Intelligent equipment full-life-cycle management and diagnosis system

Abstract

The invention discloses a smart device full life cycle management diagnosis system, which belongs to the technical field of device fault diagnosis, and comprises a device operation monitoring module for monitoring a device and acquiring operation data of the device in a device operation process, a cycle fault detection module for establishing a fault judgment model and performing cycle detection on the operation data of the device, and a fault diagnosis module for performing fault diagnosis on the operation data of the device. And the fault analysis and diagnosis module analyzes and diagnoses the operation data, judges the type of the equipment fault and gives a fault diagnosis suggestion, and the predictive warning notification module sends warning information to a manager when the equipment has a fault, so that the fault diagnosis of the mechanical equipment can be realized, the diagnosis cost is saved, and the fault diagnosis efficiency is improved. And equipment state maintenance, predictive maintenance and full-life-cycle equipment management are realized.

Description

Technical Field

[0001] The present application belongs to the technical field of equipment fault diagnosis, and specifically relates to a full life cycle management and diagnosis system for smart equipment. Background Art

[0002] With the rapid development of industrial Internet, big data and artificial intelligence technologies, intelligent maintenance management has gradually become a new direction in mechanical equipment management. It uses big data analysis and artificial intelligence algorithms to analyze and process the equipment's operating data, timely detect and warn of equipment failures, and provide scientific decision-making support for equipment maintenance and repair through intelligent diagnosis and prediction technologies.

[0003] Currently, most fault diagnosis methods for mechanical equipment only target a certain type of fault. There is no universal intelligent diagnostic method, and there is a lack of real-time monitoring and analysis of the actual operating status of the equipment, which can easily lead to over-maintenance or under-maintenance, thereby increasing the operating cost and failure risk of the equipment. Although condition-based maintenance has improved the pertinence and effectiveness of maintenance to a certain extent, its maintenance decisions still mainly rely on experience judgment and simple rule analysis. It is difficult to fully utilize the deep-level information in the equipment operation data, resulting in insufficient accuracy and timeliness of fault diagnosis, and the inability to achieve predictive maintenance and full life cycle equipment management. Summary of the Invention

[0004] To address the above-mentioned problems and technical deficiencies, this application adopts the following technical solution: a smart device full lifecycle management and diagnosis system, comprising: The equipment operation monitoring module is used to monitor the equipment and obtain the equipment operation data during the equipment operation process; The periodic fault detection module is used to establish a fault judgment model, perform periodic detection on the operation data of the equipment, and determine whether the operation data of the equipment is fault data; The fault analysis and diagnosis module is used to analyze and diagnose the operating data of the equipment when it is fault data, determine the type of equipment fault, and provide fault diagnosis suggestions; The predictive warning notification module is used to send warning information to management personnel when a device fails. The warning information includes: fault data, fault category and fault diagnosis suggestions.

[0005] Preferably, it also includes a data storage module, which is used to establish a historical database and an expert fault library. The historical database stores historical operation data of the equipment, and the expert fault library stores historical fault cases and expert diagnostic suggestions for each fault case.

[0006] Furthermore, the process of determining whether the operating data of the device is fault data is as follows: Acquire historical operation data corresponding to the equipment model from a historical database, select normal operation data from the historical operation data, convert the normal operation data, and obtain a normal operation curve graph; Converting the operating data of the equipment to obtain a target operating curve, and comparing the target operating curve with the normal operating curve; The comparison result is judged to determine whether the target operation curve graph is fault data.

[0007] Furthermore, the process of comparing the target operation curve graph with the normal operation curve graph is as follows: According to the equipment operation status, the target operation curve graph and the normal operation curve graph are time-aligned, and multiple detection points are evenly set on the time coordinate axis; A detection comparison window is preset, and the target operation curve graph and the normal operation curve graph are passed through the detection comparison window; Whenever a detection point enters the detection comparison window, a comparison is started between the target operation curve and the part of the normal operation curve entering the detection comparison window. When the detection point leaves the detection comparison window, the comparison ends, and a total of multiple comparison results matching the detection point are obtained.

[0008] Preferably, the process of judging the comparison result is as follows: First, a first ratio threshold and a second ratio threshold are preset. Each time the comparison result is abnormal, a first assignment value is obtained. Each time the comparison result is normal, a second assignment value is obtained. When the proportion of the first assigned value in all assigned values is greater than or equal to a first proportion threshold, determining that the target operation curve graph is fault operation data; When the proportion of the second assigned value in all assigned values is greater than or equal to a second proportion threshold, it is determined that the target operation curve graph is operating normally; Otherwise, the target operation curve is judged to be suspicious operation data, and the suspicious operation data is manually judged to be faulty operation data.

[0009] Furthermore, data cleaning is required before converting the historical operating data and the equipment operating data, and abnormal vibration signals are automatically filtered using signal processing and machine learning algorithms.

[0010] Preferably, the fault analysis and diagnosis module is provided with an AI intelligent diagnosis model, which is trained through historical operation data, historical fault cases and expert diagnosis suggestions, and performs AI intelligent diagnosis on the fault data. Based on the vector machine principle, it combines expert systems, artificial neural networks and fuzzy logic to generate fault diagnosis suggestions. The fault diagnosis suggestions include key monitoring values, waveform diagrams, spectrum diagrams, diagnostic conclusions and hazard suggestions.

[0011] Preferably, the predictive warning notification module needs to perform data reconstruction processing when sending fault data, input the fault data to be reconstructed into the autoencoder, use the encoder to convert the fault data into a hidden state vector, compress it into a low-dimensional space, and then use the decoder to convert the low-dimensional vector into high-dimensional reconstructed data.

[0012] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the contents of the smart device full life cycle management and diagnostic system as described above are implemented.

[0013] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the contents of the smart device full life cycle management and diagnosis system as described above.

[0014] Compared with the prior art, the present invention has the following advantages: This application monitors the equipment, obtains the equipment's operating data during the equipment's operation, establishes a fault judgment model to perform periodic detection on the equipment's operating data, analyzes and diagnoses the operating data, determines the type of equipment fault, and gives fault diagnosis suggestions. The equipment is controlled according to the fault diagnosis suggestions. By diagnosing faults in mechanical equipment, diagnostic costs are saved, and equipment status inspection, predictive maintenance, and equipment management throughout the entire life cycle are achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In the attached figure: Figure 1 A schematic diagram of the system structure of an embodiment of the present application; Figure 2 This is a flowchart of abnormal data cleaning in an embodiment of the present application; Figure 3 This is a diagram of the speed change steps of an embodiment of the present application; Figure 4 This is a schematic diagram of principal component analysis of an embodiment of the present application; Figure 5 This is a fault diagnosis flowchart of an embodiment of the present application; Figure 6 This is a schematic diagram of service life prediction for an embodiment of the present application; Figure 7 This is a schematic diagram of the device structure of an embodiment of the present application. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Generally, the components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations. Example

[0017] like Figure 1 As shown, a smart device full life cycle management and diagnosis system includes: The equipment operation monitoring module is used to monitor the equipment and obtain the equipment operation data during the equipment operation process; The periodic fault detection module is used to establish a fault judgment model, perform periodic detection on the operation data of the equipment, and determine whether the operation data of the equipment is fault data; It also includes a data storage module for establishing a historical database and an expert fault database. The historical database stores historical operating data of the equipment, and the expert fault database stores historical fault cases and expert diagnostic suggestions for each fault case.

[0018] The process of determining whether the equipment's operating data is fault data is as follows: Acquire historical operation data corresponding to the equipment model from a historical database, select normal operation data from the historical operation data, convert the normal operation data, and obtain a normal operation curve graph; Converting the operating data of the equipment to obtain a target operating curve, and comparing the target operating curve with the normal operating curve; The comparison result is judged to determine whether the target operation curve graph is fault data.

[0019] The process of comparing the target operating curve graph with the normal operating curve graph is as follows: According to the equipment operation status, the target operation curve graph and the normal operation curve graph are time-aligned, and multiple detection points are evenly set on the time coordinate axis; A detection comparison window is preset, and the target operation curve graph and the normal operation curve graph are passed through the detection comparison window; Whenever a detection point enters the detection comparison window, a comparison is started between the target operation curve and the part of the normal operation curve entering the detection comparison window. When the detection point leaves the detection comparison window, the comparison ends, and a total of multiple comparison results matching the detection point are obtained.

[0020] The process of judging the comparison results is as follows: First, a first ratio threshold and a second ratio threshold are preset. Each time the comparison result is abnormal, a first assignment value is obtained. Each time the comparison result is normal, a second assignment value is obtained. When the proportion of the first assigned value in all assigned values is greater than or equal to a first proportion threshold, determining that the target operation curve graph is fault operation data; When the proportion of the second assigned value in all assigned values is greater than or equal to a second proportion threshold, it is determined that the target operation curve graph is operating normally; Otherwise, the target operation curve is judged to be suspicious operation data, and the suspicious operation data is manually judged to be faulty operation data.

[0021] Data cleaning is required before converting historical operating data and equipment operating data, and abnormal vibration signals are automatically filtered using signal processing and machine learning algorithms.

[0022] The fault analysis and diagnosis module is used to analyze and diagnose the operating data of the equipment when it is fault data, determine the type of equipment fault, and provide fault diagnosis suggestions; The fault analysis and diagnosis module is equipped with an AI intelligent diagnosis model, which is trained through historical operation data, historical fault cases and expert diagnosis suggestions. It performs AI intelligent diagnosis on fault data and generates fault diagnosis suggestions based on the principle of vector machines, combined with expert systems, artificial neural networks and fuzzy logic. The fault diagnosis suggestions include key monitoring values, waveform diagrams, spectrum diagrams, diagnostic conclusions and hazard suggestions.

[0023] Predictive warning notification module, used to send warning information to management personnel when equipment fails. The warning information includes: fault data, fault category and fault diagnosis suggestions; When the predictive warning notification module sends fault data, data reconstruction processing is required. The fault data to be reconstructed is input into the autoencoder, and the encoder is used to convert the fault data into a hidden state vector, which is compressed into a low-dimensional space. The decoder is then used to convert the low-dimensional vector into high-dimensional reconstructed data. Example

[0024] Abnormal vibration data usually contains impact signals. The amplitude of the impact signal is much larger than the normal working condition signal. The peak-to-peak value (the difference between the maximum and minimum values) can be used as the threshold to determine whether it is an abnormal sample. However, some faults such as imbalance (1 times the frequency) and misalignment (2 times the frequency) will also produce large amplitudes. Therefore, directly using the peak-to-peak value to determine whether it is an abnormal sample is likely to exclude common fault signals. Therefore, there are many problems in the actual application of this method.

[0025] In order to reduce false alarms, it is necessary to identify and process abnormal points in the data. Data abnormal point removal (data cleaning) is selected as a data preprocessing method to identify erroneous data in the collected vibration data, to ensure the correctness of the data before use as much as possible, and to avoid its interference with the real fault signal, so as to improve the accuracy of early warning. The impact signal is divided into periodic impact and non-periodic impact caused by early faults. The characteristic of abnormal signal is that it contains non-periodic impact. According to this characteristic, the signal is sliced and processed, and the correlation of each signal spectrum is compared. This method can identify non-periodic signals, filter out the change signals caused by non-equipment faults, and reduce false alarm interference. The abnormal data cleaning process is described as follows: Figure 2 As shown: Example

[0026] Perform non-stationary analysis of vibration signals under variable speed (load) conditions, and display the signal frequency distribution diagram through time-frequency domain characteristics (wavelet transform spectrum, Hilbert-Huang transform spectrum, Wieger-Verein distribution spectrum, etc.).

[0027] During the operation of the equipment, the running speed is not a stable variable, which ultimately causes its vibration performance to have non-stationary characteristics.

[0028] Wavelet packet is a global decomposition method based on wavelet decomposition. The signal is decomposed into approximate signal and detail signal. The two signals are further decomposed at the next decomposition level. After N layers of wavelet packet decomposition, the signal is decomposed into 2 N The energy distribution of different frequency bands can represent the information characteristic pattern in independent orthogonal frequency bands. There are usually three steps, such as Figure 3 shown. Example

[0029] Perform statistics (including principal component analysis, linear discriminant analysis, manifold learning, etc.) and display of fault vibration characteristic signals to facilitate graphical display of the characteristics of different faults.

[0030] Traditional rolling bearing intelligent fault diagnosis based on statistical pattern recognition requires calculating features from the raw rolling bearing time-domain vibration signal. This is computationally intensive and requires specialized signal processing knowledge to extract features. Rotating equipment has complex structures and a wide variety of fault types, leading to a corresponding increase in the number of fault states and features.

[0031] In the actual early warning process, in order to make the early warning accurate and reliable, it is always necessary to collect as much fault sample data as possible to obtain sufficient fault information. While a large amount of data provides usable information, it also increases the difficulty of effectively utilizing this data. Useful knowledge will be submerged in redundant data, which increases the difficulty of feature extraction.

[0032] Through the function of feature selection, effective features are extracted and invalid features are eliminated to achieve the purpose of reducing the data dimension. Feature selection adopts the method of linear transformation to achieve the purpose of dimensionality reduction through linear combination of features, such as principal component analysis. Principal component analysis is a data compression and information extraction technology based on multivariate statistical analysis. It reduces the dimension of the original data space by constructing a new set of variables, and then extracts the main change information from the new mapping space, extracts statistical feature vectors, and reflects the data characteristics of the original data space.

[0033] The basic idea of principal component analysis is to find the optimal projection subspace of high-dimensional data in the sense of minimizing the reconstruction error in low-dimensional space, so as to characterize the global information of the sample with a smaller number of eigenvectors, and ultimately achieve the purpose of reducing the dimension of the original space.

[0034] Assume the sample matrix is ( samples, features), firstly the sample matrix Mean centering, PCA hopes to find a direction , so that the sample is projected into the direction The upper difference is the largest (main direction), and the optimization goal is: ; Before extraction The load matrix is constructed based on the principal directions , then the result after extracting the original sample features is: ,like Figure 4 shown. Example

[0035] Using learning models such as expert systems, neural networks, support vector machines, Bayesian learning, and deep learning (convolutional neural networks, long short-term memory networks, deep belief networks, etc.), diagnostic conclusions are given to guide equipment maintenance.

[0036] The AI diagnostic model applies artificial intelligence algorithms such as classification and clustering to the fault diagnosis of rotating equipment, conducts knowledge mining on equipment fault information, obtains diagnostic rules related to faults, and then identifies the equipment fault status in order to formulate maintenance strategies. The science of using big data to identify the health status of mechanical equipment is based on the acquisition of data by sensor systems, the accumulation of experience and knowledge by machine learning, and the purpose of intelligently judging the health status of equipment to ensure the reliability of mechanical equipment operation. It mainly uses machine learning algorithms such as support vector machines, logistic regression, neural networks, Bayesian networks, and K-nearest neighbors to establish fault diagnosis models. The process is as follows: Figure 5 shown.

[0037] Support vector machine is a commonly used qualitative analysis algorithm. It is a general machine learning algorithm developed by Vapnik et al. based on statistical learning theory and the principle of structural risk minimization (SRM). Based on limited sample information, it seeks the best compromise between model complexity and learning ability in order to achieve the best generalization ability. Its basic principle is as follows: Assume that the training sample set ,in is the i-th sample, yes The label is the category information corresponding to the i-th sample. In the binary classification problem, it is recorded as If the training samples are linearly separable or approximately linearly separable, then an optimal hyperplane can be found to correctly separate the two types of samples and maximize the distance between the two types of samples and the hyperplane. Let the optimal hyperplane be: ; The hyperplane correctly separates the two types of samples, which means: ; in is the slack variable introduced, which is: ; That is, solving optimization problems; ; in Is a penalty factor that controls the degree of penalty for misclassified samples. Introducing Lagrange multiplier , transforming the above optimization problem into its dual form (quadratic programming problem): ; The optimal classification surface formula is as follows: ; in, is the optimal solution to the quadratic programming problem, For bias.

[0038] For nonlinear classification problems, support vector machines use nonlinear transformations to map linearly inseparable data in the input space to a high-dimensional feature space. By constructing a linear discriminant function in the high-dimensional space, nonlinear classification in the original space is achieved. The inner product in the high-dimensional space can be represented by a kernel function. Common kernel functions include RBF kernels, polynomial kernels, and Sigmoid kernels. The corresponding optimization objective function is as follows: ; Accordingly, the optimal classification face change formula is as follows: ; Example

[0039] By acquiring key operating data curves for rotor bearings, a baseline value for rotor bearing life is established. Using AI technology (machine learning / deep learning), a life prediction model is established to determine the service life of rotor bearings. This allows for timely replacement before rotor bearing failure, minimizing economic losses and safety hazards.

[0040] The prediction of rotor bearing degradation trend is to obtain key data curves through product testing or normal operation, find and establish equipment life trend, learn the degradation trend based on the change of indicator data of each measuring point over time, establish a life prediction model, and give early warning information, such as Figure 6 shown.

[0041] The training dataset consists of time series formed by vibration data from multiple different measuring points of rotating equipment. Each time series corresponds to different degrees of wear and degradation of the components. At the same time, the data is contaminated by sensor noise. Variables with a variance of 0 in the data (which have no effect on the prediction results) are removed, default values are filled in, and maximum and minimum data are normalized.

[0042] The goal of regression is to obtain a model f(x) that can best fit the training set samples. The commonly used method is to construct a loss function between the sample label and the model prediction value, and minimize the loss function to determine the model f(x). Linear regression is a commonly used statistical method that linearly models the relationship between independent variables and dependent variables. Assume that the rotating equipment vibration data matrix is ( samples, feature), which Sample data categories ( ), construct the remaining life matrix of each rotating equipment For the samples, if they belong to Class, then ,otherwise , construct the following optimization formula: ; but For predicted vibration data ,calculate .but correspond remaining life. Example

[0043] To address the specific conditions of the power plant, a combination of wired and wireless measurement points was employed to add 68 monitoring points. Furthermore, 30 sets of mobile intelligent inspection equipment were deployed using 5G IoT cards and APN (Access Point Name) technology. This system facilitates intelligent inspection and fault diagnosis and perception, enabling interoperability of underlying data between intelligent inspection and monitoring. The system platform automatically analyzes and processes collected signals to generate unit diagnostic reports and maintenance recommendations, providing decision support for operations and maintenance, and digitally archiving inspection data. Example

[0044] like Figure 7 As shown, from a hardware perspective, the present application provides an embodiment of an electronic device that implements all or part of the contents of a smart device full life cycle management and diagnostic system, wherein the electronic device includes a service processor and a distributed memory, wherein the service processor is connected to the memory, wherein the distributed memory stores a service self-management program configured to store machine-readable instructions, and the service processor executes the service self-management program, and when the instructions are executed by the processor, the smart device full life cycle management and diagnostic system as described above is implemented.

[0045] From a hardware perspective, in order to effectively improve the flexibility, versatility, and efficiency of data collection, this application provides embodiments of electronic devices that implement all or part of the smart device lifecycle management and diagnostic system. The electronic devices specifically include the following: A processor, a memory, a communications interface, and a bus; wherein the processor, the memory, and the communications interface communicate with each other via the bus; the communications interface is used to realize information transmission between a data acquisition device based on a distributed model and related equipment such as a core business system, a user terminal, and related databases; the logic controller can be a desktop computer, a tablet computer, a mobile terminal, etc., but this embodiment is not limited thereto.

[0046] In this embodiment, the logic controller can be implemented with reference to the embodiment of the smart device full life cycle management and diagnosis system in the embodiment, as well as the embodiment of the data acquisition device based on the distributed model. Their contents are merged here and the repeated parts are not repeated.

[0047] It is understandable that the user terminal may include a smart phone, a tablet electronic device, a network set-top box, a portable computer, a desktop computer, a personal digital assistant (PDA), a vehicle-mounted device, a smart wearable device, etc. Among them, the smart wearable device may include smart glasses, a smart watch, a smart bracelet, etc.

[0048] In actual applications, the above-described electronic device side execution can also be completed by the client device. The specific selection can be based on the processing capabilities of the client device and the limitations of the user's usage scenario, and this application does not limit this. If all operations are completed in the client device, the client device may also include a processor.

[0049] The above-mentioned client device may have a communication module (i.e., a communication unit), which can communicate with a remote server to realize data transmission with the server. The server may include a server on the task scheduling center side, and other implementation scenarios may also include a server on an intermediate platform, such as a server on a third-party server platform that has a communication link with the task scheduling center server. The server may include a single computer device, or a server cluster consisting of multiple servers, or a server structure of a distributed device. Example

[0050] The embodiments of the present application also provide a computer-readable storage medium that can implement all the contents of the smart device full life cycle management and diagnostic system in the above-mentioned embodiments, in which the execution subject is a server or a client. The computer-readable storage medium stores a computer program, which, when executed by a processor, implements all the contents of the smart device full life cycle management and diagnostic system in the above-mentioned embodiments, in which the execution subject is a server or a client.

[0051] The embodiments of the present application may be provided as methods, apparatuses, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects.

[0052] Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0053] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (apparatus), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0054] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0055] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.

[0056] The above-described embodiments merely represent preferred implementations of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person skilled in the art would be able to make numerous variations, improvements, and substitutions without departing from the spirit of the present application, all of which fall within the scope of protection of the present application.

Claims

1. A smart device full life cycle management and diagnosis system, characterized by: include: The equipment operation monitoring module is used to monitor the equipment and obtain the equipment operation data during the equipment operation process; The periodic fault detection module is used to establish a fault judgment model, perform periodic detection on the operation data of the equipment, and determine whether the operation data of the equipment is fault data; The fault analysis and diagnosis module is used to analyze and diagnose the operating data of the equipment when it is fault data, determine the type of equipment fault, and provide fault diagnosis suggestions; The predictive warning notification module is used to send warning information to management personnel when a device fails. The warning information includes: fault data, fault category and fault diagnosis suggestions.

2. A smart device full life cycle management and diagnosis system according to claim 1, characterized in that: It also includes a data storage module, which is used to establish a historical database and an expert fault library. The historical database stores historical operating data of the equipment, and the expert fault library stores historical fault cases and expert diagnostic suggestions for each fault case.

3. A smart device full life cycle management and diagnosis system according to claim 2, characterized in that: The process of determining whether the operating data of the device is fault data is as follows: Acquire historical operation data corresponding to the equipment model from a historical database, select normal operation data from the historical operation data, convert the normal operation data, and obtain a normal operation curve graph; Converting the operating data of the equipment to obtain a target operating curve, and comparing the target operating curve with the normal operating curve; The comparison result is judged to determine whether the target operation curve graph is fault data.

4. A smart device full life cycle management and diagnosis system according to claim 3, characterized in that: The process of comparing the target operation curve graph with the normal operation curve graph is as follows: According to the equipment operation status, the target operation curve graph and the normal operation curve graph are time-aligned, and multiple detection points are evenly set on the time coordinate axis; A detection comparison window is preset, and the target operation curve graph and the normal operation curve graph are passed through the detection comparison window; Whenever a detection point enters the detection comparison window, a comparison is started between the target operation curve and the part of the normal operation curve entering the detection comparison window. When the detection point leaves the detection comparison window, the comparison ends, and a total of multiple comparison results matching the detection point are obtained.

5. A smart device full life cycle management and diagnosis system according to claim 4, characterized in that: The process of judging the comparison result is as follows: First, a first ratio threshold and a second ratio threshold are preset. Each time the comparison result is abnormal, a first assignment value is obtained. Each time the comparison result is normal, a second assignment value is obtained. When the proportion of the first assigned value in all assigned values is greater than or equal to a first proportion threshold, determining that the target operation curve graph is fault operation data; When the proportion of the second assigned value in all assigned values is greater than or equal to a second proportion threshold, it is determined that the target operation curve graph is operating normally; Otherwise, the target operation curve is judged to be suspicious operation data, and the suspicious operation data is manually judged to be faulty operation data.

6. A smart device full life cycle management and diagnosis system according to claim 3, characterized in that: Before converting the historical operating data and the equipment operating data, data cleaning is required, and abnormal vibration signals are automatically filtered using signal processing and machine learning algorithms.

7. The smart device full life cycle management and diagnosis system according to claim 1, characterized in that: The fault analysis and diagnosis module is equipped with an AI intelligent diagnosis model, which is trained through historical operation data, historical fault cases and expert diagnosis suggestions, and performs AI intelligent diagnosis on fault data. Based on the principle of vector machines, it combines expert systems, artificial neural networks and fuzzy logic to generate fault diagnosis suggestions. The fault diagnosis suggestions include key monitoring values, waveform diagrams, spectrum diagrams, diagnostic conclusions and hazard suggestions.

8. The smart device full life cycle management and diagnosis system according to claim 1, characterized in that: When the predictive warning notification module sends fault data, data reconstruction processing is required. The fault data to be reconstructed is input into the autoencoder, and the encoder is used to convert the fault data into a hidden state vector, which is compressed into a low-dimensional space. Then, the decoder is used to convert the low-dimensional vector into high-dimensional reconstructed data.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the content of the smart device full life cycle management and diagnosis system described in claim 1 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the content of the smart device full life cycle management and diagnosis system described in claim 1 is implemented.

Images