Device operation status identification method, device, equipment and storage medium

Abstract

The present invention discloses a method, device, equipment and storage medium for identifying the operation status of equipment. Aiming at the problems of low reliability and high misrecognition rate of the existing method for identifying the operation status of equipment by relying on manually set thresholds, a semi-supervised learning model based on K-means clustering algorithm and on-site business status annotation is created, historical data of equipment status is clustered to obtain multiple clustering states, the clustering states of the equipment are matched with the business states, the clustering centers and the distance algorithm are determined, and the recognition range during state switching is limited based on the on-site state switching of the equipment and the principle of the equipment, and the state after switching is predicted under a finite probability set according to the historical state switching data, so as to improve the accuracy of state recognition.

Description

Technical Field

[0001] The present invention belongs to the technical field of equipment monitoring, and in particular, relates to a method, device, equipment and storage medium for identifying equipment operating status. Background Art

[0002] With the rapid development of the medical industry, a variety of medical devices have emerged. Usually, a hospital has tens of thousands of medical devices. The current state recognition of large active devices (such as ultrasound machines) mainly relies on manually set thresholds. When the state of the device is frequently switched or switched quickly, it will impact the manually set threshold, resulting in missed or misidentified states. Therefore, the method of relying on manually set thresholds to identify the operating state of the device has low reliability and high misidentification rate. Summary of the invention

[0003] The purpose of the present invention is to provide a method, device, equipment and storage medium for identifying the operation status of an equipment. By creating an equipment operation status identification model based on a K-means clustering algorithm and combining it with a finite state automaton in automaton theory, the accuracy and reliability of equipment operation status identification are improved.

[0004] To solve the above problems, the technical solution of the present invention is:

[0005] A method for identifying a device operating state, comprising:

[0006] Create an unsupervised learning model based on the K-means clustering algorithm to cluster the historical data of equipment status and obtain multiple cluster states;

[0007] Collect the business status data of the equipment, define and classify the business status;

[0008] Cluster and annotate the service status data, match the cluster status with the service status of the device, determine the cluster center and distance algorithm, and obtain a semi-supervised learning model based on the K-means clustering algorithm and service status annotation;

[0009] Based on the operation data and operation principle of the equipment, combined with the finite state automaton, the migration of the equipment's business state is marked, the equipment's state migration rules are defined, and the equipment operation state recognition model is obtained;

[0010] Obtain the operating data of the device to be identified, input the operating data into the device operating status identification model, and output the operating status of the device.

[0011] According to an embodiment of the present invention, the step of creating an unsupervised learning model based on a K-means clustering algorithm further includes:

[0012] a. Obtain historical data of n device states to form a sample set {x1,x2,x3,...,x n}, randomly select k sample points from the sample set to serve as each sample cluster {c1,c2,...,c k} of the center points {μ1,μ2,...,μ k};

[0013] b. Calculate the distance between all sample points and the center of each cluster, and assign the sample points to the cluster with the closest distance;

[0014] c. Recalculate the cluster center based on the existing sample points in the cluster

[0015] d. Repeat steps b and c until the center of the cluster no longer migrates.

[0016] According to an embodiment of the present invention, the collecting of service status data of the device and defining and classifying the service status further comprises:

[0017] Collect business status data of different types of equipment, define and classify business status;

[0018] Record working hours under different business conditions;

[0019] The business status data of different types of equipment are unified in data format and stored.

[0020] According to an embodiment of the present invention, clustering and labeling the service status data and matching the cluster status with the service status of the device further includes:

[0021] Based on the low-density separation hypothesis, cluster and label the business status data;

[0022] The clustered data clusters under the theoretical state are matched with the actual business state of the device.

[0023] According to an embodiment of the present invention, based on the operation data and operation principle of the device, combined with a finite state automaton, the migration of the service state of the device is marked, and the state migration rule of the device is further defined, including:

[0024] The business status of the equipment is divided into standby status, power-on status, multiple working status and power-off status;

[0025] When the device switches from the off state to the on state, it must first switch from the off state to the standby state, and then switch from the standby state to the on state;

[0026] When the device switches from one working state to another working state, it must first switch to the standby state, and then switch from the standby state to another working state.

[0027] According to an embodiment of the present invention, the step of inputting the operation data into the device operation status recognition model and outputting the operation status of the device further includes:

[0028] Collecting status data of different devices of the same type and using the status data as prediction data;

[0029] Several time points are randomly selected to verify the equipment operation status recognition model, and the operation status output by the model is compared with the predicted data. If the accuracy of the comparison result does not meet the preset standard, the predicted data is incorporated into the training data of the model to continue training until the prediction accuracy reaches the preset standard.

[0030] A device for identifying a device operating state, comprising:

[0031] The initial model creation module is used to create an unsupervised learning model based on the K-means clustering algorithm, cluster the historical data of the equipment status, and obtain multiple clustering states;

[0032] The data collection module is used to collect the business status data of the equipment and define and classify the business status;

[0033] The clustering and annotation module is used to cluster and annotate the business status data, match the cluster status with the business status of the device, determine the cluster center and distance algorithm, and obtain a semi-supervised learning model based on the K-means clustering algorithm and business status annotation;

[0034] The state transition marking module is used to mark the migration of the business state of the equipment based on the operation data and operation principle of the equipment, define the state transition rules of the equipment, and obtain the equipment operation state identification model;

[0035] The state recognition module is used to obtain the operating data of the device in the state to be recognized, input the operating data into the device operating state recognition model, and output the operating state of the device.

[0036] According to an embodiment of the present invention, the device for identifying the operation status of a device further includes:

[0037] The model verification module is used to collect status data of different devices of the same type and use the status data as prediction data; and randomly select several time points to verify the device operation status recognition model.

[0038] A device for identifying a device operating status, comprising:

[0039] A memory and at least one processor, wherein instructions are stored in the memory, and the memory and the at least one processor are interconnected via a line;

[0040] The at least one processor calls the instructions in the memory so that the device for identifying the device operating status executes the method for identifying the device operating status in an embodiment of the present invention.

[0041] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, a method for identifying a device operating state in an embodiment of the present invention is implemented.

[0042] Due to the adoption of the above technical solution, the present invention has the following advantages and positive effects compared with the prior art:

[0043] The device operation status identification method in one embodiment of the present invention aims to solve the problems of low reliability and high misrecognition rate of the existing method for device operation status identification relying on manually set thresholds. By creating a semi-supervised learning model based on K-means clustering algorithm and business status annotation, the historical data of the device status is clustered to obtain multiple cluster states, the cluster states of the device are matched with the business states, the cluster centers and distance algorithms are determined, and the recognition range during state switching is limited based on the on-site state switching of the device and the principle of the device. The state after switching is predicted under a finite probability set based on the historical state switching data, thereby improving the accuracy of state recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 A flow chart of a method for identifying a device operating state in one embodiment of the present invention;

[0045] Figure 2 A schematic diagram of the effect of the K-means clustering algorithm in one embodiment of the present invention;

[0046] Figure 3 is a distribution diagram of average silhouette coefficients in one embodiment of the present invention;

[0047] Figure 4 is a device status data curve diagram in one embodiment of the present invention;

[0048] Figure 5 A clustering labeling diagram in one embodiment of the present invention;

[0049] Figure 6 A schematic diagram of a state transition mark in an embodiment of the present invention;

[0050] Figure 7 is a flow chart of model verification in one embodiment of the present invention;

[0051] Figure 8 A block diagram of a device operating status identification apparatus according to an embodiment of the present invention;

[0052] Fig. 9 Schematic diagram of a device operating status identification device in one embodiment of the present invention. DETAILED DESCRIPTION

[0053] The following is a further detailed description of a device operating status identification method, apparatus, device and storage medium proposed by the present invention in conjunction with the accompanying drawings and specific embodiments. The advantages and features of the present invention will become more apparent from the following description and claims.

[0054] Embodiment 1

[0055] like Figure 1 As shown, the present invention provides a method for identifying a device operating state, comprising:

[0056] S1: Create an unsupervised learning model based on the K-means clustering algorithm to cluster the historical data of the equipment status and obtain multiple cluster states;

[0057] S2: Collect the service status data of the equipment, define and classify the service status;

[0058] S3: Cluster and annotate the service status data, match the cluster status with the service status of the device, determine the cluster center and distance algorithm, and obtain a semi-supervised learning model based on the K-means clustering algorithm and service status annotation;

[0059] S4: Based on the operation data and operation principle of the equipment, combined with the finite state automaton, the migration of the business state of the equipment is marked, the state migration rules of the equipment are defined, and the equipment operation state recognition model is obtained;

[0060] S5: Acquire the operating data of the device in the state to be identified, input the operating data into the device operating state identification model, and output the operating state of the device.

[0061] Specifically, in step S1, an unsupervised learning model based on a K-means clustering algorithm is created to cluster the historical data of the device status to obtain multiple clustering states. The unsupervised learning model is a machine learning module.

[0062] Machine learning for classification problems can be divided into supervised learning, unsupervised learning, and semi-supervised learning.

[0063] Supervised learning: refers to the process of adjusting the parameters of the classifier using a set of samples of known categories to achieve the required performance, also known as supervised training or learning with a teacher. Supervised learning is the machine learning task of inferring a function from labeled training data. Supervised learning is the machine learning task of inferring a function from labeled training data. The training data includes a set of training examples. In supervised learning, each instance consists of an input object (usually a vector) and a desired output value (also called a supervisory signal). The supervised learning algorithm analyzes the training data and produces an inferred function that can be used to map out new instances. An optimal solution will allow the algorithm to correctly determine the class labels of unseen instances. This requires that the learning algorithm is formed in a "reasonable" way from training data to unseen situations.

[0064] Unsupervised learning: Solving various problems in pattern recognition based on training samples of unknown categories (not labeled) is called unsupervised learning. At present, unsupervised learning in deep learning is mainly divided into two categories. One is the deterministic autoencoding method and its improved algorithm, whose main goal is to restore the original data from the abstracted data as losslessly as possible; the other is the probabilistic restricted Boltzmann machine and its improved algorithm, whose main goal is to maximize the probability of the original data appearing when the restricted Boltzmann machine reaches a stable state.

[0065] Clustering is a typical example of unsupervised learning. The purpose of clustering is to group similar things together, and we don’t care what the category is. Therefore, a clustering algorithm usually only needs to know how to calculate similarity to start working.

[0066] Semi-supervised learning (SSL) is a key issue in the field of pattern recognition and machine learning. It is a learning method that combines supervised learning with unsupervised learning. Semi-supervised learning uses a large amount of unlabeled data and labeled data to perform pattern recognition. When using semi-supervised learning, it will require as few people as possible to do the work, while at the same time, it can bring relatively high accuracy. Therefore, semi-supervised learning is receiving more and more attention.

[0067] The main algorithm strategy of this invention is based on the semi-supervised learning idea of ​​clustering hypothesis, that is, when two samples are in the same cluster, they have the same class label with a high probability. The equivalent definition of this hypothesis is the Low Density Separation Assumption, that is, the classification decision boundary should pass through the sparse data area and avoid dividing the samples in the dense data area to both sides of the decision boundary.

[0068] Among them, the specific classification clustering algorithm used is the K-means clustering algorithm. The k in the K-means clustering algorithm represents the number of clusters, and means represents the mean of the data objects in the cluster (this mean is a description of the center of the cluster). Therefore, the K-means algorithm is also called the K-means algorithm. The K-means algorithm is a partition-based clustering algorithm that uses distance as the standard for measuring the similarity between data objects. That is, the smaller the distance between data objects, the higher their similarity, and the more likely they are in the same cluster. The K-means algorithm usually uses Euclidean distance to calculate the distance between data objects. The specific steps of the K-means algorithm are as follows:

[0069] a. Obtain historical data of n device states to form a sample set {x1,x2,x3,...,x n}, randomly select k sample points from the sample set to serve as each sample cluster {c1,c2,...,c k} of the center points {μ1,μ2,...,μ k};

[0070] b. Calculate the distance between all sample points and the center of each cluster, and assign the sample points to the cluster with the closest distance;

[0071] c. Recalculate the cluster center based on the existing sample points in the cluster

[0072] d. Repeat steps b and c until the center of the cluster no longer migrates.

[0073] The final effect diagram of the K-means clustering algorithm is as follows: Figure 2 As shown. Specifically for the present invention, the device status is divided into piles (clusters) from the perspective of data distribution through learning and training the historical data of the device. According to the actual operating state of the device, it can generally be considered as the shutdown state, standby state, power-on state and working state, among which the working state can be divided into a variety of different working states. For example, a gynecological color ultrasound imaging diagnostic device with a model number of HHH under the H0 brand of the Obstetrics and Gynecology Department of a certain hospital and a serial number of HHH, after reading the product description of the device HHH, it is expected that its device status can be roughly divided into 5 categories, namely {shutdown, standby, working-function A, working-function B, working-function C}; the current, voltage, and power data of the historical time flow of the device HHH are selected and exported; the data is processed and feature transformed to generate secondary feature data such as "maximum current per unit time" and "power difference per unit time". Based on the above data, K-means algorithm modeling is performed.

[0074] First, perform preliminary modeling according to the above steps a to d, then select other k values, such as values ​​from 2 to 8, re-model and calculate the average silhouette coefficient under different k values, that is, all sample points x i The average distance between other sample points in the same cluster. The value range of the average silhouette is [-1, 1], and the closer the distance between samples in a cluster, the farther the distance between samples between clusters, the larger the average silhouette coefficient, and the better the clustering effect. Then, the k value with a larger average silhouette coefficient is the optimal number of clusters.

[0075] After the above calculation, the distribution diagram of the average silhouette coefficient is obtained, as shown in Figure 3 As shown in the figure, the theoretical optimal value of k is 2, but for device HHH, its actual recognition state requires more than 5 states, so the k value here is 7, that is, there are 7 clusters, corresponding to 7 clustering states.

[0076] In step S2, the business status data of the equipment is collected, and the business status is defined and classified. The business status data of different types of equipment can be collected, and the business status can be defined and classified; the working time periods under different business statuses can be recorded; and the business status data of different types of equipment can be unified in data format and stored.

[0077] In this embodiment, the operating data of the device HHH is collected for one day, as shown in the following table.

[0078]

[0079]

[0080] The table lists some of the collected data. Plot a curve based on the device status. Figure 4 shown.

[0081] In step S3, the business status data is clustered and labeled, the cluster status is matched with the business status of the device, the cluster center and the distance algorithm are determined, and a semi-supervised learning model based on the K-means clustering algorithm and business status labeling is obtained.

[0082] After the equipment data is collected, the equipment data is matched with the cluster data. The cluster data clusters in the theoretical state are matched one by one to the actual equipment state on site; the matching accuracy is required to reach more than 90%, but the data that is often significantly mismatched can be traced (such as Figure 5 As shown by mark a in the figure, during a certain period of time, the collector may not be on site, resulting in the omission of a certain state. By using the cluster annotation in this embodiment, the theoretical equipment operation state will not be mistakenly matched to the "standby state" actually collected), so the matching accuracy can often reach more than 95%. Figure 5The curve in the figure is the equipment data, and the bar chart below the curve is the clustering annotation of each operating status. Figure 5 After grayscale processing, the labels of various operating states are difficult to distinguish, but in practical applications, Figure 5 It is a color image, and different operating states are represented by different colors.

[0083] After the business status data is clustered and labeled, the unsupervised learning model based on the K-means clustering algorithm in step S1 is transformed into a semi-supervised learning model.

[0084] In step S4, based on the operation data and operation principle of the equipment and in combination with a finite state automaton, the migration of the service state of the equipment is marked, the state migration rules of the equipment are defined, and a device operation state recognition model is obtained.

[0085] In terms of limiting the state migration rules of the device, it specifically means: when the device switches from the off state to the on state, it must first switch from the off state to the standby state, and then switch from the standby state to the on state; when the device switches from one working state to another, it must first switch to the standby state, and then switch from the standby state to another working state.

[0086] Specifically in this embodiment, after completing the cluster marking, the state transition marking of the device HHH is performed, and the state switching restriction is performed on the actual running state of the device HHH, such as Figure 6 As shown (C5-1 and C10-3V are ultrasound probes). The state switching path of the device is restricted by a finite state automaton to prevent errors such as the result recognition that C10-3V directly switches to the eL18-4 state, thereby improving the accuracy of matching.

[0087] After completing the state transition annotation, the equipment operation state identification model is completed. In order to illustrate the accuracy of the equipment operation state identification model, it is necessary to verify the accuracy of the equipment operation state identification model. The verification process is as follows: Figure 7 As shown in the figure, parameter training is performed with a semi-supervised learning model to obtain model A0. At this time, new data collection is performed on different devices of the same type once or more to obtain the 2nd to nth new device data. This part of the data is then used as prediction data, and several time points are randomly selected for verification. The operating status output by the model is compared with the prediction data. If the prediction accuracy meets the expectation, the model is applied. If it does not meet the expectation, the new prediction data is combined with the previous training data into new training data to continue training the model parameters to obtain A1,…Ax. The previous process is repeated until the prediction accuracy meets the expectation.

[0088] After completing the accuracy verification of the equipment operation status identification model, the state of the equipment under test can be identified and the operation status of the equipment under test can be monitored in real time. As described in step S5, the operation data of the equipment under test, including current, voltage, power, etc., is input, and the actual operation status of the equipment is output through model identification.

[0089] Embodiment 2

[0090] This embodiment provides a device for identifying the operating status of a device. Figure 8 As shown, the device for identifying the operation status of the equipment includes:

[0091] The initial model creation module 1 is used to create an unsupervised learning model based on the K-means clustering algorithm, cluster the historical data of the equipment status, and obtain multiple clustering states;

[0092] Data collection module 2, used to collect business status data of equipment, define and classify business status;

[0093] Clustering and labeling module 3, used to cluster and label the service status data, match the cluster status with the service status of the device, determine the cluster center and distance algorithm, and obtain a semi-supervised learning model based on K-means clustering algorithm and service status labeling;

[0094] The state transition marking module 4 is used to mark the migration of the service state of the device based on the operation data and operation principle of the device, define the state transition rules of the device, and obtain the device operation state identification model;

[0095] Model verification module 5 is used to collect status data of different devices of the same type and use the status data as prediction data; randomly select several time points to verify the accuracy of the device operation status recognition model;

[0096] The state identification module 6 is used to obtain the operating data of the device to be identified, input the operating data into the device operating state identification model, and output the operating state of the device.

[0097] The specific contents and implementation methods of each module in the above-mentioned device operating status identification device are as described in the first embodiment and will not be repeated here.

[0098] Embodiment 3

[0099] The above-mentioned second embodiment describes in detail the device for identifying the operation status of the device of the present invention from the perspective of modular functional entities. The following describes in detail the device for identifying the operation status of the device of the present invention from the perspective of hardware processing.

[0100] Please see Fig. 9The device operation status recognition device 500 may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 510 (for example, one or more processors) and a memory 520, and one or more storage media 530 (for example, one or more mass storage devices) storing application programs 533 or data 532. Among them, the memory 520 and the storage medium 530 can be temporary storage or permanent storage. The program stored in the storage medium 530 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the device operation status recognition device 500.

[0101] Furthermore, the processor 510 may be configured to communicate with the storage medium 530 and execute a series of instruction operations in the storage medium 530 on the device operation status recognition device 500 .

[0102] The device operation status identification device 500 may also include one or more power supplies 540, one or more wired or wireless network interfaces 550, one or more input and output interfaces 560, and / or one or more operating systems 531, such as Windows Serve, Vista, etc.

[0103] Those skilled in the art will understand that Fig. 9 The structure of the device operating status identification device shown does not constitute a limitation on the device operating status identification device, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0104] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions, which, when executed on a computer, cause the computer to execute the steps of the device operating status identification method in Embodiment 1.

[0105] If the module in the second embodiment is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of software. The computer software is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program codes.

[0106] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices and equipment can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0107] The embodiments of the present invention are described in detail above in conjunction with the accompanying drawings, but the present invention is not limited to the above embodiments. Even if various changes are made to the present invention, if these changes fall within the scope of the claims of the present invention and their equivalents, they still fall within the protection scope of the present invention.

Claims

1. A method for identifying the operation status of a device, characterized in that: include: Create an unsupervised learning model based on the K-means clustering algorithm to cluster the historical data of equipment status and obtain multiple cluster states; Collect the business status data of the equipment, define and classify the business status; Cluster and annotate the service status data, match the cluster status with the service status of the device, determine the cluster center and distance algorithm, and obtain a semi-supervised learning model based on the K-means clustering algorithm and service status annotation; Based on the operation data and operation principle of the equipment, combined with the finite state automaton, the migration of the equipment's business state is marked, the equipment's state migration rules are defined, and the equipment operation state recognition model is obtained; Obtaining the operating data of the device to be identified, inputting the operating data into the device operating status identification model, and outputting the operating status of the device; The step of creating an unsupervised learning model based on a K-means clustering algorithm further includes: a. Obtain historical data of n device states to form a sample set , randomly select k sample points in the sample set to serve as each sample cluster The center point ; b. Calculate the distance between all sample points and the center of each cluster, and assign the sample points to the cluster with the closest distance; c. Recalculate the cluster center based on the existing sample points in the cluster ; d. Repeat steps b and c until the center of the cluster no longer migrates; The clustering and labeling of the service status data and matching the cluster status with the service status of the device further includes: Based on the low-density separation hypothesis, cluster and label the business status data; Match the clustered data clusters in the theoretical state with the actual business state of the device; Based on the operation data and operation principle of the device, combined with the finite state automaton, the migration of the service state of the device is marked, and the state migration rules of the device are further defined, including: The business status of the equipment is divided into standby status, power-on status, multiple working status and power-off status; When the device switches from the off state to the on state, it must first switch from the off state to the standby state, and then switch from the standby state to the on state; When the device switches from one working state to another working state, it must first switch to the standby state, and then switch from the standby state to another working state.

2. The device operation status identification method according to claim 1, characterized in that: The collecting of the service status data of the device and defining and classifying the service status further include: Collect business status data of different types of equipment, define and classify business status; Record working hours under different business conditions; The business status data of different types of equipment are unified in data format and stored.

3. The device operation status identification method according to claim 1, characterized in that: After obtaining the equipment operation status identification model, the following further includes: Collecting status data of different devices of the same type and using the status data as prediction data; According to the preset rules, several time points are sampled to verify the equipment operation status recognition model, and the operation status output by the model is compared with the predicted data. If the accuracy of the comparison result does not meet the preset standard, the predicted data is incorporated into the training data of the model to continue training until the prediction accuracy reaches the preset standard.

4. A device for identifying the operation status of equipment, characterized in that: include: The initial model creation module is used to create an unsupervised learning model based on the K-means clustering algorithm, cluster the historical data of the equipment status, and obtain multiple clustering states; The data collection module is used to collect the business status data of the equipment and define and classify the business status; The clustering and annotation module is used to cluster and annotate the service status data, match the cluster status with the service status of the device, determine the cluster center and distance algorithm, and obtain a semi-supervised learning model based on the K-means clustering algorithm and service status annotation; The state transition marking module is used to mark the migration of the business state of the equipment based on the operation data and operation principle of the equipment, define the state transition rules of the equipment, and obtain the equipment operation state identification model; A state recognition module is used to obtain the operation data of the device to be identified, input the operation data into the device operation state recognition model, and output the operation state of the device; The step of creating an unsupervised learning model based on a K-means clustering algorithm further includes: a. Obtain historical data of n device states to form a sample set , randomly select k sample points in the sample set to serve as each sample cluster The center point ; b. Calculate the distance between all sample points and the center of each cluster, and assign the sample points to the cluster with the closest distance; c. Recalculate the cluster center based on the existing sample points in the cluster ; d. Repeat steps b and c until the center of the cluster no longer migrates; The clustering and labeling of the service status data and matching the cluster status with the service status of the device further includes: Based on the low-density separation hypothesis, cluster and label the business status data; Match the clustered data clusters in the theoretical state with the actual business state of the device; Based on the operation data and operation principle of the device, combined with the finite state automaton, the migration of the service state of the device is marked, and the state migration rules of the device are further defined, including: The business status of the equipment is divided into standby status, power-on status, multiple working status and power-off status; When the device switches from the off state to the on state, it must first switch from the off state to the standby state, and then switch from the standby state to the on state; When the device switches from one working state to another working state, it must first switch to the standby state, and then switch from the standby state to another working state.

5. The device for identifying the operation status of equipment according to claim 4, characterized in that: Also includes: A model verification module is used to collect status data of different devices of the same type and use the status data as prediction data; Several time points are sampled according to preset rules to verify the equipment operation status identification model.

6. A device for identifying the operation status of a device, characterized in that: include: A memory and at least one processor, wherein instructions are stored in the memory, and the memory and the at least one processor are interconnected via a line; The at least one processor calls the instruction in the memory so that the device operating status identification device executes the device operating status identification method as described in any one of claims 1-3.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for identifying the operating status of a device as described in any one of claims 1 to 3 is implemented.

Images