CN113408068A - Random forest classification machine pump fault diagnosis method and device - Google Patents
Random forest classification machine pump fault diagnosis method and device Download PDFInfo
- Publication number
- CN113408068A CN113408068A CN202110679005.1A CN202110679005A CN113408068A CN 113408068 A CN113408068 A CN 113408068A CN 202110679005 A CN202110679005 A CN 202110679005A CN 113408068 A CN113408068 A CN 113408068A
- Authority
- CN
- China
- Prior art keywords
- frequency domain
- fault
- fault state
- pump
- domain signal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/17—Mechanical parametric or variational design
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/02—Gearings; Transmission mechanisms
- G01M13/022—Power-transmitting couplings or clutches
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/02—Gearings; Transmission mechanisms
- G01M13/028—Acoustic or vibration analysis
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/04—Bearings
- G01M13/045—Acoustic or vibration analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/08—Probabilistic or stochastic CAD
Abstract
The embodiment of the application discloses a machine pump fault diagnosis method and device based on random forest classification, wherein the method comprises the steps of obtaining Z-axis vibration signal components of a machine pump in a fault state and a normal state; preprocessing the Z-axis vibration component to obtain a frequency domain signal; training the frequency domain signals by adopting a random forest method, and establishing a relation between the frequency domain signals and a fault state to obtain a fault diagnosis model; and inputting the frequency domain signal of the pump to be diagnosed into the fault diagnosis model, and judging the fault state of the pump. By using random forest classification, the method is convenient to calculate and high in diagnosis speed, and the real-time performance and accuracy of fault diagnosis are improved; the cost of fault diagnosis is reduced, and the method has good practical value.
Description
Technical Field
The invention relates to a mechanical pump maintenance technology, in particular to a mechanical pump fault diagnosis method and device based on random forest classification.
Background
Device status monitoring and fault diagnosis techniques were initiated in the 60's of the 20 th century. Under the large background that the industrial production is gradually updated and automated, the research and development aiming at the rotary mechanical failure diagnosis technology is abnormally active and rapidly developed.
The equipment investment is large in process industry, the continuous production process is long, the loss caused by the fault of mechanical equipment is large, the state evaluation of the fault equipment is effectively, quickly and accurately realized by means of fault diagnosis, the abnormal condition of the equipment is found as early as possible, and the abnormal condition or the fault of the equipment is timely and effectively processed.
The existing industrial water pump fault diagnosis technology is based on data signal acquisition of various vibration temperature sensors in different communication modes, utilizes a signal analysis theory to obtain various characteristic vectors of a system in a deep level in a time domain and a frequency domain, and utilizes the relation between the characteristic vectors and a system fault source to judge the position of the fault source, so that fault early warning and fault diagnosis are carried out.
In recent years, artificial intelligence research is deepened, and an artificial intelligence method is applied to fault diagnosis.
The existing water pump fault diagnosis algorithm mainly utilizes a BP neural network to carry out nonlinear relation modeling of each operation characteristic value and fault type of the water pump so as to predict faults. The disadvantage is that sufficient learning samples are required to ensure the reliability of the diagnosis. When a complex system is diagnosed, the practicability of the neural network is often reduced due to the problems of excessively large network scale, too long learning training time and the like. When used in industrial fields, the feedback may not be timely or the diagnosis may not be accurate, so that the purpose cannot be achieved.
Disclosure of Invention
The embodiment of the invention aims to provide a machine pump diagnosis method and device based on random forest classification, and aims to solve the technical problem that quick and accurate diagnosis of machine pump faults cannot be realized in the related technology.
According to a first aspect of embodiments of the present application, there is provided a method for diagnosing a pump fault by random forest classification, including:
acquiring Z-axis vibration signal components of the pump in a fault state and a normal state;
preprocessing the Z-axis vibration component to obtain a frequency domain signal;
training the frequency domain signals by adopting a random forest method, and establishing a relation between the frequency domain signals and a fault state to obtain a fault diagnosis model;
and inputting the frequency domain signal of the pump to be diagnosed into the fault diagnosis model, and judging the fault state of the pump.
Further, the fault condition of the pump includes at least: the rotor unbalance fault state, the rotor misalignment fault state, the ball fault state, the bearing fault state, the base looseness fault state, the coupling fault state and the normal state.
Further, after preprocessing the acquired Z-axis vibration signal component, a frequency domain signal is obtained, which includes:
carrying out fast Fourier transform on the Z-axis vibration signal component to obtain a frequency domain oscillogram;
and calculating 0.5 frequency multiplication amplitude, 1 frequency multiplication amplitude, 1.5 frequency multiplication amplitude, 2 frequency multiplication amplitude, 2.5 frequency multiplication amplitude, 3 frequency multiplication amplitude, 3.5 frequency multiplication amplitude, 4 frequency multiplication amplitude, kurtosis, skewness and total vibration value in the frequency domain oscillogram, and taking the calculated values together as frequency domain signals.
Further, training the frequency domain signal by adopting a random forest, establishing a relation between a Z-axis vibration signal component and a fault state, and obtaining a fault diagnosis model, wherein the method comprises the following steps:
the random forest is composed of a plurality of cart classification trees, each node selects a Gini index to select the optimal characteristics and corresponding values for segmentation, and for a frequency domain signal D, the definition of the Gini coefficient is
Where K is the various divisions of the tree, CkRepresents a set of samples belonging to class k in the frequency domain signal D;
and (3) dividing the frequency domain signal according to the characteristic A, wherein the Gini index of the divided sample set is as follows:
d1 and D2 are subsets of the data set after segmentation, and the characteristic A is used as an optimal partition scheme of the cart tree by minimizing the Kiney index;
constructing different training sets to increase the difference between the classification models so as to improve the sequence of the combined classification models, and then forming a multi-classification model system by using the sequence; and under the condition of giving a Z-axis vibration signal component, voting is carried out on each decision tree classification model to determine a final fault state, so that a fault diagnosis model is obtained.
According to a second aspect of the embodiments of the present application, there is provided a machine pump fault diagnosis apparatus for random forest classification, including:
the acquisition module is used for acquiring Z-axis vibration signal components in a fault state and a normal state of the pump;
the preprocessing module is used for preprocessing the Z-axis vibration component to obtain a frequency domain signal;
the modeling module is used for training the frequency domain signals by adopting a random forest method, and establishing a relation between the frequency domain signals and a fault state to obtain a fault diagnosis model;
and the diagnosis module is used for inputting the frequency domain signal of the pump to be diagnosed into the fault diagnosis model and judging the fault state of the pump.
According to a third aspect of embodiments of the present application, there is provided an electronic apparatus, including: one or more processors; a memory for storing one or more programs; when executed by the one or more processors, cause the one or more processors to implement a method as described in the first aspect.
According to a fourth aspect of embodiments herein, there is provided a computer readable storage medium having stored thereon computer instructions which, when executed by a processor, implement the steps of the method according to the first aspect.
The technical scheme provided by the embodiment of the application can have the following beneficial effects:
according to the embodiments, the method and the device for diagnosing the machine pump fault are provided, the method comprises the steps of firstly obtaining Z-axis vibration signal components in a fault state and a normal state of the machine pump, then preprocessing the Z-axis vibration components to obtain frequency domain signals, then training the frequency domain signals by adopting a random forest method, and establishing a relation between the frequency domain signals and the fault state to obtain a fault diagnosis model; the method of random forest is adopted, the problem of low training speed by using a neural network method in the prior art is solved, and the training can be efficiently completed. The method has better robustness, overcomes the uncertainty of industrial field data acquisition during data processing, and can obtain a test set for diagnosis by utilizing source data processing with missing items and heavy items.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the application.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and together with the description, serve to explain the principles of the application.
FIG. 1 is a flow diagram illustrating a method for machine pump diagnosis for random forest classification according to an exemplary embodiment.
FIG. 2 is a flow diagram illustrating random forest training resulting in a fault diagnosis model in accordance with an exemplary embodiment.
Fig. 3 is a schematic diagram illustrating a structure of a random forest classification pump diagnosis apparatus according to an exemplary embodiment.
Detailed Description
Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used in this application and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.
Fig. 1 is a flow chart illustrating a method for diagnosing a pump by random forest classification according to an exemplary embodiment, where the method is applied to a terminal and may include the following steps:
step S11, acquiring Z-axis vibration signal components of the pump in a fault state and a normal state;
step S12, preprocessing the Z-axis vibration component to obtain a frequency domain signal;
step S13, training the frequency domain signal by adopting a random forest method, establishing a relation between the frequency domain signal and a fault state, and obtaining a fault diagnosis model;
and step S14, inputting the frequency domain signal of the pump to be diagnosed into the fault diagnosis model, and judging the fault state of the pump.
According to the embodiment, the Z-axis vibration signal components of the pump in the fault state and the normal state are obtained firstly, then the Z-axis vibration components are preprocessed to obtain frequency domain signals, then the frequency domain signals are trained by adopting a random forest method, the relationship between the frequency domain signals and the fault state is established, a fault diagnosis model is obtained, the training can be efficiently completed, and the problem that the training speed is low by utilizing a neural network method in the prior art is solved; the method has better robustness, overcomes the uncertainty of industrial field data acquisition during data processing, and can obtain a test set for diagnosis by utilizing source data processing with missing items and heavy items.
In the specific implementation of the step S11, the Z-axis vibration signal components in the failure state and the normal state of the pump are obtained;
specifically, in the present embodiment, six of the most common fault conditions (rotor imbalance fault condition, rotor misalignment fault condition, ball fault condition, bearing fault condition, base loosening fault condition, coupling fault condition) and normal operating conditions were selected for simulation and data collection. The distance from the three sampling points to the fault occurrence position is different, the vibration acceleration in the Z-axis direction is acquired by the sensor, and the sampling rate is 3200 Hz. In this embodiment, data of three pump rotation speeds are collected: 1800rpm, 2400rpm, 3000rpm are all the commonly used rotational speeds in industrial field, can make the simulation result more accurate. Each fault condition and normal condition was collected for 500s at each speed. This embodiment collects enough data for the following data processing and training.
In the specific implementation of step S12, after the Z-axis vibration component is preprocessed, a frequency domain signal is obtained;
specifically, since the vibration signals obtained in step S11 are all time domain signals, in this embodiment, a frequency domain signal that can better represent the fault characteristics is obtained by using Fast Fourier Transform (FFT) and a series of calculations. Specifically, the step S12 includes:
carrying out fast Fourier transform on the Z-axis vibration signal component to obtain a frequency domain oscillogram; and calculating 0.5 frequency multiplication amplitude, 1 frequency multiplication amplitude, 1.5 frequency multiplication amplitude, 2 frequency multiplication amplitude, 2.5 frequency multiplication amplitude, 3 frequency multiplication amplitude, 3.5 frequency multiplication amplitude, 4 frequency multiplication amplitude, kurtosis, skewness and total vibration value in the frequency domain oscillogram, and taking the calculated values together as frequency domain signals.
Specifically, for every 1024 pairs of data, a spectrogram of 0-1600 Hz is obtained by FFT. Calculating spectral line values of 0.5X, 1.0X, 1.5X, 2.0X, 2.5X, 3.0X, 3.5X, 4.0X and >4.0X power frequency by taking 50Hz as power frequency;
if the first eight types of the spectral lines are directly calculated, only one spectral line is provided, the data value is very small, in order to ensure the accuracy, the sum of the spectral line values of 8 nearby spectral lines and the spectral line is calculated in the embodiment as the spectral line value of the spectral line, and the data are ensured to be in an order of magnitude;
and simultaneously, calculating a kurtosis value, a skewness value, a total vibration value 1 of the target spectral line and a total vibration value 2 of all spectral lines by using the spectrogram, wherein the four parameters and the 9 spectral line values are used as 13 parameter variables of the frequency domain signal.
In the specific implementation of the step S13, training the frequency domain signal by using a random forest method, and establishing a relationship between the frequency domain signal and a fault state to obtain a fault diagnosis model;
specifically, the random forest is composed of a plurality of cart classification trees, each node selects a kini index to select the optimal feature and the corresponding value to carry out segmentation, and for the frequency domain signal D, the definition of the kini coefficient is as follows
Where K is the various divisions of the tree, CkRepresents a set of samples belonging to class k in the frequency domain signal D;
and (3) dividing the frequency domain signal according to the characteristic A, wherein the Gini index of the divided sample set is as follows:
d1 and D2 are subsets of the data set after segmentation, and the characteristic A is used as an optimal partition scheme of the cart tree by minimizing the Kiney index;
constructing different training sets to increase the difference between the classification models so as to improve the sequence of the combined classification models, and then forming a multi-classification model system by using the sequence; and under the condition of giving a Z-axis vibration signal component, voting is carried out on each decision tree classification model to determine a final fault state, so that a fault diagnosis model is obtained.
More specifically, as shown in fig. 2, a frequency domain signal is used as a training set, training samples are randomly extracted from the frequency domain signal by using a windowing method by using a bagging method, and the extraction is performed to obtain a divided training set (the divided training sets are independent from each other, and elements may be repeated). Randomly extracting partial frequency domain signal characteristics as characteristics to be selected, determining test characteristics in the characteristics to be selected by utilizing GINI index, and judging whether to generate nodes and whether to stop growing the decision tree. And continuously repeating the steps until the decision tree meets the requirements, and generating a final random forest fault diagnosis model. By using the bagging method, high-dimensional data can be processed without feature selection.
In the specific implementation of step S14, the frequency domain signal of the pump to be diagnosed is input into the fault diagnosis model, and the fault state of the pump is determined.
Specifically, historical data of faults of the industrial field devices can be acquired on an industrial internet platform, fault states of some faults are fed back, and diagnosis of known real fault states can be carried out by using the data. Meanwhile, real-time data of some normally working equipment can be derived for diagnosis.
In this embodiment, a set of normal state pump data is derived according to historical failure data of three known pump fed back by platform staff.
No. 6 water pump-bearing fault state
No. 10 water pump-bearing fault state
2# combustion-supporting fan-coupling fault state
Circulating water pump-derived real-time data (Normal state)
In this embodiment, the data are sequentially subjected to data preprocessing and fault diagnosis, and the specific implementation steps are as follows:
the industrial field data is not identical to the laboratory bench data, and due to uncertainty of the industrial field, NULL phase (NULL) and repeated items (possibly sensor faults) exist in the data, so that data cleaning is required to be carried out firstly, and the data are deleted and deleted.
And after the processing is finished, installing the step of data preprocessing in the S12 to process data to obtain the frequency domain signal of the pump to be diagnosed.
Inputting the frequency domain signal of the pump to be diagnosed into the fault diagnosis model obtained in the step S13 for diagnosis, wherein the diagnosis result is as follows:
the embodiment verifies that the accuracy of the fault diagnosis model reaches more than 95%.
Corresponding to the embodiment of the method for diagnosing the pump fault based on random forest classification, the application also provides an embodiment of a device for diagnosing the pump fault based on random forest classification.
FIG. 3 is a block diagram illustrating a random forest classification pump fault diagnosis apparatus according to an exemplary embodiment. Referring to FIG. 3, the apparatus includes
The acquisition module 21 is used for acquiring Z-axis vibration signal components in a fault state and a normal state of the pump;
the preprocessing module 22 is configured to preprocess the Z-axis vibration component to obtain a frequency domain signal;
the modeling module 23 is configured to train the frequency domain signal by using a random forest method, establish a relationship between the frequency domain signal and a fault state, and obtain a fault diagnosis model;
and the diagnosis module 24 is configured to input the frequency domain signal of the pump to be diagnosed into the fault diagnosis model, and determine a fault state of the pump.
With regard to the apparatus in the above-described embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment related to the method, and will not be elaborated here.
For the device embodiments, since they substantially correspond to the method embodiments, reference may be made to the partial description of the method embodiments for relevant points. The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the scheme of the application. One of ordinary skill in the art can understand and implement it without inventive effort.
Correspondingly, the present application also provides an electronic device, comprising: one or more processors; a memory for storing one or more programs; when executed by the one or more processors, cause the one or more processors to implement a method of machine pump fault diagnosis for random forest classification as described above.
Accordingly, the present application also provides a computer readable storage medium having stored thereon computer instructions, wherein the instructions, when executed by a processor, implement the method for diagnosing a pump fault in random forest classification as described above.
Other embodiments of the present application will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure disclosed herein. This application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the application and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the application being indicated by the following claims.
It will be understood that the present application is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the application is limited only by the appended claims.
Claims (10)
1. A method for diagnosing the fault of a pump classified by random forests is characterized by comprising the following steps:
acquiring Z-axis vibration signal components of the pump in a fault state and a normal state;
preprocessing the Z-axis vibration component to obtain a frequency domain signal;
training the frequency domain signals by adopting a random forest method, and establishing a relation between the frequency domain signals and a fault state to obtain a fault diagnosis model;
and inputting the frequency domain signal of the pump to be diagnosed into the fault diagnosis model, and judging the fault state of the pump.
2. The method of claim 1, wherein the fault condition of the pump comprises at least: the rotor unbalance fault state, the rotor misalignment fault state, the ball fault state, the bearing fault state, the base looseness fault state, the coupling fault state and the normal state.
3. The method of claim 1, wherein preprocessing the acquired Z-axis vibration signal component to obtain a frequency domain signal comprises:
carrying out fast Fourier transform on the Z-axis vibration signal component to obtain a frequency domain oscillogram;
and calculating 0.5 frequency multiplication amplitude, 1 frequency multiplication amplitude, 1.5 frequency multiplication amplitude, 2 frequency multiplication amplitude, 2.5 frequency multiplication amplitude, 3 frequency multiplication amplitude, 3.5 frequency multiplication amplitude, 4 frequency multiplication amplitude, kurtosis, skewness and total vibration value in the frequency domain oscillogram, and taking the calculated values together as frequency domain signals.
4. The method of claim 1, wherein training the frequency domain signal with a random forest to establish a relationship between a Z-axis vibration signal component and a fault condition to obtain a fault diagnosis model comprises:
the random forest is composed of a plurality of cart classification trees, each node selects a Gini index to select the optimal characteristics and corresponding values for segmentation, and for a frequency domain signal D, the definition of the Gini coefficient is
Where K is the various divisions of the tree, CkRepresents a set of samples belonging to class k in the frequency domain signal D;
and (3) dividing the frequency domain signal according to the characteristic A, wherein the Gini index of the divided sample set is as follows:
d1 and D2 are subsets of the data set after segmentation, and the characteristic A is used as an optimal partition scheme of the cart tree by minimizing the Kiney index;
constructing different training sets to increase the difference between the classification models so as to improve the sequence of the combined classification models, and then forming a multi-classification model system by using the sequence; and under the condition of giving a Z-axis vibration signal component, voting is carried out on each decision tree classification model to determine a final fault state, so that a fault diagnosis model is obtained.
5. The utility model provides a categorised machine pump fault diagnosis device of random forest which characterized in that includes:
the acquisition module is used for acquiring Z-axis vibration signal components in a fault state and a normal state of the pump;
the preprocessing module is used for preprocessing the Z-axis vibration component to obtain a frequency domain signal;
the modeling module is used for training the frequency domain signals by adopting a random forest method, and establishing a relation between the frequency domain signals and a fault state to obtain a fault diagnosis model;
and the diagnosis module is used for inputting the frequency domain signal of the pump to be diagnosed into the fault diagnosis model and judging the fault state of the pump.
6. The apparatus of claim 5, wherein the fault condition of the pump comprises at least: the rotor unbalance fault state, the rotor misalignment fault state, the ball fault state, the bearing fault state, the base looseness fault state, the coupling fault state and the normal state.
7. The apparatus of claim 5, wherein the pre-processing of the acquired Z-axis vibration signal component to obtain a frequency domain signal comprises:
carrying out fast Fourier transform on the Z-axis vibration signal component to obtain a frequency domain oscillogram;
and calculating 0.5 frequency multiplication amplitude, 1 frequency multiplication amplitude, 1.5 frequency multiplication amplitude, 2 frequency multiplication amplitude, 2.5 frequency multiplication amplitude, 3 frequency multiplication amplitude, 3.5 frequency multiplication amplitude, 4 frequency multiplication amplitude, kurtosis, skewness and total vibration value in the frequency domain oscillogram, and taking the calculated values together as frequency domain signals.
8. The apparatus of claim 5, wherein the frequency domain signal is trained using a random forest, and a relationship between a Z-axis vibration signal component and a fault state is established to obtain a fault diagnosis model, comprising:
the random forest is composed of a plurality of cart classification trees, each node selects a Gini index to select the optimal characteristics and corresponding values for segmentation, and for a frequency domain signal D, the definition of the Gini coefficient is
Where K is the various divisions of the tree, CkRepresents a set of samples belonging to class k in the frequency domain signal D;
and (3) dividing the frequency domain signal according to the characteristic A, wherein the Gini index of the divided sample set is as follows:
d1 and D2 are subsets of the data set after segmentation, and the characteristic A is used as an optimal partition scheme of the cart tree by minimizing the Kiney index;
constructing different training sets to increase the difference between the classification models so as to improve the sequence of the combined classification models, and then forming a multi-classification model system by using the sequence; and under the condition of giving a Z-axis vibration signal component, voting is carried out on each decision tree classification model to determine a final fault state, so that a fault diagnosis model is obtained.
9. An electronic device, comprising:
one or more processors;
a memory for storing one or more programs;
when executed by the one or more processors, cause the one or more processors to implement the method of any one of claims 1-4.
10. A computer-readable storage medium having stored thereon computer instructions, which, when executed by a processor, carry out the steps of the method according to any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110679005.1A CN113408068A (en) | 2021-06-18 | 2021-06-18 | Random forest classification machine pump fault diagnosis method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110679005.1A CN113408068A (en) | 2021-06-18 | 2021-06-18 | Random forest classification machine pump fault diagnosis method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113408068A true CN113408068A (en) | 2021-09-17 |
Family
ID=77681540
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110679005.1A Pending CN113408068A (en) | 2021-06-18 | 2021-06-18 | Random forest classification machine pump fault diagnosis method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113408068A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114492588A (en) * | 2021-12-29 | 2022-05-13 | 中国大唐集团科学技术研究总院有限公司华东电力试验研究院 | Method, system, equipment and storage medium for identifying faults of auxiliary equipment of thermal power plant |
| CN115270896A (en) * | 2022-09-28 | 2022-11-01 | 西华大学 | Intelligent diagnosis method for identifying loosening fault of main bearing of aircraft engine |
| CN115614292A (en) * | 2022-11-02 | 2023-01-17 | 昆明理工大学 | Vibration monitoring device and method for vertical water pump unit |
| WO2023035869A1 (en) * | 2022-03-15 | 2023-03-16 | 中国长江三峡集团有限公司 | Gearbox fault diagnosis model training method and gearbox fault diagnosis method |
| CN116992365A (en) * | 2023-08-02 | 2023-11-03 | 广东海洋大学 | Fault diagnosis method and system under random impact interference |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106441946A (en) * | 2016-11-15 | 2017-02-22 | 重庆工商大学 | Fault recognition method and fault recognition system for hydraulic shock absorber of vehicle based on vibration signal |
| CN108680348A (en) * | 2018-05-14 | 2018-10-19 | 国网山东省电力公司莱芜供电公司 | A kind of breaker mechanical fault diagnosis method and system based on random forest |
| WO2019153388A1 (en) * | 2018-02-12 | 2019-08-15 | 大连理工大学 | Power spectral entropy random forest-based aeroengine rolling bearing fault diagnosis method |
| CN110375987A (en) * | 2019-06-24 | 2019-10-25 | 昆明理工大学 | One kind being based on depth forest machines Bearing Fault Detection Method |
| CN111046931A (en) * | 2019-12-02 | 2020-04-21 | 北京交通大学 | Turnout fault diagnosis method based on random forest |
| CN111458145A (en) * | 2020-03-30 | 2020-07-28 | 南京机电职业技术学院 | Cable car rolling bearing fault diagnosis method based on road map characteristics |
| CN112364928A (en) * | 2020-11-18 | 2021-02-12 | 浙江工业大学 | Random forest classification method in transformer substation fault data diagnosis |
-
2021
- 2021-06-18 CN CN202110679005.1A patent/CN113408068A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106441946A (en) * | 2016-11-15 | 2017-02-22 | 重庆工商大学 | Fault recognition method and fault recognition system for hydraulic shock absorber of vehicle based on vibration signal |
| WO2019153388A1 (en) * | 2018-02-12 | 2019-08-15 | 大连理工大学 | Power spectral entropy random forest-based aeroengine rolling bearing fault diagnosis method |
| CN108680348A (en) * | 2018-05-14 | 2018-10-19 | 国网山东省电力公司莱芜供电公司 | A kind of breaker mechanical fault diagnosis method and system based on random forest |
| CN110375987A (en) * | 2019-06-24 | 2019-10-25 | 昆明理工大学 | One kind being based on depth forest machines Bearing Fault Detection Method |
| CN111046931A (en) * | 2019-12-02 | 2020-04-21 | 北京交通大学 | Turnout fault diagnosis method based on random forest |
| CN111458145A (en) * | 2020-03-30 | 2020-07-28 | 南京机电职业技术学院 | Cable car rolling bearing fault diagnosis method based on road map characteristics |
| CN112364928A (en) * | 2020-11-18 | 2021-02-12 | 浙江工业大学 | Random forest classification method in transformer substation fault data diagnosis |
Non-Patent Citations (1)
| Title |
|---|
| 裔隽，张怿檬，张目清等著: "《Python机器学习实战》", 28 February 2018, 北京：科学技术文献出版社, pages: 186 - 187 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114492588A (en) * | 2021-12-29 | 2022-05-13 | 中国大唐集团科学技术研究总院有限公司华东电力试验研究院 | Method, system, equipment and storage medium for identifying faults of auxiliary equipment of thermal power plant |
| WO2023035869A1 (en) * | 2022-03-15 | 2023-03-16 | 中国长江三峡集团有限公司 | Gearbox fault diagnosis model training method and gearbox fault diagnosis method |
| GB2616970A (en) * | 2022-03-15 | 2023-09-27 | China Three Gorges Corp | Gearbox fault diagnosis model training method and gearbox fault diagnosis method |
| CN115270896A (en) * | 2022-09-28 | 2022-11-01 | 西华大学 | Intelligent diagnosis method for identifying loosening fault of main bearing of aircraft engine |
| CN115614292A (en) * | 2022-11-02 | 2023-01-17 | 昆明理工大学 | Vibration monitoring device and method for vertical water pump unit |
| CN115614292B (en) * | 2022-11-02 | 2023-10-27 | 昆明理工大学 | Vibration monitoring device and method for vertical water pump unit |
| CN116992365A (en) * | 2023-08-02 | 2023-11-03 | 广东海洋大学 | Fault diagnosis method and system under random impact interference |
| CN116992365B (en) * | 2023-08-02 | 2024-03-08 | 广东海洋大学 | Fault diagnosis method and system under random impact interference |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113408068A (en) | Random forest classification machine pump fault diagnosis method and device | |
| Wu et al. | Induction machine fault detection using SOM-based RBF neural networks | |
| CN110276416B (en) | Rolling bearing fault prediction method | |
| CN103471841B (en) | A kind of rotating machinery vibrating failure diagnosis method | |
| CN109613428A (en) | It is a kind of can be as system and its application in motor device fault detection method | |
| CN109781411A (en) | A kind of combination improves the Method for Bearing Fault Diagnosis of sparse filter and KELM | |
| CN111122191B (en) | Equipment health alarm threshold setting method based on EWMA control | |
| CN110006552B (en) | Method for detecting abnormal temperature of equipment | |
| CN113188794B (en) | Gearbox fault diagnosis method and device based on improved PSO-BP neural network | |
| CN110399986A (en) | A kind of generation method of pumping plant unit fault diagnosis system | |
| CN112729834B (en) | Bearing fault diagnosis method, device and system | |
| CN106663086A (en) | Apparatus and method for ensembles of kernel regression models | |
| CN111190349A (en) | Method, system and medium for monitoring state and diagnosing fault of ship engine room equipment | |
| CN115576284A (en) | Clothing workshop intelligent management method and system | |
| CN115614292A (en) | Vibration monitoring device and method for vertical water pump unit | |
| CN112465124A (en) | Twin depth space-time neural network model acquisition/fault diagnosis method and device | |
| CN107036808A (en) | Gearbox of wind turbine combined failure diagnostic method based on SVMs probability Estimation | |
| CN115329986A (en) | Wind turbine generator anomaly detection and positioning method based on interpretable graph neural network | |
| De Lima et al. | On fault classification in rotating machines using fourier domain features and neural networks | |
| CN110007660B (en) | Online soft measurement method for transient equivalent thermal stress of steam turbine set of thermal power plant | |
| CN115793590A (en) | Data processing method and platform suitable for system safety operation and maintenance | |
| CN113237619B (en) | Fault early warning method, device, equipment and storage medium for variable-speed rotating machinery vibration | |
| CN109990803A (en) | The method, apparatus of method, apparatus and the sensor processing of detection system exception | |
| CN115659249B (en) | Abnormality detection method for intelligent station measurement control system | |
| CN111623905A (en) | Wind turbine bearing temperature early warning method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20210917 |