CN117451347A - Portal crane reduction gearbox intelligent health monitoring system based on fault characteristic model - Google Patents
Portal crane reduction gearbox intelligent health monitoring system based on fault characteristic model Download PDFInfo
- Publication number
- CN117451347A CN117451347A CN202311407596.2A CN202311407596A CN117451347A CN 117451347 A CN117451347 A CN 117451347A CN 202311407596 A CN202311407596 A CN 202311407596A CN 117451347 A CN117451347 A CN 117451347A
- Authority
- CN
- China
- Prior art keywords
- fault
- reduction gearbox
- equipment
- time
- monitoring system
- 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
- 238000012544 monitoring process Methods 0.000 title claims abstract description 86
- 230000036541 health Effects 0.000 title claims abstract description 75
- 230000009467 reduction Effects 0.000 title claims abstract description 69
- 238000012423 maintenance Methods 0.000 claims abstract description 24
- 238000012545 processing Methods 0.000 claims abstract description 18
- 238000003745 diagnosis Methods 0.000 claims abstract description 14
- 230000003750 conditioning effect Effects 0.000 claims abstract description 4
- 238000001914 filtration Methods 0.000 claims abstract description 4
- 238000000034 method Methods 0.000 claims description 32
- 238000010586 diagram Methods 0.000 claims description 16
- 101100098918 Arabidopsis thaliana TFCA gene Proteins 0.000 claims description 14
- 238000004458 analytical method Methods 0.000 claims description 14
- 238000012281 transfemoral cerebral angiography Methods 0.000 claims description 14
- 230000002457 bidirectional effect Effects 0.000 claims description 12
- 230000008901 benefit Effects 0.000 claims description 11
- 238000000605 extraction Methods 0.000 claims description 9
- 238000012549 training Methods 0.000 claims description 9
- 230000002159 abnormal effect Effects 0.000 claims description 6
- 238000013528 artificial neural network Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 6
- 238000003062 neural network model Methods 0.000 claims description 4
- 238000012360 testing method Methods 0.000 claims description 4
- 238000005299 abrasion Methods 0.000 claims description 3
- 230000009471 action Effects 0.000 claims description 3
- 238000013459 approach Methods 0.000 claims description 3
- 238000004891 communication Methods 0.000 claims description 3
- 238000001228 spectrum Methods 0.000 claims description 3
- 238000001931 thermography Methods 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 claims description 2
- 230000003321 amplification Effects 0.000 claims 1
- 238000003199 nucleic acid amplification method Methods 0.000 claims 1
- 238000011160 research Methods 0.000 abstract description 17
- 230000032683 aging Effects 0.000 abstract description 2
- 238000007689 inspection Methods 0.000 abstract description 2
- 238000004590 computer program Methods 0.000 description 8
- 230000006872 improvement Effects 0.000 description 8
- 230000007246 mechanism Effects 0.000 description 5
- 238000003860 storage Methods 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 4
- 230000004927 fusion Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 208000032368 Device malfunction Diseases 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 241001391944 Commicarpus scandens Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000001427 coherent effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 230000005236 sound signal Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000033772 system development Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/02—Gearings; Transmission mechanisms
-
- 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
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/02—Gearings; Transmission mechanisms
- G01M13/028—Acoustic or vibration analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mathematical Physics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Computing Systems (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Acoustics & Sound (AREA)
- Bioinformatics & Computational Biology (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention relates to the technical field of port machine equipment maintenance and inspection, in particular to an intelligent health monitoring system for a portal crane reduction gearbox based on a fault characteristic model. The system comprises a state sensing module, a health monitoring module and a state monitoring platform; the state sensing module is used for sensing the situation of key parts of the door machine reduction gearbox and processing equipment data through signal conditioning, amplifying and filtering; the health monitoring module is used for analyzing and predicting the received equipment data; the state monitoring platform is used for displaying the running state of the reduction gearbox, fault alarming and historical record data in real time. The intelligent health monitoring system for the port machine equipment is designed and built based on the fault feature model research, and the aging diagnosis and fault diagnosis of the reduction gearbox are realized by developing the intelligent curve model research based on the noise sensing equipment, the running state of the equipment is mastered in real time, the equipment faults are actively prevented, and the predictive maintenance of the reduction gearbox is realized.
Description
Technical Field
The invention relates to the technical field of port machine equipment maintenance and inspection, in particular to an intelligent health monitoring system for a portal crane reduction gearbox based on a fault characteristic model.
Background
With the deep convergence of world economy, ports serve as portals for exchanging external commodities, and the importance of the born economic development is increasing. Generally, the service period of the port machine equipment is longer, the application environment is worse, the reduction gearbox needs to be subjected to the test of high temperature, high humidity, high salinity and other environments for a long time, and the internal parts are easy to break down or even damage. The detection difficulty is high during maintenance, the economic loss of equipment unscheduled fault shutdown is high, the maintenance mode is low in efficiency due to the fact that the maintenance mode is highly dependent on experience inheritance of field experts, and the port operation and maintenance cost is high.
Typically, if the sensing device malfunctions or requires maintenance, the operation and handling of the dock may be affected. For example, if the sensing device malfunctions during loading and unloading of cargo, the cargo circulation speed may be reduced, thereby increasing the waiting time of the ship. Furthermore, any error in port and dock operation may result in loss of cargo, and failure of the sensing device may exacerbate this risk. Therefore, if the health degree of the sensing equipment can be monitored, the equipment can be maintained timely according to the feedback information, and the operation efficiency of the wharf can be ensured.
The health guarantee is to diagnose, repair and maintain the running state of the reduction gearbox in real time, and predict the 'health' state of the production equipment in advance so as to develop the formulation of predictive maintenance strategies, and strive for the predictive maintenance and the maintenance management of the equipment in an economic and effective way on the premise of meeting the normal operation of the equipment. The following problems generally exist in the current door machine reduction gearbox health protection system:
1. the reasonable distribution, layout, quantity and sensor type of the monitoring sensors lack theoretical guidance, and the experience of the actual successful engineering project is less;
2. the cost and the precision of the monitoring sensor are difficult to balance, so that the cost of system equipment is often too high to achieve an effective diagnosis effect, and the whole system lacks economic benefits;
3. the research on the algorithms of sensor collection and signal processing strengthens the collection of the state and operation and maintenance data of the reduction gearbox, and the built data mining and analysis model has insufficient capability.
In the existing door machine reduction gearbox health care system, a predictive maintenance algorithm, a machine learning algorithm and a sensor fusion algorithm are mainly applied to ensure the normal operation of equipment. However, these methods have some limitations: relying on large amounts of data and computing resources, failing to accurately predict all faults, the new data is identified with low accuracy, is subject to noise and interference, and the like. These factors all affect the overall effectiveness of the door motor reduction gearbox health care system. In view of the above, we provide a door machine reduction gearbox intelligent health monitoring system based on fault feature model research.
Disclosure of Invention
The invention aims to provide an intelligent health monitoring system for a portal crane reduction gearbox based on a fault characteristic model, which aims to solve the problems in the background technology.
In order to solve the technical problems, one of the purposes of the invention is to provide an intelligent health monitoring system of a door machine reduction gearbox based on fault feature model research, which comprises a state sensing module, a health monitoring module and a state monitoring platform which are sequentially connected in a communication way; wherein:
the state sensing module is used for firstly determining the types, positions and channel quantity of the sensors by sensing the situation of key parts of the portal crane reduction gearbox, and then transmitting the device data to the health monitoring module after processing such as signal conditioning, amplifying and filtering;
the health monitoring module is used for analyzing and predicting the received equipment data, providing data support for functions/subsystems such as equipment safety evaluation, operation and maintenance management and the like, and sending the analysis and prediction results to the state monitoring platform for display;
the state monitoring platform is used for displaying the running state of the reduction gearbox, fault alarming, historical record and other data in real time.
As a further improvement of the technical scheme, the state sensing module is mainly connected into a system through monitoring means such as temperature, wen Zhen, noise, infrared thermal imaging, current and voltage and the like to monitor the action state of a door machine reduction gearbox; the selection of the measuring points needs to comprehensively consider factors such as key parts, equipment running states, benefit cost ratio and the like so as to ensure the accuracy and stability of the monitoring system and improve the safety and service life of the monitoring system; wherein, the selection of the measuring point needs to consider the following aspects:
(1) The key part of the reduction gearbox is selected from the measuring points;
(2) The measuring point should select a parameter capable of reflecting the running state of the equipment;
(3) The number and variety of stations should balance benefit and cost while satisfying function.
As a further improvement of the technical scheme, the types of faults which are easy to occur in the input stage, the intermediate stage and the output stage of the reduction gearbox are faults of a rotor, a shafting, a bearing, collision and abrasion and looseness, so that the state sensing module mainly monitors parameters such as temperature, vibration, noise and the like; the state sensing module at least comprises a temperature sensor, a Wen Zhen sensor and a noise sensor;
the running state of the equipment is estimated by installing a Wen Zhenchuan sensor at the bottom of the reduction gearbox and monitoring the temperature and vibration condition of the equipment in real time, and data is transmitted to a monitoring system for analysis and processing, so that the faults and abnormal conditions of the equipment are found in time; the noise sensor is arranged to monitor the noise generated during the running of the equipment in real time and store the noise data in different time periods, and whether the running state of the equipment is abnormal or not can be judged by comparing the noise data in different time periods.
As a further improvement of the technical scheme, the process of analyzing and predicting the received equipment data by the health monitoring module mainly comprises three parts of feature extraction, prediction learning and health diagnosis.
As a further improvement of the technical scheme, in the feature extraction, taking vibration parameters as an example, the fault type of the reduction gearbox is mainly damage of an inner ring/an outer ring of a bearing; these fault characteristics are all reflected in the vibration signal; when the door machine actually operates, the noise signal generated by the reduction gearbox is a typical non-stable vibration signal;
in general, when the traditional time-frequency coherence method (TFCA) is applied to bearing fault monitoring, two groups of different signals need to be acquired for time-frequency coherence analysis, and only one sensor (namely a single-channel acquisition signal) can be installed at the optimal acquisition position during monitoring, so that the single-channel acquisition signal cannot be analyzed by utilizing the traditional TFCA method, and the limitation is larger; based on the characteristics of acoustic emission signals and the advantages of time-frequency coherence, an average estimation operator is introduced, a time-frequency coherence method is improved, single-channel acquisition signals are analyzed, and information characteristics are extracted;
introducing an average estimation operator, wherein the expression is as follows:
where x (N) is a random discrete time sequence and N is its length, and when the length of x (N) is long enough, the average estimateCan accurately approach the real mean mu x ;
The random discrete time sequence x (n) is subjected to a windowed fourier transform (FFT) expressed as:
FFT xi (k)=FFT[x(n)·h(i-n)],k=1,2,3,...,N (2)
where h (i-n) is a sliding window, i is the position currently sliding to the discrete time sequence x (n), FFT xi (k) Fourier transform windowed for x (n);
introducing an average estimation operator, and calculating an average value of the FFT of the current segment:
wherein n is the current positionIs a length of (2); FFT of the current position xi (k) Is +.>Then an FFT can be derived xi (k) The method comprises the steps of carrying out a first treatment on the surface of the I.e. an improved short-time fourier transform is obtained, as represented by the following formula (4):
the short-time fourier time-frequency coherence that introduces the average estimation operator can be redefined as:
in the formula, STFT x (i, k) andare all conventional short-time Fourier transforms, ">STFT for introducing short-time Fourier transform of average estimation operator xx (i, k) is the time-frequency cross-power spectrum of the single channel signal; by introducing an average estimation operator, a TFCA method can be used for realizing single-channel signal analysis.
As a further improvement of the technical scheme, the prediction learning specifically comprises:
collecting fault signals at the signal collecting position of the reduction gearbox, wherein each fault signal collects X samples, so that the normal state, the inner ring fault and the outer ring fault form a group of samples by 3X signals;
performing feature processing by adopting an improved TFCA method to obtain a time-frequency diagram, randomly selecting 70% as a training set and 30% as a test set, and then transmitting the training set into a neural network model for prediction learning;
the neural network structure is two layers of bidirectional LSTM layers and two layers of dense layers and is used for regularization and dropout processing;
and inputting the time-frequency diagram data set obtained after time-frequency coherence into a bidirectional LSTM layer network to obtain and observe a loss value curve and an accuracy curve of model training, so that fault identification can be well realized.
As a further improvement of the technical scheme, the health diagnosis comprises two parts of fault alarming and health degree calculating.
As a further improvement of the technical scheme, the fault alarm specifically comprises:
comparing noise history data of the reduction gearbox during normal operation with vibration noise data acquired in real time, when signals acquired by the system accord with fault characteristics, giving out fault alarm to the system, giving out alarm information, displaying fault content, time, specific positions, causing results and solutions through a state monitoring platform, displaying corresponding help documents or operation demonstration, and facilitating timely processing by maintenance personnel; in addition, the monitoring system takes the reserved historical data as fault information, so that maintenance personnel can conveniently search and determine fault reasons later, and data is provided for optimizing equipment in the future.
As a further improvement of the technical scheme, in the health degree calculation, the system calculates the health degree of the equipment based on the occurrence frequency of equipment faults; the degree of health of the times of faults is defined as the proportion of normal operation of equipment in a period of time; the following formula is a method for calculating the health of a device according to the occurrence frequency of faults:
this equation may better reflect the health of the device because if the device frequently fails, its health will be degraded.
The second object of the present invention is to provide a monitoring system operation platform device, which includes a processor, a memory, and a computer program stored in the memory and running on the processor, wherein the processor is used for implementing the operation steps of the intelligent health monitoring system of the gantry crane reduction gearbox based on the fault feature model research when executing the computer program.
The third object of the present invention is to provide a computer readable storage medium, where a computer program is stored, and when the computer program is executed by a processor, the operating steps of the door machine reduction gearbox intelligent health monitoring system based on fault feature model research are implemented.
Compared with the prior art, the invention has the beneficial effects that:
1. in the intelligent health monitoring system of the portal crane reduction gearbox based on fault feature model research, an average estimation operator is introduced, and the traditional TFCA method is improved to solve the defect that a single channel acquisition signal cannot be analyzed, so that the analysis of the single channel acquisition signal of a sensor is realized;
2. in the intelligent health monitoring system of the portal crane reduction gearbox based on fault feature model research, 5G edge calculation is used for processing the data of each sensor, so that the problem of difficult sensor data fusion is solved; multiple and difficult faults such as bearing faults, unbalance, misalignment, mechanical looseness and the like can be timely and accurately judged;
3. in the intelligent health monitoring system of the portal crane reduction gearbox based on the fault feature model research, the state stability index of the equipment is comprehensively evaluated based on the artificial neural network, and the structural change feedback of the equipment is more accurate;
4. in the intelligent health monitoring system of the portal crane reduction gearbox based on fault feature model research, the problems of insufficient feature extraction and low prediction accuracy of the existing fault diagnosis method are solved through a bidirectional LSTM neural network technology; the real-time data is optimally learned through a bidirectional LSTM model which introduces an attention mechanism, and the system early warning is more intelligent; the system can continuously learn data and provide personalized fault prediction information for different port machine equipment.
Drawings
FIG. 1 is a diagram of an exemplary system architecture in accordance with the present invention;
FIG. 2 is a diagram of exemplary system development steps in the present invention;
FIG. 3 is a diagram of an exemplary hoisting mechanism reduction gearbox monitoring point in the present invention;
FIG. 4 is a diagram of an exemplary luffing and rotating mechanism reduction gearbox monitoring point in accordance with the present invention;
FIG. 5 is a diagram of an exemplary luffing and rotating mechanism reduction gearbox monitoring point in accordance with the present invention;
FIG. 6 is an exemplary outer-circle noise time-domain plot (left) and frequency-domain plot (right) of the present invention;
FIG. 7 is an exemplary inner circle noise time domain plot (left) and frequency domain plot (right) of the present invention;
FIG. 8 is a graph comparing exemplary outer ring failure time-frequency coherence (left) with inner ring failure time-frequency coherence (right) in the present invention;
FIG. 9 is a graph of exemplary bidirectional LSTM network accuracy in accordance with the present invention;
fig. 10 is a block diagram of an exemplary electronic computer platform device according to the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Example 1
As shown in fig. 1-9, the embodiment provides a gantry crane reduction gearbox intelligent health monitoring system based on fault feature model research, which is used for establishing a port machine equipment intelligent health monitoring system based on fault feature model research and can be used for solving the problem of high fault detection and maintenance difficulty of the existing equipment; establishing an expert diagnosis model based on probability statistics and data fusion by developing intelligent curve model research based on noise sensing equipment, realizing ageing diagnosis and fault diagnosis of the reduction gearbox, grasping the running state of the equipment in real time, actively preventing equipment faults and realizing predictive maintenance of the reduction gearbox; as shown in fig. 1, the system comprises a state sensing module 100, a health monitoring module 200 and a state monitoring platform 300 which are sequentially connected in a communication manner; wherein:
the state sensing module 100 is configured to determine the type, position and channel number of the sensor by performing situation sensing on the key part of the door machine reduction gearbox, and then process the device data through signal conditioning, amplifying, filtering and the like, and send the device data to the health monitoring module 200;
the health monitoring module 200 is used for analyzing and predicting the received device data, providing data support for functions/subsystems such as device security assessment, operation and maintenance management and the like, and sending the analysis and prediction results to the state monitoring platform 300 for display;
the state monitoring platform 300 is used for displaying data such as running conditions, fault alarms, historical records and the like of the reduction gearbox in real time.
The system is developed by taking a portal crane reduction gearbox as a research object, and the development steps of the system are shown in figure 2.
In this embodiment, the state sensing module 100 is mainly connected to the system through monitoring means such as temperature, wen Zhen, noise, infrared thermal imaging, current and voltage, etc., to monitor the action state of the door motor reduction gearbox; the selection of the measuring points needs to comprehensively consider factors such as key parts, equipment running states, benefit cost ratio and the like so as to ensure the accuracy and stability of the monitoring system and improve the safety and service life of the monitoring system; wherein, the selection of the measuring point needs to consider the following aspects:
(1) The key part of the reduction gearbox is selected from the measuring points;
(2) The measuring point should select a parameter capable of reflecting the running state of the equipment;
(3) The number and variety of stations should balance benefit and cost while satisfying function.
Further, the types of faults that easily occur in the input stage, the intermediate stage and the output stage of the reduction gearbox are faults of a rotor, a shaft system, a bearing, collision and abrasion and looseness, so that the state sensing module 100 mainly monitors parameters such as temperature, vibration and noise, and the measuring points are selected as shown in fig. 3-5; the state sensing module 100 at least includes a temperature sensor, a Wen Zhen sensor, and a noise sensor;
the state sensing module 100 processes data acquired by each sensor by using 5G edge calculation, so as to solve the problem of difficult fusion of sensor data;
the running state of the equipment is estimated by installing a Wen Zhenchuan sensor at the bottom of the reduction gearbox and monitoring the temperature and vibration condition of the equipment in real time, and data is transmitted to a monitoring system for analysis and processing, so that the faults and abnormal conditions of the equipment are found in time; the noise sensor is arranged to monitor the noise generated during the running of the equipment in real time and store the noise data in different time periods, and whether the running state of the equipment is abnormal or not can be judged by comparing the noise data in different time periods.
In this embodiment, the process of analyzing and predicting the received device data by the health monitoring module 200 mainly includes three parts, namely feature extraction, prediction learning and health diagnosis.
Firstly, in the feature extraction, taking vibration parameters as an example, the fault type of a reduction gearbox is mainly damage of an inner ring/an outer ring of a bearing and the like; these fault characteristics are all reflected in the vibration signal; when the door machine actually operates, the noise signal generated by the reduction gearbox is a typical non-stable vibration signal;
in general, when the traditional time-frequency coherence method (TFCA) is applied to bearing fault monitoring, two groups of different signals need to be acquired for time-frequency coherence analysis, and only one sensor (namely a single-channel acquisition signal) can be installed at the optimal acquisition position during monitoring, so that the single-channel acquisition signal cannot be analyzed by utilizing the traditional TFCA method, and the limitation is larger; based on the characteristics of acoustic emission signals and the advantages of time-frequency coherence, an average estimation operator is introduced, a time-frequency coherence method is improved, single-channel acquisition signals are analyzed, and information characteristics are extracted;
introducing an average estimation operator, wherein the expression is as follows:
where x (N) is a random discrete time sequence and N is its length, and when the length of x (N) is long enough, the average estimateCan accurately approach the real mean mu x ;
The random discrete time sequence x (n) is subjected to a windowed fourier transform (FFT) expressed as:
FFT xi (k)=FFT[x(n)·h(i-n)],k=1,2,3,...,N (2)
where h (i-n) is a sliding window, i is the position currently sliding to the discrete time sequence x (n), FFT xi (k) Fourier transform windowed for x (n);
introducing an average estimation operator, and calculating an average value of the FFT of the current segment:
wherein n is the current positionIs a length of (2); FFT of the current position xi (k) Is +.>Then an FFT can be derived xi (k) The method comprises the steps of carrying out a first treatment on the surface of the I.e. an improved short-time fourier transform is obtained, as represented by the following formula (4):
the short-time fourier time-frequency coherence that introduces the average estimation operator can be redefined as:
in the formula, STFT x (i, k) andare all conventional short-time Fourier transforms, ">STFT for introducing short-time Fourier transform of average estimation operator xx (i, k) is the time-frequency cross-power spectrum of the single channel signal; by introducing an average estimation operator, a TFCA method can be used for realizing single-channel signal analysis.
The system is used for collecting sound signals of the reduction gearbox through the noise sensor, the sampling frequency is 3MHz, and the sampling time is 80ms. Typically, the theoretical characteristic frequency of the outer ring faults is 52Hz, and the fault frequency of the inner ring is 77.5Hz. The outer-circle noise signal time-domain diagram and the frequency-domain diagram are shown in fig. 6, and the inner-circle noise signal time-domain diagram and the frequency-domain diagram are shown in fig. 7. And performing time-frequency coherent analysis on the single-channel noise signals acquired by the sensor in the MATLAB based on the improved TFCA method, extracting fault characteristics, and obtaining fault signal time-frequency distribution as shown in figure 8.
As can be seen from fig. 8, the outer ring fault extraction cycle time interval is T 0 19.17ms, i.e. the outer ring failure frequency is f 0 = 52.16Hz; the time interval of the fault period of the inner ring is T i =12.81 ms, failure frequency f 0 = 78.06Hz. Consistent with the theoretical failure frequency. The improved TFCA method is utilized to extract the characteristics of the original signal, and the periodic impact component in the signal can be extracted well.
Secondly, prediction learning is specifically:
collecting fault signals at the signal collecting position of the reduction gearbox, wherein each fault signal collects 400 samples, so that 1200 signals are combined into a group of samples in a normal state, an inner ring fault and an outer ring fault;
performing feature processing by adopting an improved TFCA method to obtain a time-frequency diagram, randomly selecting 70% as a training set and 30% as a test set, and then transmitting the training set into a neural network model for prediction learning;
the neural network model adopts a bidirectional LSTM model which introduces an attention mechanism, and the real-time data is optimally learned by the bidirectional LSTM neural network technology, so that the problems of insufficient feature extraction and low prediction accuracy of the existing fault diagnosis method can be solved, and personalized fault prediction information can be provided for different port machine equipment by continuously carrying out data learning;
the neural network structure is two layers of bidirectional LSTM layers and two layers of dense layers and is used for regularization and dropout processing;
specifically, the network configuration parameters are shown in table 1:
table 1 bidirectional LSTM network structure parameters
And inputting the time-frequency diagram data set obtained after the time-frequency coherence into a bidirectional LSTM layer network to obtain and observe a loss value curve and an accuracy curve of model training, as shown in figure 9. From the graph, when iteration is performed for 50 times, the loss value curve and the precision curve are gradually stable, the accuracy reaches 0.95, and fault identification can be realized very much.
Third, the health diagnosis comprises two parts of fault alarm and health degree calculation.
Specifically, the fault alarm specifically includes:
comparing noise history data of the reduction gearbox during normal operation with vibration noise data acquired in real time, when signals acquired by the system accord with fault characteristics, giving out fault alarm to the system, giving out alarm information, displaying fault content, time, specific positions, causing results and solutions through the state monitoring platform 300, displaying corresponding help documents or operation demonstration, and facilitating timely processing by maintenance personnel; in addition, the monitoring system takes the reserved historical data as fault information, so that maintenance personnel can conveniently search and determine fault reasons later, and data is provided for optimizing equipment in the future.
Further, in the health calculation, the system calculates the health of the device based on the frequency at which the device failure occurs; the degree of health of the times of faults is defined as the proportion of normal operation of equipment in a period of time; the following formula is a method for calculating the health of a device according to the occurrence frequency of faults:
this equation may better reflect the health of the device because if the device frequently fails, its health will be degraded.
As shown in fig. 10, the present embodiment also provides a monitoring system operating platform apparatus, which includes a processor, a memory, and a computer program stored in the memory and running on the processor.
The processor comprises one or more than one processing core, the processor is connected with the memory through a bus, the memory is used for storing program instructions, and the running steps of the intelligent health monitoring system of the gantry crane reduction gearbox based on fault feature model research are realized when the processor executes the program instructions in the memory.
Alternatively, the memory may be implemented by any type or combination of volatile or nonvolatile memory devices such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disk.
In addition, the invention also provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program realizes the operation steps of the intelligent health monitoring system of the door machine reduction gearbox based on the fault characteristic model research when being executed by a processor.
Optionally, the present invention further provides a computer program product containing instructions that, when run on a computer, cause the computer to perform the steps of operating the gantry crane reduction gearbox intelligent health monitoring system according to the above aspects based on the fault feature model study.
It will be appreciated by those of ordinary skill in the art that the processes for implementing all or part of the steps of the above embodiments may be implemented by hardware, or may be implemented by a program for instructing the relevant hardware, and the program may be stored in a computer readable storage medium, where the above storage medium may be a read-only memory, a magnetic disk or optical disk, etc.
The foregoing has shown and described the basic principles, principal features and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the above-described embodiments, and that the above-described embodiments and descriptions are only preferred embodiments of the present invention, and are not intended to limit the invention, and that various changes and modifications may be made therein without departing from the spirit and scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (9)
1. A door machine reducing gear box intelligent health monitoring system based on a fault characteristic model is characterized in that: the health monitoring system comprises a state sensing module (100), a health monitoring module (200) and a state monitoring platform (300) which are sequentially connected in a communication way; wherein:
the state sensing module (100) is used for firstly sensing the situation of key parts of the portal crane reduction gearbox, determining the types, positions and channel quantity of the sensors, and then transmitting the device data to the health monitoring module (200) after processing of signal conditioning, amplification and filtering;
the health monitoring module (200) is used for analyzing and predicting the received equipment data, providing data support for the functions of equipment safety evaluation and operation and maintenance management, and sending the analysis and prediction results to the state monitoring platform (300) for display;
the state monitoring platform (300) is used for displaying the running state of the reduction gearbox, fault alarming and historical record data in real time.
2. The intelligent health monitoring system for a portal crane gearbox based on a fault signature model as claimed in claim 1, wherein: the state sensing module (100) is connected into the system mainly through monitoring means of temperature, wen Zhen, noise, infrared thermal imaging and current and voltage, and monitors the action state of a door machine reduction gearbox; the selection of the measuring points needs to comprehensively consider factors of key parts, equipment running states and benefit cost ratio so as to ensure the accuracy and stability of the monitoring system and improve the safety and service life of the monitoring system; wherein, the selection of the measuring point needs to consider the following aspects:
(1) The key part of the reduction gearbox is selected from the measuring points;
(2) The measuring point should select a parameter capable of reflecting the running state of the equipment;
(3) The number and variety of stations should balance benefit and cost while satisfying function.
3. The intelligent health monitoring system for door machine reduction gearbox based on fault characteristic model as claimed in claim 2, wherein: the types of faults which easily occur in the input stage, the intermediate stage and the output stage of the reduction gearbox are faults of a rotor, a shafting, a bearing, collision and abrasion and looseness, so that the state sensing module (100) mainly monitors parameters of temperature, vibration and noise; the state sensing module (100) at least comprises a temperature sensor, a Wen Zhen sensor and a noise sensor;
the running state of the equipment is estimated by installing a Wen Zhenchuan sensor at the bottom of the reduction gearbox and monitoring the temperature and vibration condition of the equipment in real time, and data is transmitted to a monitoring system for analysis and processing, so that the faults and abnormal conditions of the equipment are found in time; the noise sensor is arranged to monitor the noise generated during the running of the equipment in real time and store the noise data in different time periods, and whether the running state of the equipment is abnormal or not can be judged by comparing the noise data in different time periods.
4. The intelligent health monitoring system for a portal crane gearbox based on a fault signature model as claimed in claim 1, wherein: the process of analyzing and predicting the received device data by the health monitoring module (200) mainly comprises three parts of feature extraction, prediction learning and health diagnosis.
5. The intelligent health monitoring system for door machine reduction gearbox based on fault signature model as claimed in claim 4, wherein: in the feature extraction, taking vibration parameters as an example, the fault type of the reduction gearbox is mainly damage of an inner ring/an outer ring of the bearing; these fault characteristics are all reflected in the vibration signal; when the door machine actually operates, the noise signal generated by the reduction gearbox is a typical non-stable vibration signal;
based on the problem that TFCA of the traditional time-frequency coherence method cannot analyze single-channel acquisition signals and has larger limitation, according to the characteristics of acoustic emission signals and the advantages of time-frequency coherence, an average estimation operator is introduced, the time-frequency coherence method is improved, analysis is carried out on the single-channel acquisition signals, and information characteristics are extracted;
introducing an average estimation operator, wherein the expression is as follows:
where x (N) is a random discrete time sequence and N is its length, and when the length of x (N) is long enough, the average estimateCan accurately approach the real mean mu x ;
Carrying out windowed Fourier transform FFT on the random discrete time sequence x (n), wherein the expression is as follows:
FFT xi (k)=FFT[x(n)·h(i-n)],k=1,2,3,...,N (2)
where h (i-n) is a sliding window, i is the position currently sliding to the discrete time sequence x (n), FFT xi (k) Fourier transform windowed for x (n);
introducing an average estimation operator, and calculating an average value of the FFT of the current segment:
wherein n is the current positionIs a length of (2); FFT of the current position xi (k) Is +.>Then an FFT can be derived xi (k) The method comprises the steps of carrying out a first treatment on the surface of the I.e. an improved short-time fourier transform is obtained, as represented by the following formula (4):
the short-time fourier time-frequency coherence that introduces the average estimation operator can be redefined as:
in the formula, STFT x (i, k) andare all conventional short-time Fourier transforms, ">STFT for introducing short-time Fourier transform of average estimation operator xx (i, k) is the time-frequency cross-power spectrum of the single channel signal; by introducing an average estimation operator, a TFCA method can be used for realizing single-channel signal analysis.
6. The intelligent health monitoring system of a portal crane reduction gearbox based on a fault feature model according to claim 4, wherein the predictive learning is specifically:
collecting fault signals at the signal collecting position of the reduction gearbox, wherein each fault signal collects X samples, so that the normal state, the inner ring fault and the outer ring fault form a group of samples by 3X signals;
performing feature processing by adopting an improved TFCA method to obtain a time-frequency diagram, randomly selecting 70% as a training set and 30% as a test set, and then transmitting the training set into a neural network model for prediction learning;
the neural network structure is two layers of bidirectional LSTM layers and two layers of dense layers and is used for regularization and dropout processing;
and inputting the time-frequency diagram data set obtained after time-frequency coherence into a bidirectional LSTM layer network to obtain and observe a loss value curve and an accuracy curve of model training, so that fault identification can be realized.
7. The intelligent health monitoring system for door machine reduction gearbox based on fault signature model as claimed in claim 4, wherein: the health diagnosis comprises two parts of fault alarming and health degree calculating.
8. The intelligent health monitoring system of a portal crane gearbox based on a fault feature model as claimed in claim 7, wherein the fault alarm is specifically:
comparing noise history data of the reduction gearbox during normal operation with vibration noise data acquired in real time, when signals acquired by the system accord with fault characteristics, giving out fault alarm to the system, giving out alarm information, displaying fault content, time, specific positions, causing results and solutions through a state monitoring platform (300), displaying corresponding help documents or operation demonstration, and facilitating timely processing by maintenance personnel; in addition, the monitoring system takes the reserved historical data as fault information, so that maintenance personnel can conveniently search and determine fault reasons later, and data is provided for optimizing equipment in the future.
9. The intelligent health monitoring system for door machine reduction gearbox based on fault signature model as claimed in claim 7, wherein: in the health degree calculation, the system calculates the health degree of the equipment based on the occurrence frequency of equipment faults; the degree of health of the times of faults is defined as the proportion of normal operation of equipment in a period of time; the following formula is a method for calculating the health of a device according to the occurrence frequency of faults:
this equation may better reflect the health of the device.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311407596.2A CN117451347A (en) | 2023-10-27 | 2023-10-27 | Portal crane reduction gearbox intelligent health monitoring system based on fault characteristic model |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311407596.2A CN117451347A (en) | 2023-10-27 | 2023-10-27 | Portal crane reduction gearbox intelligent health monitoring system based on fault characteristic model |
Publications (1)
Publication Number | Publication Date |
---|---|
CN117451347A true CN117451347A (en) | 2024-01-26 |
Family
ID=89595984
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202311407596.2A Pending CN117451347A (en) | 2023-10-27 | 2023-10-27 | Portal crane reduction gearbox intelligent health monitoring system based on fault characteristic model |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117451347A (en) |
-
2023
- 2023-10-27 CN CN202311407596.2A patent/CN117451347A/en active Pending
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Camci et al. | Feature evaluation for effective bearing prognostics | |
CN109186813A (en) | A kind of temperature sensor self-checking unit and method | |
Yan et al. | Fisher’s discriminant ratio based health indicator for locating informative frequency bands for machine performance degradation assessment | |
Dou et al. | A rule-based intelligent method for fault diagnosis of rotating machinery | |
CN112766342A (en) | Abnormity detection method for electrical equipment | |
CN107291991A (en) | A kind of Wind turbines early defect method for early warning based on dynamic network mark | |
CN112629905A (en) | Equipment anomaly detection method and system based on deep learning and computer medium | |
CN111457958A (en) | Port machine equipment situation monitoring method and device, computer equipment and storage medium | |
CN117171657A (en) | Wind power generation equipment fault diagnosis method and device, electronic equipment and storage medium | |
Zegrari et al. | Data-driven Diagnostics for Electric Traction Systems: A Study of Induction Motor | |
Hu et al. | Mutual information-based feature disentangled network for anomaly detection under variable working conditions | |
CN117192369A (en) | Traction motor monitoring and diagnosing method based on digital twin technology | |
Zhang et al. | Fault diagnosis method of belt conveyor idler based on sound signal | |
CN117451347A (en) | Portal crane reduction gearbox intelligent health monitoring system based on fault characteristic model | |
CN113516023B (en) | Method and system for diagnosing equipment vibration abnormality | |
CN116226719A (en) | Bearing fault diagnosis method based on multidimensional steady-state vibration characteristics and related components | |
CN114492636B (en) | Transformer winding state signal acquisition system | |
KR20230102431A (en) | Oil gas plant equipment failure prediction and diagnosis system based on artificial intelligence | |
Waters et al. | Vibration Anomaly Detection using Deep Autoencoders for Smart Factory | |
Hoffmann et al. | Embedding anomaly detection autoencoders for wind turbines | |
Zheng | Predicting remaining useful life using continuous wavelet transform integrated discrete teager energy operator with degradation model | |
wahhab Lourari et al. | An ANFIS-based Framework for the Prediction of Bearing’s Remaining Useful Life | |
Randrianandraina et al. | Wind turbine generator bearing anomaly detection and explanation using rrcf approach | |
CN117466153B (en) | Fault detection method, device, computer equipment and readable storage medium | |
Latuny | A sensitivity comparison of Neuro-fuzzy feature extraction methods from bearing failure signals |
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 |