CN112816195A - Reciprocating mechanical equipment fault diagnosis method and device - Google Patents
Reciprocating mechanical equipment fault diagnosis method and device Download PDFInfo
- Publication number
- CN112816195A CN112816195A CN202110003526.5A CN202110003526A CN112816195A CN 112816195 A CN112816195 A CN 112816195A CN 202110003526 A CN202110003526 A CN 202110003526A CN 112816195 A CN112816195 A CN 112816195A
- Authority
- CN
- China
- Prior art keywords
- fault
- data
- equipment
- vibration
- sample
- 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
- 238000000034 method Methods 0.000 title claims abstract description 59
- 238000003745 diagnosis Methods 0.000 title claims abstract description 40
- 238000012549 training Methods 0.000 claims abstract description 36
- 230000033001 locomotion Effects 0.000 claims abstract description 29
- 238000012360 testing method Methods 0.000 claims description 32
- 230000002159 abnormal effect Effects 0.000 claims description 21
- 238000003860 storage Methods 0.000 claims description 15
- 238000012795 verification Methods 0.000 claims description 12
- 238000002405 diagnostic procedure Methods 0.000 claims description 5
- 239000000523 sample Substances 0.000 description 69
- 230000001133 acceleration Effects 0.000 description 23
- 238000004891 communication Methods 0.000 description 18
- 230000000875 corresponding effect Effects 0.000 description 13
- 238000012544 monitoring process Methods 0.000 description 11
- 239000013598 vector Substances 0.000 description 11
- 238000010586 diagram Methods 0.000 description 9
- 238000001514 detection method Methods 0.000 description 7
- 239000007789 gas Substances 0.000 description 7
- 208000037408 Device failure Diseases 0.000 description 5
- 230000008901 benefit Effects 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 206010063385 Intellectualisation Diseases 0.000 description 4
- 230000008859 change Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000006835 compression Effects 0.000 description 3
- 238000007906 compression Methods 0.000 description 3
- 238000010606 normalization Methods 0.000 description 3
- 230000000737 periodic effect Effects 0.000 description 3
- 230000002093 peripheral effect Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000007635 classification algorithm Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000009776 industrial production Methods 0.000 description 2
- 239000007788 liquid Substances 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 230000007723 transport mechanism Effects 0.000 description 2
- VGGSQFUCUMXWEO-UHFFFAOYSA-N Ethene Chemical compound C=C VGGSQFUCUMXWEO-UHFFFAOYSA-N 0.000 description 1
- 239000005977 Ethylene Substances 0.000 description 1
- 238000005299 abrasion Methods 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000005336 cracking Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 239000012530 fluid Substances 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000001939 inductive effect Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000005272 metallurgy Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000003345 natural gas Substances 0.000 description 1
- 239000003208 petroleum Substances 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Images
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
-
- 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
- G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
- G01H17/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves, not provided for in the preceding groups
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention discloses a fault diagnosis method for reciprocating mechanical equipment, which is executed in computing equipment and comprises the following steps: obtaining vibration data of equipment to be tested in a motion cycle, wherein the equipment to be tested is reciprocating mechanical equipment; inputting the vibration data into a preset self-coding model so that the self-coding model outputs reconstruction data of the vibration data, wherein the self-coding model is obtained by training vibration data of equipment with the same type as the equipment to be tested in normal operation as a training sample; and calculating a reconstruction error between the vibration data and the reconstruction data, and judging that the equipment to be tested breaks down when the reconstruction error is greater than a preset error threshold value. The invention also discloses corresponding computing equipment.
Description
Technical Field
The invention relates to the technical field of health state monitoring of industrial equipment, in particular to a fault diagnosis method and device of reciprocating mechanical equipment.
Background
Reciprocating mechanical equipment is a common mechanical equipment, and is a mechanical equipment that a crank is driven by an electric motor, the crank drives a connecting rod, the rotary motion of the electric motor is changed into reciprocating motion, and finally a piston (a diaphragm) is pushed by the motion of the connecting rod to do reciprocating motion. The reciprocating mechanical equipment is used as a general machine, has the advantages of wide range of air inlet and exhaust pressure, no limitation on flow, high energy efficiency, wide application range and the like, and is widely applied to the industrial fields of petroleum, chemical industry, metallurgy, power and the like. The reciprocating mechanical equipment has the disadvantages of complex internal work, high pressure, more parts which are easy to fail and high failure rate in actual production. In industrial production, most reciprocating mechanical equipment is a reciprocating compressor and is mainly responsible for compression, transportation and other work of flammable and explosive gases such as ethylene, natural gas and the like and other dangerous media, once a fault occurs, if the fault cannot be found and eliminated in time, major safety accidents are easily caused, and casualties and production loss are caused. Based on this, fault diagnosis and predictive maintenance for reciprocating mechanical equipment become a non-negligible part of industrial production. Therefore, the system for on-line monitoring and fault diagnosis of reciprocating mechanical equipment is researched, the monitoring system is used for continuously and effectively monitoring the running state of the unit, and the system has great significance in fault alarming and diagnosis.
At present, in the on-line monitoring of reciprocating machinery equipment, a sensor is mounted to acquire information of the equipment, such as temperature, sound, vibration, etc., and the equipment is subjected to fault diagnosis by analyzing the state information. The method mainly comprises a working condition thermodynamic parameter method, a vibration detection method, a gas leakage detection method and an oil analysis method. The vibration detection method is the most important fault diagnosis method for the reciprocating mechanical equipment at the present stage due to the advantages of low cost, simple flow, multiple faults detection and the like.
The existing vibration detection method adopts a vibration sensor to collect vibration signals of reciprocating mechanical equipment, extracts corresponding fault characteristic parameters such as vibration acceleration peak value, acceleration effective value (RMS), acceleration waveform kurtosis, acceleration waveform skewness and the like from the vibration signals, sets an alarm threshold value aiming at the fault characteristic parameters manually, and considers that the equipment has faults when the characteristic parameter values exceed the alarm threshold value.
The time domain waveform of the vibration of the reciprocating mechanical device is highly correlated with the operating characteristics of the reciprocating mechanical device. The operation process of compressing gas can be divided into four processes of expansion, suction, compression and exhaust. The piston (diaphragm) is in continuous reciprocating motion in the cylinder, so that the cylinder can suck and discharge gas in a reciprocating cycle. The opening and closing of the inlet and outlet valves and the movement of the piston (diaphragm) itself cause an impact during each reciprocation of the piston (diaphragm). The impacts are represented as periodic impact waveforms with similar magnitudes in a vibration time domain waveform, and the periodic impact waveforms are important basis for vibration monitoring of reciprocating mechanical equipment. However, in practical situations, due to the influence of various factors such as temperature, gas medium, and working environment, and due to a large amount of noise in the vibration data collected by the sensor, the amplitude of the impact on the waveform fluctuates within a certain range, the time of the impact is advanced or delayed to a certain extent, and the whole vibration waveform cannot clearly represent all periodic impacts. For example, fig. 1 shows the vibration time-domain waveforms in two periods, and the impact amplitude and the impact time of the two periods are different.
When a fault occurs, part of the impact is also represented by the change of the impact amplitude and the advance and the delay of the impact, so that the fault and the normal operation limit cannot be clearly distinguished by using the threshold value of the fault characteristic parameter. Meanwhile, the change of the working condition can make the high point of the fault characteristic parameter change continuously, and the setting of the threshold value is difficult. Due to the reasons, fault diagnosis for the reciprocating mechanical equipment still depends on manual experience at the present stage, namely, the alarm threshold value of the fault characteristic parameter is changed through expert experience according to different working conditions, so that misjudgment and missed judgment of faults exist in vibration detection for the reciprocating mechanical equipment.
Disclosure of Invention
To this end, the present invention provides a reciprocating machine fault diagnosis method and apparatus in an attempt to solve or at least alleviate the above-identified problems.
According to a first aspect of the present invention, there is provided a reciprocating machine equipment fault diagnosis method, executed in a computing device, comprising: obtaining vibration data of equipment to be tested in a motion cycle, wherein the equipment to be tested is reciprocating mechanical equipment; inputting the vibration data into a preset self-coding model so that the self-coding model can output reconstruction data of the vibration data, wherein the self-coding model is obtained by training vibration data of equipment with the same type as the equipment to be tested in normal operation as a training sample; and calculating a reconstruction error between the vibration data and the reconstruction data, and judging that the equipment to be tested breaks down when the reconstruction error is larger than a preset error threshold value.
Optionally, in the fault diagnosis method for the reciprocating mechanical device according to the present invention, a vibration sensor and a key phase sensor are disposed on the device under test, the key phase sensor is disposed on a rotating component of the device under test, and the step of acquiring vibration data of the device under test in one motion cycle includes: acquiring a vibration signal acquired by a vibration sensor and a pulse signal output by a key phase sensor; determining the starting time and the ending time of one rotation of the rotating part according to the pulse signal; and intercepting the vibration signal according to the starting time and the ending time, and taking the intercepted vibration signal as vibration data of the equipment to be tested in a motion period.
Optionally, in the method for diagnosing the fault of the reciprocating mechanical equipment according to the present invention, the training samples are normalized before participating in the training of the self-coding model, so that the values in the training samples are within a preset interval.
Alternatively, in the fault diagnosis method of the reciprocating mechanical equipment according to the invention, the error threshold is determined according to the following steps: inputting a plurality of verification samples into the self-coding model so that the self-coding model outputs reconstruction data of each verification sample, wherein the verification samples comprise a plurality of normal samples and a plurality of abnormal samples, the normal samples are vibration data of equipment of the same type as the equipment to be tested during normal operation, and the abnormal samples are vibration data of the equipment of the same type as the equipment to be tested during failure; and respectively calculating the reconstruction error of each verification sample and the reconstruction data thereof, and determining an error threshold according to the reconstruction error of the normal sample and the reconstruction error of the abnormal sample, wherein the error threshold is greater than or equal to the reconstruction error of the normal sample and less than or equal to the reconstruction error of the abnormal sample.
Optionally, in the fault diagnosis method for the reciprocating mechanical equipment according to the present invention, the method further includes: when the equipment to be tested is judged to have a fault, respectively calculating the distance between fault data and fault samples, and taking a preset number of fault samples with the minimum distance as neighbor samples, wherein the fault data are vibration data according to which the equipment to be tested is judged to have a fault, and the fault samples comprise vibration data of equipment of the same type as the equipment to be tested in different fault types; and determining the fault type of the equipment to be tested according to the fault type of each neighbor sample and the distance from the neighbor sample to the fault data.
Alternatively, in the fault diagnosis method of a reciprocating mechanical device according to the present invention, the step of calculating the distances between the fault data and the fault samples, respectively, includes: determining a fault component according to the position of a vibration sensor for collecting fault data on the equipment to be tested; acquiring a fault sample corresponding to the fault component; and respectively calculating the distance between the fault data and each obtained fault sample.
Optionally, in the method for diagnosing a fault of a reciprocating mechanical device according to the present invention, the step of determining the fault type of the device under test according to the fault type to which each neighbor sample belongs and the distance to the fault data includes: determining the weight of each neighbor sample according to the distance from each neighbor sample to the fault data, wherein the smaller the distance from each neighbor sample to the fault data is, the larger the weight is; calculating the confidence coefficient of the fault data belonging to each fault type, wherein the confidence coefficient of the fault type is the weighted sum of the distances of all the neighbor samples belonging to the fault type; and taking the fault type with the maximum confidence coefficient as the fault type of the equipment to be tested.
Alternatively, in the fault diagnosis method of the reciprocating mechanical equipment according to the present invention, the weights of the neighbor samples are determined according to the following formula:
wherein, wiIs the weight of the neighbor sample i, diThe distance from the neighbor sample i to the fault data is calculated, and c is a preset normal number;
alternatively, the weights of the neighbor samples are determined according to the following formula:
wherein, wiIs the weight of the neighbor sample i, diThe distance from the neighbor sample i to the fault data is shown, e is a natural constant, and a and b are preset normal constants.
According to a second aspect of the invention, there is provided a computing device comprising: at least one processor and a memory storing program instructions; the program instructions, when read and executed by the processor, cause the computing device to perform the reciprocating machine fault diagnostic method described above.
According to a third aspect of the present invention, there is provided a readable storage medium storing program instructions which, when read and executed by a computing device, cause the computing device to perform the above-described reciprocating machine equipment fault diagnosis method.
The technical scheme of the invention adopts a self-coding model to identify the fault of the reciprocating mechanical equipment. The self-coding model comprises an encoder and a decoder, wherein the encoder is a nonlinear compression method and can learn the characteristics in the vibration data, reduce the dimension of the vibration data and compress the vibration data into low-dimension codes; the decoder reconstructs the code after dimension reduction, and restores the low-dimension code to be close to the original input expression as possible to obtain reconstructed data.
The technical scheme of the invention trains the automatic encoder network by accumulated vibration data of a large number of reciprocating mechanical equipment during normal operation to generate a self-encoding model. The method comprises the steps of compressing and reconstructing current vibration data of equipment to be detected by using a trained model to obtain reconstructed data, calculating a reconstruction error between the vibration data and the reconstructed data, and identifying equipment faults by comparing the reconstruction error with an error threshold value, so that automation and intellectualization of fault identification of reciprocating mechanical equipment are realized. Meanwhile, the abstract characteristics of the vibration data are extracted from the coding model for fault recognition, an alarm threshold value does not need to be set through manual experience, and fault misjudgment and misjudgment caused by improper setting of the alarm threshold value are avoided.
Further, after the fault data are identified, the fault data are compared with accumulated fault samples of the reciprocating mechanical equipment by adopting a KNN algorithm, the first K fault samples which are most similar to the fault data in the sample set are used as neighbor samples, the fault types of the neighbor samples and the distance from the neighbor samples to the fault data are integrated, the confidence coefficient that the fault data belong to the fault types is predicted, the fault type with the highest confidence coefficient is used as the fault type of the equipment to be detected, and automatic classification based on historical fault data is realized.
The equipment fault diagnosis scheme of the invention adopts the self-coding model to carry out abnormity identification on the vibration data acquired by the reciprocating mechanical equipment, and after the fault data is identified, fault classification is carried out on the fault data by using a classification algorithm such as KNN (K nearest neighbor) and the like, so that automation and intellectualization of fault diagnosis of the reciprocating mechanical equipment are realized.
The foregoing description is only an overview of the technical solutions of the present invention, and the embodiments of the present invention are described below in order to make the technical means of the present invention more clearly understood and to make the above and other objects, features, and advantages of the present invention more clearly understandable.
Drawings
To the accomplishment of the foregoing and related ends, certain illustrative aspects are described herein in connection with the following description and the annexed drawings, which are indicative of various ways in which the principles disclosed herein may be practiced, and all aspects and equivalents thereof are intended to be within the scope of the claimed subject matter. The above and other objects, features and advantages of the present disclosure will become more apparent from the following detailed description read in conjunction with the accompanying drawings. Throughout this disclosure, like reference numerals generally refer to like parts or elements.
FIG. 1 is a diagram illustrating a vibration time domain waveform in two cycles in the prior art;
FIG. 2 shows a schematic diagram of an equipment fault diagnostic system 100 according to one embodiment of the present invention;
FIG. 3 shows a schematic diagram of a computing device 200, according to one embodiment of the invention;
FIG. 4 illustrates a network architecture diagram of a self-coding model according to one embodiment of the invention;
figure 5 illustrates a flow diagram of a reciprocating machine equipment fault diagnostic method 500 according to one embodiment of the present invention.
Detailed Description
Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
Aiming at the problems in the prior art, the invention provides a fault diagnosis method for reciprocating mechanical equipment, so as to realize intelligent and automatic fault diagnosis for the reciprocating mechanical equipment.
FIG. 2 shows a schematic diagram of an equipment fault diagnosis system 100 according to one embodiment of the present invention. As shown in fig. 2, the device failure diagnosis system 100 includes a device under test 110, a vibration sensor 120, a key phase sensor 130, and a computing device 200.
It should be noted that the equipment failure diagnostic system 100 shown in fig. 2 is merely exemplary. In a specific practical situation, the device failure diagnosis system may include different numbers of devices to be tested, vibration sensors, key phase sensors, and computing devices, and the present invention does not limit the numbers of the devices to be tested, the vibration sensors, the key phase sensors, and the computing devices included in the device failure diagnosis system.
The device fault diagnosis system 100 is used to identify a fault state of a device under test. And when the equipment to be tested is identified to be in a fault state, further determining the fault type of the equipment to be tested.
In an embodiment of the present invention, device under test 110 is a reciprocating mechanical device that moves periodically, including but not limited to a reciprocating compressor, a reciprocating vacuum pump, and the like.
The device under test 110 is provided with at least one measuring point (or called monitoring point, monitoring part/component), and each measuring point is provided with a vibration sensor 120 for acquiring a vibration signal (the vibration signal may be, for example, an acceleration signal, a velocity signal, etc.) of the corresponding measuring point. In the embodiment shown in fig. 2, two measuring points are provided on the device under test 110, and each measuring point is provided with a vibration sensor 120.
It should be noted that the positions and the number of the measuring points of the device under test can be set by those skilled in the art according to the actual situation, and the present invention is not limited to this. Generally, a position with stronger rigidity on the device under test 110 is selected as a measuring point, such as a supporting position, a carrying area, etc. of the device. The number of measurement points can be set according to the size and structural characteristics of the device to be measured, for example, and vibration sensors 120 in different numbers are arranged on the motor driving end, the motor non-driving end, the crankcase, the crosshead, and each cylinder. When some equipment is large in size and complex in structure, the vibration condition of the equipment cannot be well monitored by only arranging one measuring point. For example, for a motor component, a measuring point may be respectively arranged at the driving end and the non-driving end of the motor component, and the vibration sensor 120 is installed, because the motor is large and the vibration transmission of the bearings at the two ends is poor, two sensors need to be arranged; the crank cases of the 2-cylinder compressor and the 4-cylinder compressor are smaller, and two ends of the crank cases and the upper surfaces of the tiles can be respectively provided with a sensor; and the like. The sensors at the driving end and the non-driving end of the motor are used for monitoring the friction of bearings at two ends of the motor; sensors on the crosshead are used to monitor crosshead gap impacts; sensors on the crankcase are to monitor crankshaft friction; the sensor on the cylinder body is used for monitoring liquid impact, air valve opening and closing impact and friction; and the like.
In addition, it should be noted that the present invention does not limit the type and model of the vibration sensor 120. For example, the vibration sensor may be of the piezoelectric, piezoresistive, capacitive, inductive, etc. type.
In addition to the vibration sensor 120, a key phase sensor 130 is provided on the rotating member of the device under test 110, and the key phase sensor 130 is used to determine the starting position of one rotation. Specifically, the key phase sensor 130 sets a key phase mark on the rotating part of the device under test, and when the key phase mark rotates with the rotating part to the probe position, the key phase sensor 130 generates a pulse signal, that is, the key phase sensor 130 generates a pulse signal every time the rotating part rotates one circle. By collecting the pulse time, the phase angle of the vibration can be determined, the starting position of one rotation of the rotating component can be determined, and accordingly, the vibration time domain waveform collected by the vibration sensor 120 can be converted into the angular domain waveform. The installation position of the key phase sensor 130 may be determined by those skilled in the art, and the present invention is not limited thereto. For example, the key phase sensor 130 may be provided on the flywheel component.
As shown in fig. 2, in the device failure diagnosis system 100, the computing device 200 is in communication connection with the vibration sensors 120 disposed at each measurement point of the device under test 110 and the key phase sensor 130 disposed on the rotating component, and is adapted to receive the vibration signals collected by each vibration sensor 120 and the pulse signals output by the key phase sensor 130, and store, analyze, display and the like the vibration signals and the pulse signals. In an embodiment of the present invention, the computing device 200 may extract angular domain vibration data from the vibration signal and the pulse signal, execute the reciprocating mechanical device fault diagnosis method 500 of the present invention to identify a fault of the device under test 110, and further determine the type of the fault when the fault is identified.
FIG. 3 shows a schematic diagram of a computing device 200, according to one embodiment of the invention. It should be noted that the computing device 200 shown in fig. 3 is only an example, and in practice, the computing device for implementing the device fault diagnosis method of the present invention may be any type of device, and the hardware configuration thereof may be the same as that of the computing device 200 shown in fig. 3 or different from that of the computing device 200 shown in fig. 3. In practice, the computing device implementing the device fault diagnosis method of the present invention may add or delete hardware components of the computing device 200 shown in fig. 3, and the present invention does not limit the specific hardware configuration of the computing device.
As shown in FIG. 3, in the basic configuration 102, the computing device 200 typically includes a system memory 206 and one or more processors 204. A memory bus 208 may be used for communication between the processor 204 and the system memory 206.
Depending on the desired configuration, the processor 204 may be any type of processing, including but not limited to: a microprocessor (μ P), a microcontroller (μ C), a Digital Signal Processor (DSP), or any combination thereof. The processor 204 may include one or more levels of cache, such as a level one cache 210 and a level two cache 212, a processor core 214, and registers 216. Example processor cores 214 may include Arithmetic Logic Units (ALUs), Floating Point Units (FPUs), digital signal processing cores (DSP cores), or any combination thereof. The example memory controller 218 may be used with the processor 204, or in some implementations the memory controller 218 may be an internal part of the processor 204.
Depending on the desired configuration, system memory 206 may be any type of memory, including but not limited to: volatile memory (such as RAM), non-volatile memory (such as ROM, flash memory, etc.), or any combination thereof. The physical memory in the computing device is usually referred to as a volatile memory RAM, and data in the disk needs to be loaded into the physical memory to be read by the processor 204. System memory 206 may include an operating system 220, one or more applications 222, and program data 224. In some implementations, the application 222 can be arranged to execute instructions on the operating system with the program data 224 by the one or more processors 204. Operating system 220 may be, for example, Linux, Windows, or the like, which includes program instructions for handling basic system services and for performing hardware-dependent tasks. The application 222 includes program instructions for implementing various user-desired functions, and the application 222 may be, for example, but not limited to, a browser, instant messenger, a software development tool (e.g., an integrated development environment IDE, a compiler, etc.), and the like. When the application 222 is installed into the computing device 200, a driver module may be added to the operating system 220.
When the computing device 200 is started, the processor 204 reads program instructions of the operating system 220 from the memory 206 and executes them. Applications 222 run on top of operating system 220, utilizing the interface provided by operating system 220 and the underlying hardware to implement various user-desired functions. When the user starts the application 222, the application 222 is loaded into the memory 206, and the processor 204 reads the program instructions of the application 222 from the memory 206 and executes the program instructions.
A network communication link may be one example of a communication medium. Communication media may typically be embodied by computer readable instructions, data structures, program modules, and may include any information delivery media, such as carrier waves or other transport mechanisms, in a modulated data signal. A "modulated data signal" may be a signal that has one or more of its data set or its changes made in such a manner as to encode information in the signal. By way of non-limiting example, communication media may include wired media such as a wired network or private-wired network, and various wireless media such as acoustic, Radio Frequency (RF), microwave, Infrared (IR), or other wireless media. The term computer readable media as used herein may include both storage media and communication media.
The computing device 200 also includes a storage interface bus 234 coupled to the bus/interface controller 230. The storage interface bus 234 is coupled to the storage device 232, and the storage device 232 is adapted for data storage. The example storage device 232 may include removable storage 236 (e.g., CD, DVD, U-disk, removable hard disk, etc.) and non-removable storage 238 (e.g., hard disk drive, HDD, etc.).
In a computing device 200 according to the present invention, the program data 224 includes a pre-trained self-encoded model, and the application 222 includes instructions for performing a reciprocating machine fault diagnosis method 500 of the present invention, which may instruct the processor 204 to perform the reciprocating machine fault diagnosis method 500 of the present invention for automated, intelligent fault diagnosis of a device under test (reciprocating machine).
The self-coding model may be trained and stored by the computing device 200, for example, or trained by another computing device, and transplanted into the computing device 200 after training is complete.
According to one embodiment, a plurality of self-coding models are deployed in the computing device 200, each self-coding model corresponding to a reciprocating machine device. For example, a reciprocating compressor corresponds to a self-encoding model1, a reciprocating vacuum pump corresponds to a self-encoding model2, and so on.
In an embodiment of the invention, the self-coding model is trained for training samples with vibration data of the reciprocating mechanical equipment during normal operation. Specifically, the training samples may be collected as follows:
when the equipment normally runs, the vibration sensor collects a vibration signal of the equipment to obtain a vibration time domain waveform, wherein the horizontal axis (x axis) of the waveform is time, and the vertical axis (y axis) of the waveform is a vibration acceleration value. The key phase sensor outputs a pulse time domain waveform, the horizontal axis of the waveform is time, the vertical axis of the waveform is a pulse value, and the time between two pulses is the time of one movement period of the equipment, wherein the first pulse is the starting time of the movement period, and the second pulse is the ending time of the movement period. And intercepting the vibration time domain waveform according to the starting time and the ending time to obtain the vibration time domain waveform of the equipment in one motion period. Furthermore, the time of one motion period can be divided into 360 equally spaced parts, each part corresponds to 1 degree of an angular domain, the vibration time domain waveform in one motion period is sampled according to the time corresponding to each 1 degree, 360 vibration acceleration values are obtained, the 360 vibration acceleration values form an angular domain vibration waveform, the horizontal axis of the waveform is an angle, and the vertical coordinate of the waveform is a vibration acceleration value. Each angular domain vibration waveform vector containing 360 points is a training sample, and a plurality of training samples form a training set.
According to an embodiment, in order to improve the convergence rate of the self-coding model during training and enable the trained self-coding model to have good accuracy and generalization performance, the training samples need to be normalized before participating in the training of the self-coding model, and the vibration acceleration values in the training samples are located within a preset interval (e.g., [0,1 ]). The normalization of the training samples also eliminates the effect of different amplitude ranges (due to the same equipment differences or different operating conditions) on fault detection.
It should be noted that the present invention is not limited to the specific method of normalization operation. According to one embodiment, for each training sample, normalization may be performed according to the following formula:
wherein x isiFor the ith vibration acceleration value, x, in the training samplemax、xminRespectively the maximum value and the minimum value, x, of each vibration acceleration value in the training samplei' is the ith vibration acceleration value in the normalized training sample.
In the embodiment of the invention, the vibration data of various types of reciprocating mechanical equipment in normal operation is taken as a training sample, and a self-coding model of each type of reciprocating mechanical equipment is generated through (unsupervised) training.
The self-coding model includes two parts, an encoder and a decoder. The encoder is used for learning features in the vibration data, reducing dimensions of the vibration data, compressing the vibration data into low-dimensional codes, and extracting more abstract features of the vibration data. The decoder reconstructs the code after dimension reduction, and restores the low-dimension code to be close to the original input expression as possible to obtain reconstructed data.
It should be noted that the present invention is not limited to a specific network structure of the self-coding model, and any self-encoder structure is within the scope of the present invention. Fig. 4 shows a network structure diagram of a self-coding model according to an embodiment of the invention. As shown in FIG. 4, the self-coding model includes an input layer, three hidden layers H1-H3, and an output layer. The input data of the model is vibration data X, and the vibration data X is an angular domain vibration acceleration vector X with the length of m [ < X > ]1，x2，x3，…，xm]. Typically, m is 360, i.e. each degree over the angular range corresponds to one vibration acceleration value.
The input layer, the hidden layer H1, and the hidden layer H2 constitute an encoder. The vibration data X is compressed into a hidden vector of length n (n < m) via a hidden layer H1; and then compressed into a concealment vector of length p (p < n) via the concealment layer H2.
The hidden layer H2, the hidden layer H3, and the output layer constitute a decoder. The p-dimensional hidden vector output by the hidden layer H2 passes through the hidden layer H3 and is reconstructed into a hidden vector with the length of n; then, the data is reconstructed into m-dimensional output data X ═ X through an output layer1’，x2’，x3’，…，xm’]。
Since the self-coding model is trained by using vibration data of the reciprocating mechanical equipment during normal operation, the data still obeys certain distribution characteristics despite the existence of noise and fluctuation during normal operation of the reciprocating mechanical equipment. The self-coding model learns these features through a large amount of training sample data. When a fault occurs, the vibration waveform data can have the changes of the impact amplitude, the increase of the impact quantity and the obvious advance or delay of the impact on an angular domain, which can change the characteristics of the data, so that the self-coding model trained by using the vibration data of the normal operation of the equipment can not reconstruct the fault data well. That is, when a failure occurs, the error of reconstructing the input data from the coding model may increase significantly. Therefore, whether the reciprocating mechanical equipment is in an abnormal state, namely a fault state can be judged by monitoring the reconstruction error of the input vibration data and the reconstruction data.
Based on the trained self-coding model, the computing device 200 may perform the reciprocating machine fault diagnosis method 500 of the present invention to implement automated and intelligent fault diagnosis for the reciprocating machine.
Figure 5 illustrates a flow diagram of a reciprocating machine equipment fault diagnostic method 500 according to one embodiment of the present invention. Method 500 is performed in a computing device, such as computing device 200 described above. As shown in fig. 5, the method 500 begins at step S510.
In step S510, vibration data of a device under test in one motion cycle is obtained, where the device under test is a reciprocating mechanical device.
As described above, the device under test is provided with the vibration sensor and the key phase sensor, and the key phase sensor is provided on the rotating member (e.g., flywheel) of the device under test. In step S510, the vibration data of the device under test in one motion cycle may further be obtained according to the following steps:
and acquiring a vibration signal acquired by the vibration sensor and a pulse signal output by the key phase sensor. The vibration signals at all the moments form a vibration time domain waveform with time on the horizontal axis and vibration acceleration on the vertical axis; the pulse signals at each time constitute a pulse time domain waveform with time on the horizontal axis and pulse values on the vertical axis. According to the pulse signal, one rotation of the rotating member, i.e., the start time and the end time of one movement cycle, is determined. In the pulse time domain waveform, the time between two pulses is the time of one motion cycle of the device, wherein the first pulse is the start time of the motion cycle and the second pulse is the end time of the motion cycle. And then, intercepting the vibration signal according to the determined starting time and the determined ending time, and taking the intercepted vibration signal segment as vibration data of the equipment to be tested in a motion period.
Furthermore, the time of one motion period can be divided into 360 equally spaced parts, each part corresponds to 1 degree of the angular domain, the vibration data in one motion period is sampled according to the time corresponding to each 1 degree, 360 vibration acceleration values are obtained, the 360 vibration acceleration values form an angular domain vibration waveform, the horizontal axis of the waveform is an angle, and the vertical axis of the waveform is a vibration acceleration value. After the above processing, step S510 obtains vibration data of the device under test in one motion cycle, where the vibration data is a vector including 360 data points, that is, the vibration data X ═ X1，x2，x3，…，x360]Each data point in the vector represents a vibration acceleration value for one rotation angle, i.e. xiThe vibration acceleration value when the rotation angle is i.
In step S520, the vibration data obtained in step S510 is input into a preset self-coding model, so that the self-coding model outputs reconstruction data of the vibration data, where the self-coding model is obtained by training a training sample with the vibration data of the device of the same type as the device to be tested during normal operation.
As described above, the computing device 200 stores self-coding models corresponding to different types of reciprocating mechanical devices. The self-coding model used in step S520 is a self-coding model corresponding to the type of the device to be tested, that is, the self-coding model used in step S520 is obtained by training, as a training sample, vibration data of the device of the same type as the device to be tested during normal operation.
The vibration data X obtained in step S510 is set to [ X ═ X1，x2，x3，…，x360]Inputting a self-coding model, and outputting reconstructed data X' ═ X of the vibration data from the self-coding model1’，x2’，x3’，…，x360’]。
Subsequently, in step S530, a reconstruction error between the vibration data and the reconstruction data is calculated, and when the reconstruction error is greater than a preset error threshold, it is determined that the device under test has a fault.
According to one embodiment, the reconstruction error between the vibration data and the reconstruction data is the euclidean distance between the two, i.e. the reconstruction error is calculated according to the following formula:
wherein x isi、xi' the ith data point (vibration acceleration value) in the vibration data and the reconstructed data, respectively.
According to one embodiment, the error threshold is determined according to the following steps: and inputting a plurality of verification samples into the self-coding model so as to output reconstruction data of each verification sample from the self-coding model, wherein the plurality of verification samples comprise a plurality of normal samples and a plurality of abnormal samples, the normal samples are vibration data of equipment of the same type as the equipment to be tested during normal operation, and the abnormal samples are vibration data of the equipment of the same type as the equipment to be tested during failure. The normal sample and the abnormal sample can be collected according to the same method as the training sample, and the collected normal sample and the training sample are both angular domain vibration waveforms of 360 degrees. And respectively calculating the reconstruction error of each verification sample and the reconstruction data thereof, and determining an error threshold according to the reconstruction error of the normal sample and the reconstruction error of the abnormal sample, wherein the error threshold is more than or equal to the reconstruction error of the normal sample and less than or equal to the reconstruction error of the abnormal sample.
It should be noted that the error threshold should be greater than or equal to the reconstruction error of the normal sample and less than or equal to the reconstruction error of the abnormal sample, but the specific value of the error threshold is not limited in the present invention. For example, the error threshold may be set to an average of a maximum value of the reconstruction error of the normal sample and a minimum value of the reconstruction error of the abnormal sample, or to an average of an average value of the reconstruction error of the normal sample and an average value of the reconstruction error of the abnormal sample, or the like. In addition, when the error threshold is determined according to the reconstruction error of the normal sample and the reconstruction error of the abnormal sample, the normal sample and the abnormal sample with obviously abnormal reconstruction errors can be removed, namely, the numerical values of the obviously abnormal reconstruction errors do not participate in the calculation of the error threshold.
In the technical solution of the present invention, when step S530 determines that the device under test has a fault, the fault type may be further identified.
According to an embodiment, the fault type may be determined according to the following steps S540, S550:
in step S540, when it is determined that the device under test has a fault, the distances between the fault data and the fault samples are respectively calculated, and a preset number of fault samples with the smallest distance are used as neighbor samples, where the fault data is vibration data according to which the device under test has a fault, that is, if it is determined that the device under test has a fault in step S530, the vibration data acquired in step S510 is the fault data.
The fault samples comprise vibration data of equipment of the same type as the equipment to be tested in different fault types, namely, each fault sample comprises a 360-dimensional angular domain vibration vector and a fault type corresponding to the vector. And the number of the fault samples of each fault type is the same or similar, so that the accuracy of subsequent fault classification is ensured.
The failure types include, for example, a variety of gas valve failures (including gas valve leakage, valve plate sticking, valve plate fracture, spring failure, and the like), piston and cylinder component failures (including support ring wear or fracture, piston cracking, cylinder liner wear, cylinder pull, cylinder crash, fluid hammer, and the like), moving component failures (including piston rod wear, piston rod fracture, crosshead shoe wear, small head shoe wear, large head shoe wear, connecting rod bolt fracture, crankshaft fracture, and the like), and the like.
Further, in order to improve the accuracy of fault classification, in step S540, the fault component may be determined first, and the fault type may be primarily screened according to the fault component. Namely: first, a faulty component is determined based on the location of the vibration sensor collecting the fault data on the device under test. And then, acquiring fault samples corresponding to the fault components, and respectively calculating the distance between the fault data and each acquired fault sample.
For example, the fault data is collected by a sensor arranged on a cylinder body, the fault component is the cylinder body, the type of fault possibly related to the cylinder body comprises liquid impact, damage of a suction/exhaust valve, cylinder pulling, cylinder sticking, cylinder hitting, cylinder abrasion and the like, and accordingly, the fault sample corresponding to the cylinder body is the fault sample with the fault type being the fault type. For example, if the failure data is acquired by a sensor provided in the crosshead, the failure point is the crosshead, and the failure type of the crosshead is the crosshead wear, the bolt loosening, and the large and small head shoe wear.
According to one embodiment, the distance between the fault data and the fault sample is the euclidean distance of the vibration acceleration vectors corresponding to the two.
In step S540, after the distance between the failure data and the failure sample is calculated, a preset number of failure samples with the minimum distance are used as neighbor samples. The value of the preset number can be set by a person skilled in the art, and the value of the preset number is not limited by the invention. After determining the neighbor samples, step S550 is performed.
In step S550, the fault type of the device under test is determined according to the fault type to which each neighbor sample belongs and the distance to the fault data.
According to one embodiment, step S550 further includes steps S552 and S554:
in step S552, the weight of each neighbor sample is determined according to the distance from the neighbor sample to the fault data, wherein the smaller the distance from the neighbor sample to the fault data, the greater the weight thereof.
It should be noted that, in the embodiment of the present invention, the setting of the weight of the neighbor sample needs to satisfy the condition that the smaller the distance to the fault data is, the larger the weight is, but the present invention does not limit the specific setting method of the weight, and any setting method of the weight is within the protection scope of the present invention.
For example, the weights of the neighbor samples may be set according to the following formula:
wherein, wiIs the weight of the neighbor sample i, diIs the distance from the neighbor sample i to the fault data, and c is a preset normal number, which is used to prevent the denominator in the above equation from being 0. The value of c may be set by one skilled in the art without limitation.
For another example, the weights of the neighbor samples may be calculated with reference to a gaussian function, the weights are the largest and 1 when the distance is 0, and the weights gradually decrease and approach to 0 as the distance increases, that is, the weights of the neighbor samples are calculated according to the following formula:
wherein, wiIs the weight of the neighbor sample i, diThe distance from the neighbor sample i to the fault data, e is a natural constant, a and b are preset normal numbers, the values of the two can be set by the technicians in the field, and the invention does not do soAnd (4) limiting.
After the weight of each neighbor sample is calculated, step S554 is executed.
In step S554, calculating a confidence that the fault data belongs to each fault type, wherein the confidence of the fault type is a weighted sum of distances of each neighbor sample belonging to the fault type; and taking the fault type with the maximum confidence coefficient as the fault type of the equipment to be tested.
For example, there are 9 neighbor samples of failure data, and the distances to the failure data are d respectively1～d9The weights are respectively w1～w9. The 9 neighbor samples relate to 3 fault types, namely type 1-type 3. Wherein, the neighbor samples 1, 4, 7 belong to the fault type1, the neighbor samples 2, 5, 8 belong to the fault type2, the neighbor samples 3, 6, 9 belong to the fault type3, and then the confidence p of the fault type in 31～p3Calculated according to the following formula respectively:
p1＝d1*w1+d4*w4+d7*w7
p2＝d2*w2+d5*w5+d8*w8
p3＝d3*w3+d6*w6+d9*w9
by calculation, p2At maximum, then p will be2And the corresponding fault type2 is used as the fault type of the equipment to be tested.
The technical scheme of the invention trains the automatic encoder network by accumulated vibration data of a large number of reciprocating mechanical equipment during normal operation to generate a self-encoding model. The method comprises the steps of compressing and reconstructing current vibration data of equipment to be detected by using a trained model to obtain reconstructed data, calculating a reconstruction error between the vibration data and the reconstructed data, and identifying equipment faults by comparing the reconstruction error with an error threshold value, so that automation and intellectualization of fault identification of reciprocating mechanical equipment are realized. Meanwhile, the abstract characteristics of the vibration data are extracted from the coding model for fault recognition, an alarm threshold value does not need to be set through manual experience, and fault misjudgment and misjudgment caused by improper setting of the alarm threshold value are avoided.
Further, after the fault data are identified, the fault data are compared with accumulated fault samples of the reciprocating mechanical equipment by adopting a KNN algorithm, the first K fault samples which are most similar to the fault data in the sample set are used as neighbor samples, the fault types of the neighbor samples and the distance from the neighbor samples to the fault data are integrated, the confidence coefficient that the fault data belong to the fault types is predicted, the fault type with the highest confidence coefficient is used as the fault type of the equipment to be detected, and automatic classification based on historical fault data is realized.
The equipment fault diagnosis scheme of the invention adopts the self-coding model to carry out abnormity identification on the vibration data acquired by the reciprocating mechanical equipment, and after the fault data is identified, fault classification is carried out on the fault data by using a classification algorithm such as KNN (K nearest neighbor) and the like, so that automation and intellectualization of fault diagnosis of the reciprocating mechanical equipment are realized.
The various techniques described herein may be implemented in connection with hardware or software or, alternatively, with a combination of both. Thus, the methods and apparatus of the present invention, or certain aspects or portions thereof, may take the form of program code (i.e., instructions) embodied in tangible media, such as removable hard drives, U.S. disks, floppy disks, CD-ROMs, or any other machine-readable storage medium, wherein, when the program is loaded into and executed by a machine, such as a computer, the machine becomes an apparatus for practicing the invention.
In the case of program code execution on programmable computers, the control terminal generally includes a processor, a storage medium readable by the processor (including volatile and non-volatile memory and/or storage elements), at least one input device, and at least one output device. Wherein the memory is configured to store program code; the processor is configured to execute the device failure diagnostic method of the present invention according to instructions in the program code stored in the memory.
By way of example, and not limitation, readable media may comprise readable storage media and communication media. Readable storage media store information such as computer readable instructions, data structures, program modules or other data. Communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media. Combinations of any of the above are also included within the scope of readable media.
In the description provided herein, algorithms and displays are not inherently related to any particular computer, virtual system, or other apparatus. Various general purpose systems may also be used with examples of this invention. The required structure for constructing such a system will be apparent from the description above. Moreover, the present invention is not directed to any particular programming language. It is appreciated that a variety of programming languages may be used to implement the teachings of the present invention as described herein, and any descriptions of specific languages are provided above to disclose preferred embodiments of the invention.
In the description provided herein, numerous specific details are set forth. It is understood, however, that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.
Similarly, it should be appreciated that in the foregoing description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure and aiding in the understanding of one or more of the various inventive aspects. However, the disclosed method should not be interpreted as reflecting an intention that: that the invention as claimed requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment. Thus, the claims following the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of this invention.
Those skilled in the art will appreciate that the modules or units or components of the devices in the examples disclosed herein may be arranged in a device as described in this embodiment or alternatively may be located in one or more devices different from the devices in this example. The modules in the foregoing examples may be combined into one module or may be further divided into multiple sub-modules.
Those skilled in the art will appreciate that the modules in the device in an embodiment may be adaptively changed and disposed in one or more devices different from the embodiment. The modules or units or components of the embodiments may be combined into one module or unit or component, and furthermore they may be divided into a plurality of sub-modules or sub-units or sub-components. All of the features disclosed in this specification (including any accompanying claims, abstract and drawings), and all of the processes or elements of any method or apparatus so disclosed, may be combined in any combination, except combinations where at least some of such features and/or processes or elements are mutually exclusive. Each feature disclosed in this specification (including any accompanying claims, abstract and drawings) may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise.
Furthermore, those skilled in the art will appreciate that while some embodiments described herein include some features included in other embodiments, rather than other features, combinations of features of different embodiments are meant to be within the scope of the invention and form different embodiments. For example, in the following claims, any of the claimed embodiments may be used in any combination.
Furthermore, some of the described embodiments are described herein as a method or combination of method elements that can be performed by a processor of a computer system or by other means of performing the described functions. A processor having the necessary instructions for carrying out the method or method elements thus forms a means for carrying out the method or method elements. Further, the elements of the apparatus embodiments described herein are examples of the following apparatus: the apparatus is used to implement the functions performed by the elements for the purpose of carrying out the invention.
As used herein, unless otherwise specified the use of the ordinal adjectives "first", "second", "third", etc., to describe a common object, merely indicate that different instances of like objects are being referred to, and are not intended to imply that the objects so described must be in a given sequence, either temporally, spatially, in ranking, or in any other manner.
While the invention has been described with respect to a limited number of embodiments, those skilled in the art, having benefit of this description, will appreciate that other embodiments can be devised which do not depart from the scope of the invention as described herein. Furthermore, it should be noted that the language used in the specification has been principally selected for readability and instructional purposes, and may not have been selected to delineate or circumscribe the inventive subject matter. Accordingly, many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the appended claims. The present invention has been disclosed in an illustrative rather than a restrictive sense with respect to the scope of the invention, as defined in the appended claims.
Claims (10)
1. A reciprocating machine equipment fault diagnosis method, executed in a computing device, comprising:
obtaining vibration data of equipment to be tested in a motion cycle, wherein the equipment to be tested is reciprocating mechanical equipment;
inputting the vibration data into a preset self-coding model so that the self-coding model can output reconstruction data of the vibration data, wherein the self-coding model is obtained by training vibration data of equipment with the same type as the equipment to be tested in normal operation as a training sample;
and calculating a reconstruction error between the vibration data and the reconstruction data, and judging that the equipment to be tested breaks down when the reconstruction error is larger than a preset error threshold value.
2. The method of claim 1, wherein a vibration sensor and a key phase sensor are provided on the device under test, the key phase sensor being provided on a rotating component of the device under test,
the step of acquiring vibration data of the device to be tested in a motion cycle comprises the following steps:
acquiring a vibration signal acquired by a vibration sensor and a pulse signal output by a key phase sensor;
determining the starting time and the ending time of one rotation of the rotating part according to the pulse signal;
and intercepting the vibration signal according to the starting time and the ending time, and taking the intercepted vibration signal as vibration data of the equipment to be tested in a motion period.
3. The method according to claim 1 or 2, wherein the training samples are normalized before participating in the training of the self-coding model, so that the values in the training samples are within a preset interval.
4. A method according to any of claims 1-3, wherein the error threshold is determined according to the following steps:
inputting a plurality of verification samples into the self-coding model so that the self-coding model outputs reconstruction data of each verification sample, wherein the verification samples comprise a plurality of normal samples and a plurality of abnormal samples, the normal samples are vibration data of equipment of the same type as the equipment to be tested during normal operation, and the abnormal samples are vibration data of the equipment of the same type as the equipment to be tested during failure;
and respectively calculating the reconstruction error of each verification sample and the reconstruction data thereof, and determining an error threshold according to the reconstruction error of the normal sample and the reconstruction error of the abnormal sample, wherein the error threshold is greater than or equal to the reconstruction error of the normal sample and less than or equal to the reconstruction error of the abnormal sample.
5. The method of any of claims 1-4, further comprising:
when the equipment to be tested is judged to have a fault, respectively calculating the distance between fault data and fault samples, and taking a preset number of fault samples with the minimum distance as neighbor samples, wherein the fault data are vibration data according to which the equipment to be tested is judged to have a fault, and the fault samples comprise vibration data of equipment of the same type as the equipment to be tested in different fault types;
and determining the fault type of the equipment to be tested according to the fault type of each neighbor sample and the distance from the neighbor sample to the fault data.
6. The method of claim 5, wherein the step of separately calculating distances between the fault data and the fault samples comprises:
determining a fault component according to the position of a vibration sensor for collecting fault data on the equipment to be tested;
acquiring a fault sample corresponding to the fault component; and
and respectively calculating the distance between the fault data and each obtained fault sample.
7. The method of claim 5 or 6, wherein the step of determining the fault type of the device under test from the fault type to which each neighbor sample belongs and the distance to the fault data comprises:
determining the weight of each neighbor sample according to the distance from each neighbor sample to the fault data, wherein the smaller the distance from each neighbor sample to the fault data is, the larger the weight is;
calculating the confidence coefficient of the fault data belonging to each fault type, wherein the confidence coefficient of the fault type is the weighted sum of the distances of all the neighbor samples belonging to the fault type;
and taking the fault type with the maximum confidence coefficient as the fault type of the equipment to be tested.
8. The method of claim 7, wherein the weights of the neighbor samples are determined according to the following formula:
wherein, wiIs the weight of the neighbor sample i, diThe distance from the neighbor sample i to the fault data is calculated, and c is a preset normal number;
alternatively, the weights of the neighbor samples are determined according to the following formula:
wherein, wiIs the weight of the neighbor sample i, diThe distance from the neighbor sample i to the fault data is shown, e is a natural constant, and a and b are preset normal constants.
9. A computing device, comprising:
at least one processor and a memory storing program instructions;
the program instructions, when read and executed by the processor, cause the computing device to perform the reciprocating machine equipment fault diagnostic method of any of claims 1-8.
10. A readable storage medium storing program instructions that, when read and executed by a computing device, cause the computing device to perform the reciprocating machine equipment fault diagnosis method of any of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110003526.5A CN112816195A (en) | 2021-01-04 | 2021-01-04 | Reciprocating mechanical equipment fault diagnosis method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110003526.5A CN112816195A (en) | 2021-01-04 | 2021-01-04 | Reciprocating mechanical equipment fault diagnosis method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112816195A true CN112816195A (en) | 2021-05-18 |
Family
ID=75857020
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110003526.5A Pending CN112816195A (en) | 2021-01-04 | 2021-01-04 | Reciprocating mechanical equipment fault diagnosis method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112816195A (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104048749A (en) * | 2014-06-24 | 2014-09-17 | 东北电力大学 | Vibration fault detection system and method for wind turbine generator units |
| CN109188502A (en) * | 2018-07-05 | 2019-01-11 | 中国科学技术大学 | A kind of beam transport network method for detecting abnormality and device based on self-encoding encoder |
| CN109409425A (en) * | 2018-10-09 | 2019-03-01 | 宁波大学 | A kind of fault type recognition method based on neighbour's constituent analysis |
| CN110057584A (en) * | 2019-04-01 | 2019-07-26 | 西安交通大学 | A kind of locomotive traction motor bearing degradation monitoring method |
| CN111412978A (en) * | 2020-04-22 | 2020-07-14 | 北京化工大学 | Reciprocating machinery abnormity detection method based on fault-free vibration signal |
| US20200234143A1 (en) * | 2019-01-23 | 2020-07-23 | MakinaRocks Co., Ltd. | Anomaly detection |
| CN111562108A (en) * | 2020-05-09 | 2020-08-21 | 浙江工业大学 | Rolling bearing intelligent fault diagnosis method based on CNN and FCMC |
| CN112070125A (en) * | 2020-08-19 | 2020-12-11 | 西安理工大学 | Prediction method of unbalanced data set based on isolated forest learning |
| CN112149757A (en) * | 2020-10-23 | 2020-12-29 | 新华三大数据技术有限公司 | Abnormity detection method and device, electronic equipment and storage medium |
-
2021
- 2021-01-04 CN CN202110003526.5A patent/CN112816195A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104048749A (en) * | 2014-06-24 | 2014-09-17 | 东北电力大学 | Vibration fault detection system and method for wind turbine generator units |
| CN109188502A (en) * | 2018-07-05 | 2019-01-11 | 中国科学技术大学 | A kind of beam transport network method for detecting abnormality and device based on self-encoding encoder |
| CN109409425A (en) * | 2018-10-09 | 2019-03-01 | 宁波大学 | A kind of fault type recognition method based on neighbour's constituent analysis |
| US20200234143A1 (en) * | 2019-01-23 | 2020-07-23 | MakinaRocks Co., Ltd. | Anomaly detection |
| CN110057584A (en) * | 2019-04-01 | 2019-07-26 | 西安交通大学 | A kind of locomotive traction motor bearing degradation monitoring method |
| CN111412978A (en) * | 2020-04-22 | 2020-07-14 | 北京化工大学 | Reciprocating machinery abnormity detection method based on fault-free vibration signal |
| CN111562108A (en) * | 2020-05-09 | 2020-08-21 | 浙江工业大学 | Rolling bearing intelligent fault diagnosis method based on CNN and FCMC |
| CN112070125A (en) * | 2020-08-19 | 2020-12-11 | 西安理工大学 | Prediction method of unbalanced data set based on isolated forest learning |
| CN112149757A (en) * | 2020-10-23 | 2020-12-29 | 新华三大数据技术有限公司 | Abnormity detection method and device, electronic equipment and storage medium |
Non-Patent Citations (3)
| Title |
|---|
| "基于变分自编码器的异常小区检测", pages 2, Retrieved from the Internet <URL:https://m.c114.com.cn/w164-1093754.html> * |
| 张彼德 等: "电力变压器故障诊断的k值自适应加权KNN算法研究", 《湖北电力》 * |
| 张彼德 等: "电力变压器故障诊断的k值自适应加权KNN算法研究", 《湖北电力》, no. 02, 30 April 2020 (2020-04-30), pages 10 - 16 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Elhaj et al. | Numerical simulation and experimental study of a two-stage reciprocating compressor for condition monitoring | |
| CN107076157B (en) | Method and system for detecting a fault in a pump assembly through a handheld communication device | |
| US9959738B2 (en) | Reciprocating machinery monitoring system and method | |
| Ali et al. | Automated valve fault detection based on acoustic emission parameters and support vector machine | |
| US8655571B2 (en) | MFCC and CELP to detect turbine engine faults | |
| US20100106458A1 (en) | Computer program and method for detecting and predicting valve failure in a reciprocating compressor | |
| CN109997025A (en) | Engine Gernral Check-up device and method | |
| CN101059130A (en) | On-line remote state monitoring and fault analysis diagnosis system of reciprocating compressor | |
| CN104712542A (en) | Reciprocating compressor sensitive characteristic extracting and fault diagnosis method based on internet of things | |
| JP2011027744A (en) | System, device, and method for shock monitoring | |
| Lv et al. | Adaptive fault diagnosis of sucker rod pump systems based on optimal perceptron and simulation data | |
| CN112443479A (en) | Compressor fault diagnosis method and device, storage medium and electronic equipment | |
| Ftoutou et al. | Internal combustion engine valve clearance fault classification using multivariate analysis of variance and discriminant analysis | |
| Ahmed et al. | Fault detection of reciprocating compressors using a model from principles component analysis of vibrations | |
| CN112816195A (en) | Reciprocating mechanical equipment fault diagnosis method and device | |
| CN111595585B (en) | Diesel engine cylinder fault diagnosis method, device and equipment and readable storage medium | |
| Pichler et al. | Fault detection in reciprocating compressor valves for steady-state load conditions | |
| Jin et al. | A vibration-based approach for diesel engine fault diagnosis | |
| Trout et al. | Reciprocating compressor valve condition monitoring using image-based pattern recognition | |
| Salah et al. | Automated valve fault detection based on acoustic emission parameters and artificial neural network | |
| WO2022069990A1 (en) | Method and system for monitoring and testing industrial valves | |
| Chesnes et al. | Condition monitoring of reciprocating compressor valve health via a statistical time-frequency approach. | |
| TWI697622B (en) | Reciprocating compressor monitoring system and method thereof | |
| Pichler et al. | On the robustness of fault detection in reciprocating compressor valves | |
| Tossavainen | Sound based fault detection system |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210518 |