CN119848741A - Automatic fault diagnosis method and device for fan equipment - Google Patents
Automatic fault diagnosis method and device for fan equipment Download PDFInfo
- Publication number
- CN119848741A CN119848741A CN202510319556.5A CN202510319556A CN119848741A CN 119848741 A CN119848741 A CN 119848741A CN 202510319556 A CN202510319556 A CN 202510319556A CN 119848741 A CN119848741 A CN 119848741A
- Authority
- CN
- China
- Prior art keywords
- data
- diagnosis
- fan
- frequency
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
- G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
- F03D—WIND MOTORS
- F03D17/00—Monitoring or testing of wind motors, e.g. diagnostics
- F03D17/005—Monitoring or testing of wind motors, e.g. diagnostics using computation methods, e.g. neural networks
- F03D17/0065—Monitoring or testing of wind motors, e.g. diagnostics using computation methods, e.g. neural networks for diagnostics
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
- F03D—WIND MOTORS
- F03D17/00—Monitoring or testing of wind motors, e.g. diagnostics
- F03D17/009—Monitoring or testing of wind motors, e.g. diagnostics characterised by the purpose
- F03D17/015—Monitoring or testing of wind motors, e.g. diagnostics characterised by the purpose for monitoring vibrations
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
- F03D—WIND MOTORS
- F03D17/00—Monitoring or testing of wind motors, e.g. diagnostics
- F03D17/009—Monitoring or testing of wind motors, e.g. diagnostics characterised by the purpose
- F03D17/018—Monitoring or testing of wind motors, e.g. diagnostics characterised by the purpose for monitoring temperature
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
- F03D—WIND MOTORS
- F03D17/00—Monitoring or testing of wind motors, e.g. diagnostics
- F03D17/027—Monitoring or testing of wind motors, e.g. diagnostics characterised by the component being monitored or tested
- F03D17/032—Bearings
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D21/00—Measuring or testing not otherwise provided for
- G01D21/02—Measuring two or more variables by means not covered by a single other subclass
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
Landscapes
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Combustion & Propulsion (AREA)
- General Physics & Mathematics (AREA)
- Chemical & Material Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Sustainable Energy (AREA)
- Physics & Mathematics (AREA)
- Mechanical Engineering (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Sustainable Development (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
本发明公开了一种风机设备故障自动诊断方法及装置，对采集到的风机设备运行数据分为低频特征和高频振动数据分别清洗、降噪、特征提取与降维处理；构建故障诊断模型并进行故障模式分类，以实现对风机设备故障的自动诊断和识别；每次诊断或测试后保存数据用于更新故障诊断模型数据集。可以快速给出多参数诊断结论，提高现场的点检效率，减少人力维修和停机成本。
The present invention discloses a method and device for automatically diagnosing fan equipment faults. The collected fan equipment operation data is divided into low-frequency features and high-frequency vibration data, and the data is cleaned, denoised, feature extracted and dimension reduced respectively; a fault diagnosis model is constructed and fault mode classification is performed to realize automatic diagnosis and identification of fan equipment faults; data is saved after each diagnosis or test to update the fault diagnosis model data set. A multi-parameter diagnosis conclusion can be quickly given, the on-site inspection efficiency is improved, and the manpower maintenance and downtime costs are reduced.
Description
Technical Field
The invention relates to the technical field of fan fault diagnosis, in particular to a fan fault diagnosis method and device based on fault signal analysis.
Background
The fan is widely applied in industrial production, and the running state of the fan directly influences the production efficiency and the safety. The fan fault can cause production interruption, equipment damage and even safety accidents, so that the fan fault diagnosis method has important significance for accurately diagnosing the fan fault. The traditional fan fault diagnosis method is mostly dependent on manual experience, faults are judged by observing the running state of equipment and simple vibration measurement, and the method has the problems of strong subjectivity, low accuracy, incapability of timely finding potential faults and the like. With the development of signal processing technology and machine learning algorithm, fan fault diagnosis methods based on vibration signal analysis gradually become research hotspots. However, the existing fault diagnosis method based on vibration signal analysis still has some problems in practical application, such as insufficient overall signal feature extraction, inaccurate distinction of different fault types, inability to adapt to complex working condition changes, and the like.
In addition, in the fault diagnosis of the rotary equipment, the diagnosis device generally used by the equipment maintainer can only measure the vibration value or display the frequency spectrum, and the data is only a part of information of the running state of the equipment. The running state of fan equipment cannot be comprehensively reflected, and the health condition conclusion of the equipment cannot be directly obtained.
Secondly, the prior art cannot realize automatic fault diagnosis and identification, cannot directly judge whether the equipment works well, and cannot clearly identify the specific fault type and position of the equipment. Therefore, when the equipment maintainer performs diagnosis, the equipment maintainer needs to rely on self experience and judgment, relies on manual analysis and judgment, has complicated process and depends on the experience level of operators. The influence of subjective factors is increased, so that misdiagnosis or missed diagnosis may be caused.
Again, most diagnostic devices fail to accurately detect operating conditions at the time of measurement, often requiring manual operation and subsequent data processing analysis, resulting in failure detection and diagnostic delays. The state of the equipment cannot be analyzed and fed back comprehensively in real time, so that the fault early warning is not timely easily caused under the working condition that the equipment needs to continuously run for a long time, and the production efficiency and the safety are further affected.
Finally, the prejudgement is insufficient. The potential faults cannot be timely early-warned and predicted, equipment is easily found after the faults are deteriorated, and maintenance and shutdown costs are increased.
Disclosure of Invention
The technical problem to be solved by the invention is to provide the automatic fault diagnosis method and the device for the fan equipment, which can rapidly give out a multi-parameter diagnosis conclusion for the fan equipment, improve spot inspection efficiency on site and reduce labor maintenance and shutdown cost. The problem that a large amount of manual intervention is needed is avoided, and the inapplicability, inaccuracy and incompleteness of the universal equipment diagnosis method and device for fan fault investigation are avoided.
In order to solve the technical problems, the invention adopts the following technical scheme:
An automatic fan equipment fault diagnosis method comprises at least one of the following diagnosis steps:
The collected operation data of the fan equipment is divided into low-frequency characteristic and high-frequency vibration data, and the low-frequency characteristic and the high-frequency vibration data are respectively cleaned, noise reduced, characteristic extracted and dimension reduced;
Constructing a fault diagnosis model and classifying fault modes to realize automatic diagnosis and identification of fan equipment faults;
Storing data after each diagnosis or test for updating the fault diagnosis model data set;
The fan equipment operation data comprise process data, vibration data, temperature data, environment parameters and rotating speed parameters.
In the technical scheme, data are stored after each diagnosis or test, samples are added into a fault diagnosis model sample library, a model updating mode is selected, and the accuracy of the fault diagnosis model is iterated step by step.
In the above technical solution, after each diagnosis or test, the validity of the test result is confirmed according to the data tag, and then the diagnosis or test data is saved to the data set.
In the above technical solution, the model updating mode includes three modes of manually triggering manual updating training, updating training based on accuracy rate, and periodically updating training based on time period.
In the technical scheme, for low-frequency characteristic value data, an outlier is identified by using a Z-score method, and linear interpolation filling data is adopted.
In the technical scheme, for the high-frequency vibration data, low-pass filtering is firstly carried out to remove high-frequency noise, and then self-adaptive filtering is carried out to further inhibit dynamic noise.
In the technical scheme, when the characteristics are extracted and dimension is reduced, the characteristics of the process data, the vibration data, the temperature data, the environment parameters and the working condition parameters are extracted, and the PCA is utilized for characteristic selection.
The vibration signals comprise vibration signals in the horizontal direction, the vertical direction and the axial direction, and the temperature signals comprise temperature signals of key parts such as the temperature of the bearing bush.
In the above technical solution, various characteristic parameters are extracted from the preprocessed vibration signal, including time domain characteristics (such as root mean square value, peak value, kurtosis value, etc.), frequency domain characteristics (such as amplitude, frequency, etc. of each frequency band), and time-frequency domain characteristics (such as wavelet transform coefficient, etc.).
In the technical scheme, the characteristic values of all signals obtained during characteristic extraction and dimension reduction comprise peak values and effective values of acceleration time domain signals per second, frequency multiplication amplitude values of acceleration signal spectrums, 2 frequency multiplication amplitude values, 3 frequency multiplication amplitude values, 4 frequency multiplication amplitude values and 5 frequency multiplication amplitude values, peak values and effective values of acceleration envelope signals per second, amplitude values of envelope signal spectrums under fault frequency, energy of acceleration signal spectrums at low frequency band, medium frequency band and high frequency band per second, peak values and effective values of current time domain signals per second, frequency multiplication amplitude values of current signal spectrums, 2 frequency multiplication amplitude values, 3 frequency multiplication amplitude values, 4 frequency multiplication amplitude values and 5 frequency multiplication amplitude values, and average values of temperature signals per second.
In the technical scheme, a fault diagnosis model is built based on the K neighbor classification method.
According to the technical scheme, the fault diagnosis model is obtained by training a large amount of historical fault data based on a support vector machine algorithm, and can identify common fan fault types such as unbalance, misalignment, rolling bearing faults, looseness of a bearing seat and the like. The model carries out classification judgment on the input characteristic parameters by learning the characteristic modes corresponding to different fault types, and outputs a preliminary fault diagnosis result.
In the above technical solution, when constructing a fault diagnosis model and classifying fault modes, the method specifically includes the following steps:
dividing the feature data set after dimension reduction, randomly selecting more than half of the feature samples from the feature data set as training points, using the rest of the feature samples as test samples, and setting a K value;
Calculating Euclidean distance from a sample to be tested to each training point;
Sorting each distance, and screening K training points nearest to a sample to be tested;
Finding out the class of the equipment fault modes affiliated to the screened K nearest neighbor training points;
and classifying the fault modes by classifying the to-be-inspected samples into the equipment fault modes with the highest proportion among the K training points according to a few rules obeying majority.
In the technical scheme, the severity and the development trend of the fault are further analyzed for judging the fault condition. On one hand, the extracted characteristic parameters are compared with a preset threshold value, and when the characteristic parameters exceed a certain threshold value, the fault degree is considered to be aggravated, on the other hand, the trend analysis is carried out on the characteristic parameters, the change trend of the characteristic parameters along with time is observed, and the development direction and the development speed of the fault are predicted. Meanwhile, the severity of the fault is comprehensively evaluated by combining the information such as the running time and the maintenance history of the fan, and a basis is provided for subsequent maintenance work.
In the technical scheme, specific maintenance suggestions and maintenance strategies are given according to the fault diagnosis result and the severity evaluation. For example, for a slight imbalance fault, a simple dynamic balance adjustment may be recommended at the machine shutdown, and for a severe bearing fault, a shutdown to replace the bearing may be scheduled as soon as possible. And finally, the information such as diagnosis results, maintenance suggestions and the like is arranged into report output and stored in a database so as to facilitate subsequent inquiry and analysis, and data support is provided for full life cycle management of the equipment.
An automatic fan equipment fault diagnosis device, comprising:
The system comprises a data acquisition module, a control module and a control module, wherein the data acquisition module is used for acquiring various operation data of fan equipment, including process data, vibration data, temperature data, environmental parameters and rotating speed, wherein the process data acquisition comprises air door opening, fan inlet and outlet pressure, fan inlet and outlet temperature, lubricating oil pressure and temperature;
the calculation module is used for cleaning and reducing noise, extracting features and reducing dimension of the acquired data, and classifying fault modes by utilizing the feature data after dimension reduction to realize automatic diagnosis and identification;
The model training and updating module is used for constructing a fault diagnosis model based on a K neighbor classification method, confirming and storing data after each diagnosis or test and updating the data into a sample library, and gradually iterating the accuracy of the fault diagnosis model in a mode of manually triggering training, updating based on accuracy or carrying out periodic training based on a time period.
The modules are closely cooperated to finish the diagnosis of fan faults.
According to the technical scheme, the data acquisition module comprises a built-in data access module, a data acquisition module and an environment parameter acquisition module, and is provided with a configuration module for a user to select proper equipment initial parameters, wherein the data access module is used for acquiring process data, the data acquisition module is used for acquiring vibration data, temperature data and rotating speed, and the environment parameter acquisition module is used for acquiring real-time environment humidity parameters.
In the above technical scheme, the configuration module is used for configuring initial parameters of equipment, including equipment type, bearing model, gear number and impeller blade number.
In the above technical solution, the configuration module is preferably set as an input device or an interface.
In the technical scheme, the data access module is arranged in the diagnosis device, is configured with the related information of the field process data server and is used for interfacing external equipment and signals according to a standard protocol and acquiring process data of the equipment in real time.
According to the technical scheme, the data acquisition module is arranged in a mode that an acceleration sensor is arranged on each bearing seat of the fan and connected to the patrol instrument through aviation plug, and equipment vibration data are acquired.
In the technical scheme, the data acquisition module is arranged in such a way that temperature sensors are arranged on bearing seats of the fan and acquire temperature data in the same mode of vibration.
In the technical scheme, the data acquisition module is used for acquiring the working rotation speed of the equipment by using the rotation speed sensor as the rotation speed parameter.
That is, the data acquisition module is configured to mount a rotation speed sensor, an acceleration sensor, a temperature sensor on each bearing seat of the blower when acquiring vibration data.
Preferably, the rotating speed sensor and the temperature sensor are arranged on the fan equipment fault diagnosis device through aviation plug so as to collect equipment vibration data.
In addition, the invention also provides a storage medium, on which a computer program is stored, which when executed by a processor can implement the steps of the fan fault diagnosis method. By installing the program in the storage medium on the equipment, the equipment can have the function of fan fault diagnosis, and real-time monitoring of the running state of the fan and automatic fault diagnosis are realized.
Compared with the prior art, the invention has the following beneficial effects:
Although vibration analysis is a mature fault detection technique, existing vibration detection devices still have a number of limitations, particularly in terms of automation and intelligence of fault diagnosis. Most existing diagnostic and detection devices can only provide basic vibration data such as acceleration, velocity, displacement, etc., or simple spectrograms. While these data provide clues to the status of the operation of the device to the expert, it is difficult to automatically and accurately determine the type of fault and its location without the involvement of a professional. Therefore, the existing equipment state monitoring technology often needs to rely on experienced professionals to perform complex spectrum analysis and judgment, and such manual diagnosis is time-consuming and labor-consuming and is easily affected by subjective factors. The device is internally provided with various characteristic value calculation modes and mechanism diagnosis rules, and can rapidly give a diagnosis conclusion from characteristic extraction to fault self-diagnosis, improve spot inspection efficiency on site and reduce manpower.
At present, better diagnosis devices such as Emerson CSI2140 series products in the market only have the functions of vibration data acquisition, vibration characteristic value calculation and waveform presentation, and specific analysis also needs to manually analyze the frequency spectrum to locate faults. In addition, as VM-63A of RION company, a popular product is used, but the product has three characteristic value display functions, only can display the vibration, and cannot realize accurate analysis of faults.
The automatic fault diagnosis device for the fan equipment is internally provided with a plurality of data acquisition and processing modes, is specially designed and developed for the fan equipment, is used for acquiring a plurality of operation data of the fan equipment, including process data, vibration data, temperature data, environment parameters and rotating speed parameters, can process the data of the opening degree of an air door, the inlet and outlet pressure of the fan, the inlet and outlet temperature of the fan, the lubricating oil pressure and the temperature of the fan equipment, and can judge and process more faults, more accuracy and more comprehensiveness. The invention can more comprehensively reflect the running state of the fan and improve the accuracy of fault diagnosis through multi-dimensional signal acquisition and comprehensive feature extraction. The method can accurately identify the fault type of the fan, evaluate the severity and development trend of the fault, give out targeted maintenance suggestions, effectively improve the operation reliability and maintenance efficiency of fan equipment, reduce the loss and risk caused by equipment faults, and has remarkable economic and social benefits, and can be widely applied to the field of fan equipment management in various industrial production.
Compared with a general intelligent diagnosis tool, the fan equipment fault automatic diagnosis device provided by the invention has the advantages that the diagnosis model aiming at the fan is set, the self-updating module is matched with the diagnosis model, the manual updating mode is set, the updating mode is based on the accuracy rate, the updating mode is selected based on the time updating mode, and the model sample library can be updated based on the test result, so that the diagnosis capability in the model library is enhanced, the equipment has a self-learning function, and the diagnosis accuracy rate can be iterated step by step.
The fault type can be judged, the severity and the development trend of the fault can be evaluated, more targeted suggestions are provided for equipment maintenance, and the equipment downtime and the maintenance cost are effectively reduced.
Drawings
The invention will be further described with reference to the accompanying drawings and examples, in which:
FIG. 1 is a flow chart of an embodiment of an automatic fan equipment failure diagnosis method of the present invention.
FIG. 2 is a schematic diagram of the operation of the data acquisition module in the fan equipment failure automatic diagnostic device of the present invention.
FIG. 3 is a schematic diagram of the operation of the calculation module in the fan equipment failure automatic diagnostic device of the present invention.
FIG. 4 is a schematic diagram of the model training and updating module in the fan equipment failure automatic diagnostic device of the present invention.
FIG. 5 is a schematic diagram of the operation of a configuration module included in the data acquisition module in the fan apparatus failure automatic diagnosis device of the present invention.
Fig. 6 is a data diagram (prior to noise reduction) of an actual diagnostic acquisition in accordance with an embodiment of the present invention.
Fig. 7 is a schematic diagram of the data de-noising of fig. 6.
Fig. 8 is a 2-dimensional data diagram of the data of fig. 7 after noise reduction.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
Example 1
The embodiment provides a fan equipment fault automatic diagnosis method, the flow of which is shown in fig. 1, and the specific steps are as follows:
And a plurality of sensors arranged on the fan are utilized to collect operation data such as vibration signals, rotating speed, temperature and humidity signals and the like. Vibration signals are collected from the horizontal direction, the vertical direction and the axial direction respectively to comprehensively reflect the vibration state of the fan, and temperature signals are used for collecting the temperatures of key parts such as bearing bushes and the like so as to discover abnormal conditions such as overheating and the like in time.
Preprocessing the acquired signals. The method comprises the steps of filtering and denoising, removing high-frequency noise and interference components in signals, improving the signal to noise ratio of the signals, normalizing, unifying data with different dimensions and measuring ranges to the same range, and facilitating subsequent feature extraction and analysis.
And extracting various characteristic parameters from the preprocessed vibration signals. In the time domain, calculating statistical characteristics such as root mean square value, peak value, kurtosis value and the like, wherein the characteristics can reflect the overall energy and impact characteristics of signals, in the frequency domain, converting the signals into the frequency domain through Fast Fourier Transform (FFT), extracting amplitude and frequency information of each frequency band to identify different vibration frequency components, in the time-frequency domain, acquiring time-frequency domain characteristics of the signals through wavelet transform and the like, and capturing transient changes and local characteristics of the signals.
The extracted characteristic parameters are input into a pre-trained fault diagnosis model, the model is obtained by training a large amount of historical fault data based on a support vector machine algorithm, and the common fan fault types such as unbalance, misalignment, rolling bearing faults, looseness of a bearing seat and the like can be identified. The model carries out classification judgment on the input characteristic parameters by learning the characteristic modes corresponding to different fault types, and outputs a preliminary fault diagnosis result.
And further analyzing the severity degree and the development trend of the fault for judging the fault condition. On one hand, the extracted characteristic parameters are compared with a preset threshold value, and when the characteristic parameters exceed a certain threshold value, the fault degree is considered to be aggravated, on the other hand, the trend analysis is carried out on the characteristic parameters, the change trend of the characteristic parameters along with time is observed, and the development direction and the development speed of the fault are predicted. Meanwhile, the severity of the fault is comprehensively evaluated by combining the information such as the running time and the maintenance history of the fan, and a basis is provided for subsequent maintenance work.
If unbalance is diagnosed, based on the ratio of 1 times and 2 times, it can be judged whether the fan rotor is unbalanced.
And judging whether the fan rotor is in poor centering or not based on the condition of the ratio of 2 times to 1 times.
Preferably, the embodiment can be additionally provided with specific maintenance suggestions and maintenance strategies according to fault diagnosis results and severity evaluation. For example, for a slight imbalance fault, a simple dynamic balance adjustment may be recommended at the machine shutdown, and for a severe bearing fault, a shutdown to replace the bearing may be scheduled as soon as possible. And finally, the information such as diagnosis results, maintenance suggestions and the like is arranged into report output and stored in a database so as to facilitate subsequent inquiry and analysis, and data support is provided for full life cycle management of the equipment.
Set suggested matching rules as in the model:
If the vibration difference of the air inlet side shaft is large, checking the condition of the air inlet side vibration probe;
The vibration difference of the air outlet side shaft is large, and the condition of the air outlet side vibration probe is checked;
the temperature difference of the bearing bush is larger, it is recommended to check the bushing temperature probe condition.
The equipment base loosens, 1, checking the fastening force of the foundation of the bearing pedestal and fastening the bolts according to the requirement, 2, checking whether the belleville springs deform and the foundation loosens, and 3, whether the frame corrodes to cause the vibration of the bearing pedestal.
Oil film whirl, checking whether the bearing bush is worn or not, and adjusting the bearing bush gap and the bearing bush oil supply pressure if necessary;
The friction of dynamic and static components is suggested to be 1, parameters are adjusted to ensure even thermal expansion of the unit, 2, concentricity of a rotor and a stator is overhauled and adjusted, 3, basic levelness is measured to eliminate sedimentation influence, and 4, the performance of the unit is checked to determine whether overhaul is carried out.
Rotor unbalance, proposal 1. Impeller steam blowing and ammonia water soaking, 2. Dynamic balancing and static balancing, and 3. Checking whether the blade is damaged or cracked.
The method for detecting the defect of the centering of the fan rotor comprises the steps of 1, detecting the centering deviation condition of a coupling, 2, detecting the installation concentricity of the shaft centers at two ends of equipment, 3, detecting whether a foundation subsides or not, and 4, screening whether a rotating shaft is bent (thermally) or not through signal characteristics.
Example 2
The fan equipment fault automatic diagnosis device is internally provided with various characteristic value calculation modes and mechanism diagnosis rules, and can rapidly give diagnosis conclusion from characteristic extraction to fault self-diagnosis, improve spot inspection efficiency on site and reduce manpower.
The invention relates to an automatic fault diagnosis device for fan equipment, which is shown in figures 1-5 and comprises the following specific steps:
The system comprises a data acquisition module, a calculation module and a model training and updating module.
The data acquisition module is shown in fig. 2, and comprises a data access module, a data acquisition module and an environmental parameter acquisition module, wherein the data acquisition module is used for acquiring various operation data of the fan equipment, including process data, vibration data, temperature data, environmental parameters and rotating speed parameters, and is provided with a configuration module for a user to select proper initial parameters of the equipment.
The data access module is arranged in the diagnosis device and is configured with relevant information of a field process data server, and is used for interfacing external equipment and signals according to a standard protocol to acquire process data of the equipment in real time, wherein the process data are acquired by air door opening, fan inlet and outlet pressure, fan inlet and outlet temperature, lubricating oil pressure and temperature.
The data acquisition module is used for acquiring vibration data, temperature data and rotating speed parameters.
Vibration data acquisition, namely, installing acceleration sensors on each bearing seat of the fan, and acquiring equipment vibration data by plugging the acceleration sensors on a patrol instrument.
Temperature data, namely, temperature sensors are arranged on bearing seats of the fan, and the temperature data are acquired in the same mode of vibration.
The working rotation speed of the equipment is collected by the rotation speed sensor to be used as a rotation speed parameter.
The environment parameter acquisition module is composed of a temperature and humidity module arranged in the diagnosis device and is used for acquiring real-time environment humidity.
The calculation module is used for cleaning and denoising the acquired data, extracting the characteristics and reducing the dimension, and classifying the failure modes by utilizing the characteristic data after the dimension reduction, so as to realize automatic diagnosis and identification, as shown in fig. 3;
The model training and updating module is shown in fig. 4, and is used for constructing a fault diagnosis model based on a K-nearest neighbor classification method, and for confirming and storing diagnosis or test data after each diagnosis or test and updating the diagnosis or test data into a sample library, and gradually iterating the accuracy of the fault diagnosis model by means of manual trigger training, updating based on accuracy or regular training based on a time period.
The configuration module is shown in fig. 5, and the fan equipment fault automatic diagnosis device forms a part of the data acquisition module through the configuration module such as an input device or an input interface, and is used for a user to select appropriate initial parameters of equipment, including equipment type, bearing model, gear number and impeller blade number.
And then, a rotating speed sensor, a temperature sensor and an acceleration sensor are arranged on each bearing seat of the fan.
Preferably, the rotation speed sensor, the temperature sensor and the acceleration sensor are arranged on the fan equipment fault diagnosis device through aviation plug so as to collect equipment vibration data.
Example 3
As shown in fig. 1 to 5, the automatic fan equipment fault diagnosis method of the present invention can identify common fan faults such as unbalance, misalignment, rolling bearing faults, looseness of a bearing seat, etc., and includes the following steps:
The collected operation data of the fan equipment is divided into low-frequency characteristic and high-frequency vibration data, and the low-frequency characteristic and the high-frequency vibration data are respectively cleaned, noise reduced, characteristic extracted and dimension reduced;
Constructing a fault diagnosis model and classifying fault modes to realize automatic diagnosis and identification of fan equipment faults;
diagnostic or test data is saved after each diagnosis or test for updating the fault diagnosis model dataset.
The data acquisition process is as follows:
The acquisition data comprise process data, vibration data, temperature data, environmental parameters and rotating speed parameters of the equipment;
The process data acquisition comprises throttle opening, inlet and outlet pressure of the fan, inlet and outlet temperature of the fan, lubricating oil pressure and temperature. The process data is realized by a data access module arranged in the diagnostic device, the data access module is configured with relevant information of a field process data server, process data of equipment is obtained in real time,
Vibration data acquisition, namely, installing acceleration sensors on each bearing seat of the fan, and acquiring equipment vibration data by plugging the acceleration sensors on a patrol instrument.
Temperature data, namely, installing temperature sensors on each bearing seat of the fan, and collecting temperature data in the same mode of vibration;
environmental parameters, namely a temperature and humidity module is arranged in the diagnosis device, and real-time environmental parameters are collected;
The working rotation speed of the equipment is collected by the rotation speed sensor and is used as a working condition parameter.
The data cleaning and noise reduction process is as follows:
Low frequency characteristic value data, because the low frequency characteristic value data belongs to slow variable data in the field, the data does not have abrupt change such as jump, etc., so that the interference (jump data) in the identification process is reduced, and the interference to the diagnosis conclusion is reduced
The following data cleaning method is provided:
Identifying outliers in the feature data using the Z-score method:
Z-score refers to the difference between a data point and the mean of the data set, expressed in standard deviation:
;
Wherein: Is a particular data point in the dataset; Is the mean of the dataset; is the standard deviation of the data set, and is generally equal to or greater than 3 or equal to or less than-3 in the statistical sense, and the data point is regarded as an abnormal value.
After the abnormal point is identified, filling data is carried out by utilizing a linear interpolation mode.
The high-frequency vibration data is generated by the interference of various strong electricity and magnetic fields on the industrial site, so that noise data in the high-frequency vibration signals are filtered by using a noise reduction method.
In the signal preprocessing stage, firstly, low-pass filtering is carried out on the acquired vibration acceleration signals to filter high-frequency noise components, and meanwhile, low-frequency information related to mechanical faults is reserved. And then, the adaptive filter further suppresses dynamic noise, and adjusts the parameters of the filter according to real-time feedback, so that the filtering effect is ensured to adapt to signal changes under different working conditions.
The feature selection and dimension reduction process is as follows:
the characteristic values of all the signals obtained comprise peak values, effective values, frequency multiplication amplitude values, 2 frequency multiplication amplitude values, 3 frequency multiplication amplitude values, 4 frequency multiplication amplitude values and 5 frequency multiplication amplitude values of the acceleration signal frequency spectrums per second, peak values, effective values of the acceleration envelope signals per second, the effective values and amplitude values of the envelope signal frequency spectrums under fault frequency, energy of the acceleration signal frequency spectrums in a low frequency band, a medium frequency band and a high frequency band per second, frequency multiplication amplitude values, 2 frequency multiplication amplitude values, 3 frequency multiplication amplitude values, 4 frequency multiplication amplitude values and 5 frequency multiplication amplitude values of the current signal frequency spectrums per second, and average values of the temperature signals per second.
The model operation and diagnosis process is as follows:
and inputting the value output in the previous step into the model, generating a final result, and assisting the operation and maintenance judgment of the point inspection personnel. The model construction is based on a K-nearest neighbor classification method.
Classifying the feature after dimension reduction by adopting a K nearest neighbor algorithm in a classifying mode, wherein the classifying method specifically comprises the following steps:
1) Dividing the feature data set after dimension reduction, randomly selecting 70% of feature samples from the feature data set as training points, namely, the prior information source of fault classification, and setting a K value, wherein the K nearest neighbor algorithm is a classification method, namely, K nearest neighbors refer to K nearest neighbors, each feature sample to be detected can be represented by K training points closest to the K nearest neighbors;
2) Calculating Euclidean distance from a sample to be tested to each training point;
3) Sorting each distance, and screening K training points nearest to a sample to be tested;
4) Finding out the class of the equipment fault modes affiliated to the screened K nearest neighbor training points;
5) And classifying the fault modes by classifying the to-be-inspected samples into the equipment fault modes with the highest proportion among the K training points according to a few rules obeying majority.
The model self-updating process is as follows:
After diagnosis or test is carried out each time, the diagnosis or test data of the time are stored, the spot inspection personnel compares the judgment result of the diagnosis device with the actual manual judgment result on site, the actual judgment result is input into the diagnosis device to form a diagnosis closed loop, and samples are added into a sample library.
The diagnostic device is internally provided with a plurality of training modes, can be manually triggered to train, is based on accuracy training or is based on a time period to carry out a regular training mechanism, and gradually iterates the accuracy of the model.
The accuracy rate can be set to different range thresholds, for example, for different fault occasions, the iteration can be performed with higher accuracy rate for more complex situations, and lower fault accuracy rate can be applied for faults with comparison rules.
Example 4
The invention relates to a fan equipment fault automatic diagnosis method and a device practical application example.
All data of the field are sent to the platform after the project sensor is installed, and the data are used for verifying that the early warning and diagnosis part data come from real data on the field device.
The data are arranged into a data set to form a standard data format, the result is generated by batch operation of the program, the result is compared with the actual data, the diagnosis model is optimized and compared in the test process, and the parameters of the diagnosis model are optimized.
The collected data is specifically diagnosed, the collected data is subjected to noise reduction treatment by using a Z-score method, and the data before and after noise reduction are respectively shown in fig. 6 and 7.
The data after the dimension reduction processing is shown in fig. 8.
And training and generating a classification model by using the dimension reduced data, and inputting the data acquired on site into the classification model.
The diagnosis device generates a diagnosis result, selects whether to be added into the sample library after manually confirming the result, and tags the data, so that the accuracy of the model can be continuously improved. The comparison can realize that the equipment continuous operation rate is 99%, the personnel spot inspection efficiency is increased by 20%, and the fault diagnosis accuracy rate is 60%.
It will be understood that modifications and variations will be apparent to those skilled in the art from the foregoing description, and it is intended that all such modifications and variations be included within the scope of the following claims.
Claims (10)
1. An automatic fan equipment fault diagnosis method is characterized by comprising at least one of the following diagnosis steps:
The collected operation data of the fan equipment is divided into low-frequency characteristic and high-frequency vibration data, and the low-frequency characteristic and the high-frequency vibration data are respectively cleaned, noise reduced, characteristic extracted and dimension reduced;
Constructing a fault diagnosis model and classifying fault modes to realize automatic diagnosis and identification of fan equipment faults;
Storing data after each diagnosis or test for updating the fault diagnosis model data set;
The fan equipment operation data comprise process data, vibration data, temperature data, environment parameters and rotating speed parameters, wherein the process data comprise air door opening degree, fan inlet and outlet pressure, fan inlet and outlet temperature, lubricating oil pressure and temperature data of the equipment.
2. The automatic diagnosis method of fan apparatus failure according to claim 1, wherein data is saved after each diagnosis or test, samples are added to a failure diagnosis model sample library, a model update mode is selected, and the accuracy of the failure diagnosis model is iterated step by step.
3. The fan apparatus malfunction automatic diagnosis method according to claim 1, wherein for the low frequency characteristic value data, an outlier is identified using a Z-score method, and the data is filled using linear interpolation.
4. The automatic diagnosis method for fan apparatus failure according to claim 1, wherein for the high-frequency vibration data, low-pass filtering is performed to remove high-frequency noise, and then dynamic noise is further suppressed by adaptive filtering.
5. The method for automatic diagnosis of fan apparatus failure according to claim 1, wherein the feature selection is performed by extracting features of the process data, vibration data, temperature data, environmental parameters, and operating mode parameters and using PCA when feature extraction and dimension reduction are performed.
6. The method for automatically diagnosing faults of fan equipment according to claim 1, wherein the characteristic values of the signals obtained during characteristic extraction and dimension reduction comprise peak values, effective values, frequency multiplication amplitude values, 2 frequency multiplication amplitude values, 3 frequency multiplication amplitude values, 4 frequency multiplication amplitude values and 5 frequency multiplication amplitude values of an acceleration time domain signal per second, peak values, effective values and amplitude values of an acceleration envelope signal per second, energy of an acceleration signal spectrum in a low frequency band, a medium frequency band and a high frequency band, peak values, effective values and frequency multiplication amplitude values, 2 frequency multiplication amplitude values, 3 frequency multiplication amplitude values, 4 frequency multiplication amplitude values and 5 frequency multiplication amplitude values of a current signal spectrum per second, and average values of a temperature signal per second.
7. The fan apparatus failure automatic diagnosis method according to claim 1, wherein the failure diagnosis model is constructed based on a K-nearest neighbor classification method.
8. An automatic fan equipment fault diagnosis device is characterized by comprising:
The system comprises a data acquisition module, a control module and a control module, wherein the data acquisition module is used for acquiring various operation data of fan equipment, including process data, vibration data, temperature data, environmental parameters and rotating speed, wherein the process data acquisition comprises air door opening, fan inlet and outlet pressure, fan inlet and outlet temperature, lubricating oil pressure and temperature;
the calculation module is used for cleaning and reducing noise, extracting features and reducing dimension of the acquired data, and classifying fault modes by utilizing the feature data after dimension reduction to realize automatic diagnosis and identification;
The model training and updating module is used for constructing a fault diagnosis model based on a K neighbor classification method, confirming and storing data after each diagnosis or test and updating the data into a sample library, and gradually iterating the accuracy of the fault diagnosis model in a mode of manually triggering training, updating based on accuracy or carrying out periodic training based on a time period.
9. The automatic fan equipment fault diagnosis device according to claim 8, wherein the data acquisition module comprises a built-in data access module, a data acquisition module and an environment parameter acquisition module, wherein the configuration module is used for selecting initial parameters of equipment, the data access module is used for acquiring process data, the data acquisition module is used for acquiring vibration data, temperature data and rotating speed, and the environment parameter acquisition module is used for acquiring environment humidity.
10. The automatic fan apparatus failure diagnosis apparatus according to claim 8, wherein the data acquisition module is configured to install a rotation speed sensor, a temperature sensor, an acceleration sensor on each bearing housing of the fan when acquiring vibration data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202510319556.5A CN119848741A (en) | 2025-03-18 | 2025-03-18 | Automatic fault diagnosis method and device for fan equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202510319556.5A CN119848741A (en) | 2025-03-18 | 2025-03-18 | Automatic fault diagnosis method and device for fan equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN119848741A true CN119848741A (en) | 2025-04-18 |
Family
ID=95360370
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202510319556.5A Pending CN119848741A (en) | 2025-03-18 | 2025-03-18 | Automatic fault diagnosis method and device for fan equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119848741A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108287529A (en) * | 2017-12-28 | 2018-07-17 | 北京必可测科技股份有限公司 | Inspection, long-range accurate diagnosis and maintenance system and a method are patrolled in integration |
| CN113632026A (en) * | 2019-08-29 | 2021-11-09 | 亿可能源科技（上海）有限公司 | Rotating machinery equipment fault diagnosis method, system and storage medium |
| WO2022261805A1 (en) * | 2021-06-15 | 2022-12-22 | 大连理工大学 | Diesel engine gearbox fault diagnosis method |
| CN118857737A (en) * | 2024-06-30 | 2024-10-29 | 中国舰船研究设计中心 | A sliding bearing fault diagnosis system and method |
| CN119393366A (en) * | 2024-10-17 | 2025-02-07 | 华能洛阳热电有限责任公司 | Multimodal fault diagnosis method for wind turbine based on voiceprint feature and vibration data fusion |
| CN119509958A (en) * | 2024-09-26 | 2025-02-25 | 东方电气集团科学技术研究院有限公司 | A common feature diagnosis system for spindle bending based on stacking integration method |
-
2025
- 2025-03-18 CN CN202510319556.5A patent/CN119848741A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108287529A (en) * | 2017-12-28 | 2018-07-17 | 北京必可测科技股份有限公司 | Inspection, long-range accurate diagnosis and maintenance system and a method are patrolled in integration |
| CN113632026A (en) * | 2019-08-29 | 2021-11-09 | 亿可能源科技（上海）有限公司 | Rotating machinery equipment fault diagnosis method, system and storage medium |
| WO2022261805A1 (en) * | 2021-06-15 | 2022-12-22 | 大连理工大学 | Diesel engine gearbox fault diagnosis method |
| CN118857737A (en) * | 2024-06-30 | 2024-10-29 | 中国舰船研究设计中心 | A sliding bearing fault diagnosis system and method |
| CN119509958A (en) * | 2024-09-26 | 2025-02-25 | 东方电气集团科学技术研究院有限公司 | A common feature diagnosis system for spindle bending based on stacking integration method |
| CN119393366A (en) * | 2024-10-17 | 2025-02-07 | 华能洛阳热电有限责任公司 | Multimodal fault diagnosis method for wind turbine based on voiceprint feature and vibration data fusion |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | A novel statistical time-frequency analysis for rotating machine condition monitoring | |
| CN107247230B (en) | Rotary motor state monitoring method based on support vector machine and data driving | |
| Huo et al. | Incipient fault diagnosis of roller bearing using optimized wavelet transform based multi-speed vibration signatures | |
| Soylemezoglu et al. | Mahalanobis-Taguchi system as a multi-sensor based decision making prognostics tool for centrifugal pump failures | |
| KR101718251B1 (en) | Method and system for monitoring rotating blade health | |
| CN117270514B (en) | Production process whole-flow fault detection method based on industrial Internet of things | |
| CN111922095A (en) | Vibration diagnosis method for abnormal torsional vibration fault of roller of cold rolling mill | |
| CN118243200B (en) | A gas turbine flowmeter fault diagnosis method and system | |
| CN115539139B (en) | Monitoring method for safety of steam turbine | |
| KR102545672B1 (en) | Method and apparatus for machine fault diagnosis | |
| CN118030409A (en) | Method and system for detecting abnormal operation performance of fan unit | |
| CN118245808B (en) | Machine learning-based speed reducer fault diagnosis method and system | |
| CN118503882B (en) | The method of operating status parameter collection and fault diagnosis of three-phase horizontal screw centrifuge | |
| CN106441843B (en) | A kind of rotating machinery fault method for waveform identification | |
| Zorchenko et al. | Technologies used by General Electric to create digital twins for energy industry | |
| CN117370879A (en) | Real-time online fault diagnosis method and system for wind turbine gearbox | |
| CN112711850A (en) | Unit online monitoring method based on big data | |
| CN119398742A (en) | An equipment operation and maintenance management system based on tobacco processing production line | |
| US11339763B2 (en) | Method for windmill farm monitoring | |
| CN119124606A (en) | A method and device for online diagnosis of turbine cooler | |
| CN119848741A (en) | Automatic fault diagnosis method and device for fan equipment | |
| Souza et al. | Applying Mahalanobis-Taguchi method to detect faults in rotating machinery | |
| Dong et al. | An early fault warning and diagnosis model based on a two-step analysis strategy for wind turbine bearings | |
| Alekseev et al. | Data measurement system of compressor units defect diagnosis by vibration value | |
| Elsamanty et al. | Rotary Machines Fault Diagnosis based on Principal Component Analysis |
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 |