CN108363896B - Fault diagnosis method for hydraulic cylinder - Google Patents
Fault diagnosis method for hydraulic cylinder Download PDFInfo
- Publication number
- CN108363896B CN108363896B CN201810441856.0A CN201810441856A CN108363896B CN 108363896 B CN108363896 B CN 108363896B CN 201810441856 A CN201810441856 A CN 201810441856A CN 108363896 B CN108363896 B CN 108363896B
- Authority
- CN
- China
- Prior art keywords
- hydraulic cylinder
- flow data
- inlet
- outlet flow
- fault
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000003745 diagnosis Methods 0.000 title claims abstract description 32
- 238000004088 simulation Methods 0.000 claims abstract description 46
- 239000013598 vector Substances 0.000 claims abstract description 23
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 15
- 238000005457 optimization Methods 0.000 claims abstract description 15
- 230000002068 genetic effect Effects 0.000 claims abstract description 12
- 238000012549 training Methods 0.000 claims abstract description 12
- 238000012360 testing method Methods 0.000 claims abstract description 10
- 238000002347 injection Methods 0.000 claims abstract description 6
- 239000007924 injection Substances 0.000 claims abstract description 6
- 238000012706 support-vector machine Methods 0.000 claims description 25
- 238000000354 decomposition reaction Methods 0.000 claims description 16
- 230000008859 change Effects 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 3
- 238000013178 mathematical model Methods 0.000 claims description 3
- 238000010295 mobile communication Methods 0.000 claims description 2
- 238000012163 sequencing technique Methods 0.000 claims description 2
- 230000008569 process Effects 0.000 abstract description 4
- 239000000243 solution Substances 0.000 abstract description 3
- 238000012795 verification Methods 0.000 abstract 1
- 230000000694 effects Effects 0.000 description 4
- 238000001595 flow curve Methods 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000009776 industrial production Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 230000035772 mutation Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000002790 cross-validation Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000012843 least square support vector machine Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention discloses a hydraulic cylinder fault diagnosis method, which comprises the steps of firstly carrying out physical modeling simulation on a hydraulic cylinder in AMESim software and completing a fault injection process to obtain fault data; then extracting wavelet packet energy of the data to construct a feature vector; then establishing a fault diagnosis model based on a high-dimensional nonlinear classifier; training a fault diagnosis model by using a training sample, calculating an optimal solution of parameters required by the model by adopting a genetic algorithm, and constructing a multi-valued classifier; and inputting the hydraulic cylinder test sample into the model for fault diagnosis. The invention effectively solves the problem of insufficient fault data in the algorithm verification process, effectively solves the problems of nonlinearity and small samples by combining the application of the genetic algorithm optimization parameter-searching high-dimensional nonlinear classifier algorithm, and improves the fault identification capability of the fault diagnosis model.
Description
Technical Field
The invention relates to the field of hydraulic cylinder fault diagnosis, in particular to a hydraulic cylinder fault diagnosis method based on a support vector machine (WPT-GASVM) optimized based on wavelet packet decomposition and genetic algorithm.
Background
The hydraulic system is widely applied to industrial production and is an important component of a complex system. The hydraulic cylinder is used as an execution part in a hydraulic servo system, is widely applied to realizing linear motion or rotary reciprocating motion, and the health state of the hydraulic cylinder directly influences the whole production activity. Once the hydraulic cylinder breaks down, equipment or load is damaged if the hydraulic cylinder breaks down, and personal and property safety is threatened if the hydraulic cylinder breaks down, so that major accidents are caused. With the trend of large-scale and complicated equipment, the traditional scheduled maintenance is difficult to meet the requirement of industrial production.
The conventional fault diagnosis method for the hydraulic cylinder often has the problem of insufficient fault data, and the data generated in the operation process of equipment is lack of fault data and has a large amount of useless data. On the other hand, in the selection of the fault diagnosis method, the conventional model-based method is difficult to be applied due to the problem of modeling accuracy, and data-driven visual maintenance is receiving more and more attention and development.
Chinese patent application No.: 201610497436.5, patent name: the invention discloses a hydraulic cylinder internal leakage fault diagnosis and evaluation method, which adopts a wavelet analysis method to extract the characteristics of a pressure signal and combines a BP neural network to realize the diagnosis and evaluation of the hydraulic cylinder internal leakage grade.
In the literature, research on a hydraulic cylinder leakage fault diagnosis method based on LS-SVM (journal name: machine tool and hydraulic pressure, volume number and other information: 2017,45(15):184-187), the hydraulic cylinder leakage degree is classified by taking the hydraulic cylinder leakage as a research object, and a least square support vector machine method is adopted to realize the diagnosis target.
Although the prior art has successfully used a data-driven method to diagnose the fault of the hydraulic cylinder, the problems of difficulty in obtaining fault data and selection of model parameters still exist.
Disclosure of Invention
The invention aims to solve the technical problem of providing a hydraulic cylinder fault diagnosis method of a support vector machine (WPT-GASVM) based on wavelet packet decomposition and genetic algorithm optimization aiming at the defects involved in the background technology.
The invention adopts the following technical scheme for solving the technical problems:
a hydraulic cylinder fault diagnosis method comprises the following specific steps:
step 1), establishing a hydraulic cylinder simulation model through AMESim software;
step 2), fault injection is carried out, and inlet and outlet flow data of N groups of hydraulic cylinder simulation models in each state of a state set are obtained respectively, wherein the state set comprises a normal working state, a hydraulic cylinder leakage fault state, a load sudden change fault state and a piston rod axis offset fault state;
step 3), carrying out four-layer decomposition on each group of inlet and outlet flow data of the hydraulic cylinder simulation model in each state by adopting a wavelet packet decomposition method to obtain decomposition signals of each group of inlet and outlet flow data under 16 frequency bands, calculating energy values of the inlet flow data and the decomposition signals of the outlet flow data of each group of inlet and outlet flow data under each frequency band to further obtain characteristic vectors of the group of inlet and outlet flow data, and finally forming a characteristic data set by combining the states of the hydraulic cylinder simulation models corresponding to each group of inlet and outlet flow data;
step 4), optimizing a penalty factor of the support vector machine model and a preset kernel function parameter by using a genetic algorithm, establishing a multi-valued support vector machine classifier on the basis, and then training the multi-valued support vector machine classifier by using a feature data set to obtain the trained multi-valued support vector machine classifier;
and 5) performing characteristic extraction on the inlet and outlet flow data of the hydraulic cylinder to be tested to obtain a characteristic vector of the hydraulic cylinder, inputting the characteristic vector into the trained multi-value support vector machine classifier to obtain the state of the hydraulic cylinder to be tested, and completing fault diagnosis.
As a further optimization scheme of the hydraulic cylinder fault diagnosis method, the detailed steps of the step 1) are as follows:
step 1.1), placing a servo amplifier, a servo valve, a hydraulic cylinder and a position sensor model in a sketch mode of AMESim software to form a sketch of a hydraulic cylinder simulation model;
step 1.2), selecting a mathematical model for a servo amplifier, a servo valve, a hydraulic cylinder and a position sensor model in a sub-model mode of AMESim software;
step 1.3), setting parameters for a servo amplifier, a servo valve, a hydraulic cylinder and a position sensor model respectively under a parameter model of AMESim software to complete the setting of a hydraulic cylinder simulation model;
and step 1.4), operating the hydraulic cylinder simulation model in an operation mode of AMESim software until the operation time is greater than a preset time threshold value so as to confirm that the hydraulic cylinder simulation model operates normally.
As a further optimization scheme of the hydraulic cylinder fault diagnosis method, the detailed steps of the step 2) are as follows:
step 2.1), operating the hydraulic cylinder simulation model in the operation mode of AMESim software to obtain the inlet and outlet flow data of the hydraulic cylinder simulation model under N groups of normal working conditions, wherein N is a natural number greater than 0;
step 2.2), selecting a hydraulic cylinder simulation model in a parameter mode of AMESim software, changing a parameter leakagecoif to form a hydraulic cylinder leakage fault, and operating the hydraulic cylinder simulation model in an operation mode of the AMESim software to obtain inlet and outlet flow data of the hydraulic cylinder simulation model under the leakage faults of N groups of hydraulic cylinders;
step 2.3), selecting a hydraulic cylinder simulation model in a parameter mode of AMESim software, changing a parameter total mass bed moved to form a load sudden-change fault, and operating the hydraulic cylinder simulation model in an operation mode of the AMESim software to obtain inlet and outlet flow data of the hydraulic cylinder simulation model under N groups of load sudden-change faults;
and 2.4) selecting a hydraulic cylinder simulation model in a parameter mode of AMESim software, changing the parameter angle rod makes with horizontal to form a piston rod axis deviation fault, and operating the hydraulic cylinder simulation model in an operation mode of the AMESim software to obtain the inlet and outlet flow data of the hydraulic cylinder simulation model under the N groups of piston rod axis deviation faults.
As a further optimization scheme of the hydraulic cylinder fault diagnosis method, the detailed steps of the step 3) are as follows:
step 3.1), selecting a wavelet basis function system to carry out discrete wavelet packet four-layer decomposition on each group of inlet and outlet flow data of the hydraulic cylinder simulation model under the normal working condition, the hydraulic cylinder leakage fault, the load sudden change fault and the piston rod axis offset fault to obtain decomposition signals of each group of inlet and outlet flow data under 16 frequency bands;
step 3.2), respectively calculating the energy values of decomposed signals of the inlet flow data and the outlet flow data in each group of inlet and outlet flow data under each frequency band, and sequencing the decomposed signals from large to small;
step 3.3), for each group of inlet and outlet flow data, selecting front k of the energy value of the inlet flow data under each frequency band1Front k of energy value of outlet flow data of the mobile communication terminal under each frequency band2Using the vector of each component as a characteristic vector;
and 3.4) forming a characteristic data set according to the characteristic vectors of the inlet and outlet flow data of each group and the states of the hydraulic cylinder simulation models corresponding to the characteristic vectors.
As a further optimization scheme of the hydraulic cylinder fault diagnosis method, the detailed steps of the step 4) are as follows:
step 4.1), setting parameter initialization values of a genetic algorithm, wherein the parameter initialization values comprise evolution algebra, population scale, selection probability, cross probability, variation probability, fitness function and individual optimization space;
step 4.2), taking a penalty factor supporting a vector machine model and a preset kernel function parameter as population individuals, and randomly generating an initial population in an optimization space;
step 4.3), iteratively solving the optimal model parameters, and combining a fitness function to obtain optimal values of the parameters through selection operation, crossover operation and variation operation;
step 4.4), constructing a multi-valued support vector machine classifier according to the obtained optimal parameter value, and randomly selecting 90% of data from the characteristic data set as a training set and the rest 10% of data as a test set to train the multi-valued support vector machine classifier;
and 4.5) repeating the steps 4.1) to 4.4) until the fault diagnosis accuracy of the test set is greater than a preset accuracy threshold.
As a further optimization scheme of the hydraulic cylinder fault diagnosis method according to the present invention, in the step 4), the preset kernel function parameter is a gaussian kernel function parameter, and the penalty factor C of the support vector machine model is 16, and the gaussian kernel function parameter γ is 3.0314.
Compared with the prior art, the invention adopting the technical scheme has the following technical effects:
1) through physical modeling and hydraulic cylinder simulation model parameter setting, multiple fault injection can be realized, batch operation can be realized, the fault injection process is simplified, and the problem of insufficient fault data is solved;
2) the support vector machine method based on genetic algorithm optimization solves the problems that the support vector machine model parameter selection is difficult and the speed is slow, and compared with a neural network method, the method has a better effect under the condition of small sample data.
Drawings
FIG. 1 is a flow chart of hydraulic cylinder fault diagnosis steps;
FIG. 2 is a flow chart of support vector machine algorithm for genetic algorithm optimization.
Detailed Description
The technical scheme of the invention is further explained in detail by combining the attached drawings:
the present invention may be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. In the drawings, components are exaggerated for clarity.
The invention provides a typical valve-controlled hydraulic cylinder position control system, which mainly comprises a servo amplifier, a servo valve, a hydraulic cylinder and a position sensor. The system mainly controls the hydraulic cylinder to move through the hydraulic pump and the three-position four-way reversing valve, then the sensor feeds back position information, and signals of the reversing valve are further adjusted and controlled, so that accurate position control is realized. As shown in fig. 1, the present invention provides a method for diagnosing a fault of a hydraulic cylinder for such a system, which is implemented according to the following steps:
step 1: system simulation model established by using AMESim Rev13 software
Step 1.1: control system model for determining position of valve-controlled hydraulic cylinder and parameter setting
The AMESim software simulation model is mainly divided into a sketch mode, a sub-model mode, a parameter mode and an operation mode. Firstly, placing each element in a sketch mode to finish a model sketch; selecting mathematical models of all elements in a sub-model Mode, and generally selecting a Primal model; setting parameters for each simulation element under a parameter model, which is shown in table 1 specifically; finally, the operating time is set for 12 seconds in the operating mode. The correctness of the built model can be judged by observing an inlet flow curve, an outlet flow curve, a piston rod displacement curve and the like of the hydraulic cylinder.
TABLE 1 actuator position control System component parameters
Step 1.2: fault injection method
In the parameter mode, selecting a hydraulic cylinder model, and changing a parameter leakage coefficient to obtain a hydraulic cylinder leakage fault; changing the total mass bed moved to obtain the load mutation fault; changing the angle rod makes with horizontal can result in a piston rod axis offset failure. Multiple groups of data can be obtained simultaneously through batch processing, and the specific operation is as follows: in the parameter mode, select the Batch parameters tab in setings in the menu bar, then double click the hydraulic cylinder model to drag the parameter, such as the leak coefficient, to the left area of the Batch parameters tab, and then input multiple sets of setting values on the right side of the tab.
Step 1.3: operation and data preservation
In the running mode, setting running time Final time and sampling step size Print interval, and selecting a Batch running mode in Run type. After the operation is finished, the hydraulic cylinder model is double-clicked, the flow rate at port 1 parameter in the Variable List dialog box is dragged to the blank position of the model interface to pop up an AMEPlot window and display an inlet flow curve, and similarly, the flow rate at port 2 parameter is dragged to the blank position of the model interface to display an outlet flow curve. In the AMEPlot window, the Save data … command in the File menu is selected to Save the data as a data File, then the File is opened in the notebook, and all the data is copied to the Excel form. Through the steps, 50 groups of inlet and outlet flow data are respectively obtained under normal working conditions and three fault modes, 200 groups of inlet and outlet flow data are obtained in total, and each group of data comprises inlet flow data and outlet flow data.
Step 2: wavelet packet energy feature extraction
Step 2.1: using db5 wavelet function system to decompose 4 layers of wavelet packet for each group of inlet flow signal, and decomposing coefficient of wavelet packet for 16 frequency bands is
Wherein
Representing the coefficient of the inlet flow signal in the nth frequency band after the 4-layer wavelet packet decomposition,
is a specific coefficient value;
step 2.2: and further extracting energy characteristics on the basis of wavelet packet decomposition and drawing an energy histogram. The specific calculation method comprises the following steps: each frequency band has an energy value of
Wherein i represents the ith frequency band, and j represents the jth coefficient under the ith frequency band; then, the normalization processing is carried out,
obtaining an energy vector
Step 2.3: and selecting a value with relatively concentrated energy from 16 energy values as a feature vector, and selecting the first 4 feature vectors as the feature vectors of the inlet flow by observing the energy histogram.
Step 2.4: according to the above steps, the same treatment is performed on the outlet flow, and the first 4 energy values are still selected. Therefore, each flow data in each working mode can obtain an 8-dimensional characteristic vector T ═ E1 E2 E3 E4 E5 E6 E7E8]In which E1 E2 E3 E4Is the characteristic value of the inlet flow energy of the hydraulic cylinder, E5 E6 E7 E8Is the characteristic value of the flow energy of the outlet of the hydraulic cylinder. Therefore, 200 sets of 8-dimensional characteristic data are obtained, and the normal working condition, the leakage fault, the load surge fault and the piston rod axis deviation fault are respectively marked as 1, 2, 3 and 4.
And step 3: and (3) building, training and testing a support vector machine model optimized based on a genetic algorithm, as shown in FIG. 2.
Step 3.1: and constructing a multi-classification Support Vector Machine (SVM). Constructing a multi-classifier by combining a plurality of binary classifiers by adopting a one-to-one method, and combining k-class data pairwise to construct k-x (k-1)/2 binary classifiers; each binary classifier only trains two types of data and judges the final category of each group of feature data by combining a voting method; in addition, since the number of data features is small and the number of samples is normal, a Gaussian kernel function k (x) can be used1，x2)＝exp(-γ||x1-x2||2) Where γ is a parameter of the gaussian kernel function.
Step 3.2: and determining a penalty factor C of basic parameters of the SVM model needing to be optimized and a Gaussian kernel function parameter gamma.
Step 3.3: the optimal parameters are selected using a Genetic Algorithm (GA). The GA method firstly generates a population containing a plurality of solutions at random, then selects a certain number of individuals from the current population as the next generation population according to the operations of crossing, variation and selection, and converges to the optimal individuals after a plurality of evolutions. Setting GA parameters: evolution algebra 200, population quantity 20, selection probability 0.9, cross probability 0.7 and mutation probability 0.7; the method comprises the steps of selecting a parameter C as an optimization space (0, 100), selecting a parameter gamma as an optimization space [0, 100], selecting the accuracy of a training set under 5-fold cross validation as a fitness function value, selecting an individual with the highest fitness function as an optimal individual after the evolution is finished, and finally obtaining the optimal parameter C which is 16 and gamma which is 3.0314.
Step 3.4: training and testing the SVM model. In this embodiment, a LibSVM toolbox is selected to establish a training environment on a Matlab platform, 3 times of experimental training and testing are performed in total, 90% of feature data sets are randomly selected each time to serve as a training set, and the remaining 10% of feature data sets are selected to serve as a testing set. The results of the three tests can show the effect of fault diagnosis through a confusion matrix and the overall accuracy, and the diagnosis results are shown in table 2.
TABLE 2 Fault diagnosis results
It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
The above-mentioned embodiments, objects, technical solutions and advantages of the present invention are further described in detail, it should be understood that the above-mentioned embodiments are only illustrative of the present invention and are not intended to limit the present invention, and any modifications, equivalents, improvements and the like made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (2)
1. A hydraulic cylinder fault diagnosis method is characterized by comprising the following specific steps:
step 1), establishing a hydraulic cylinder simulation model through AMESim software;
step 1.1), placing a servo amplifier, a servo valve, a hydraulic cylinder and a position sensor model in a sketch mode of AMESim software to form a sketch of a hydraulic cylinder simulation model;
step 1.2), selecting a mathematical model for a servo amplifier, a servo valve, a hydraulic cylinder and a position sensor model in a sub-model mode of AMESim software;
step 1.3), setting parameters for a servo amplifier, a servo valve, a hydraulic cylinder and a position sensor model respectively under a parameter model of AMESim software to complete the setting of a hydraulic cylinder simulation model;
step 1.4), operating the hydraulic cylinder simulation model in an operation mode of AMESim software until the operation time is greater than a preset time threshold value so as to confirm that the hydraulic cylinder simulation model operates normally;
step 2), fault injection is carried out, and inlet and outlet flow data of N groups of hydraulic cylinder simulation models in each state of a state set are obtained respectively, wherein the state set comprises a normal working state, a hydraulic cylinder leakage fault state, a load sudden change fault state and a piston rod axis offset fault state;
step 2.1), operating the hydraulic cylinder simulation model in the operation mode of AMESim software to obtain the inlet and outlet flow data of the hydraulic cylinder simulation model under N groups of normal working conditions, wherein N is a natural number greater than 0;
step 2.2), selecting a hydraulic cylinder simulation model in a parameter mode of AMESim software, changing a parameter leakage coeffient to form a hydraulic cylinder leakage fault, and operating the hydraulic cylinder simulation model in an operation mode of the AMESim software to obtain inlet and outlet flow data of the hydraulic cylinder simulation model under the leakage faults of the N groups of hydraulic cylinders;
step 2.3), selecting a hydraulic cylinder simulation model in a parameter mode of AMESim software, changing a parameter total mass bed moved to form a load sudden-change fault, and operating the hydraulic cylinder simulation model in an operation mode of the AMESim software to obtain inlet and outlet flow data of the hydraulic cylinder simulation model under N groups of load sudden-change faults;
step 2.4), selecting a hydraulic cylinder simulation model in a parameter mode of AMESim software, changing a parameter angle pole makes with horizontal to form a piston rod axis deviation fault, and operating the hydraulic cylinder simulation model in an operation mode of the AMESim software to obtain inlet and outlet flow data of the hydraulic cylinder simulation model under N groups of piston rod axis deviation faults;
step 3), carrying out four-layer decomposition on each group of inlet and outlet flow data of the hydraulic cylinder simulation model in each state by adopting a wavelet packet decomposition method to obtain decomposition signals of each group of inlet and outlet flow data under 16 frequency bands, calculating energy values of the inlet flow data and the decomposition signals of the outlet flow data of each group of inlet and outlet flow data under each frequency band to further obtain characteristic vectors of the group of inlet and outlet flow data, and finally forming a characteristic data set by combining the states of the hydraulic cylinder simulation models corresponding to each group of inlet and outlet flow data;
step 3.1), selecting a wavelet basis function system to carry out discrete wavelet packet four-layer decomposition on each group of inlet and outlet flow data of the hydraulic cylinder simulation model under the normal working condition, the hydraulic cylinder leakage fault, the load sudden change fault and the piston rod axis offset fault to obtain decomposition signals of each group of inlet and outlet flow data under 16 frequency bands;
step 3.2), respectively calculating the energy values of decomposed signals of the inlet flow data and the outlet flow data in each group of inlet and outlet flow data under each frequency band, and sequencing the decomposed signals from large to small;
step 3.3), for each group of inlet and outlet flow data, selecting front k of the energy value of the inlet flow data under each frequency band1Front k of energy value of outlet flow data of the mobile communication terminal under each frequency band2Using the vector of each component as a characteristic vector;
step 3.4), forming a characteristic data set according to the characteristic vectors of the inlet and outlet flow data of each group and the states of the hydraulic cylinder simulation models corresponding to the characteristic vectors;
step 4), optimizing a penalty factor of the support vector machine model and a preset kernel function parameter by using a genetic algorithm, establishing a multi-valued support vector machine classifier on the basis, and then training the multi-valued support vector machine classifier by using a feature data set to obtain the trained multi-valued support vector machine classifier;
step 4.1), setting parameter initialization values of a genetic algorithm, wherein the parameter initialization values comprise evolution algebra, population scale, selection probability, cross probability, variation probability, fitness function and individual optimization space;
step 4.2), taking a penalty factor supporting a vector machine model and a preset kernel function parameter as population individuals, and randomly generating an initial population in an optimization space;
step 4.3), iteratively solving the optimal model parameters, and combining a fitness function to obtain optimal values of the parameters through selection operation, crossover operation and variation operation;
step 4.4), constructing a multi-valued support vector machine classifier according to the obtained optimal parameter value, and randomly selecting 90% of data from the characteristic data set as a training set and the rest 10% of data as a test set to train the multi-valued support vector machine classifier;
step 4.5), repeating the steps 4.1) to 4.4) until the fault diagnosis accuracy of the test set is greater than a preset accuracy threshold;
and 5) performing characteristic extraction on the inlet and outlet flow data of the hydraulic cylinder to be tested to obtain a characteristic vector of the hydraulic cylinder, inputting the characteristic vector into the trained multi-value support vector machine classifier to obtain the state of the hydraulic cylinder to be tested, and completing fault diagnosis.
2. The hydraulic cylinder fault diagnosis method according to claim 1, wherein in the step 4), the preset kernel function parameter is a gaussian kernel function parameter, and a penalty factor C =16 and a gaussian kernel function parameter γ =3.0314 support a vector machine model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810441856.0A CN108363896B (en) | 2018-05-10 | 2018-05-10 | Fault diagnosis method for hydraulic cylinder |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810441856.0A CN108363896B (en) | 2018-05-10 | 2018-05-10 | Fault diagnosis method for hydraulic cylinder |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108363896A CN108363896A (en) | 2018-08-03 |
| CN108363896B true CN108363896B (en) | 2021-12-21 |
Family
ID=63012215
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810441856.0A Active CN108363896B (en) | 2018-05-10 | 2018-05-10 | Fault diagnosis method for hydraulic cylinder |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108363896B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109783959B (en) * | 2019-01-23 | 2023-10-10 | 保定市博宏高科控制技术有限公司 | Valve-based combined multi-model simulation platform and method |
| CN110135066B (en) * | 2019-05-15 | 2021-04-13 | 北京交通大学 | Fault diagnosis method for pilot type overflow valve of power shifting gearbox |
| CN111753465B (en) * | 2020-06-05 | 2022-08-16 | 长江三峡通航管理局 | Method for diagnosing leakage fault in buffering oil cylinder of anti-collision buffering hydraulic system of ship lift |
| CN112036456A (en) * | 2020-08-19 | 2020-12-04 | 阳光电源股份有限公司 | Photovoltaic fault data generation method and device and computer readable storage medium |
| CN112650146B (en) * | 2020-12-18 | 2021-11-19 | 国家机床质量监督检验中心 | Fault diagnosis optimization method, system and equipment of numerical control machine tool under multiple working conditions |
| CN112926400A (en) * | 2021-01-29 | 2021-06-08 | 华南理工大学 | Intelligent diagnosis method and system for leakage fault in hydraulic cylinder based on data driving |
| CN112861437B (en) * | 2021-02-22 | 2022-05-03 | 四川大学 | Fault diagnosis method and device for aileron actuator based on AMESim model |
| CN113656985B (en) * | 2021-09-01 | 2024-03-15 | 上海仁童电子科技有限公司 | Modeling method, fault diagnosis method and device for valve element in braking system |
| CN116150676B (en) * | 2023-04-19 | 2023-09-26 | 山东能源数智云科技有限公司 | Equipment fault diagnosis and identification method and device based on artificial intelligence |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7106903B2 (en) * | 2001-09-24 | 2006-09-12 | Vima Technologies, Inc. | Dynamic partial function in measurement of similarity of objects |
| CN101122585A (en) * | 2007-09-12 | 2008-02-13 | 天津大学 | Automatic identification method for supersonic phased array for detecting oil gas pipeline girth weld defect type |
| CN106682298A (en) * | 2016-12-22 | 2017-05-17 | 西北工业大学 | Aviation hydraulic steering engine system fault simulation model base building method |
| CN107255785A (en) * | 2017-04-28 | 2017-10-17 | 南京邮电大学 | Based on the analog-circuit fault diagnosis method for improving mRMR |
| CN107291997A (en) * | 2017-05-31 | 2017-10-24 | 南京航空航天大学 | A kind of cold rolling hydraulic AGC system design of Fault Diagnosis Strategy method |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104061208A (en) * | 2014-07-02 | 2014-09-24 | 北京机械设备研究所 | Online fault diagnosis method for hydraulic system |
| CN106053067A (en) * | 2016-05-24 | 2016-10-26 | 广东石油化工学院 | Bearing fault diagnosis method based on quantum genetic algorithm optimized support vector machine |
-
2018
- 2018-05-10 CN CN201810441856.0A patent/CN108363896B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7106903B2 (en) * | 2001-09-24 | 2006-09-12 | Vima Technologies, Inc. | Dynamic partial function in measurement of similarity of objects |
| CN101122585A (en) * | 2007-09-12 | 2008-02-13 | 天津大学 | Automatic identification method for supersonic phased array for detecting oil gas pipeline girth weld defect type |
| CN106682298A (en) * | 2016-12-22 | 2017-05-17 | 西北工业大学 | Aviation hydraulic steering engine system fault simulation model base building method |
| CN107255785A (en) * | 2017-04-28 | 2017-10-17 | 南京邮电大学 | Based on the analog-circuit fault diagnosis method for improving mRMR |
| CN107291997A (en) * | 2017-05-31 | 2017-10-24 | 南京航空航天大学 | A kind of cold rolling hydraulic AGC system design of Fault Diagnosis Strategy method |
Non-Patent Citations (2)
| Title |
|---|
| 基于小波分析的电力系统故障时空检测与诊断;李刚 等;《系统仿真学报》;20151231;第27卷(第12期);第3018-3024页 * |
| 基于小波包与遗传算法和支持向量机的液压泵故障诊断;曹斌 等;《组合机床与自动化加工技术》;20141231(第12期);第115-118页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108363896A (en) | 2018-08-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108363896B (en) | Fault diagnosis method for hydraulic cylinder | |
| CN112149316B (en) | Aero-engine residual life prediction method based on improved CNN model | |
| CN109141847B (en) | Aircraft system fault diagnosis method based on MSCNN deep learning | |
| CN106555788B (en) | Application based on the deep learning of Fuzzy Processing in hydraulic equipment fault diagnosis | |
| Saxena et al. | Evolving an artificial neural network classifier for condition monitoring of rotating mechanical systems | |
| CN102707256B (en) | Fault diagnosis method based on BP-Ada Boost nerve network for electric energy meter | |
| CN112001110B (en) | Structural damage identification monitoring method based on vibration signal space real-time recurrent graph convolutional neural network | |
| CN108256173A (en) | A kind of Gas path fault diagnosis method and system of aero-engine dynamic process | |
| CN108921230A (en) | Method for diagnosing faults based on class mean value core pivot element analysis and BP neural network | |
| Wu et al. | A transformer-based approach for novel fault detection and fault classification/diagnosis in manufacturing: A rotary system application | |
| CN113743016B (en) | Engine residual life prediction method based on self-encoder and echo state network | |
| CN110647911A (en) | Bearing fault diagnosis method based on principal component analysis and deep belief network | |
| CN112766283B (en) | Two-phase flow pattern identification method based on multi-scale convolution network | |
| CN108805195A (en) | A kind of motor group method for diagnosing faults based on two-value deep-neural-network | |
| CN114462459A (en) | Hydraulic machine fault diagnosis method based on 1DCNN-LSTM network model | |
| CN116226739A (en) | Map convolution network industrial process fault diagnosis method based on space-time fusion | |
| CN113159225B (en) | Multivariable industrial process fault classification method | |
| CN110554667A (en) | convolutional Neural Network (CNN) based intermittent industrial process fault diagnosis | |
| CN116383757B (en) | Bearing fault diagnosis method based on multi-scale feature fusion and migration learning | |
| Wang | Research on the fault diagnosis of mechanical equipment vibration system based on expert system | |
| CN114841063A (en) | Aero-engine residual life prediction method based on deep learning | |
| CN113971489A (en) | Method and system for predicting remaining service life based on hybrid neural network | |
| CN114638421A (en) | Method for predicting requirement of generator set spare parts | |
| CN114218991A (en) | Rotor fault diagnosis method based on weighted BP-AdaBoost | |
| CN113435321A (en) | Method, system and equipment for evaluating state of main shaft bearing and readable storage medium |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |