CN113344087A

CN113344087A - Method for establishing sample set and computing equipment

Info

Publication number: CN113344087A
Application number: CN202110667955.2A
Authority: CN
Inventors: 张神林; 宋海峰; 贾维银
Original assignee: Anhui Ronds Science & Technology Inc Co
Current assignee: Anhui Ronds Science & Technology Inc Co
Priority date: 2021-06-16
Filing date: 2021-06-16
Publication date: 2021-09-03
Anticipated expiration: 2041-06-16
Also published as: CN113344087B

Abstract

The invention discloses a method for establishing a sample set, which is executed in computing equipment and comprises the following steps: acquiring vibration data and equipment attribute information of one or more kinds of equipment; inputting the vibration data into a pre-training model, and outputting corresponding labeled data after the vibration data is processed by the pre-training model; generating undetermined samples based on the vibration data and the corresponding labeling data, and generating an undetermined sample set based on a plurality of undetermined samples; and correcting a plurality of samples to be determined in the sample set to be determined to generate a target sample set so as to train the pre-training model based on the target sample set. The invention also discloses corresponding computing equipment. According to the method for establishing the sample set, a large number of more comprehensive samples can be generated, the generation efficiency of the samples is high, the model based on the trained samples can meet the monitoring requirements on various states of the equipment, and the intelligent level of equipment state monitoring is favorably improved.

Description

Method for establishing sample set and computing equipment

Technical Field

The invention relates to the technical field of intelligent monitoring, in particular to a method and computing equipment for establishing a sample set.

Background

Under the background of big data and artificial intelligence, the intelligent degree is promoted in all walks of life. In the process of researching intelligent algorithms and engineering practices, experimental samples are important tools for training and evaluating the intelligent algorithms, and the scale of the experimental samples is even more important than the delicate design of algorithm models. At present, the intelligent level of the equipment state monitoring field is still low due to the lack of effective, comprehensive and large amount of related samples for reflecting the equipment fault state in the equipment state monitoring field.

The most common and mature device condition monitoring means at the present stage is vibration-based device condition monitoring. In the prior art, training samples used in the vibration-based intelligent monitoring technology for the equipment state are generally obtained through the following two ways:

one is to simulate some faults in a laboratory manually, for example, by damaging a normal bearing inner ring, and collect data of the type for demonstrating vibration signal characteristics when the bearing inner ring is in fault, wherein the type is mostly used for research in colleges and universities. This solution has the following drawbacks: the simulated equipment is single, and the running states of all equipment in the industrial field with complicated equipment types cannot be simulated; the operation condition of the equipment is too simple compared with the actual operation condition of the industrial field equipment; the operation state of the device which can be simulated is single, and the state data of the whole life cycle of the device is difficult to obtain. Therefore, the objectivity and representativeness of the vibration state data of the equipment and equipment obtained through the laboratory are insufficient.

In another scheme, a corresponding sample is formed through manual marking based on characteristic data of partial consistent equipment operation states accumulated by manufacturers engaged in equipment state monitoring in the industry through a long-term nursing process. The defects of the scheme are that the monitoring equipment quantity of each manufacturer in the industry is small at present, and the sample source is insufficient; related manufacturers in the industry do not have service teams for monitoring the state of corresponding equipment, the tracking and closed loop of the state of the equipment in the monitoring process are not timely, the running state of the equipment cannot be timely known, and further the running state of the equipment cannot be timely combined with the corresponding state data characteristics; at present, samples formed by related manufacturers in the industry are often biased to a certain running state, the types of the samples are not rich enough, and at present, each manufacturer is biased to create various damaged samples, such as bearing defect samples, and less samples are created in other links in the equipment state monitoring process, such as related samples (such as signals error, start-stop, alarm, diagnosis and the like) possibly involved in the equipment state data acquisition process, so that the sample comprehensiveness is poor.

For this reason, a method for establishing a sample set is needed to solve the problems in the above technical solutions.

Disclosure of Invention

To this end, the present invention provides a method of establishing a sample set in an attempt to solve, or at least alleviate, the problems presented above.

According to an aspect of the present invention, there is provided a method of establishing a sample set, executed in a computing device, comprising the steps of: acquiring vibration data and equipment attribute information of one or more kinds of equipment; inputting the vibration data into a pre-training model, and outputting corresponding labeled data after the vibration data is processed by the pre-training model; generating undetermined samples based on the vibration data and the corresponding labeling data, and generating an undetermined sample set based on a plurality of undetermined samples; and correcting a plurality of samples to be determined in the sample set to be determined to generate a target sample set so as to train the pre-training model based on the target sample set.

Optionally, in the method of creating a sample set according to the present invention, the pre-training model is a pre-training startup and shutdown state determination model, and the annotation data includes a startup state or a shutdown state.

Optionally, in the method for creating a sample set according to the present invention, after generating the target sample set, the method further includes the steps of: the target sample set is partitioned according to device attribute information to generate a sample subset corresponding to each device attribute information.

Optionally, in the method of creating a sample set according to the present invention, the device attribute information includes one or more of a device type, a device name, a device identifier, device vendor information, and structural component information.

Optionally, in the method for creating a sample set according to the present invention, the step of dividing the target sample set according to the device attribute information includes: the target set of samples is partitioned according to device type to generate a subset of samples corresponding to each device type.

Optionally, in the method for establishing a sample set according to the present invention, the step of dividing the target sample set according to the device attribute information further includes: for the devices of the same device type, dividing the sample subsets corresponding to the device type according to the structural components of the devices to generate the sample subsets corresponding to the structural components.

Optionally, in the method for creating a sample set according to the present invention, before inputting the vibration data into the pre-training model, the method further includes the steps of: one or more sample sets and a tag type corresponding to each sample set are defined.

Optionally, in the method of establishing a sample set according to the present invention, the sample set includes one or more of a false signal sample set, a start-stop sample set, an alarm sample set, and a diagnosis sample set.

Optionally, in the method of creating a sample set according to the present invention, the obtaining vibration data of one or more devices comprises: vibration data of one or more devices over a predetermined period of time is obtained, the vibration data including trend data or spectral data.

Optionally, in the method of creating a sample set according to the present invention, before acquiring vibration data of one or more devices, the method includes the steps of: vibration data for one or more devices is collected by a data collection system to obtain vibration data from the data collection system.

According to an aspect of the present invention, there is provided a computing device comprising: at least one processor; and a memory storing program instructions, wherein the program instructions are configured to be executed by the at least one processor, the program instructions comprising instructions for performing the method of establishing a sample set as described above.

According to an 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 method as described above.

According to the technical scheme, the invention provides a method for establishing a sample set, which comprises the steps of obtaining actual vibration data of various types of equipment, generating a to-be-determined sample set after preliminarily marking the vibration data by using a pre-training model, and further generating a target sample set after manually correcting a plurality of to-be-determined samples in the to-be-determined sample set. Therefore, a large number of samples covering various equipment states can be generated, and the generation efficiency and accuracy of the samples can be improved by combining the pre-training model with manual correction processing. The model is trained based on more comprehensive and accurate samples, and the trained model can meet the monitoring requirements on multiple states of the equipment in the actual operation process of the equipment, so that the intelligent level of equipment state monitoring is favorably improved.

Further, the invention generates a sample subset corresponding to each equipment type by subdividing the sample set, and trains the pre-training model by adopting the sample subset corresponding to the equipment type. In this way, when the device state is detected based on the trained model, the trained model corresponding to the device type can be used to judge the device state for each type of device, which is beneficial to more accurately determining the device state.

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 shows a schematic diagram of a computing device 100, according to one embodiment of the invention; and

fig. 2 shows a flow diagram of a method 200 of establishing a sample set according to one embodiment of the 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.

Fig. 1 is a schematic block diagram of an example computing device 100.

As shown in FIG. 1, in a basic configuration 102, a computing device 100 typically includes a system memory 106 and one or more processors 104. A memory bus 108 may be used for communication between the processor 104 and the system memory 106.

Depending on the desired configuration, the processor 104 may be any type of processing, including but not limited to: a microprocessor (UP), a microcontroller (UC), a digital information processor (DSP), or any combination thereof. The processor 104 may include one or more levels of cache, such as a level one cache 110 and a level two cache 112, a processor core 114, and registers 116. The example processor core 114 may include an Arithmetic Logic Unit (ALU), a Floating Point Unit (FPU), a digital signal processing core (DSP core), or any combination thereof. The example memory controller 118 may be used with the processor 104, or in some implementations the memory controller 118 may be an internal part of the processor 104.

Depending on the desired configuration, system memory 106 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. System memory 106 may include an operating system 120, one or more applications 122, and program data 124. In some implementations, the application 122 can be arranged to execute instructions on an operating system with program data 124 by one or more processors 104.

Computing device 100 also includes a storage device 132, storage device 132 including removable storage 136 and non-removable storage 138.

Computing device 100 may also include a storage interface bus 134. The storage interface bus 134 enables communication from the storage devices 132 (e.g., removable storage 136 and non-removable storage 138) to the basic configuration 102 via the bus/interface controller 130. At least a portion of the operating system 120, applications 122, and data 124 may be stored on removable storage 136 and/or non-removable storage 138, and loaded into system memory 106 via storage interface bus 134 and executed by the one or more processors 104 when the computing device 100 is powered on or the applications 122 are to be executed.

Computing device 100 may also include an interface bus 140 that facilitates communication from various interface devices (e.g., output devices 142, peripheral interfaces 144, and communication devices 146) to the basic configuration 102 via the bus/interface controller 130. The example output device 142 includes a graphics processing unit 148 and an audio processing unit 150. They may be configured to facilitate communication with various external devices, such as a display or speakers, via one or more a/V ports 152. Example peripheral interfaces 144 may include a serial interface controller 154 and a parallel interface controller 156, which may be configured to facilitate communication with external devices such as input devices (e.g., keyboard, mouse, pen, voice input device, touch input device) or other peripherals (e.g., printer, scanner, etc.) via one or more I/O ports 158. An example communication device 146 may include a network controller 160, which may be arranged to facilitate communications with one or more other computing devices 162 over a network communication link via one or more communication ports 164.

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 a manner that encodes 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.

Computing device 100 may be implemented as a personal computer including both desktop and notebook computer configurations. Of course, computing device 100 may also be implemented as part of a small-form factor portable (or mobile) electronic device such as a cellular telephone, a digital camera, a Personal Digital Assistant (PDA), a personal media player device, a wireless web-watch device, a personal headset, an application specific device, or a hybrid device that include any of the above functions. And may even be implemented as a server, such as a file server, a database server, an application server, a WEB server, and so forth. The embodiments of the present invention are not limited thereto.

In an embodiment in accordance with the invention, the computing device 100 is configured to perform a method 200 of establishing a sample set in accordance with the invention. The application 122 of the computing device 100 includes a plurality of program instructions for executing the method 200 of creating a sample set according to the present invention, which can be read and executed by the computing device 100, so that the computing device 100 executes the method 200 of creating a sample set according to the present invention.

As shown in fig. 2, the method 200 begins at step S210. In step S210, vibration data and device attribute information for one or more devices are acquired. It should be noted that the vibration data of the device acquired here is vibration data collected during actual operation of the device, and the vibration data may include, for example, trend data, waveform data, spectrum data, and the like. The device attribute information includes, for example, one or more of a device type, a device name, a device identification, device vendor information, and structural component information. However, the present invention is not limited to the specific types of vibration data and device attribute information.

In one embodiment, for example, vibration data may be obtained for one or more devices operating over a predetermined period of time. It should be noted that the vibration data over the predetermined period of time can reflect various operating conditions of the device, so that the model can be trained on a sample basis to meet the monitoring requirements for various conditions of the device. Here, the specific duration of the predetermined time period is not limited in the present invention, and the predetermined time period may be set by a person skilled in the art according to actual needs. In one embodiment, the predetermined period of time may be one-half year.

It should be noted that for each device, multiple sets of vibration data may be acquired for the device over multiple different predetermined time periods in order to acquire a large number of vibration data to form a sufficient number of samples. It should be appreciated that by acquiring vibration state data of a plurality of devices, a sample set covering a plurality of device types can be subsequently created based on the vibration state data of the plurality of devices, so that a model trained based on the sample set can be used for monitoring the state of various types of devices to meet various monitoring requirements.

In one embodiment, vibration data of one or more devices during actual operation may be collected by the data collection system before obtaining vibration data of the one or more devices, such that, in step S210, the computing device may obtain vibration data of the actual operation of the devices from the data collection system.

Subsequently, in step S220, the obtained vibration data is input into the pre-training model, and the vibration data is processed by the pre-training model to output corresponding labeled data. Here, the pre-training model may preliminarily determine a corresponding device state based on the vibration data, and output corresponding labeled data.

It should be noted that the pre-training model is a model that is preliminarily established and can preliminarily determine the state of the device, and the state of the device can be preliminarily determined by inputting vibration data of the device into the pre-training model. The pre-training model is used for preliminarily judging the states of the devices, so that the model is not distinguished for different types of devices and can be used for preliminarily judging the states of the devices of different types. It should be noted that in order for the model to be useful for accurately determining the state of the device, the pre-trained model also needs to be further trained based on a large number of samples. Here, the present invention does not limit the kind of the pre-training model and the specific judgment logic for the device status.

Subsequently, in step S230, a to-be-determined sample is preliminarily generated based on the vibration data and the corresponding labeling data, and a to-be-determined sample set is generated based on the plurality of to-be-determined samples. Here, since multiple sets of vibration data of multiple types of devices can be acquired in step S210, accordingly, multiple pending samples can be generated in step S230, and a pending sample set can be generated based on the multiple pending samples. It should be understood that the labeled data in the sample to be determined is preliminarily determined by using the pre-training model, and therefore, the labeled data is not necessarily accurate, and the sample formed based on the labeled data is the sample to be determined which needs to be further subjected to verification and correction processing. It should also be appreciated that labeling the data based on the pre-trained model facilitates the rapid generation of multiple pending samples to improve the efficiency of sample generation.

Finally, in step S230, a plurality of samples to be determined in the sample set to be determined are modified to generate a target sample set. In this way, the pre-trained model described above may be trained based on a set of target samples. Specifically, the labeling data of each undetermined sample in the undetermined sample set is manually checked, the undetermined sample with correct labeling data can be directly used as a target sample, and the sample with wrong labeling data can be used as the target sample after being manually corrected, so that a plurality of target samples based on manual checking and correction processing can be combined to obtain the target sample set.

According to one embodiment, before inputting the vibration data into the pre-training model and establishing the pending sample set, the following steps are further performed: one or more sample sets are predefined, i.e., one or more sample set types are defined, and a tag type corresponding to each sample set is defined. Specifically, the sample set in the present invention may include one or more of a false signal sample set, a start-stop sample set, an alarm sample set, and a diagnosis sample set, for example, and is not limited to the listed sample set types. Thus, from the defined set or sets of samples, a corresponding type of target set of samples may be created according to the above-described steps of the invention. For example, when the target sample set is a start-up and shut-down sample set, the tag types corresponding to the start-up and shut-down sample set include a start-up state and a shut-down state, and a pre-trained start-up and shut-down state determination model may be trained based on the start-up and shut-down sample set, so as to determine that the device is the start-up state or the shut-down state by using the trained start-up and shut-down state determination model.

It should also be noted that the source of the vibration data may be different for different types of sample sets. For example, for vibration error signal data, part of the error signal data is labeled in the equipment monitoring process, some data of the labeled data in the equipment monitoring process can be directly obtained to serve as undetermined samples of the error signal, and an undetermined sample set is formed on the basis of a plurality of undetermined samples, so that preliminary labeling of the vibration data is not needed on the basis of a pre-training model.

In one embodiment, the pre-training model may be a pre-training start-stop state determination model for preliminary determination of the start-stop state (start-up state or stop state) of the plant, which may be used to accurately determine the start-stop state or stop state of the plant after training the model. The invention is not limited herein to the specific structure and algorithmic logic of the pre-trained startup and shutdown state determination model.

In one embodiment, the vibration data for the plant is input into a pre-trained shutdown state determination model that first pre-processes the raw vibration data to remove some outlier data in the vibration data. Furthermore, the pre-training startup and shutdown state determination model can judge whether the vibration data is in a unified state according to a preset startup and shutdown threshold value, wherein the startup and shutdown threshold value comprises an acceleration maximum value of 5m/s ^2 and an acceleration minimum value of 0.5m/s ^ 2. According to the set acceleration threshold, if all the vibration data are greater than the acceleration maximum value, the model can determine that the device is in the start-up state. If all of the vibration data is less than the acceleration minimum, the model may determine that the plant is in a shutdown state. And if the state of the vibration data is not single, judging the clustering effect, specifically, determining the classification condition of the vibration data by adopting a multi-cluster clustering K-Means algorithm, and outputting the starting and stopping states of the equipment according to corresponding threshold values when the classification effect is good, namely, the score is high. And if the clustering effect of the vibration data is not good, the starting and stopping state of the equipment can be further determined according to whether the frequency spectrum index meets the corresponding starting and stopping threshold value.

Based on the method in the above embodiment, the pre-training startup and shutdown state determination model may preliminarily determine the startup and shutdown state of the device according to the vibration data, and mark the vibration data to form a startup and shutdown pending sample set including a plurality of startup and shutdown pending samples.

According to one embodiment, after the target sample is generated, the target sample set may be further subdivided. Specifically, the target samples may be divided according to the device attribute information to generate a subset of samples corresponding to each device attribute information. Here, the device attribute information includes, for example, one or more of a device type, a device name, a device identification, device vendor information, and structural component information, but the present invention is not limited thereto.

According to one embodiment, the target sample set may be divided according to device types to generate sample subsets corresponding to each device type, that is, each finally generated sample subset corresponds to a device. In this way, after the pre-training model is trained by using the sample subset corresponding to the device type, for each type of device, the trained model corresponding to the device type can be used to judge the state of the device, which is beneficial to more accurately determining the state of the device.

For example, the vibration data of different types of equipment at startup and shutdown are significantly different, for example, in a petrochemical site, the difference between the vibration data of a fan and the vibration data of a conventional centrifugal pump is large, and even in many cases, the vibration amplitude of the fan at startup is smaller than that of the centrifugal pump at shutdown. In order to train the pre-training model more specifically, the invention further divides the target sample set according to the type of the device, for example, the target sample set can be divided into a pump sample subset, a fan sample subset, and the like. In one embodiment, for determining the start-up and shut-down states of the device, the sample set may be implemented as a start-up and shut-down sample set, and the target sample set may be further divided into a pump start-up and shut-down sample subset, a fan start-up and shut-down sample subset, and the like according to the device type.

Further, in addition to dividing the target sample set according to the device type, when the number of samples of each type is large enough, for a device of the same device type, the sample set (the sample subset corresponding to the device type) can be further subdivided according to each structural component under the device of the type, so that after the model is trained based on the sample set, a more targeted model can be selected for monitoring the state of the device according to different structural components of the device, and the state of the device can be determined more finely and accurately. For example, for a pump, the structural components thereof include a centrifugal pump, a magnetic pump, a canned motor pump, a gear pump, a high-speed pump, etc., and based on this, the pump start-stop sample subset can be further divided into a centrifugal pump sample subset, a magnetic pump sample subset, a canned motor pump sample subset, a gear pump sample subset, a high-speed pump sample subset, etc., according to the structural components of the centrifugal pump.

According to the method 200 for establishing the sample set, the actual vibration data of various types of equipment is obtained, the pre-training model is utilized to carry out primary labeling on the vibration data to generate the undetermined sample set comprising a plurality of undetermined samples, and then the undetermined samples in the undetermined sample set are manually corrected to generate the target sample set. Therefore, a large number of samples covering various equipment states can be generated, and the generation efficiency and accuracy of the samples can be improved by combining the pre-training model with manual correction processing. The model is trained based on more comprehensive and accurate samples, and the trained model can meet the monitoring requirements on multiple states of the equipment in the actual operation process of the equipment, so that the intelligent level of equipment state monitoring is favorably improved. Furthermore, the invention further subdivides the sample set to generate a sample subset corresponding to each equipment type, and trains the pre-training model by adopting the sample subset corresponding to the equipment type. In this way, when the device state is detected based on the trained model, the trained model corresponding to the device type can be used to judge the device state for each type of device, which is beneficial to more accurately determining the device state.

A9, the method of any one of A1-A8, wherein acquiring vibration data of one or more devices comprises: vibration data of one or more devices over a predetermined period of time is obtained, the vibration data including trend data or spectral data.

A10, the method as claimed in any one of A1-A9, wherein before obtaining vibration data of one or more devices, comprising the steps of: vibration data for one or more devices is collected by a data collection system to obtain vibration data from the data collection system.

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 computing device will generally include 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 multilingual spam-text recognition method of the present invention according to instructions in said 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 the best mode 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, and the scope of the present invention is defined by the appended claims.

Claims

1. A method of creating a sample set, executed in a computing device, comprising the steps of:

acquiring vibration data and equipment attribute information of one or more kinds of equipment;

inputting the vibration data into a pre-training model, and outputting corresponding labeled data after the vibration data is processed by the pre-training model;

generating undetermined samples based on the vibration data and the corresponding labeling data, and generating an undetermined sample set based on a plurality of undetermined samples; and

and correcting a plurality of samples to be determined in the sample set to be determined to generate a target sample set so as to train the pre-training model based on the target sample set.

2. The method of claim 1, wherein the pre-trained model is a pre-trained startup and shutdown determination model, the annotation data comprising a startup state or a shutdown state.

3. The method of claim 1, wherein after generating the target sample set, further comprising the steps of:

the target sample set is partitioned according to device attribute information to generate a sample subset corresponding to each device attribute information.

4. The method of any of claims 1-3, wherein the device attribute information includes one or more of a device type, a device name, a device identification, device vendor information, structural component information.

5. The method of any of claims 1-4, wherein partitioning the target sample set according to device attribute information comprises:

the target set of samples is partitioned according to device type to generate a subset of samples corresponding to each device type.

6. The method of claim 5, wherein partitioning the target sample set according to device attribute information further comprises:

for the devices of the same device type, dividing the sample subsets corresponding to the device type according to the structural components of the devices to generate the sample subsets corresponding to the structural components.

7. The method of any one of claims 1-6, wherein prior to inputting the vibration data into the pre-trained model, further comprising the steps of:

one or more sample sets and a tag type corresponding to each sample set are defined.

8. The method of claim 7, wherein,

the sample set comprises one or more of a false signal sample set, a start-stop sample set, an alarm sample set and a diagnosis sample set.

9. A computing device, comprising:

at least one processor; and

a memory storing program instructions, wherein the program instructions are configured to be adapted to be executed by the at least one processor, the program instructions comprising instructions for performing the 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 method of any of claims 1-8.