METHOD FOR PREDICTING THE REMAINING SERVICE LIFE OF A SEALING ARRANGEMENT OF A PISTON COMPRESSOR

Description

[0001] The present invention relates to a method for determining the degree of wear and for predicting the remaining service life of a sealing arrangement of a piston compressor, a preventive maintenance system for a piston compressor, a piston compressor comprising a preventive maintenance system, a computer-readable storage medium, and the use of the method for monitoring the degradation of a sealing arrangement of a piston compressor and / or for predicting the remaining service life of a sealing arrangement of a piston compressor and / or for adjusting the leakage occurring at a sealing arrangement of a piston compressor.

[0002] Condition monitoring is the process of monitoring a condition parameter (vibration, temperature, etc.) in machines to identify significant changes that indicate a developing fault. It is a key component of predictive maintenance, also known as preventive maintenance or proactive maintenance. Preventive maintenance allows for the planning of maintenance work or other measures to avoid unexpected machine downtime or consequential damage and its consequences. This improves the availability of production facilities while simultaneously reducing equipment downtime and maintenance costs.

[0003] Compressors are commonly used to compress fluids such as gases or aerosols. In a reciprocating compressor, the rotary motion of a crankshaft, generated by a drive unit, is converted by connecting rods into a reciprocating motion of one or more pistons. This motion is used to compress the intake gas or aerosol in a compressor unit. Due to its design, the piston is subjected to the forces that accelerate it from zero velocity at the end of each stroke to a high velocity in the middle of the stroke and back to zero again. For at least part of the compression cycle, the compression chamber is sealed by closed cylinder valves before the compressed gas or aerosol is released to a discharge device through the opening of at least one cylinder valve. These valves are subject to a certain amount of wear due to their movements.To seal the crank-side working chamber, piston rod sealing systems, so-called packings, are used on the moving piston rod. Due to the contact of the sealing elements with the oscillating piston rod, they are subject to constant wear.

[0004] The unwanted escape of the compressed gas or aerosol from the compressor unit, particularly due to wear-related leakage of the sealing elements and / or cylinder valves described above, is commonly referred to as leakage. To measure leakage—and thus assess the functionality of the compressor unit—appropriate sensors can be used, ideally positioned on each individual compressor unit or stage. However, for cost reasons, such sensors are often not installed on the compressor and / or do not always automatically detect and transmit the leakage values ​​they measure.

[0005] From EP 2 458 214 A1, a method for monitoring the operation of a reciprocating compressor using a pressure sensor and a vibration sensor is known. Based on the detected pressure within the device and the detected vibrations of the compressor's reciprocating mechanism, the protection system described therein calculates a stiffness value of the cylinder assembly, thus enabling the assessment of its structural integrity during operation. The protection system also allows the compressor to be shut down after it has been determined that the compressor's condition deviates from a predefined standard condition.

[0006] EP 3 436 877 A1 relates to a method and a device for carrying out a method for vibration diagnostic monitoring and assessment of individual machine parts. At least one time signal of the machine's vibration is recorded and evaluated by means of frequency analysis using a frequency analyzer. Frequency maxima extracted from the frequency spectrum are then assigned to specific machine parts, which are compared with setpoint data, model data, and / or reference data of the respective machine parts. If the extracted frequency maxima, which are determined from the currently measured vibration data, deviate from the setpoint data, the model data, and / or the reference data of the machine parts, an assessment message or warning is issued.

[0007] The publication: Loukopoulos Panagiotis et al.: "Abrupt fault remaining useful life estimation using measurements from a reciprocating compressor valve failure", in Mechanical Systems and Signal Processing, Vol. 121, November 24, 2018, pages 359-372, discloses a method for determining the degree of wear and for predicting the remaining service life of a sealing arrangement of a piston compressor.

[0008] A disadvantage of the aforementioned methods and systems is that the user is only notified when one of the calculated values ​​deviates from the ideal value. Real-time transmission of the remaining service life and an indication of the wear level of the compressor's sealing elements are not possible with these methods and systems. Consequently, there remains a need for methods and systems that can also make accurate predictions about the future condition and the point of failure of machine parts such as sealing elements. Furthermore, in preventive maintenance, it frequently happens that parts with a long remaining service life are replaced. Such premature replacement incurs unnecessary costs and is not sustainable.

[0009] The article is also state-of-the-art. "The study "Abrupt fault remaining useful life estimation using measurement from a reciprocating compressor valve failure" by Loukopoulos et al. (Mechanical Systems and Signal Processing 2019, 121, 359-372) is well-known. It compares the predictive performance of various methods for estimating the remaining useful life (RUL) of reciprocating compressors with respect to the accuracy and precision of the compared methods. The predictive methods discussed are data-driven, modeling the degradation process using historical information. The data used for the calculations were retrieved from an industrial reciprocating compressor via a server. Therefore, they are not live raw data from sensors attached to or measuring on the compressor.

[0010] However, a disadvantage of such methods is that they reach their limits when faced with new error events, i.e., cases for which they have not been trained, and that the accuracy of the prediction generally depends on the quantity and quality of the available data.

[0011] Based on the aforementioned prior art, the present invention aims to overcome such and other disadvantages of the prior art and, in particular, to provide a method by which the degree of wear of a sealing arrangement of a piston compressor can be determined and its remaining service life can be predicted cost-effectively and reliably.

[0012] The problem is solved by a method, a preventive maintenance system for a piston compressor, a piston compressor, a computer-readable storage medium, and the use of the method for monitoring the degradation of a sealing assembly of a piston compressor and / or for predicting the remaining service life of a sealing assembly of a piston compressor and / or for adjusting the leakage occurring in a sealing assembly of a piston compressor according to the independent claims. Advantageous embodiments and further developments are the subject of the dependent claims.

[0013] The problem is solved in particular by a computer-implemented method for determining the degree of wear and predicting the remaining service life of a sealing assembly of a piston compressor. The method comprises the following steps a) to h): a) Provision of vibration data of the reciprocating part of the piston compressor. Alternatively, it is also conceivable that features calculable from vibration data are used in the inventive method. The vibration data or the features calculable from it must be traceable to specific positions (segments M) of the movement of the reciprocating part of the piston compressor. b) Creation of an input matrix from the vibration data or the features calculable from it. The input matrix has a dimension (M x N) determined by the number of segments M and the number of sampling points N. c) Execution of a simulation comprising a plurality of simulation steps, applying a model that can describe error propagation in a closed system to the input matrix. Each simulation step generates an output matrix.The output matrix of the nth simulation step is used as the input matrix of the n+1th simulation step. d) Optionally, the Laplace matrix is ​​formed for each output matrix, and the eigenvalues ​​of the respective Laplace matrix are calculated. The Laplace matrix can be obtained, in particular, by multiplying the respective output matrix by its transpose. e) Optionally, the normalized distance of the output matrix to the output matrix of the previous simulation step is calculated. f) Exactly one parameter is selected, either from values ​​of the output matrices, from the calculated eigenvalues ​​of the Laplace matrices formed in optional step d), or from the normalized distances calculated in optional step e). The selected parameter represents a measure of the synchronization of the vibration data traceable to the segments M or of features calculable therefrom.g) Determining the number of simulation steps required until the selected parameter falls below a predefined threshold close to zero. This result of the simulation, i.e., the determined number of simulation steps, is converted into a unit of time based on the sampling rate of the vibration data used in the input matrix or the features calculable from it. h) Repeating the above-described steps a) to g) at regular discrete time intervals to obtain a large number of simulation results. i) After each new simulation, a curve is fitted to the obtained simulation results, taking into account the previous simulation results, in particular by means of local linear regression or another suitable method, and the future time at which the curve intersects a predefined lower bound is determined.The lower limit is defined in particular by the abscissa (x-axis) passing through the origin of a Cartesian coordinate system. The calculated time at which the curve, fitted to the simulation results, intersects the lower limit is considered an indicator of the occurrence of a malfunction, especially a total failure, and the predicted end of the service life of the sealing assembly.

[0014] Within the scope of the present invention, the term "sampling time" is understood to mean the time at which a value, in particular a vibration data value or the value of a feature calculable therefrom, is measured or calculated.

[0015] Within the scope of the present invention, the term "sealing arrangement" refers in particular to a piston rod packing sealing a piston rod and / or a cylinder valve sealing a compression chamber.

[0016] The size of the predefined threshold depends on the input data, i.e., the vibration data or the features that can be calculated from it, and is close to zero, since the approach of the second smallest eigenvalues ​​to zero suggests a high system instability.

[0017] The present invention is based on the finding that leaks affect the vibrations measured at the compressor. In particular, broken packing or guide rings, in addition to increased leakage, also lead to a displacement or deviation of the piston rod and piston from their ideal motion, increased friction, and vibrations in the affected segments of the piston movement. Since the acceleration forces on the piston are greatest in the middle of the piston movement, the displacement is also greater in this area and manifests itself through increased vibrations. When increased leakage occurs, this displacement increases further and is accompanied by increased vibrations that propagate across several adjacent segments of the piston movement towards the end of the stroke.

[0018] This is reflected in the simulation phase of the present invention, in which the propagation speed of a fault towards the end of the stroke is simulated. Assigning the measured vibrations to segments of the motion of the reciprocating part of the piston compressor, and in particular to a specific position of the piston, is a necessary preprocessing step to ensure that the input matrix is ​​a two-dimensional representation of the vibrations measured over a full cycle of the reciprocating device. Without being limited to this, it is assumed that the vibration pattern, via the motion of the reciprocating device, and especially via the piston movement, represents the thermodynamic processes inside the cylinder.This hypothesis is supported by information from manual machine diagnostics experts who report that a possible leakage or increased friction in segments is accompanied by a uniform and parallel increase in vibration peaks in these segments.

[0019] Within the scope of the present invention, the vibrations occurring in the respective segments of the piston movement and their propagation to neighboring segments are considered as interacting particles. Due to the temporal evolution of the vibrations, the time-dependent Ginzburg-Landau model (TDGL), for example, can be used to describe the dynamics between the particles. This model is a modification of the kinetic Ising model, in which discrete variables represent the states of the magnetic spins (-1, +1) of the particles. The free energy of such a system is estimated by observing the change in the order parameter (G. Gaspari, Physical Review B 1986, 33(5), 3295-3305).

[0020] The inventive method essentially represents a hybrid digital sensor that uses vibration data to simulate the function or measurement of a leakage sensor. In other words, the inventive method makes it possible to simulate leakage that is not directly measured via another parameter, namely vibration data or features calculable from it, and to predict packing or valve damage. The chosen method does not require additional pressure or temperature sensors, or the calculation of dynamic pressures, and is therefore cost-effective with regard to the necessary sensor technology. Furthermore, the inventive method has proven to be particularly robust with respect to the accuracy of the predicted end of the service life of the sealing arrangement, for example, with regard to changing compressor load or other fluctuations that are not attributable to wear.

[0021] In a preferred embodiment of the method according to the invention, the second smallest eigenvalue of the output matrix is ​​selected as the characteristic value. The predefined threshold value in this case is approximately zero, for example, an amount of 0.000025.

[0022] In a preferred embodiment of the method according to the invention, process step d) is carried out and the second smallest eigenvalue of the Laplace matrix is ​​selected as the characteristic value. The predefined threshold value is approximately zero in this case, for example, a value of 0.000025.

[0023] The second smallest eigenvalues ​​of the Laplace matrix are called the algebraic connectivity or Fiedler eigenvector (M. Fiedler, Czechoslovak Mathematical Journal 1973, 23(2), 298-305). This metric plays a crucial role in the synchronization of coupled oscillators and the robustness of networks, including cascading failures (S. Strogatz, Physica D 2000, 143, 1-20). If the algebraic connectivity of a graph reaches zero, the graph is split into more than one connected component, meaning a dynamic process will never synchronize (Hernändez et al., Journal of Complex Networks 2014, 2(3), 272-287).

[0024] The inventive method can be carried out remotely, whereby the vibration data measured at the compressor are sent to an independent location and evaluated there.

[0025] Alternatively, in a further embodiment of the method according to the invention, it is also conceivable that the provided vibration data or the features calculable therefrom are obtained as part of the method being carried out. For this purpose, at least one time signal x(t) of the vibration of a reciprocating part of the piston compressor is detected by at least one vibration sensor. Optionally, the features can be calculated from the detected time signal x(t). The detected time signal x(t) of the vibration or the optionally calculated features is then divided into a plurality of time intervals, which always correspond to the same position of the reciprocating device, in particular of the piston or piston rod. For each time interval, the mean value of the time signal x(t) detected in the respective time interval or of the optionally calculated features is then calculated in order to obtain averaged vibration values ​​or averaged features.Furthermore, the position of the reciprocating part of the piston compressor is detected by at least one position sensor, and the detected position data is provided for carrying out the method. Based on the detected position data, the movement of the reciprocating part is divided into a plurality of segments M, which are preferably of equal size. Finally, the averaged vibration values ​​or averaged characteristics are assigned to the segments M of the movement in such a way that the respective averaged vibration values ​​or averaged characteristics always correspond to the same segments M of the movement of the reciprocating part, and thus a leak at a specific point is reflected accordingly in the vibration data. For example, the time interval is one minute, i.e.,Within one minute, a certain number of measurement points are obtained in a specific segment M of the movement of the reciprocating part, which measurement points are averaged and provided as the minute average of the respective segment M for the input matrix.

[0026] In a preferred embodiment of the inventive method, the prediction of the time at which the selected parameter falls below the predefined threshold is only made after the parameter has reached a predetermined criticality value. The criticality value marks the onset of an irreversible deterioration process in the sealing arrangement.

[0027] This significantly reduces the variability of the simulation results and thus increases the accuracy of the predicted time at which the sealing assembly will actually malfunction. Furthermore, this shifts the start of the predictions closer to the actual failure point, which increases the accuracy of the predictions and reduces the risk of false positives.

[0028] In a preferred embodiment of the method according to the invention, the model applied to the input matrix is ​​the time-dependent Ginzburg-Landau model (TDGL).

[0029] Within the scope of the present invention, it was found that the time-dependent Ginzburg-Landau model (TGDL) as a generalization of the n-vector and 2D Ising models (Gaspari et al., Physical Review B 1986, 33(5), 3295-3305) is particularly well suited to the present experimental context.

[0030] The use of piston compressors often occurs under conditions that have an external influence on the vibrations of the reciprocating part of the piston compressor. For example, when piston compressors are used on seagoing vessels, the prevailing wave pattern and the associated ship movements can themselves cause vibrations that can be superimposed on the vibrations detected at the piston compressor.

[0031] To reduce such effects, the provided vibration data or features can be additionally denoised.

[0032] In a preferred embodiment of the method according to the invention, the provided and optionally denoised vibration data or features are normalized between -1 and +1, which is advantageous depending on the selected model and distribution of the vibration data.

[0033] In a preferred embodiment, the inventive method further comprises a data transmission step for transmitting the vibration data or characteristics to at least one ground station. Additionally or alternatively, it is also conceivable that the inventive method comprises a data transmission step for transmitting the predicted remaining service life of the sealing arrangement to a user interface of the piston compressor. In both cases, at least the simulation steps and the prediction of the time at which the selected parameter falls below a predefined threshold are performed wholly or partially on the ground using the ground station.

[0034] Transmitting the vibration data measured at the piston compressor to a ground station enables remote monitoring of the compressor's condition without requiring all the computing power necessary for carrying out the inventive method to be located at or near the compressor. Furthermore, transmitting the predicted remaining service life of the sealing assembly to a user interface of the piston compressor, such as a control panel or indicator light, prevents users from being surprised by a potentially unforeseen compressor failure. This can be particularly advantageous when monitoring piston compressors used on seagoing vessels, as an unforeseen compressor failure outside of ports can pose a significant safety risk.

[0035] In a preferred embodiment, the inventive method further comprises transmitting the identified time and / or the predicted end of the service life of the sealing arrangement to a user. This transmission preferably occurs in real time. Additionally or alternatively, it is also conceivable according to the invention to estimate the degree of wear of the sealing arrangement from the behavior of the characteristic parameter over the number of simulation steps performed and to transmit the resulting degree of wear to a user. This transmission also preferably occurs in real time.

[0036] This enables the user of the piston compressor to be informed about its condition and, if necessary, to take timely precautions for any required maintenance interruption.

[0037] In a preferred embodiment of the inventive method, the sealing arrangement is a piston rod packing sealing against a piston rod or a cylinder valve delimiting a compression chamber. Piston rod packings and cylinder valves are among the parts of a piston compressor whose damage or even failure leads to a serious impairment of the compression process and thus reduces the performance of the piston compressor.

[0038] In a preferred embodiment of the method according to the invention, the execution of the method is started automatically, either at a time selectable by a user or at fixed time intervals.

[0039] This automation of the process significantly reduces the risk of an unforeseen failure of the piston compressor. Furthermore, it also allows for monitoring the execution of maintenance work prescribed by the manufacturer and / or the user.

[0040] In principle, all of the aforementioned embodiments can be combined with each other.

[0041] The task is further solved by a preventive maintenance system for use with a piston compressor. The preventive maintenance system comprises a computing unit with at least one processor. The computing unit is configured to receive position data of a reciprocating part of the piston compressor, to receive vibration data or computable features of the reciprocating part of the piston compressor, and to assign these vibration data or features to segments M of the movement of the reciprocating part of the piston compressor.The computational device is further configured to create an input matrix from the received vibration data or characteristics and to perform a multitude of simulations, each comprising a plurality of simulation steps, on the input matrix using a model that can describe fault propagation in a closed system. This yields an output matrix in each case, with the output matrix of the nth simulation step being used as the input matrix of the n+1th simulation step. The interval between two simulations corresponds to a defined unit of time. In particular, the interval between two simulations can be between one minute and six hours, preferably between one minute and four hours, and most preferably between one minute and sixty minutes.The smaller the time interval between two simulations, the smoother the results, although the time required to obtain the results increases. The computational unit can optionally be configured to construct the Laplace matrix of the respective output matrix and calculate the eigenvalues ​​of that Laplace matrix, or to calculate the normalized distance between the output matrix of the n+1th simulation step and the output matrix of the nth simulation step. The computational unit can also be configured to select a characteristic value from the values ​​of the output matrices, from the calculated eigenvalues ​​of the optionally constructed Laplace matrices, or from the optionally calculated normalized distances. The selected characteristic value represents a measure of the synchronization of the vibration data traceable to the segments M or of the features calculable from them.The calculation unit is configured to determine, as a result of each simulation, the number of simulation steps required until the selected parameter falls below a predefined threshold close to zero, and to output this number in a unit of time. The time unit, or simulation time, is calculated from the time interval between the (averaged) values ​​entered into the input matrix, which are generated from the vibration data. The calculation unit is also configured to fit a curve to the obtained simulation results and, using a suitable method (local linear regression, for example), to calculate the future point in time at which the curve intersects a predefined lower bound, in particular the x-axis, thus obtaining the failure time as a date.The calculated intersection point serves as an indicator of a malfunction and the predicted end of the sealing assembly's service life. Ultimately, the calculation device is also configured to output the remaining service life of the piston compressor's sealing assembly based on this calculation.

[0042] The advantages of the preventive maintenance system according to the invention essentially result from the advantages already described for the method according to the invention.

[0043] In an advantageous embodiment, the preventive maintenance system according to the invention further comprises at least one position sensor, which is designed to detect the position of the reciprocating part of the piston compressor and is communicatively coupled to the calculation unit. In this embodiment, the preventive maintenance system further comprises at least one vibration sensor, which is designed to detect a time signal x(t) of the vibration of a reciprocating part of the piston compressor and is communicatively coupled to the calculation unit. The calculation unit is configured to divide the detected time signal x(t) of the vibration into a plurality of time intervals and to calculate the average vibration values ​​of the respective time intervals.Furthermore, the calculation device is configured to divide the movement of the reciprocating part into a plurality of, preferably equally sized, segments M based on the determined position and to assign the averaged vibration values ​​to these segments M of the movement.

[0044] This allows the condition of the piston compressor's sealing assembly to be monitored even more precisely and the remaining service life to be predicted more accurately.

[0045] In a preferred embodiment of the preventive maintenance system according to the invention, comprising at least one position sensor, at least one of the vibration sensors is arranged on the cylinder head or crosshead of the piston compressor, which cylinder head or crosshead is assigned to the reciprocating part of the piston compressor.

[0046] It has been found that placing the vibration sensor on the cylinder head allows for particularly reliable statements regarding the condition and remaining service life of the cylinder valves, especially the exhaust valve. Conversely, it has been found that the condition and remaining service life of a piston rod packing can be obtained particularly accurately by placing the vibration sensor on the crosshead.

[0047] The task is further solved by a piston compressor and a preventive maintenance system as described above.

[0048] In a preferred embodiment, the piston compressor is designed as a piston compressor with vertical piston movement. The vertical piston movement ensures low wear of the sealing and guide elements. Such piston compressors are frequently used on marine vessels.

[0049] The task is further accomplished by means of a computer-readable storage medium. The computer-readable storage medium embodies a computer program, which comprises computer-readable program code. The program code is configured, when executed by at least one processor of a computer, to cause the processor to carry out the procedure described herein.

[0050] Within the scope of the present invention, the term "storage medium" includes in particular cloud storage, flash storage and embedded storage.

[0051] The problem is ultimately solved by using the method described herein to monitor the degradation of a piston compressor's sealing assembly. Alternatively or additionally, the problem is solved by using the method described herein to predict the remaining service life, in particular the remaining operating hours or the replacement date, of a piston compressor's sealing assembly. As a further alternative or in addition to the uses mentioned above, the problem is solved by using the method described herein to simulate leakage occurring in a piston compressor's sealing assembly.

[0052] Various embodiments of the invention are explained in more detail below with reference to the drawings, wherein identical or corresponding elements are generally provided with the same reference numerals. The drawings show: Fig. 1 Flowchart showing a process flow according to the invention; Fig. 2 Graphical representation of output matrices 5, 5' after n simulation steps; Fig. 3 Course of the real parts Re of the smallest eigenvalues ​​6d over the n simulation steps of a simulation; Fig. 4a Simulation results 9, 9' over all started simulations; Fig. 4b Local linear regression of the simulation results 9, 9' from Fig. 4a Fig. 5: Criticality index (CI) over time; Fig. 6: Actual and predicted remaining lifetime from the onset of the irreversible degradation process. Fig. 5 ;

[0053] Figure 1Figure 1 shows a flowchart illustrating the sequence of a possible embodiment of the inventive method for determining the degree of wear and predicting the remaining service life of a sealing arrangement of a piston compressor. Following the vibration data 1a provided in step a), which is taken into account by at least one of the vibration sensors belonging to the piston compressor, or the features 1b calculated therefrom, an input matrix 3 is created in step b). For example, this includes generating the data structure from the measured vibration data 1a or the features 1b calculated therefrom such that one column in the input matrix 3 represents the vibration data 1a or the features 1b calculated therefrom at a specific position of the piston, and the rows of the input matrix 3 depict the change in the vibration data 1a or the features 1b over time.This results in either an N × N matrix 3 or an N × M matrix 3. Different positions of the vibration sensor(s) on the piston compressor allow conclusions to be drawn about the criticality of the degradation and the type of sealing arrangement. Following step b), a simulation comprising a plurality of simulation steps is started in step c). For this purpose, the state-field-based simulation of the vibration dynamics is performed in step c.1), in this example using the time-dependent Ginzburg-Landau model 4. When applied correctly, the results of the multitude of existing state-field models 4, with which fault propagation in closed systems can be described, are equivalent. To solve the partial differential equation, the discrete version of the Laplace operator Δ is applied recursively in each simulation step to the output matrix 5 of the previous simulation step. In the Figure 1In the example shown, the output matrix 5 of the first simulation step c.1) is used as the input matrix 3' of the second simulation step c.2), which in turn yields output matrix 5'. Figure 2 The output matrices 5, 5' are shown as a heatmap after a varying number of simulation steps. The figure corresponding to time n = 1 shows a plurality of white and black dots, which represent the values ​​(states) of the output matrix 5 after the first simulation step. After a certain number of simulation steps, the output matrix changes only marginally. In the Figure 1In the illustrated embodiment of the method according to the invention, after each of steps c.1) to cn), the Laplace matrix 6a, 6a', 6a'' is formed from the obtained output matrix 5, 5', 5''. This can be done, for example, by multiplying the respective output matrix 5, 5', 5'' by its transpose (not shown). Subsequently, the eigenvalues ​​6b, 6b', 6b'' of the respective Laplace matrices 6a, 6a', 6a'' of a simulation step are calculated, which is shown in the

[0054] Figure 1 The flowchart for process steps d.1), d.2), and dn) is explicitly shown. For each simulation step, exactly one characteristic value 6c is selected from the calculated eigenvalues ​​6b, 6b', 6b", as exemplified by process steps f.1), f.2), and fn). In the example described here, the selected characteristic value 6c is the second smallest eigenvalue of each simulation step. Figure 3Figure 1 shows, as an example, the behavior of the real parts Re of the second smallest eigenvalues ​​6d over the n simulation steps of a simulation. In procedure step g), the simulation step at which the second smallest eigenvalue first falls below a predefined threshold 8a close to zero is identified. This simulation step is returned as simulation result 9 in simulation time, in particular in hours, minutes, or seconds, and optionally as a future point in time, in particular as a calendar date.

[0055] The simulation time is calculated from the interval between measured values, i.e., the number N of sampling points of the input matrix 3, which is generated from the vibration data 1a or the features 1b that can be calculated from it. For example, when using minute averages, the time it takes for the predefined threshold 8a to fall below the threshold after 750 simulation steps n can be converted into a time unit of approximately half a day by dividing 750 by 1440 minutes (24h x 60min). Subsequently, according to procedure step h), further simulations are performed, repeating procedure steps b), creating the input matrix 3, c), performing a simulation step 4 and generating the Laplace matrix 6a, d) calculating the eigenvalues ​​6b, f) selecting and storing the second smallest eigenvalue 6c, and g) identifying the number of simulation steps required until the threshold is close to zero and returning this value in simulation time.Finally, in process step i), a curve is fitted to the obtained simulation results 9, 9', 9'' of the performed simulations using a suitable method, and the time 7 at which the curve intersects a predefined lower bound 8c is calculated. Figure 4a All started simulations (x-axis) are plotted against the number of simulation steps until the predefined threshold value 8a is undershot (y-axis). Figure 4b The fitting of the obtained simulation results 9, 9', 9" by local linear regression using Loess smoothing over the entire set of simulation results 9, 9', 9'' shown in Figure 4a is given as an example. The failure time or remaining service life is calculated from the intersection of the regression curve with the predefined lower bound 8c, which in this example is the x-axis.

[0056] Figure 5This shows the progression of the criticality index (CI) over time. Based on this progression, it can be determined at which point an irreversible deterioration process begins in the sealing assembly, triggering the calculation of the remaining service life. This point in time can be determined from the statistical distribution of the simulation results, particularly if a certain proportion of the simulation results fall below a threshold value, which is defined in Figure 5 has a value of 1 and is marked by the white line after approximately 92 days (damage onset).

[0057] Figure 6 shows the progression of the prediction error precision over a period of six weeks prior to the actual end of the sealing assembly's service life, i.e., the occurrence of a malfunction or total failure of the sealing assembly, in calendar days. As shown from Figure 6As can be seen, the difference between the predicted end of the service life and the actual occurrence of the damage event is very small, especially in the last day before the actual failure of the sealing arrangement, which is an indicator of a very high accuracy of the predicted end of the service life.

Claims

1. A computer-implemented method for determining the degree of wear and predicting the remaining service life of a sealing arrangement (10) of a reciprocating compressor (11), the method comprising the following steps: a) Providing vibration data (1a), or features (1b) which can be calculated therefrom, of the reciprocating part (12) of the reciprocating compressor (11), which can be traced back to segments M of the movement of the reciprocating part (12) of the reciprocating compressor (11); b) Generating an input matrix (3) from the vibration data (1a) or the features (1b) which can be calculated therefrom, wherein the input matrix (3) has a dimension (M × N) determined by the number of segments M and the number of sampling times N; c) Performing a simulation comprising a plurality of simulation steps by applying a model (4), with which the error propagation in a closed system can be described, on the input matrix (3), wherein each simulation step generates an output matrix (5, 5'), and wherein the output matrix (5) of the nth simulation step is used as the input matrix (3') of the n+1th simulation step; d) Optionally, forming the Laplacian matrix (6a, 6a') for each output matrix (5, 5'), in particular by multiplying the respective output matrix (5, 5') by its transposed matrix (5a, 5a'), and calculating the eigenvalues (6b, 6b') of the respective Laplacian matrix (6a, 6a'); e) Optionally, calculating the normalized distance of the output matrix (5') of the n+1th simulation step to the output matrix (5) of the nth simulation step; f) Selecting exactly one parameter (6c) from the values of the output matrices (5, 5') or from the calculated eigenvalues (6b, 6b') of the optionally formed Laplace matrices (6a, 6a') or from the optionally calculated normalized distances; g) Determining the number of simulation steps that must be performed until the selected parameter (6c) falls below a predefined threshold value (8a) close to zero as the result (9) of the simulation performed, and outputting this number in a time unit; h) repeating the method steps a) to g) at regular discrete time intervals to obtain a plurality of simulation results (9, 9'); and i) Fitting a curve to the simulation results (9, 9') obtained, in particular by local linear regression, and calculating the time (7) at which the curve intersects a predefined lower boundary (8c); wherein the selected parameter (6c) represents a measure for the synchronization of the vibration data (1a) traceable to the segments M or features (1b) calculable therefrom, and wherein the calculated intersection point is regarded as an indicator for the occurrence of a malfunction, in particular for a total failure, and the predicted end of the service life of the sealing arrangement (10).

2. The method according to claim 1, wherein the second smallest eigenvalue (6d) of the Laplace matrix (6a) is selected as the parameter (6c), and wherein the predefined threshold value (8a) is approximately zero, in particular the threshold value (8a) has an amount of 0.000025.

3. The method according to claim 1 or 2, wherein the vibration data (1a) provided or the features (1b) which can be calculated therefrom are obtained by the following sub-steps: - Detecting at least one time signal x(t) of the vibration of a reciprocating part (12) of the reciprocating compressor (11) by at least one vibration sensor (13); - Optionally, calculating the features (1b) from the recorded time signal x(t); - Dividing the detected time signal x(t) of the vibration or the optionally calculated features (1b) into a plurality of time segments (1d) and forming the average values of the detected time signal x(t) or the optionally calculated features (1b) in the respective time segments (1d) to obtain averaged vibration values (2a) or averaged features (2b); - Detecting the position of the reciprocating part (12) of the reciprocating compressor (11) by at least one position sensor (14) and providing the detected position data (1c); - dividing the movement of the reciprocating part (12) into a plurality of, preferably equally sized, segments M, based on the detected position data (1c); and - assigning the averaged vibration values (2a) or the averaged features (2b) to the segments M of the movement in such a way that the respective averaged vibration values (2a) or the respective averaged features (2b) always correspond to the same segments M of the movement of the reciprocating part (12).

4. The method according to one of the preceding claims, wherein the prediction of the time (7) at which the selected parameter (6c) falls below the predefined threshold value (8a) is made only after the parameter (6c) reaches a predetermined criticality value (8b) marking the onset of an irreversible deterioration process in the sealing arrangement (10).

5. The method according to any one of the preceding claims, wherein the model (4) applied to the input matrix (3) is the time-dependent Ginzburg-Landau model (TDGL).

6. The method according to one of the preceding claims, further comprising a data transmission step for transmitting the vibration data (1a) or features (1b) to at least one ground station (15) and / or for transmitting the predicted remaining service life of the sealing arrangement (10) to a user interface (11a) of the reciprocating compressor (11), wherein at least the simulation steps and the prediction of the time (7) at which the selected parameter (6c) falls below a predefined threshold value (8a) are carried out in whole or in part on the ground with the ground station (15).

7. The method according to one of the preceding claims, wherein the sealing arrangement (10) is a piston rod packing (10a) sealing against a piston rod, or a cylinder valve (10b) delimiting a compression chamber.

8. The method according to one of the preceding claims, wherein the execution of the method is started automatically at a time selectable by a user or at fixed time intervals.

9. The method according to any one of the preceding claims, further comprising the sub-step: - Transmitting the identified time (7) and / or the predicted end of the service life of the sealing arrangement (10) to a user, in particular in real time; and / or - Estimating the degree of wear of the sealing arrangement (10) from the behavior of the parameter (6c) over the number of simulation steps carried out and transmitting the degree of wear thus obtained to a user, in particular in real time.

10. A preventive maintenance system (20) for use with a reciprocating compressor (11), the preventive maintenance system (20) having a calculating means (21) comprising at least one processor (22), the calculating means (21) being configured: - to receive position data (1c) of a reciprocating part (12) of the reciprocating compressor (11); - to receive vibration data (1a), or features (1b) that can be calculated therefrom, of the reciprocating part (12) of the reciprocating compressor (11); - to assign the vibration data (1a) or features (1b) to segments M of the movement of the reciprocating part (12) of the reciprocating compressor (11); - to generate an input matrix (3) from the received vibration data (1a) or features (1b); - to perform a plurality of simulations each comprising a plurality of simulation steps by applying a model (4), with which the error propagation in a closed system can be described, on the input matrix (3), wherein each simulation step generates an output matrix (5), and wherein the output matrix (5) of the nth simulation step is used as the input matrix (3') of the n+1th simulation step; - Optionally, to form the Laplace matrix (6a) of the respective output matrix (5) and calculate the eigenvalues (6b) of the respective Laplace matrix (6a); - Optionally, to calculate the normalized distance of the output matrix (5') of the n+1th simulation step to the output matrix (5) of the nth simulation step; - To select a parameter (6c) from the values of the output matrices (5, 5') or from the calculated eigenvalues (6b, 6b') of the optionally formed Laplace matrices (6a, 6a') or from the optionally calculated normalized distances, wherein the selected parameter (6c) represents a measure for the synchronization of the vibration data (1a) traceable to the segments M or features (1b) calculable therefrom; - To determine, as a result (9, 9') of the simulation carried out in each case, the number of simulation steps that must be carried out until the selected parameter (6c) falls below a predefined threshold value (8a) close to zero, and output this number in a time unit; - To fit a curve to the obtained simulation results (9, 9'), in particular by local linear regression, and calculate the time (7) at which the curve intersects a predefined lower boundary (8c), the calculated intersection point being regarded as an indicator for the occurrence of a malfunction and the predicted end of the service life of the sealing arrangement; and - To output the remaining service life of the sealing arrangement (10) of the reciprocating compressor (11).

11. The preventive maintenance system (20) according to claim 10, further comprising at least one position sensor (14) which is designed to detect the position of the reciprocating part (12) of the reciprocating compressor (11) and is communicatively coupled to the calculation device (21), and at least one vibration sensor (13), which is designed to detect a time signal x(t) of the vibration of a reciprocating part (12) of the reciprocating compressor (11) and is communicatively coupled to the calculation device (21), the calculation device (21) being configured to divide the detected time signal x(t) of the vibration into a plurality of time segments (1d) and form the averaged vibration values (2a) of the respective time segments (1d), to divide the movement of the reciprocating part (12) into a plurality of segments M, preferably of equal size, on the basis of the determined position, and to assign the averaged vibration values (2a) to these segments M of the movement in such a way that the respective averaged vibration values (2a) always correspond to the same segments M of the movement of the reciprocating part (12).

12. The preventive maintenance system (20) according to claim 11, wherein at least one of the vibration sensors (13) is arranged on the cylinder head (16) or crosshead (17) of the reciprocating compressor (11) associated with the reciprocating part (12).

13. A reciprocating compressor (11) comprising a preventive maintenance system (20) according to any one of claims 10 to 12.

14. The reciprocating compressor (11) according to claim 14, designed as a reciprocating compressor (11) with vertical piston movement.

15. A computer-readable storage medium (30) embodying a computer program (31), the computer program (31) comprising computer-readable program code (32) which, when executed by at least one processor (22) of a computer, is configured to cause the processor (22) to perform the method according to any one of claims 1 to 9.

16. Use of the method according to one of claims 1 to 9 for monitoring the degradation of a sealing arrangement (10) of a reciprocating compressor (11) and / or for predicting the remaining service life, in particular predicting the remaining operating hours or the replacement date, of a sealing arrangement (10) of a reciprocating compressor (11) and / or for reproducing the leakage occurring at a sealing arrangement (10) of a reciprocating compressor (11).