, faulty belts or chains, blockages, etc.) etc. may affect the balance and cause eccentric accelerations (vibrations). Such vibrations can be measured by an accelerometer. Figure 3 The outer extremes of the deviation are shown by the dashed line 32. However, the deviation may not be circular, but rather the machine may vibrate along an elliptical pattern.

[0032] Figure 4 An example of an elliptical pattern and some of its parameters that can be used to monitor the operation of a machine is shown in FIG. The ellipse parameters may include parameters such as the x-axis radius a, the y-axis radius b, and the center point (h, k). Parameters such as the rotation angle may also be determined. The determination and use of these parameters will be explained below.

[0033] According to an example, the raw acceleration data is measured in the directions of the x and y axes. Figure 5A The raw acceleration data is shown as two 90 degree shifted sine waveforms in the time domain. One of the sine waves is for the x-axis and the other is for the y-axis. Figure 5B Shown according to Figure 5A A sinusoidal waveform is plotted on an ellipsoid in the xy coordinate system.

[0034] The information of the ellipsoid can be used to provide an automatic, low-cost and reliable way to continuously monitor machine conditions based on accelerometer data. Multiple conditions related to the machine, or even all conditions of interest, can be identified based on a sensor device including a single multi-axis accelerometer. Alternatively, a single-axis accelerometer or multiple single-axis accelerometers can be used as the source or accelerometer data.

[0035] The ability to identify and monitor machine conditions based on accelerometer data can be based on analysis of accelerometer data in the time domain. The analysis can include calculation and analysis of ellipsoidal parameters. In some applications, further parameters derived from the ellipsoidal parameters can be analyzed.

[0036] The analysis may also be provided in the frequency domain. The analysis may be provided only in the frequency domain. The analysis may also be based on both the time domain and the frequency domain. The frequency domain analysis may be based on, for example, a power spectral density analysis. The frequency domain based analysis may be used to supplement the analysis results from the time domain analysis.

[0037] Figure 61 is a flow chart according to an example of a method for monitoring a machine having at least one rotating component. In the method, accelerometer data collected for a machine is received by a processor device at 100. Then, at 102, the device determines an ellipse describing movement of the machine in a time domain based on the received accelerometer data. At 104, at least one ellipse parameter for defining the ellipse is defined. Then, at 106, the at least one ellipse parameter can be used to analyze a condition of the machine.

[0038] Determination of at least one ellipse parameter may include rotating the ellipse and / or moving the zero center point of the ellipse. Further parameters may be derived from the at least one ellipse parameter. Further parameters may include, for example, parameters indicating total weight and / or imbalance.

[0039] A change in at least one condition of the machine may be detected. In response to the detection, a control instruction may be generated to react to the detected change. For example, the machine may be stopped, slowed, reduced in load, further diagnostics performed, etc.

[0040] The accelerometer data may be divided into rotation sets. The smallest rotation set includes a single rotation. Ellipses describing the corresponding rotation sets may be calculated, and ellipse parameters may be determined that define the ellipses for these sets.

[0041] The ellipses can be compared to analyze changes in machine operation.

[0042] A more detailed example is in Figure 7 The evaluation of the accelerometer data is based on the analysis of the time domain. As shown in the figure, the frequency domain can also be analyzed. The frequency analysis can be performed on a selected part of the data based on the power spectral density analysis.

[0043] The raw data from the accelerometer may be preprocessed prior to analysis. For example, at block 200, a low pass filter with a predetermined appropriate cutoff frequency may be applied to avoid violating the sampling theorem. The cutoff frequency may depend on the environment. For example, the frequency may be approximately 200 Hz. The data acquisition may already have an analog anti-aliasing filter. To ensure appropriate cutoff, an additional digital filter may be applied. Preprocessing of the raw data depends on the application and setup and is not required in all implementations.

[0044] After filtering, the input data can be divided into different operating phases of the machine. For example, the settling phase can be set to the first thirty minutes, during which the machine stabilizes itself. The length of the stable state can be defined, for example, based on earlier measurements. Then there is a stable phase in which the movement stabilizes. The swing phase occurs after the machine has been commanded to stop but is still moving. The following example focuses on the stable phase (after the end of the settling phase at 202), that is, the phase between the settling phase and the command for engine stop.

[0045] The evaluation of data in the time domain focuses on extracting information about how the machine moves and how it changes when it moves. This analysis can be based on the analysis of acceleration and displacement.

[0046] At 204, the acceleration analysis operation may divide the acceleration data set into individual rotations. When dividing, the cutting points may be determined based on low-pass filtering or searching for local maxima.

[0047] After the division, an ellipse is fitted for each rotation at 206. At this stage, the ellipse is created from the accelerometer data representing the movement of the machine. An example of the calculation to generate the ellipse is given below.

[0048] At 208, an ellipse parameter (or parameters) is extracted from each created ellipse. The parameter (or parameters) for each rotation or set of rotations gives a series of comparable parameters.

[0049] At 210, an average value may be created for each ellipse parameter over a predefined time period. A plurality of rotations may be grouped into rotation sets, and an average ellipse may be generated for each set.

[0050] As an alternative to or in addition to the acceleration analysis, a displacement analysis may also be provided. The displacement analysis has an additional step 203 in which a double integration is provided. The displacement analysis gives the distance (e.g. in millimeters or fractions thereof) instead of in m / s 2 The time aspect is thus eliminated from the displacement analysis.

[0051] Frequency domain based analysis can be provided in parallel with time domain analysis.

[0052] At 212, the output can be used by the analysis data processor entity to search for patterns indicating various conditions. The analysis can be based on a comparison of the determined pattern with a database of patterns indicating various conditions. The condition can be an existing condition or a developing condition. The detected condition can be determined to require a control action. The control action can be immediate (e.g., an alarm, a shutdown, a reduction in speed and / or load, an increase in cooling and / or lubrication). It can also make changes in operating parameters that may take effect over a longer period of time and monitor the impact of these changes.

[0053] In the following, an example of a possible mathematical basis for the determination of ellipse parameters is explained. An ellipse can be described as a polynomial function:

[0054] c1*x 2 +c2*xy+c3*y 2 +c4*x+c5*y=1 (1)

[0055] According to one example, the fitting of the elliptic function may be based on the application of the least squares method.

[0056] ||A*xb|| 2 (2)

[0057] The ellipse parameters a, b, center point (h, k) and rotation angle may not be directly identifiable in the above original ellipse equation (1). In order to extract this data, the ellipse is rotated using a major axis transformation and then the ellipse is moved to the zero center point. The ellipse equation can be formed by diagonalizing the matrix A. The matrix A is symmetric and can be diagonalized without further inspection. The transformation matrix can be used to transfer the quadratic surface to another coordinate system.

[0058] x T Ax+u T x=1 (3)

[0059] This can be expanded to

[0060] x T TDT T x+u T TT T x=1 (4)

[0061] in

[0062] D is the diagonalized matrix of A, and

[0063] T is the transformation matrix.

[0064] For coordinate transformation, another coordinate is needed:

[0065] x′=T T x (5)

[0066] Included in the original equation, this leads to

[0067] x′ T Dx′+u T Tx′=1 (6)

[0068] In the new coordinate system, x' is the rotation of the eliminated ellipse, and after calculating D and T, the ellipse parameters can then be calculated. For example, the center point on the x' axis and the original x axis and the angle of the transformation can be calculated.

[0069] If the angle of the transformation is small, the calculation may result in numerical problems that cause unnecessary complications without any real impact on the result. This can be addressed by appropriate checking routines if necessary. The check may indicate that the calculation can be simplified because there is no real need for rotation.

[0070] Accelerometer-based monitoring arranged according to the principles described herein has been tested for a machine including a belt-driven flour sifter stack. Measurements included variables of different weights on the sifter (0 Kg to 140 Kg), symmetric and asymmetric loads, and different belt tensions. In addition, tests were conducted on the effects of conditions such as door vibration, lack of preload, and special measurements for situations such as tilted bars, loose bars, and cut bars.

[0071] In these measurements it was found that the behavior of the ellipse parameters a and b is essentially similar, for example both parameters show a clear negative and similar linear trend with increasing screen stack weight. Asymmetric loads also map well to a linear behavior. The coordinates of the center point of the ellipse also show a similar behavior of the a and b coordinates. Analysis of the weight imbalance shows a clear deviation from a normal weight increase.

[0072] In general, the evaluation in the time domain shows that the ellipse parameters have the ability to indicate the condition of the machine and changes therein. In particular, the ellipse radii a and b are considered informative. Further optimization is achieved by further condition parameters such as total weight and imbalance. Based on the results, the total weight has an impact on the changing diameter of the circular movement. On the other hand, the imbalance affects the ellipse rate of the movement.

[0073] The influence can be better divided by defining a weight factor and an imbalance factor. The weight factor can make it possible to calculate what can be the minimum detectable weight by taking into account the standard deviation of all data points. The imbalance factor can be used to detect an unbalanced weight distribution. An observation about the imbalance factor in the test is that in case of an even distribution of weight on the sieve, the deviation is negligible. In theory, all measurement points should establish a horizontal line that indicates that a symmetrically loaded machine performs an almost perfect circular movement. Any imbalance can be detected as its variation.

[0074] Accelerometer data may be received from a plurality of machines at a central data processing entity. The machines may be included in a machine system, such as a production plant. The accelerometer data from the plurality of machines may be analyzed, and the operation of the plant may be monitored based at least in part on the analysis of the accelerometer data from the plurality of sensors and associated machines.

[0075] The analysis may be based on data provided by the monitoring and data from a data storage device. For example, data collected for a particular type of machine may be used. The data storage device may include a library of maintenance data associated with one or more types of machines, and the analysis may include finding data corresponding to at least one condition and / or at least one elliptical parameter determined for one or more machines of that type based on the received accelerometer data. Any deviation from expectations may be processed and necessary corrective action taken.

[0076] Thus, for example, a production line comprising a plurality of machines can be monitored and the entire process and / or its maintenance can be optimized.

[0077] To more easily partition functions based on acceleration data (which can be a superposition of all vibrations), other sensor technologies can be added. For example, a power meter for measuring drive current (mass flow change, drive status) and temperature probes on the drive and bearings can be provided. For example, temperature information can be helpful because information about temperature is easy to obtain and process, and temperature increase is one of the easily indicative parameters that is understood and used when determining the state of the machine.

[0078] The machine to be monitored may include any machine having at least one rotating component. The principles disclosed herein may be used for a variety of monitoring applications, such as providing drive monitoring, belt monitoring (e.g., belt status, belt tension, etc.), drive chain monitoring, bearing monitoring, gear monitoring, pump or fan blade monitoring, imbalance and weight distribution monitoring, machine tilt monitoring, rod loss monitoring, mass flow change monitoring, loose component monitoring, etc.

[0079] An efficient and reliable way of identifying and measuring machine conditions may be provided by the arrangements disclosed herein. A low cost sensor arrangement for measuring and identifying relevant conditions of one or more machines may be provided. Early notification of potential machine failure may be provided. Overall maintenance costs may be reduced, and predictive maintenance and maintenance planning may become more efficient. The ability to analyze data locally at the location of the machine and centrally at a remote location may be provided. The ability to react to changes in machine conditions in a timely manner may be provided. A deeper understanding of the process and the conditions affecting the machines involved in the process may be provided. Process optimization may be provided.

[0080] It should be noted that although the above detailed examples have been described with reference to certain applications and machines, several changes and modifications may be made to the disclosed solutions without departing from the scope of the present invention. In particular, different embodiments have been described as examples. Different features from different embodiments may be combined.

[0081] In general, some embodiments may be implemented in hardware or dedicated circuits, software, logic, or any combination thereof. For example, some aspects may be implemented in hardware, while other aspects may be implemented in firmware or software that may be executed by a controller, microprocessor, or other computing device, although the embodiments are not limited thereto. Although various embodiments may be illustrated and described as block diagrams, flow charts, or using some other graphical representation, it is well understood that, as non-limiting examples, the blocks, devices, systems, techniques, or methods described herein may be implemented in hardware, software, firmware, dedicated circuits or logic, general purpose hardware or controllers or other computing devices, or some combination thereof.

[0082] These embodiments can be realized by computer software that is stored in the memory and can be executed by at least one data processor, or by hardware, or by a combination of software and hardware. In addition, in this respect, it should be noted that any of the above processes can represent program steps, or interconnected logic circuits, blocks and functions, or a combination of program steps and logic circuits, blocks and functions. Software can be stored on a physical medium such as a memory chip or a memory block realized in a processor, on a magnetic medium such as a hard disk or a floppy disk, and on an optical medium such as, for example, a DVD and its data variant, a CD. The memory can be any type suitable for a local technical environment, and can be realized using any suitable data storage technology, such as a semiconductor-based memory device, a magnetic memory device and system, an optical memory device and system, a fixed memory and a removable memory. A data processor can be any type suitable for a local technical environment, and as a non-limiting example, can include one or more of a general-purpose computer, a special-purpose computer, a microprocessor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a gate-level circuit and a processor based on a multi-core processor architecture. Although certain aspects of the invention may be illustrated and described as block diagrams, flow charts, or using some other schematic pictorial representations, it is well understood that, as non-limiting examples, these blocks, devices, systems, techniques and methods described herein may be implemented at least in part in hardware, software, firmware, special purpose circuits or logic, general purpose hardware or controllers or other computing devices, or some combination thereof.

[0083] The foregoing description provides a comprehensive and informative description of exemplary embodiments of the present invention by way of exemplary and non-limiting examples. However, various modifications and adaptations may become apparent to those skilled in the relevant art in view of the foregoing description when read in conjunction with the accompanying drawings and the appended claims. All such and similar modifications of the teachings of the present invention are intended to fall within the scope of the present invention.

Claims

1. A method for monitoring a machine having at least one rotating component by means of a data processing device, the method comprising: receiving accelerometer data collected for the machine, determining an ellipse describing movement of the machine in the time domain and / or frequency domain based on the accelerometer data, determining at least one ellipse parameter defining the ellipse, and The at least one ellipsoidal parameter is used in monitoring the machine. 2 . The method of claim 1 , wherein the determining of the at least one ellipse parameter comprises rotating the ellipse and moving a zero center point of the ellipse.

3. The method according to claim 1 or 2, wherein the at least one ellipse parameter comprises the radius of the ellipse in the x-coordinate and / or the y-coordinate, the center point of the ellipse in the x-coordinate and / or the y-coordinate and / or the transformation angle of the ellipse.

4. A method according to any preceding claim, comprising deriving further parameters from the at least one ellipsoidal parameter, wherein the further parameters relate to weight and / or imbalance.

5. A method according to any preceding claim, comprising detecting a change in at least one condition of the machine based on the monitoring, and generating a control instruction in response to the detection.

6. A method according to any preceding claim, the method comprising dividing the accelerometer data into one or more sets of rotations, calculating ellipses describing corresponding sets of rotations, and determining ellipse parameters, the ellipse parameters defining the ellipse describing the corresponding sets of rotations.

7. The method of claim 6, wherein the partitioning comprises finding a rotational cutting point for the partitioning based on low-pass filtering, wherein every other crossing of a zero point is determined as a cutting point, or searching for a local maximum to identify a cutting point, and at the determined cutting points, the accelerometer data is partitioned into sets.

8. A method according to any preceding claim, wherein the accelerometer data of the machine is generated by a single multi-axis accelerometer.

9. The method of any one of claims 1 to 7, wherein the accelerometer data of the machine is generated by at least two single-axis accelerometers.

10. A method according to any preceding claim, the method comprising receiving accelerometer data from a plurality of machines comprised in a system of machines, analysing the accelerometer data from the plurality of machines, and monitoring operation of the system based on the analysis.

11. A method according to any preceding claim, comprising performing further analysis based on data provided by the monitoring and data from a data storage device.

12. The method according to claim 11, wherein: The data storage device includes a record of data associated with the type of machine, and the further analysis includes mapping the data from the storage device to at least one condition and / or the at least one elliptical parameter determined for the machine based on the received accelerometer data.

13. A method according to any preceding claim, wherein the monitoring comprises monitoring at least one of a drive, belt, bearings, gears, chain, blades and / or imbalance of the machine, weight distribution of the machine, changes in mass flow, inclination of the machine and / or a lever.

14. A monitoring device, the monitoring device comprising at least one processor and at least one memory including computer program code, the at least one memory and the computer program code being configured to, with the at least one processor, cause the device to at least: receiving accelerometer data collected for a machine having at least one rotating component, determining an ellipse describing movement of the machine in the time domain and / or frequency domain based on the accelerometer data, determining at least one ellipse parameter defining the ellipse, and The at least one ellipsoidal parameter is used in monitoring the machine.

15. The apparatus according to claim 14, configured to perform the method according to any one of claims 2 to 13. 16 . A computer-readable medium comprising a program code for causing a processor to execute instructions for the method according to claim 1 .