DE202023106044U1 - A system for carrying out the performance assessment of rolling element bearings - Google Patents

Abstract

A system for performing performance degradation assessment of rolling bearings, comprising:
a signal detection unit that uses a bearing unit to detect bearing vibration signals;
a decomposition processing unit for decomposing the detected bearing vibration signal into product functions (PFs), the decomposition processing unit using dedicated hardware configured using local mean decomposition (LMD);
a feature extraction module that uses a configured processor to extract envelope signals from product functions, the processor being provided with decomposed product functions as input and the processor then extracting appropriate defect features or envelope signals from input product functions;
a main processing unit with a dedicated processor that is trained by using the extracted features from the product functions as input for training, resulting in the development of a trained SC-PI (Spectral Clustering-Parting Index) model with optimal number of clusters;
a real-time bearing condition assessment module in conjunction with the main processing unit, configured to evaluate the condition of the rolling bearing using a special processor trained to do so based on extracted features used as input for training the main processing unit become; And
a user interface for displaying the results of the final condition assessment of the rolling bearing and for displaying a confidence value calculated by the real-time bearing condition assessment module to indicate the reliability of the assessment.

Description

FIELD OF THE INVENTION

The present disclosure relates to a system for performing performance degradation evaluation of rolling bearings, particularly to a system that uses local averaging and special clustering to perform condition evaluation of rolling bearings in mechanical machines.

BACKGROUND OF THE INVENTION

Rolling bearings (REBs) are critical components in rotating machinery and their failure poses a significant threat to performance, reliability and safety and results in significant financial losses. Statistics show that bearing failures are one of the leading causes of engine and aircraft engine failures. Given this fact, the need for effective condition assessment of REBs becomes clear to prevent malfunctions and reduce downtime.

Over the years, efforts have been made to develop prognostic modules for assessing REB condition. These modules use advanced artificial intelligence tools, signal processing techniques and nonlinear theories. In this context, there is a need for an invention that takes advantage of Local Mean Decomposition (LMD) and Spectral Clustering (SC) to create an assessment model and degradation index for bearings. The invention must address the critical need for accurate REB condition assessment and aim to mitigate the risks associated with bearing failures in machines, increase reliability and minimize financial losses by using innovative techniques such as LMD and SC to provide a solution for this urgent problem.

From the above discussion, it is clear that there is a need for a system that uses local averaging and special clustering to assess the condition of rolling bearings in mechanical machines.

SUMMARY OF THE INVENTION

The present disclosure relates to a system for performing performance degradation evaluation of rolling bearings. The invention presents a novel system for assessing the condition of rolling bearings, critical components in mechanical systems. The system first acquires vibration signals from a bearing test bench and then decomposes them into product functions (PFs) using Local Mean Decomposition (LMD). Envelope signals corresponding to the PFs are then extracted to highlight impulsive information. A feature space is constructed using the PFs envelope signals, and features such as Singular Value (SV) and Singular Value Envelope (SVE) are used for condition assessment due to their exceptional proficiency. The features derived from LMD are classified using Spectral Clustering with Parting Index (SC-PI) to build a scoring model, where the optimal number of clusters is determined by PI. A comparative analysis with existing indices demonstrates the superior performance of this system in evaluating the optimal cluster numbers and contributes to a more accurate and reliable assessment of bearing condition in various mechanical systems.

The present disclosure aims to provide a system for performing performance degradation evaluation of rolling bearings. The system includes: a signal acquisition unit that uses a bearing test unit to acquire bearing vibration signals; a decomposition processing unit for decomposing the detected bearing vibration signal into product functions (PFs), the decomposition processing unit using dedicated hardware configured using local mean decomposition (LMD); a feature extraction module that uses a configured processor to extract envelope signals from product functions, the processor being provided with decomposed product functions as input and the processor then extracting appropriate defect features or envelope signals from input product functions; a main processing unit with a dedicated processor that is trained by using the extracted features from the product functions as input for training, resulting in the development of a trained SC-PI (Spectral Clustering-Parting Index) model with optimal number of clusters; a real-time bearing condition assessment module in conjunction with the main processing unit, configured to evaluate the condition of the rolling bearing using a special processor trained to do so based on extracted features used as input for training the main processing unit become; and a user interface for displaying the results of the final condition assessment of the rolling bearing and for displaying a confidence value calculated by the real-time bearing condition assessment module to indicate the reliability of the assessment.

In one embodiment, the bearing test stand includes components, namely a speed control unit, a first accelerometer, a load disk, a second accelerometer, a first bearing housing, a dog clutch, a proximity sensor, an AC motor, a computer system and a data acquisition unit, second bearing housing and a shaft.

In one embodiment, feature extraction is performed on obtained features from training the model for evaluation over the entire failure history of a degraded bearing and some features of a healthy bearing.

The system for performing performance degradation evaluation of rolling bearings further includes: a clustering module with a programmed processor configured to perform spectral clustering (SC) to group and classify the extracted defect features based on similarity, which the design a feature space that is then used to train the SC-PI (Spectral Clustering-Parting Index) model; and a split index processing module configured to calculate the optimal number of clusters in spectral clustering based on the synergic association index (SAI) and the separation index (SI), resulting in improved precision of condition assessment.

In one embodiment, the feature space created by performing spectral clustering includes features, namely singular value (SV) and singular value entropy (SVE) of the extracted envelope signals from product functions, and are used to train the SC-PI (Spectral Clustering-Parting) index ) model in the main processing unit.

In one embodiment, the real-time bearing condition assessment module, in conjunction with the already trained SC-PI (Spectral Clustering-Parting Index) model, takes a test feature vector (one at a time) and clusters it to the trained SC-PI (Spectral) model. Parting Index) model that calculates the distance from a normal state cluster and then calculates the confidence value (CV), which ranges between 0 and 1, where a value close to 1 indicates a healthy state and a value close to 0 indicates deterioration Rolling bearing indicates.

In one embodiment, the user interface includes a display for displaying the calculated confidence value and the condition of rolling bearings classified as healthy, faulty and failed bearings.

An object of the present disclosure is to provide a system for performing performance degradation evaluation of rolling bearings.

A further aim of the present disclosure is to use techniques, namely local averaging and special clustering, to assess the condition of rolling bearings in mechanical machines.

Another goal of the present disclosure is to improve the accuracy and reliability of condition assessment of rolling bearings in mechanical systems and thus address the challenge of high-dimensional data.

Another objective of the present disclosure is to provide a means for effectively visualizing pulse information in vibration signals, thereby making it easier to identify potential bearing failures.

Another goal of the present disclosure is to use a novel technique called "parting index" to determine the optimal number of clusters for spectral clustering, which provides a more systematic approach to analyzing bearing condition data.

Another goal of the present disclosure is to demonstrate the superior performance of the proposed system in evaluating optimal cluster counts compared to existing evaluation techniques and ultimately contribute to safer and more reliable mechanical systems.

In order to further illustrate the advantages and features of the present disclosure, a more detailed description of the invention will be made with reference to specific embodiments thereof shown in the accompanying drawings. It is understood that these drawings represent only typical embodiments of the invention and are therefore not to be viewed as limiting its scope. The invention is described and explained in more detail and in greater detail with reference to the accompanying drawings.

BRIEF DESCRIPTION OF THE FIGURES

• 1 veranschaulicht ein Blockdiagramm eines Systems zur Durchführung einer Zustandsbewertung von Wälzlagern gemäß einer Ausführungsform der vorliegenden Offenbarung; und
• 2 veranschaulicht den Arbeitsrahmen des vorgeschlagenen Systems gemäß einer Ausführungsform der vorliegenden Offenbarung.

These and other features, aspects and advantages of the present disclosure will be better understood by reading the following detailed description with reference to the accompanying documents th drawings, in which like reference numerals represent like parts throughout the drawings, where:

• 1 illustrates a block diagram of a system for performing condition assessment of rolling bearings in accordance with an embodiment of the present disclosure; and
• 2 illustrates the working framework of the proposed system according to an embodiment of the present disclosure.

Additionally, experienced craftsmen will recognize that elements in the drawings are presented for convenience and may not necessarily have been drawn to scale. For example, the flowcharts illustrate the method through key steps that help improve understanding of aspects of the present disclosure. Additionally, in view of the construction of the device, one or more components of the device may have been represented in the drawings by conventional symbols, and the drawings may show only the specific details relevant to understanding the embodiments of the present disclosure around the drawings not to be obscured by details that would be readily apparent to one of ordinary skill in the art who would benefit from the description herein.

DETAILED DESCRIPTION:

In order to promote understanding of the principles of the invention, reference will now be made to the embodiment shown in the drawings and specific language will be used to describe the same. It is to be understood, however, that this is not intended to limit the scope of the invention, since changes and further modifications to the system illustrated and further applications of the principles of the invention set forth therein are contemplated as would normally occur to one skilled in the art Technology to which the invention relates.

It will be understood by those skilled in the art that the foregoing general description and the following detailed description are exemplary and illustrative of the invention and are not intended to limit the same.

References in this specification to “an aspect,” “another aspect,” or similar language mean that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present disclosure is included. Therefore, the phrases “in one embodiment,” “in another embodiment,” and similar phrases in this specification may, but not necessarily, refer to the same embodiment.

The terms “includes,” “comprising,” or other variations thereof are intended to cover non-exclusive inclusion, such that a process or method that includes a list of steps not only includes those steps, but may include other steps not specifically listed or following this inherent in the process or method. Likewise, one or more devices or subsystems or elements or structures or components prefixed with "comprises...a" do not exclude, without further limitation, the existence of other devices or other subsystems or other elements or other structures from other components or additional devices or additional subsystems or additional elements or additional structures or additional components.

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as generally understood by one of ordinary skill in the art to which this invention pertains. The system, methods and examples provided herein are for illustrative purposes only and are not intended to be limiting.

Embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.

The functional units described in this specification are identified as devices. A device may be implemented in programmable hardware devices such as processors, digital signal processors, central processing units, field programmable gate arrays, programmable array logic, programmable logic devices, cloud processing systems, or the like. The devices can also be implemented in software for execution by different types of processors. An identified device may contain executable code and, for example, include one or more physical or logical blocks of computer instructions, which may be organized, for example, as an object, procedure, function, or other construct. However, the executable file of an identified device does not have to be physically located together, but may consist of different instructions stored in different locations, which, when logically put together, form the device fulfill the stated purpose of the device.

In fact, a device or module's executable code could be a single instruction or multiple instructions, and even distributed across several different code segments, between different applications and across multiple storage devices. Similarly, operational data may be identified and represented herein within the device and may be embodied in any suitable form and organized into any suitable type of data structure. The operational data may be collected as a single data set or distributed across various locations, including various storage devices, and may exist, at least in part, as electronic signals in a system or network.

References in this specification to “a selected embodiment,” “an embodiment,” or “an embodiment” mean that a particular feature, structure, or feature described in connection with the embodiment is disclosed in at least one embodiment of the Item is included. Therefore, the expressions “a selected embodiment,” “in one embodiment,” or “in one embodiment” used elsewhere in this specification do not necessarily refer to the same embodiment.

Additionally, the features, structures, or properties described may be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to provide a thorough understanding of embodiments of the disclosed subject matter. However, one skilled in the art will recognize that the disclosed subject matter may be practiced without one or more of the specific details or with other methods, components, materials, etc. In other cases, well-known structures, materials, or processes may not be shown or described in detail so as not to obscure aspects of the subject matter disclosed.

According to exemplary embodiments, the disclosed computer programs or modules may be executed in many exemplary ways, for example, as an application residing in a device's memory or as a hosted application running on and communicating with a server via a device application or browser Set of standard protocols such as TCP/IP, HTTP, XML, SOAP, REST, JSON and other sufficient protocols. The disclosed computer programs may be written in exemplary programming languages that are executed from the device's memory or from a hosted server, such as BASIC, COBOL, C, C++, Java, Pascal, or scripting languages such as JavaScript, Python, Ruby, PHP, Perl or other suitable programming languages.

Some of the disclosed embodiments involve or otherwise involve data transmission over a network, such as the transmission of various inputs or files over the network. The network may include, for example, one or more of the following: Internet, Wide Area Networks (WANs), Local Area Networks (LANs), analog or digital wired and wireless telephone networks, e.g. B. a PSTN, Integrated Services Digital Network (ISDN), a cellular network and Digital Subscriber Line (xDSL), radio, television, cable, satellite and / or other transmission or tunneling mechanisms for data transmission. The network may include multiple networks or subnetworks, each of which may include, for example, a wired or wireless data path. The network may include a circuit-switched voice network, a packet-switched data network, or any other network capable of transmitting electronic communications. The network may include, for example, networks based on Internet Protocol (IP) or Asynchronous Transfer Mode (ATM), and may support voice using, for example, VoIP, Voice over ATM, or other comparable protocols used for voice data communications. In one implementation, the network includes a cellular network configured to enable the exchange of text or SMS messages.

Examples of the network include a Personal Area Network (PAN), a Storage Area Network (SAN), a Home Area Network (HAN), a Campus Area Network (CAN) and a Local Area Network (LAN), a Wide Area Network (WAN), a Metropolitan Area Network (MAN), a Virtual Private Network (VPN), an Enterprise Private Network (EPN), Internet, a Global Area Network (GAN) and so on.

1 shows a block diagram of a system (100) for carrying out a condition assessment of rolling bearings according to an embodiment of the present disclosure. The system (100) comprises a signal acquisition unit (102) which uses a bearing test unit (102a) to acquire bearing vibration signals.

In one embodiment, a decomposition processing unit (104) is used to decompose the detected bearing vibration signal into product functions (PFs) (104a), the decomposition processing unit (104) using dedicated hardware configured using local mean decomposition (LMD). .

In one embodiment, a feature extraction module (106) uses a configured processor to extract envelope signals from the product function (104a), wherein decomposed product functions (104a) are provided as input to the processor and the processor then extracts appropriate error features (106a) or envelope signals from input product functions (104a). 104a).

In one embodiment, a main processing unit (108) with a dedicated processor is trained by using the extracted features (106a) from the product features (104a) as input for training, resulting in the development of a trained SC-PI (Spectral Clustering) Parting index model (108a) with optimal number of clusters.

In one embodiment, a real-time bearing condition assessment module (110), in conjunction with the main processing unit (108), is configured to assess the condition of the rolling bearing using a special processor trained to do so based on the extracted data characteristics (106a), which are used as input for training the main processing unit (108).

In one embodiment, a user interface (112) for displaying the results of the final assessment of the condition of the rolling bearing and also for displaying a confidence value calculated by the real-time bearing condition assessment module to indicate the reliability of the assessment.

In one embodiment, the bearing unit (102a) includes components, namely a speed control unit, a first accelerometer, a load disk, a second accelerometer, a first bearing housing, a dog clutch, a proximity sensor, an AC motor, a computer system and data recording unit, a second bearing housing and a shaft .

In one embodiment, feature extraction to obtain features for training the model for evaluation is performed over the entire failure history of a degraded bearing and some features of a healthy bearing.

group and classify extracted error features (106a) based on similarity, facilitating the construction of a feature space, which are then used to train the SC-PI (Spectral Clustering-Parting Index) model; and a split index processing module (116) configured to calculate the optimal number of clusters in the spectral clustering based on the synergic association index (SAI) and the separation index (SI), resulting in improved precision of the condition assessment . The feature space created by spectral clustering includes features called Singular Value (SV) and Singular Value Entropy (SVE) of the extracted envelope signals from product functions, and is used to train the SC-PI (Spectral Clustering-Parting Index) model Main Processing Unit (108).

In one embodiment, the real-time bearing condition assessment module (110), in conjunction with the already trained SC-PI (Spectral Clustering-Parting Index) model, takes a test feature vector (one at a time) and feeds it to the trained SC-PI model. Spectral Clustering-Parting Index (PI) model, which calculates the distance to a normal state cluster and then calculates the confidence value (CV), which ranges between 0 and 1, with a value close to 1 indicating “healthy” and a value close to 0 indicates a deterioration in rolling behavior element bearing.

In one embodiment, the user interface (112) includes a display (112a) for displaying the calculated confidence value and the condition of rolling bearings classified as healthy, faulty and failed bearings.

In one embodiment, the signal acquisition unit (102), the decomposition processing unit (104), the feature extraction module (106), the main processing unit (108) and a real-time bearing condition assessment module (110) may be programmably implemented hardware devices such as processors, digital signal processors, central processing units, field programmable gates. Arrays, programmable array logic, programmable logic devices, cloud processing systems or the like.

2 illustrates the working framework of the proposed system according to an embodiment of the present disclosure.

Referring to Bearing vibration signals are first obtained, with feature extraction then being carried out from the signals obtained is carried out, with the further functioning of the system being described step by step below.

Step 1: The acquired bearing vibration signal is decomposed into product functions (PFs) by implementing Local Mean Decomposition (LMD). The notation PF 1 ,..., PFn denotes multiple sets of PFs. The first six PFs in this publication have been awarded. Each level of signal degradation reveals the presence of the first 6 PFs, which are shown to be highly correlated with the original signal. From the entropic interpretation it is clear that these first 6 PFs contain more than 90% of the total energy of the signal, which is enough to analyze the characteristics of the operating component.

Step 2: The product features are used for feature extraction, where relevant defect features related to the condition of the bearing are extracted from the product features obtained by decomposing the original signal.

1. 1. Berechnen Sie die Hilbert-Transformation des erhaltenen PF gemäß Gleichung (1). $$h_i\left(t\right)=H\left[P{F}_i\left(t\right)\right]=\frac{1}{\pi }{\displaystyle {\int }_{-\infty }^{+\infty }\frac{P{F}_i\left(\tau \right)}{t-\tau }}d\tau$$
   
   Dabei ist PFi (t) der i-te PF und hi(t) stellt die Hilbert-Transformation desselben dar.
2. 2. Berechnen Sie die Hüllkurve PF env jedes analytischen Signals hi(t), indem Sie dessen Absolutwert nehmen.

PF envelope signals: Due to the periodic excitation caused by the repeated contact between the defective surfaces and the rollers, the bearing signals are highly modulated, making it difficult to isolate the sensitive component that can indicate the performance. When the envelope signal is examined, impulsive information can be identified more clearly. To calculate the envelope signal, the following measures must be taken:

1. 1. Calculate the Hilbert transform of the obtained PF according to equation (1). $$H_i\left(t\right)=H\left[P{F}_i\left(t\right)\right]=\frac{1}{\pi }{\displaystyle {\int }_{-\infty }^{+\infty }\frac{P{F}_i\left(\tau \right)}{t-\tau }}d\tau$$
   
   Here PFi (t) is the ith PF and hi(t) represents the Hilbert transformation of the same.
2. 2. Calculate the envelope PF env of each analytical signal hi(t) by taking its absolute value.

Singular value entropy (SVE) of the _PF environment: A general measure of disorder is the entropy-centered feature, which is calculated from an arbitrary preset set of values. Higher energy content is indicated by lower individual value entropy because organization in a signal requires more energy. SV's consistency over fragmented signals is its main selling point. Fluctuations in noise levels can be quickly detected. Therefore, as shown in Eq. (2) The matrix PF _{env_n} of the PF envelope signals is decomposed into three matrices: $$P{F}_{en{v}_k\times n}={\left[P{F}_{envp}\right]}_{k\times k}{\left[P{F}_{envs}\right]}_{k\times n}{\left[P{F}_{envq}\right]}_{n\times n}^T$$

In the above relationship, [PF _envs ] _k×n represents the desired matrix comprising singular values, which can be further represented as _SV = _{[S V1} , S _V2 , ... , S _Vt ], where n is the sample length and k stands for the number of PFs taken into account. In this case, t is less than k and n. In addition to PF envs, two orthogonal matrices are created, PF _envp and PF _envq .

wobei w _i = Svi /S das Segment der Gesamtenergie S aller Proben im i-ten PF ist. Die Feature-Sets werden wie folgt zusammengesetzt: fs = [S _V1 , ... , S _Vn, Sv _en1, ... , Sv _en n ] für Klassifizierungszwecke.The singular value entropy can be calculated using Eq. can be calculated from the SVs. (3) as stated below. $$S{v}_{en}=-{\displaystyle \sum _{i=1}^k{\omega }_i\text{log}{\omega }_i}$$

where w _i = Svi /S is the segment of the total energy S of all samples in the i-th PF. The feature sets are composed as follows: fs = [S _V1 , ... , S _Vn , Sv _en1 , ... , Sv _en n ] for classification purposes.

The above features are calculated via PFs and taken into account during the learning phase. Feature extraction is performed for the entire fault history of a defective bearing as well as for some attributes of the healthy bearing.

Step 3: Perform spectral clustering to group and classify the extracted defect features based on their similarity, enabling structured bearing condition analysis.

Step 4: Parting Index (PI) is used to determine the optimal number of clusters in spectral clustering based on a Synergy Association Index (SAI) and Separation Index (SI), thereby improving the precision of condition assessment.

Step 5: Develop an artificial intelligent model for evaluation, providing the features for developing the Spectral Clustering-Parting Inxed (SC-PI) model with the optimal number of clusters as input, based on the grouped features the Model is trained to assess the condition of the rolling bearing. The trained SC-PI model includes centers and indices associated with the fault state, intermediate state, and normal state. The feature vectors corresponding to the assigned indices become grouped and their corresponding cluster centers are calculated.

wobei Xi den Merkmalsvektor darstellt und ce das Clusterzentrum eines ausgewählten gesunden Vektors darstellt. Dieser SC-QE stellt den gewünschten HI dar, der weiter als Konfidenzwertkurve (CV) gemäß Gl. dargestellt werden kann. (5) unten angegeben. Die CV-Werte liegen zur einfacheren Visualisierung von Gesundheitszuständen zwischen 0 und 1, wobei Werte näher an 1 einen normalen und gesunden Zustand anzeigen, während Werte näher an 0 eine Verschlechterung anzeigen. $$CV=\text{exp}\left(-\left(QE\right)/c\right)$$

Step 6: Real-time testing of the trained model is performed using a test feature vector, where after developing the model, the test feature vectors are taken (one at a time) and compared with the trained SC-PI model. Its distance from the center of the normal state cluster is calculated according to Equation. (4) as stated below. $$SC-QE={\Vert x_i-c{e}_j\Vert }^2$$

where Xi represents the feature vector and ce represents the cluster center of a selected healthy vector. This SC-QE represents the desired HI, which is further expressed as a confidence value curve (CV) according to Eq. can be displayed. (5) given below. CV values range from 0 to 1 for easier visualization of health conditions, with values closer to 1 indicating a normal and healthy condition, while values closer to 0 indicate deterioration. $$Cv=\text{exp}\left(-\left(QE\right)/c\right)$$

Here c represents the scaling factor, which can be varied to get the desired idea.

In one embodiment, validation of the evaluation approach was demonstrated by presenting vibration signal samples from a failure test on a test stand. An adjustable or movable load plate is used to change the load distribution and a robust cast frame eliminates unwanted vibrations. Rapid increases in the RMS value of the bearing signal sample can be used to predict when a bearing is nearing the end of its useful life. The calculation uses 2560 samples or 0.1 seconds of the signals with a sampling frequency of 25.6 kHz. An interval of 10 minutes is maintained during the collection of vibration data. For the purpose of the study, the downtimes for two bearings are assumed to be 12250 and 14990 minutes. NI Lab VIEW and MATLAB R2015b are the platforms used for data acquisition and calculation, and the workstation is equipped with an Intel® Core i5-45705@2.90 GHz CPU.

In one embodiment, a feature set is created by grouping twelve features calculated from a signal sample. For training and testing purposes, this set was split into two sets. The training set includes features derived from samples taken over the life of Bearing B1 (1225 samples) and healthy signal samples taken over the life of Bearing B2 (500 samples). The SC-PI model is trained with the range of functions obtained in this way. The result is a normalized set of eigenvectors. The k-medoids clustering method further categorizes these vectors. The SAI-based criterion is used to estimate the number of optimal clusters, which in our case is three. The results of the test phase show that the classes are clearly separated from each other without any influence from different families. The centers and indices associated with error, intermediate, and normal states are provided. The cluster centers for the feature vectors corresponding to the assigned indices are determined as "Knormal", "Kintermediate" and "Kfailure". The distance between the test feature and the center of the normal state cluster is determined when the test features are projected onto the feature space. In order to know the status of the labor camp in real time, this process is carried out online. Note that each test function is calculated in the same way as in the training phase. This process allows the operator to assess the situation in real time and maximize accuracy. When calculating the degradation index in the multi-dimensional feature space, the test features are projected individually to maximize precision and reduce interference. The results of the test showed that for bearing 1, the CV decrease was found to be significant at a time step of 830. The SC-CV1 was between 1 and 0.9995, which is a clear indicator of normality. After an 830-fold step, there is a noticeable variation with continuous CV increase and decrease, a sign of crack propagation, as well as the continuous formation of an abraded surface and the flattening of the fault by the rolls. The fluctuation became significant, reflecting the severe degradation level of the operation when the time step reached 1100. If this condition is reached, the bearing will eventually fail, as shown at time step 1222. The development curve of bearing deterioration was successfully seen and identified for bearing 2 as well as displacement in different levels.

Camp 2 began to deteriorate slightly around time step 1289, and by 1350 the camp had reached a serious condition. However, the collapse occurred in 1499. Similar to Camp 1, the proposed system development of the SC-CV2 curve indicates the various degrees of deterioration and eliminates any errors lerability in determining the occurrence of an impending fault.

In one embodiment, a comparison is made between existing techniques and the proposed system. For numerous reasons, a comparison of the spectrum approach and degradation index assessment with the existing methods is carried out, as well as a comparison of various assessment indices for an ideal number of clusters with the proposed index. The SC-PI model is first compared to classic time domain functions. To evaluate the storage performance, a comparison is made between the sequence and well-known classification-based techniques using K-medoids, SOM and a Gaussian mixture modal. Performance, accuracy and consistency comparisons are used to show how the proposed model is preferable to the current model. The findings obtained with the proposed system demonstrate its superiority over the widely used PDA storage methods. Experimental findings demonstrate the feasibility of the method in terms of initial fault detection, prediction accuracy, clarity during the healthy phase, and subsequent condition detection. The proposed PI has been compared to a number of other current indices and is superior in several respects, including monotonicity, noise effect, density effect and the presence of subclusters.

To create a reliable evaluation model, the feature space collected from all failure signal samples of a bearing is required. Other rotating machine parts, including gears and turbines, can have PDA using the same developed system. In addition, the proposed system has the advantage over comparable systems in that it requires less computing time and can train with even smaller amounts of data.

• • Erhöhte Genauigkeit und Zuverlässigkeit bei der Bewertung von Wälzlagern.
• • Reduzierung der Dimensionsprobleme durch die Kombination von LMD und SC.
• • Optimierung des Clustering-Prozesses mithilfe des Parting Index (PI).
• • Verbesserte Sicherheit und Leistung mechanischer Systeme, die die Erfindung integrieren.

Advantages:

• • Increased accuracy and reliability when evaluating rolling bearings.
• • Reducing dimensional problems by combining LMD and SC.
• • Optimization of the clustering process using the Parting Index (PI).
• • Improved safety and performance of mechanical systems incorporating the invention.

The disclosed system provides a significant advance in rolling bearing assessment by addressing the challenges of high-dimensional data and providing a more accurate and reliable way to determine bearing condition. The invention finds wide application in mechanical engineering and ensures the integrity and security of various systems.

The drawings and the description above provide examples of embodiments. Those skilled in the art will recognize that one or more of the elements described can certainly be combined into a single functional element. Alternatively, certain elements can be divided into several functional elements. Elements of one embodiment may be added to another embodiment. For example, the orders of the processes described herein may be changed and are not limited to the manner described herein. Additionally, the actions of a flowchart do not have to be implemented in the order shown; Not all actions necessarily have to be carried out. Even those actions that are not dependent on other actions can be carried out in parallel with the other actions. The scope of the embodiments is in no way limited by these specific examples. Numerous variations, whether explicitly stated in the specification or not, such as: B. Differences in structure, dimensions and material use are possible. The scope of the embodiments is at least as broad as indicated by the following claims.

Advantages, other benefits, and solutions to problems have been described above with respect to specific embodiments. However, the advantages, benefits, solutions to problems and any components that may cause a benefit, advantage or solution to occur or become more pronounced should not be construed as a critical, necessary or essential function or component of any or all of the claims.

CREDENTIALS

100

A system for assessing the condition of rolling bearings.

102

A signal acquisition unit

104

Decomposition processing unit

104a

Product Features (Pfs)

106

Feature extraction module

106a

Extracted fault features

108

Main processing unit

108a

Trained SC-PI model (Spectral Clustering-Parting Index)

110

Real-time bearing condition assessment module

112

user interface

112a

Advertisement

114

Clustering module

116

A partial index processing module

202

Bearing signal

204

Product function

206

Selection Of Pfs With High Energy Content

208

Selected PF

210

Extracted features

212

Training functions

214

Executing Spectral Clustering

216

Clustered Normalized Vectors (N)

218

Mapping the cluster index of N to training features

220

Clustered training set

222

Calculation of cluster centers

224

Degradation model

226

Assessment of deterioration

228

Test A Range Of Features

Claims (7)

A system for performing performance degradation assessment of rolling bearings, comprising: a signal detection unit that uses a bearing unit to detect bearing vibration signals; a decomposition processing unit for decomposing the detected bearing vibration signal into product functions (PFs), the decomposition processing unit using dedicated hardware configured using local mean decomposition (LMD); a feature extraction module that uses a configured processor to extract envelope signals from product functions, the processor being provided with decomposed product functions as input and the processor then extracting appropriate defect features or envelope signals from input product functions; a main processing unit with a dedicated processor that is trained by using the extracted features from the product functions as input for training, resulting in the development of a trained SC-PI (Spectral Clustering-Parting Index) model with optimal number of clusters; a real-time bearing condition assessment module in conjunction with the main processing unit, configured to evaluate the condition of the rolling bearing using a special processor trained to do so based on extracted features used as input for training the main processing unit become; And a user interface for displaying the results of the final condition assessment of the rolling bearing and for displaying a confidence value calculated by the real-time bearing condition assessment module to indicate the reliability of the assessment. System after Claim 1 , wherein the bearing test unit includes components, namely a speed control unit, a first accelerometer, a load disk, a second accelerometer, a first bearing housing, a dog clutch, a proximity sensor, an AC motor, a computer system, a data acquisition unit, a second bearing housing and a shaft. System after Claim 1 , where feature extraction is performed on obtained features from training the model for evaluation over the entire failure history of a degraded bearing and some features of a healthy bearing. System after Claim 1 , further comprising: a clustering module with a programmed processor configured to perform spectral clustering (SC) to group and classify the extracted defect features based on similarity, facilitating the construction of a feature space that is then used to Training of SC-PI (Spectral Clustering) is used. model with separation index; And a division index processing module configured to calculate the optimal number of clusters in spectral clustering based on the synergic association index (SAI) and the separation index (SI), resulting in improved precision of condition assessment. System after Claim 4 , where the feature space created by performing spectral clustering contains features, namely singular value (SV) and singular value entropy (SVE) of the extracted envelope signals from product functions, and is used to train the SC-PI (Spectral Clustering Parting Index) -Model in the main processing unit. System after Claim 1 , where the real-time bearing condition assessment module in conjunction with takes a test feature vector (one at a time) from the already trained SC-PI model (Spectral Clustering-Parting Index) and feeds it to the trained system SC-PI model (Spectral Clustering-Parting Index), where the distance to a normal state cluster and then the confidence value (CV) is calculated, which is between 0 and 1, where a value close to 1 indicates healthy and a value close to 0 indicates deterioration of the rolling bearing. System after Claim 1 , wherein the user interface includes a display for displaying the calculated confidence value and the condition of rolling bearings classified as good, faulty and failed bearings.

Images