Intelligent Factory Operation and Maintenance Management System Driven by Multi-Dimensional Data

Abstract

The present invention discloses an intelligent plant operation and maintenance management system based on multi-dimensional data-driven, which relates to the technical field of intelligent operation and maintenance management. It includes a knowledge graph construction module that constructs a plant operation and maintenance knowledge graph in the RDF standard format based on plant operation and maintenance characteristics; a model construction module that constructs an intelligent operation and maintenance prediction model based on a deep reinforcement learning model and a long short-term memory network model; a fault risk value prediction module that obtains a fault risk prediction value through the intelligent operation and maintenance prediction model based on the plant operation and maintenance knowledge graph; and a risk assessment and early warning module that generates an early warning notice and formulates maintenance measures based on the fault risk prediction value. By integrating multi-dimensional data collection and preprocessing, intelligent feature extraction, knowledge graph construction, and an intelligent operation and maintenance prediction model that combines deep reinforcement learning and long short-term memory network, the present invention significantly improves the efficiency and reliability of factory operation.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent operation and maintenance management, in particular to an intelligent plant operation and maintenance management system driven by multi-dimensional data. Background Art

[0002] With the continuous advancement of the industrialization process, plant operation and maintenance, as an important part of ensuring the efficient operation of enterprises, has received increasing attention. Plant operation and maintenance management involves multiple links such as equipment installation, commissioning, operation, maintenance, and overhaul. Its goal is to ensure the long-term efficient and stable operation of equipment and avoid economic losses caused by downtime due to failures. In recent years, with the continuous development of informatization and digital technologies, intelligent plant operation and maintenance management driven by multi-dimensional data has gradually become a key means to improve operation and maintenance efficiency and prediction accuracy. The application of advanced technologies such as big data technology, Internet of Things technology, and artificial intelligence algorithms has gradually shifted plant operation and maintenance management from traditional manual experience-based to data-based intelligent decision-making. In particular, the introduction of technologies such as deep learning, natural language processing, and principal component analysis (PCA) not only enhances the analysis ability of multi-dimensional plant operation and maintenance data but also provides strong support for equipment fault prediction and maintenance strategy optimization.

[0003] However, although many current intelligent plant operation and maintenance management systems have made progress in prediction accuracy and fault warning through data collection and analysis means, there are still many limitations in the existing technologies. First, traditional operation and maintenance management systems mostly rely on a single data source, lacking effective integration and intelligent analysis of multi-dimensional data, and it is difficult to comprehensively reflect the equipment operation status and potential risks. Second, although existing intelligent prediction models can make predictions based on historical data, they often ignore the in-depth correlation analysis of multi-source heterogeneous data. Especially in dynamic behavior prediction and complex relationship modeling, it is difficult to handle the complex interactions and long-term dependencies between system states and behaviors during the plant operation and maintenance process. This results in insufficient accuracy and real-time performance of current prediction models and cannot fully meet the requirements of the increasingly complex industrial environment. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the intelligent plant operation and maintenance management system driven by multi-dimensional data provided by the present invention solves the problem of the deficiency of traditional operation and maintenance management systems in multi-dimensional data fusion and in-depth correlation analysis.

[0006] To solve the above technical problems, the present invention provides the following technical solutions:

[0007] The present invention provides an intelligent plant operation and maintenance management system based on multi-dimensional data-driven, which includes a data collection and preprocessing module for collecting multi-dimensional plant operation and maintenance data and preprocessing the multi-dimensional plant operation and maintenance data; a feature extraction module for extracting features through natural language processing technology and principal component analysis algorithm based on the preprocessed multi-dimensional plant operation and maintenance data to obtain plant operation and maintenance features; a knowledge graph construction module for constructing a plant operation and maintenance knowledge graph based on the plant operation and maintenance features through the RDF standard format; a model construction module for constructing an intelligent operation and maintenance prediction model based on a deep reinforcement learning model and a long short-term memory network model; a fault risk value prediction module for obtaining a fault risk prediction value through the intelligent operation and maintenance prediction model based on the plant operation and maintenance knowledge graph; and a risk assessment and early warning module for generating an early warning notice and formulating maintenance measures based on the fault risk prediction value.

[0008] As a preferred solution of the intelligent plant operation and maintenance management system based on multi-dimensional data-driven described in the present invention, wherein: the multi-dimensional plant operation and maintenance data includes vibration frequency, power consumption, operation logs, and fault reports;

[0009] The preprocessing of the multi-dimensional plant operation and maintenance data includes preliminary cleaning, format conversion, and standardization processing.

[0010] As a preferred solution of the intelligent plant operation and maintenance management system based on multi-dimensional data-driven described in the present invention, wherein: the steps of extracting plant operation and maintenance features through natural language processing technology and principal component analysis algorithm based on the preprocessed multi-dimensional plant operation and maintenance data are as follows:

[0011] Based on the preprocessed multi-dimensional plant operation and maintenance data, extract plant text data and plant numerical data through a data parsing method;

[0012] Convert the plant text data into a word vector form through a word embedding method;

[0013] Based on the plant text data in word vector form, aggregate through an aggregation method to obtain a global context feature vector of the plant text data;

[0014] Based on the plant numerical data, analyze through a statistical method to obtain a covariance matrix;

[0015] Perform eigenvalue decomposition on the covariance matrix through an eigenvalue decomposition method to extract the principal component eigenvalues and corresponding principal component eigenvectors in the plant numerical data;

[0016] Based on the extracted principal component eigenvalues sorted by size, use the principal component eigenvector corresponding to the largest principal component eigenvalue as the feature principal component;

[0017] Map the factory operation numerical data to the feature principal components to obtain the factory operation features after dimensionality reduction;

[0018] Based on the global context feature vector of the factory operation text data and the factory operation features after dimensionality reduction, connect them in the vector order through the splicing method to obtain the factory operation and maintenance features.

[0019] As a preferred solution of the intelligent factory operation and maintenance management system based on multi-dimensional data-driven of the present invention, wherein: based on the factory operation and maintenance features, construct a factory operation and maintenance knowledge graph through the RDF standard format. The specific steps are as follows,

[0020] Based on the factory operation and maintenance features, define the factory operation and maintenance entity types, factory operation and maintenance entity attributes, and factory operation and maintenance entity relationships of the knowledge graph through feature parsing and semantic abstraction methods to obtain the semantic structure of the factory operation and maintenance knowledge graph;

[0021] Based on the factory operation text data and factory operation numerical data in the multi-dimensional factory operation and maintenance data, map them to the attribute values of the factory operation and maintenance entities through natural language processing methods and data analysis methods;

[0022] Based on the semantic structure of the factory operation and maintenance knowledge graph and the attribute values of the factory operation and maintenance entities, construct the triple of the factory operation and maintenance knowledge graph through the RDF standard format;

[0023] Based on the triple of the factory operation and maintenance knowledge graph, organize it into an RDF graph structure through GraphDB to obtain the factory operation and maintenance knowledge graph.

[0024] As a preferred solution of the intelligent factory operation and maintenance management system based on multi-dimensional data-driven of the present invention, wherein: based on the deep reinforcement learning model and the long short-term memory network model, construct an intelligent operation and maintenance prediction model. The specific steps are as follows,

[0025] Through the graph embedding method and the time series annotation method, convert the historical factory operation and maintenance knowledge graph into the time series data format;

[0026] Use the long short-term memory network model as the basic model;

[0027] The input layer of the LSTM model receives the historical factory operation and maintenance knowledge graph converted into the time series data format;

[0028] The LSTM layer selectively remembers and forgets information through the memory unit and the gate mechanism, and captures the long-term dependencies in the time series;

[0029] The fully connected layer extracts features and reduces dimensions based on the output of the LSTM layer, and maps them to the output layer;

[0030] The output layer outputs the preliminary prediction value;

[0031] The deep reinforcement learning layer learns through interaction with the environment, optimizes the decision-making process using the reward function, adjusts the parameters of the deep reinforcement learning model, and optimizes the output results of the LSTM model;

[0032] The deep reinforcement learning output layer outputs the final predicted value;

[0033] A preliminary intelligent operation and maintenance prediction model is constructed;

[0034] Through the backpropagation algorithm and the reinforcement learning algorithm, the preliminary intelligent operation and maintenance prediction model is trained, evaluated, and optimized, and finally an intelligent operation and maintenance prediction model is constructed.

[0035] As a preferred solution of the intelligent plant operation and maintenance management system based on multi-dimensional data-driven of the present invention, wherein: through the backpropagation algorithm and the reinforcement learning algorithm, the preliminary intelligent operation and maintenance prediction model is trained, evaluated, and optimized, and finally an intelligent operation and maintenance prediction model is constructed. The specific steps are as follows.

[0036] By the random division method, the historical plant operation and maintenance knowledge graph converted into the time series data format is divided into a training set and a validation set;

[0037] Based on the training set, the prediction performance of the LSTM model is evaluated using the minimized loss function, and the weights and parameters of the LSTM model are optimized through the backpropagation algorithm;

[0038] Based on the predicted values output by the LSTM model, the deep reinforcement learning model uses the policy gradient method to adjust the decision-making process according to the reward signal, adjusts the output of the LSTM model, and optimizes the fault risk prediction accuracy;

[0039] Using the validation set, the prediction performance of the preliminary intelligent operation and maintenance prediction model is evaluated by analyzing the accuracy rate, recall rate, and F1-score, and the reward function, the parameters of the LSTM layer, and the reinforcement learning strategy are fine-tuned based on the evaluation results;

[0040] The k-fold cross-validation method is used to validate the preliminarily trained intelligent operation and maintenance prediction model, and finally the intelligent operation and maintenance prediction model is output.

[0041] As a preferred solution of the intelligent plant operation and maintenance management system based on multi-dimensional data-driven of the present invention, wherein: based on the plant operation and maintenance knowledge graph, through the intelligent operation and maintenance prediction model, the fault risk prediction value is obtained. The specific steps are as follows.

[0042] Based on the plant operation and maintenance knowledge graph converted into the time series data format, the long short-term memory network model is used to process the time series data, and combined with the deep reinforcement learning model based on the plant operation and maintenance knowledge graph, the complex dynamic relationships in the plant operation and maintenance are captured;

[0043] Based on complex dynamic relationships, a multi-dimensional data-driven approach is used to obtain the predicted values of fault risks at each time point, and the expression is:

[0044] ;

[0045] Among them, represents the predicted value of the fault risk at time point , represents the weights of the deep reinforcement learning model, represents the predicted value output by the DRL model based on the plant operation and maintenance knowledge graph at time point and the plant operation and maintenance status at time point and time point , represents the weights of the long short-term memory network model, represents the preliminary predicted value output by the LSTM model based on the plant operation and maintenance knowledge graph converted into a time series data format from time point to time point , represents the length of the time period, represents the operation and maintenance feedback data at time point , represents the index variable of the time point.

[0046] As an optimized solution of the intelligent plant operation and maintenance management system based on multi-dimensional data driving described in the present invention, among them: the preliminary predicted value output by the LSTM model based on the plant operation and maintenance knowledge graph converted into a time series data format from time point to time point is as follows,

[0047] The LSTM model analyzes the plant operation and maintenance knowledge graph converted into a time series data format, predicts the plant operation and maintenance status at the current moment, and obtains the preliminary predicted value by weighted averaging the prediction results of historical moments. The expression is:

[0048] ;

[0049] Among them, represents the number of historical moments, represents the index variable of the historical moment, represents the th weighted coefficient of the LSTM model at the historical moment, represents to The time interval of the plant operation and maintenance knowledge graph converted into a time series data format, represents time point shifted forward moments plus the time period , represents a time point shifted forward time instants.

[0050] As a preferred solution of the intelligent plant operation and maintenance management system based on multi-dimensional data-driven of the present invention, wherein: the DRL model is based on the plant operation and maintenance knowledge graph at the time point and the plant operation and maintenance status at the time point to output a predicted value. The specific steps are as follows.

[0051] The deep reinforcement learning model utilizes the plant operation and maintenance knowledge graph, combines the complex dynamic relationship between the status and behavior in the historical data, continuously improves the prediction ability of the plant operation and maintenance behavior through the decision optimization process, and captures the dynamic relationship between the plant operation and maintenance status and behavior by analyzing the multi-dimensional plant operation and maintenance data at past time instants, and obtains the predicted value by weighted averaging the prediction results at historical time instants. The expression is:

[0052] ;

[0053] wherein, represents the weighted coefficient of the DRL model at the th historical time instant, represents the plant operation and maintenance knowledge graph at the time point , represents the plant operation and maintenance status at the time point .

[0054] As a preferred solution of the intelligent plant operation and maintenance management system based on multi-dimensional data-driven of the present invention, wherein: generating a warning notice and formulating a maintenance measure based on the predicted value of the fault risk. The specific steps are as follows.

[0055] Based on the historical fault risk data, clustering is performed by the clustering analysis method, and a risk upper limit threshold and a risk lower limit threshold are set;

[0056] Based on the predicted value of the fault risk , through the risk upper limit threshold and the risk lower limit threshold , the fault risk level of the plant operation and maintenance is evaluated;

[0057] When , it indicates that the fault risk of the plant operation and maintenance is in the low fault risk area;

[0058] When , it indicates that the fault risk of the plant operation and maintenance is in the medium fault risk area;

[0059] When it is the case, it indicates that the failure risk of plant operation and maintenance is in the high failure risk area;

[0060] Based on the failure risk level of plant operation and maintenance, the operation status of equipment and potential failure risks are tracked in real time and decision support is provided, and early warning notifications are automatically generated;

[0061] Based on the failure risk prediction value and early warning notification, and combined with the equipment operation environment and equipment status, corresponding maintenance measures are formulated.

[0062] The beneficial effects of the present invention are as follows: By integrating multi-dimensional data collection and preprocessing, intelligent feature extraction, knowledge graph construction, and an intelligent operation and maintenance prediction model that combines deep reinforcement learning and long short-term memory network, the comprehensive monitoring of plant operation and maintenance status and accurate failure risk prediction are realized. It not only improves the automation and intelligence level of operation and maintenance management, but also effectively prevents equipment failures through early warning and optimized maintenance measures, reduces unplanned downtime, thus significantly improving the efficiency and reliability of factory operation, reducing maintenance costs, and enhancing the stability and security of the production system. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0064] Figure 1 It is a schematic diagram of the intelligent plant operation and maintenance management system based on multi-dimensional data drive in Embodiment 1;

[0065] Figure 2 It is a schematic diagram of constructing a plant operation and maintenance knowledge graph in Embodiment 1. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0066] In order to make the above objects, features, and advantages of the present invention more obvious and understandable, the specific embodiments of the present invention will be described in detail below with reference to the drawings in the specification.

[0067] Many specific details are set forth in the following description in order to provide a thorough understanding of the present invention, but the present invention may be implemented in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0068] Second, the "one embodiment" or "embodiment" referred to herein means a specific feature, structure, or characteristic that may be included in at least one implementation manner of the present invention. The appearances of "in one embodiment" in different places in this specification do not all refer to the same embodiment, nor are they separate or alternative embodiments that exclude each other with other embodiments.

[0069] Embodiment 1, referring to Figure 1 and Figure 2 , is the first embodiment of the present invention. This embodiment provides an intelligent plant operation and maintenance management system based on multi-dimensional data-driven, including the following steps:

[0070] Data collection and preprocessing module, which collects multi-dimensional plant operation and maintenance data and preprocesses the multi-dimensional plant operation and maintenance data.

[0071] The multi-dimensional plant operation and maintenance data includes vibration frequency, power consumption, operation logs, and fault reports.

[0072] Preprocessing the multi-dimensional plant operation and maintenance data includes preliminary cleaning, format conversion, and standardization processing.

[0073] It should be noted that the multi-dimensional plant operation and maintenance data is preprocessed through the preprocessing methods of preliminary cleaning, format conversion, and standardization processing. The specific process is as follows:

[0074] Preliminary cleaning: First, identify and remove outliers and noise in the multi-dimensional plant operation and maintenance data, handle missing values - fill in the blanks using interpolation methods or reasonably estimate according to business logic, and perform a deduplication operation to eliminate duplicate records, ensuring that the finally obtained data set is accurate, complete, and highly consistent;

[0075] Format conversion: Parse and uniformly convert the multi-dimensional plant operation and maintenance data from different sources and with various original formats (such as CSV, JSON, XML, or database records) into an internal standard format, ensuring that the structures of all data items are consistent and the semantics are clear. At the same time, encode non-numerical multi-dimensional plant operation and maintenance data (such as text labels or category information) into numerical forms;

[0076] Standardization processing: Adjust the distribution of the multi-dimensional plant operation and maintenance data of each feature to a common scale. Usually, calculate statistical parameters such as the mean and standard deviation of each feature, and then apply techniques such as min-max normalization or Z-score standardization to map the values to a specific interval or convert them into a standard normal distribution with zero mean and unit variance, thereby eliminating the influence of dimensions.

[0077] Feature extraction module, based on the preprocessed multi-dimensional plant operation and maintenance data, performs feature extraction through natural language processing technology and principal component analysis algorithm to obtain plant operation and maintenance features.

[0078] Based on the preprocessed multi-dimensional plant operation and maintenance data, plant text data and plant numerical data are extracted through data parsing methods.

[0079] It should be noted that by using specific parsing algorithms and rules, in-depth analysis is carried out on the standardized multi-dimensional plant operation and maintenance data to identify and separate plant text data containing descriptive information (such as natural language descriptions in operation logs and fault reports) and plant numerical data of quantitative indicators (such as measurement values like temperature readings and vibration frequencies).

[0080] Through the word embedding method, the plant text data is transformed into the form of word vectors.

[0081] It should be noted that by using a pre-trained or self-trained word embedding model (such as Word2Vec, GloVe or FastText), each word is mapped to a fixed-length real number vector in a high-dimensional space, that is, a word vector.

[0082] Based on the plant text data in the form of word vectors, aggregation is carried out through the aggregation method to obtain the global context feature vector of the plant text data.

[0083] It should be noted that first, for each word in each text record, its corresponding word vector representation is used; then, a suitable aggregation strategy is selected, such as simply averaging or weighted averaging the word vectors of all words, using max / min pooling to select the most representative word vector features, or adopting a more complex attentive pooling method to assign different weights according to the importance of words in the sentence; finally, a fixed-length global context feature vector that can reflect the overall semantics and context information of the entire text paragraph is generated through the selected aggregation method.

[0084] Based on the plant numerical data, analysis is carried out through statistical methods to obtain the covariance matrix.

[0085] It should be noted that the covariance between each pair of plant numerical data features is calculated, which measures the degree of mutual relationship between the changes of two eigenvalue; then, a square matrix is constructed, where the value at each position reflects the covariance between the corresponding two features, and the elements on the diagonal represent the variances of each feature itself; finally, the obtained covariance matrix;

[0086] It should also be noted that the plant numerical data features come from the quantitative measurement values in the preprocessed multi-dimensional plant operation and maintenance data set, such as vibration frequency, temperature and power consumption, etc.

[0087] The covariance matrix is eigen-decomposed through the eigenvalue decomposition method to extract the principal component eigenvalues and the corresponding principal component eigenvectors in the plant numerical data.

[0088] It should be noted that, first, a covariance matrix is calculated based on the preprocessed plant numerical data; then, a mathematical algorithm (such as power iteration, QR algorithm, or Jacobi method) is applied to perform eigenvalue decomposition on the covariance matrix to solve a set of eigenvalues and their corresponding eigenvectors, where the eigenvalues represent the magnitudes of the data variances explained by the respective principal components, and the eigenvectors indicate the projection patterns of the data in the directions of these principal components.

[0089] Based on the sorted principal component eigenvalues extracted by size, the principal component eigenvector corresponding to the largest principal component eigenvalue is taken as the characteristic principal component.

[0090] It should be noted that after completing the eigenvalue decomposition of the covariance matrix, a set of eigenvalues and their corresponding eigenvectors are obtained; then, they are sorted in descending order according to the magnitudes of the eigenvalues because the larger eigenvalues represent more variance information in the data; then, the eigenvector with the largest eigenvalue is selected, and this vector indicates the direction with the largest variance in the dataset, i.e., the first principal component.

[0091] The plant numerical data is mapped onto the characteristic principal component to obtain the dimension-reduced plant characteristics.

[0092] It should be noted that, first, the characteristic principal component representing the main variance direction of the data is found by using the eigenvalue decomposition of the covariance matrix; then, one or several of the most important principal components (usually those eigenvectors corresponding to the largest eigenvalues) are selected; then, by performing matrix multiplication on the original plant numerical data matrix and the selected principal component vectors, the data is projected onto the new principal component coordinate axes; finally, the dimension-reduced plant characteristics are obtained.

[0093] Based on the global context feature vector of the plant text data and the dimension-reduced plant characteristics, they are concatenated in the vector order through the concatenation method to obtain the plant operation and maintenance characteristics.

[0094] It should be noted that, first, it is ensured that the global context feature vector reflecting the text content and the numerical feature vector dimension-reduced by principal component analysis have been obtained; then, these two vectors are horizontally concatenated together in a predetermined order (usually text features first and then numerical features or vice versa) to form a longer comprehensive feature vector; finally, the plant operation and maintenance characteristics are obtained.

[0095] The knowledge graph construction module constructs a plant operation and maintenance knowledge graph based on the plant operation and maintenance characteristics in the RDF standard format.

[0096] Based on the plant operation and maintenance characteristics, the plant operation and maintenance entity types, plant operation and maintenance entity attributes, and plant operation and maintenance entity relationships of the knowledge graph are defined through feature parsing and semantic abstraction methods to obtain the semantic structure of the plant operation and maintenance knowledge graph.

[0097] It should be noted that, first, the extracted plant operation and maintenance features are deeply analyzed to identify representative entity types (such as equipment, operators, maintenance events, etc.); then, the key attributes of each entity are determined (such as equipment model, operating status, fault code, etc.), and high-level concepts and categories are refined through semantic abstraction; then, according to the interactions and associations between entities in the multi-dimensional plant operation and maintenance data, the relationships between entities are defined (such as "equipment fails", "operator performs maintenance", etc.); finally, these entity types, attributes and relationships are organized in a standard format such as RDF (Resource Description Framework) to construct a knowledge graph semantic structure that can express complex relationships and dynamic changes in the field of plant operation and maintenance.

[0098] Based on the plant text data and plant numerical data in the multi-dimensional plant operation and maintenance data, they are mapped to the attribute values of plant operation and maintenance entities through natural language processing methods and data analysis methods.

[0099] It should be noted that, first, natural language processing technology is used to parse the plant text data, extract key information and convert it into structured semantic tags or classifications (such as fault types, operation instructions, etc.); at the same time, statistical analysis is performed on the plant numerical data to calculate descriptive statistics (such as mean, standard deviation) or other characteristic values to quantify equipment performance or operating status; then, the information extracted from the text and numerical data is associated with the corresponding plant operation and maintenance entities as their attribute values, for example, specific fault codes or average energy consumption are assigned to a certain equipment entity.

[0100] Based on the knowledge graph semantic structure of plant operation and maintenance and the attribute values of plant operation and maintenance entities, knowledge graph triples of plant operation and maintenance are constructed through the RDF standard format.

[0101] It should be noted that, first, entities and their relationships are determined according to the defined semantic structure, and then each entity and its attribute values are represented as resources (Subject) in RDF; then, appropriate predicates (Predicate) are selected or defined to describe the relationships between entities or the attributes of entities; finally, the attribute values of the entities are used as objects (Object), organized in the triple form of "subject-predicate-object", and encoded using the RDF standard syntax (such as Turtle, N-Triples, etc.), thus forming a series of knowledge graph triples that describe the objects and their mutual relationships in the field of plant operation and maintenance.

[0102] Based on the knowledge graph triples of plant operation and maintenance, they are organized into an RDF graph structure through GraphDB to obtain the plant operation and maintenance knowledge graph.

[0103] It should be noted that, first, the constructed RDF triples are imported into a graph database such as GraphDB; then, GraphDB automatically parses these triples and organizes them into a coherent graph structure according to the relationships between the subject, predicate, and object, where entities are used as nodes and relationships are used as edges; next, the query language provided by GraphDB (such as SPARQL) is used for data retrieval and complex pattern matching to verify and optimize the graph structure; finally, the complete plant operation and maintenance knowledge graph is accessed and displayed through a visualization tool or API interface.

[0104] The model construction module constructs an intelligent operation and maintenance prediction model based on a deep reinforcement learning model and a long short-term memory network model.

[0105] The historical plant operation and maintenance knowledge graph is converted into a time series data format through a graph embedding method and a time series annotation method.

[0106] It should be noted that, first, graph embedding techniques (such as GraphSAGE, Node2Vec, etc.) are used to convert the nodes and edges in the knowledge graph into low-dimensional vector representations to capture the semantic information of entities and their relationships; then, these embedding vectors are sorted according to timestamps or the order of event occurrence, and time series annotations are added to identify the time points of each state; next, the embedding vectors with time tags are serialized to form a continuous time series dataset.

[0107] Taking the long short-term memory network model as the basic model;

[0108] The input layer of the LSTM model receives the historical plant operation and maintenance knowledge graph converted into a time series data format.

[0109] The LSTM layer, through memory units and gate mechanisms, selectively remembers and forgets information in the LSTM model to capture long-term dependencies in the time series.

[0110] The fully connected layer extracts features and reduces dimensions based on the output of the LSTM layer and maps them to the output layer.

[0111] The output layer outputs preliminary prediction values.

[0112] The deep reinforcement learning layer learns through interaction with the environment, uses a reward function to optimize the decision-making process, adjusts the parameters of the deep reinforcement learning model, and optimizes the output results of the LSTM model.

[0113] The deep reinforcement learning output layer outputs the final prediction values.

[0114] An initial intelligent operation and maintenance prediction model is constructed.

[0115] The preliminary intelligent operation and maintenance prediction model is trained, evaluated, and optimized through backpropagation algorithm and reinforcement learning algorithm, and finally the intelligent operation and maintenance prediction model is constructed.

[0116] It should be noted that the advantage of constructing an intelligent operation and maintenance prediction model based on the long short-term memory network (LSTM model) combined with the deep reinforcement learning (DRL) model is that: the LSTM model can effectively capture the long-term dependencies in time series data and handle the complex patterns that change over time in plant operation and maintenance; while deep reinforcement learning continuously optimizes the decision-making process through interaction with the environment, improving the accuracy of fault prediction and maintenance strategies. The combination of the two not only enhances the model's learning ability for historical data, improves the prediction accuracy of future fault risks, but also can adaptively adjust maintenance measures to achieve more intelligent and automated operation and maintenance management, thus effectively reducing equipment failure rates and maintenance costs.

[0117] The preliminary intelligent operation and maintenance prediction model is trained, evaluated, and optimized through backpropagation algorithm and reinforcement learning algorithm, and finally the intelligent operation and maintenance prediction model is constructed.

[0118] The historical plant operation and maintenance knowledge graph converted into time series data format is divided into a training set and a validation set by the random division method.

[0119] It should be noted that first, ensure that the time series data is arranged in chronological order and remains complete; then, according to a predetermined ratio (such as 70% training set and 30% validation set), use the random sampling method to select samples from the entire dataset, while ensuring that the continuity and context dependence of the time series are not destroyed; then, when dividing, pay attention to avoiding future data leakage into the training set. Usually, the practice is to extract data from an earlier time period as the training set and data from a later time period as the validation set; finally, generate independent training and validation sets.

[0120] Based on the training set, the prediction performance of the LSTM model is evaluated by minimizing the loss function, and the weights and parameters of the LSTM model are optimized through the backpropagation algorithm.

[0121] It should be noted that the training data is input into the LSTM model for forward propagation, and the loss value (such as mean square error or cross entropy) between the output of the LSTM model and the true label is calculated; then the backpropagation algorithm is used to update the weight matrix, bias term, and parameters of the gating unit of the LSTM model according to the gradient of the loss with respect to each parameter to gradually reduce the loss function.

[0122] Through the predicted values output by the LSTM model, the deep reinforcement learning model uses the policy gradient method to adjust the decision-making process according to the reward signal, adjust the output of the LSTM model, and optimize the prediction accuracy of fault risks.

[0123] It should be noted that the LSTM model first generates preliminary fault risk prediction values based on the time series data transformed from the historical plant operation and maintenance knowledge graph. Subsequently, these prediction values are input into the deep reinforcement learning (DRL) model as the environmental state. The DRL model executes actions according to the current policy and receives reward or punishment signals based on the actual operation and maintenance effects. Then, the DRL model uses the policy gradient method to adjust its decision-making policy parameters according to the reward signal, and at the same time feeds the optimized decision back to the LSTM model, prompting the LSTM to update its weights and parameters through the backpropagation algorithm to correct the prediction output.

[0124] The validation set is used to evaluate the prediction performance of the preliminary intelligent operation and maintenance prediction model by analyzing the accuracy, recall rate, and F1-score, and the reward function, the parameters of the LSTM layer, and the reinforcement learning policy are fine-tuned based on the evaluation results.

[0125] It should be noted that when using the validation set to evaluate the performance of the preliminary intelligent operation and maintenance prediction model, the validation data is first input into the trained model to generate fault risk prediction results, which are compared with the actual fault labels in the validation set. The accuracy (the proportion of correct predictions), recall rate (the proportion of actual positive classes and correct predictions), and F1-score (the harmonic mean of accuracy and recall rate) are calculated to comprehensively evaluate the prediction ability of the preliminary intelligent operation and maintenance prediction model. Based on the results of these evaluation metrics, the memory units and gate mechanism parameters of the LSTM layer are fine-tuned to optimize its ability to capture long-term dependencies in time series; at the same time, the reward function and other related parameters in the reinforcement learning policy are adjusted to improve the decision-making process.

[0126] The k-fold cross-validation method is used to validate the preliminarily trained intelligent operation and maintenance prediction model, and finally the intelligent operation and maintenance prediction model is output.

[0127] It should be noted that the multi-dimensional plant operation and maintenance data set is divided into k mutually exclusive subsets; each time, one subset is selected as the validation set, and the remaining data is used to evaluate the performance of the preliminarily trained intelligent operation and maintenance prediction model; for each subset, the performance metrics of the preliminarily trained intelligent operation and maintenance prediction model on this validation set, such as accuracy, recall rate, and F1-score, are recorded; the results of all k validations are summarized, and the average performance is calculated to evaluate the overall stability and generalization ability of the preliminarily trained intelligent operation and maintenance prediction model; according to the cross-validation results, the configuration or parameter settings of the preliminarily trained intelligent operation and maintenance prediction model are selected, and finally the verified and optimized intelligent operation and maintenance prediction model is output.

[0128] The fault risk value prediction module obtains the fault risk prediction value through the intelligent operation and maintenance prediction model based on the plant operation and maintenance knowledge graph.

[0129] Based on the plant operation and maintenance knowledge graph converted into a time series data format, the long short-term memory network model is used to process the time series data, and combined with the deep reinforcement learning model based on the plant operation and maintenance knowledge graph, the complex dynamic relationships in plant operation and maintenance are captured.

[0130] It should be noted that, first, the plant operation and maintenance knowledge graph is converted into a time series data format to reflect the operation and maintenance status changing over time; then, the LSTM model is used to analyze these time series data to capture long-term dependencies and historical patterns; at the same time, the DRL model makes decisions based on the current and past operation and maintenance status and continuously optimizes its strategy through interaction with the environment; the LSTM model and the DRL model work together, the LSTM model provides accurate time series predictions, and the DRL model adjusts the operation and maintenance strategy according to the prediction results, and the two work together to identify and adapt to complex dynamic changes.

[0131] Based on the complex dynamic relationships, a multi-dimensional data-driven method is used to obtain the fault risk prediction values at each time point, and the expression is:

[0132] ;

[0133] Where, represents the fault risk prediction value at time point , represents the weight of the deep reinforcement learning model, represents the prediction value output by the DRL model based on the plant operation and maintenance knowledge graph at time point and the plant operation and maintenance status at time point and time point , represents the prediction value output by the long short-term memory network model, represents the weight of the long short-term memory network model, represents the preliminary prediction value output by the LSTM model based on the plant operation and maintenance knowledge graph converted into a time series data format from time point to time point , represents the length of the time period, represents the operation and maintenance feedback data at time point , represents the index variable of the time point.

[0134] It should be noted that the operation and maintenance feedback data is collected and integrated from multi-source information such as the device operation status, fault alarms, maintenance activity results, and performance indicators through a real-time monitoring system and a database storing historical data related to device operation and maintenance;

[0135] The operation and maintenance feedback data includes real-time or historical records such as device operation status, fault alarms, maintenance activity results, and performance indicators;

[0136] It should be noted that the plant operation and maintenance status refers to a set of data reflecting the operating conditions of factory equipment at a specific time point, including but not limited to the working parameters of the equipment (such as temperature, pressure, vibration frequency, etc.), operating status (such as startup, shutdown, maintenance in progress), fault information, and environmental conditions, etc.;

[0137] It should also be noted that the plant operation and maintenance status comes from multiple data sources, mainly including the physical parameters collected in real time by the sensor network, the operation logs recorded by the equipment's built-in monitoring system, the manual inspection reports input by maintenance personnel, and historical fault and repair records, etc.

[0138] The LSTM model is based on the preliminary predicted values output from the plant operation and maintenance knowledge graph converted into a time series data format from time point to time point .

[0139] The LSTM model analyzes the plant operation and maintenance knowledge graph converted into a time series data format to predict the plant operation and maintenance status at the current moment, and obtains the preliminary predicted value by weighted averaging the prediction results of historical moments. The expression is:

[0140] ;

[0141] where represents the number of historical moments, represents the index variable of the historical moment, represents the weighted coefficient of the LSTM model at the th historical moment, represents to the plant operation and maintenance knowledge graph converted into a time series data format for the time interval, represents the time point shifted forward by moments plus the time period , represents the time point shifted forward by moments.

[0142] It should be noted that the calculation process of the above expression is as follows:

[0143] First, for the past time points, each time point corresponds to a historical time period, and the time series data of this time period is input into the LSTM model to generate predicted values; then, each predicted value is multiplied by its corresponding weighted coefficient, and these weighted coefficients reflect the importance of different historical moments; next, all the weighted predicted values are summed and divided by the sum of all the weighted coefficients to calculate a weighted average preliminary predicted value.

[0144] The DRL model is based on time points of the plant operation and maintenance knowledge graph and time points of the plant operation and maintenance status to output predicted values.

[0145] The deep reinforcement learning model utilizes the plant operation and maintenance knowledge graph, combines the complex dynamic relationship between states and actions in historical data, continuously improves the prediction ability of plant operation and maintenance actions through the decision optimization process, and captures the dynamic relationship between plant operation and maintenance states and actions by analyzing multi-dimensional plant operation and maintenance data at past moments. The predicted value is obtained by weighted averaging the prediction results at historical moments, and the expression is:

[0146] ;

[0147] where, represents the weighted coefficient of the DRL model at the th historical moment, represents the plant operation and maintenance knowledge graph at time point , represents the plant operation and maintenance status at time point .

[0148] It should be noted that the calculation process of the above expression is as follows:

[0149] First, for the past time points, each time point corresponds to a plant operation and maintenance knowledge graph and a plant operation and maintenance status at a historical moment. These data are input into the deep reinforcement learning (DRL) model to generate predicted values; then, each predicted value is multiplied by its corresponding weighted coefficient, and these weights reflect the importance of the prediction results at different historical moments; next, all weighted predicted values are summed and divided by the sum of all weighted coefficients to calculate a weighted average predicted value.

[0150] The risk assessment and early warning module generates early warning notifications and formulates maintenance measures based on the predicted fault risk values.

[0151] Based on historical fault risk data, clustering is performed through the clustering analysis method, and the upper risk threshold and the lower risk threshold are set;

[0152] Based on the predicted fault risk value , through the upper risk threshold and the lower risk threshold , the fault risk level of plant operation and maintenance is evaluated.

[0153] It should be noted that, first, historical fault risk data of plant operations and maintenance are collected and sorted out; then, clustering algorithms (such as K-means, hierarchical clustering, etc.) are applied to divide these data into several clusters according to similarity, and each cluster represents a different level of risk situation; next, the characteristics of each cluster are analyzed to determine the boundary points that can distinguish low, medium, and high risks, and accordingly, a risk lower limit threshold (identifying the upper limit of the low-risk interval) and a risk upper limit threshold (identifying the lower limit of the high-risk interval) are set.

[0154] When it indicates that the fault risk of plant operations and maintenance is in the low fault risk area;

[0155] When it indicates that the fault risk of plant operations and maintenance is in the medium fault risk area;

[0156] When it indicates that the fault risk of plant operations and maintenance is in the high fault risk area;

[0157] Based on the fault risk level of plant operations and maintenance, real-time tracking and decision support are carried out on the equipment operation status and potential fault risks, and early warning notifications are automatically generated.

[0158] It should be noted that multi-dimensional plant operations and maintenance data are continuously monitored and collected, advanced data analysis and machine learning algorithms are used to evaluate the fault risk in real time, the current risk level is automatically judged according to the preset threshold, the development trend of potential faults is predicted, optimized maintenance suggestions are provided in combination with historical fault data, and once it is detected that the risk exceeds the set threshold, an early warning notification is immediately triggered and relevant personnel are notified in a timely manner through channels such as emails and text messages to ensure rapid response and the adoption of appropriate preventive or corrective measures.

[0159] Based on the fault risk prediction value and early warning notification, and in combination with the equipment operation environment and equipment status, corresponding maintenance measures are formulated.

[0160] It should be noted that corresponding maintenance measures are formulated as follows:

[0161] Maintenance measures in the low-risk area:

[0162] Regular inspection: According to the equipment operation cycle, arrange regular inspections and maintenance, such as cleaning, lubrication, replacement of vulnerable parts, etc., to ensure that the equipment remains in good condition;

[0163] Optimized maintenance: For historical operation and maintenance data, optimize the maintenance plan, such as adjusting the maintenance cycle or optimizing resource allocation, to improve the reliability and lifespan of the equipment.

[0164] Monitoring improvement: In the case of low risk, enhance the real-time monitoring of the equipment to facilitate early identification when potential risks occur;

[0165] Maintenance measures for medium-risk areas:

[0166] Enhanced monitoring: Strengthen the real-time monitoring of equipment, especially predictive failure indicators (such as temperature, pressure, etc.), and judge the trend of potential failures through data analysis means;

[0167] Predictive maintenance: According to historical data and prediction models, perform maintenance or replacement of components prone to problems in advance to avoid the expansion of risks;

[0168] Extra detection: Conduct extra detection or load testing on equipment with a higher probability of failure to verify whether the equipment is within the acceptable risk range;

[0169] Maintenance measures for high-risk areas:

[0170] Immediate maintenance: Activate the emergency plan, perform emergency shutdown of the equipment, repair, replace faulty components or take other measures to prevent further damage to the equipment or cause a greater failure;

[0171] Shutdown processing: If the equipment failure has reached an irreparable level, it may be necessary to perform shutdown processing and start the spare part replacement or emergency repair procedure;

[0172] Report generation: Generate a detailed risk report to relevant departments or management, analyze the root cause of the problem, formulate long-term improvement measures, and avoid the recurrence of similar failures.

[0173] In summary, the present invention realizes the comprehensive monitoring of the plant operation and maintenance status and the accurate prediction of fault risks by integrating multi-dimensional data acquisition and preprocessing, intelligent feature extraction, knowledge graph construction, and an intelligent operation and maintenance prediction model that combines deep reinforcement learning and long short-term memory network. It not only improves the automation and intelligence level of operation and maintenance management, but also effectively prevents equipment failures through early warning and optimized maintenance measures, reduces unplanned downtime, thus significantly improving the efficiency and reliability of factory operation, reducing maintenance costs, and enhancing the stability and security of the production system.

[0174] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not restrictive. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. Intelligent factory operation and maintenance management system driven by multi-dimensional data, characterized by: include, Data collection and preprocessing module, collects multi-dimensional plant operation and maintenance data, and preprocesses the multi-dimensional plant operation and maintenance data; The feature extraction module extracts features based on the pre-processed multi-dimensional plant operation and maintenance data through natural language processing technology and principal component analysis algorithm to obtain plant operation and maintenance features; The knowledge graph construction module builds the factory operation and maintenance knowledge graph based on the factory operation and maintenance characteristics in the RDF standard format; Model building module, which builds an intelligent operation and maintenance prediction model based on deep reinforcement learning model and long short-term memory network model; The fault risk value prediction module obtains the fault risk prediction value based on the plant operation and maintenance knowledge graph through the intelligent operation and maintenance prediction model, as follows: Based on the plant operation and maintenance knowledge graph converted into time series data format, the time series data is processed through the long short-term memory network model, and the deep reinforcement learning model is combined with the plant operation and maintenance knowledge graph to capture the complex dynamic relationships in plant operation and maintenance; Based on complex dynamic relationships, a multidimensional data-driven approach is used to obtain the failure risk prediction value at each time point. The expression is: ； in, Indicates time point The failure risk prediction value of represents the weight of the deep reinforcement learning model, Indicates that the DRL model is based on time points Factory Operation and Maintenance Knowledge Graph and time point Factory operation and maintenance status The predicted value of the output, represents the weight of the long short-term memory network model, Indicates that the LSTM model is based on time points To the time point The initial prediction value of the plant operation and maintenance knowledge graph output converted into time series data format, Indicates the length of the time period, Indicates time point Operation and maintenance feedback data, An index variable representing a time point; The risk assessment and early warning module generates early warning notifications and formulates maintenance measures based on the failure risk prediction value.

2. The intelligent factory operation and maintenance management system based on multi-dimensional data drive according to claim 1 is characterized in that: The multi-dimensional plant operation and maintenance data includes vibration frequency, power consumption, operation logs and fault reports; The preprocessing of the multi-dimensional plant operation and maintenance data includes preliminary cleaning, format conversion and standardization.

3. The intelligent factory operation and maintenance management system based on multi-dimensional data drive according to claim 2 is characterized in that: Based on the pre-processed multi-dimensional plant operation and maintenance data, feature extraction is performed through natural language processing technology and principal component analysis algorithm to obtain plant operation and maintenance features. The specific steps are as follows: Based on the pre-processed multi-dimensional factory operation and maintenance data, factory text data and factory numerical data are extracted through data analysis methods; Through the word embedding method, the factory affairs text data is converted into word vector form; Based on the factory affairs text data in the form of word vectors, the global context feature vector of the factory affairs text data is obtained by aggregation through aggregation method; Based on the factory numerical data, the covariance matrix is ​​obtained by statistical analysis; The covariance matrix is ​​decomposed by the eigenvalue decomposition method to extract the principal component eigenvalues ​​and corresponding principal component eigenvectors in the factory numerical data; Based on the order of the extracted principal component eigenvalues, the principal component eigenvector corresponding to the largest principal component eigenvalue is used as the characteristic principal component; Map the factory affairs numerical data to the characteristic principal components to obtain the factory affairs characteristics after dimensionality reduction; Based on the global context feature vector of factory affairs text data and the factory affairs features after dimensionality reduction, the factory affairs operation and maintenance features are obtained by connecting them in vector order through the splicing method.

4. The intelligent factory operation and maintenance management system based on multi-dimensional data drive according to claim 3 is characterized by: Based on the factory operation and maintenance characteristics, the factory operation and maintenance knowledge graph is constructed through the RDF standard format. The specific steps are as follows: Based on the factory operation and maintenance characteristics, the factory operation and maintenance entity types, factory operation and maintenance entity attributes, and factory operation and maintenance entity relationships of the knowledge graph are defined through feature analysis and semantic abstraction methods to obtain the semantic structure of the factory operation and maintenance knowledge graph; Based on the factory affairs text data and factory affairs numerical data in the multi-dimensional factory affairs operation and maintenance data, they are mapped into attribute values ​​of factory affairs operation and maintenance entities through natural language processing methods and data analysis methods; Based on the semantic structure of the plant operation and maintenance knowledge graph and the attribute values ​​of the plant operation and maintenance entities, the plant operation and maintenance knowledge graph triples are constructed using the RDF standard format; Based on the plant operation and maintenance knowledge graph triples, GraphDB is used to organize them into an RDF graph structure to obtain the plant operation and maintenance knowledge graph.

5. The intelligent factory operation and maintenance management system based on multi-dimensional data drive according to claim 4 is characterized by: The intelligent operation and maintenance prediction model is constructed based on the deep reinforcement learning model and the long short-term memory network model. The specific steps are as follows: Through graph embedding method and time series annotation method, the historical plant operation and maintenance knowledge graph is converted into time series data format; The model is based on the long short-term memory network model; The input layer of the LSTM model receives the historical plant operation and maintenance knowledge graph converted into a time series data format; LSTM layer uses memory cells and gate mechanisms. The LSTM model selectively remembers and forgets information, capturing long-term dependencies in time series. The fully connected layer extracts features and reduces the dimension based on the output of the LSTM layer and maps it to the output layer; The output layer outputs preliminary prediction values; The deep reinforcement learning layer optimizes the decision-making process by using the reward function through interactive learning with the environment, adjusts the parameters of the deep reinforcement learning model, and optimizes the output results of the LSTM model; The deep reinforcement learning output layer outputs the final prediction value; Construct a preliminary intelligent operation and maintenance prediction model; Through the back propagation algorithm and reinforcement learning algorithm, the preliminary intelligent operation and maintenance prediction model is trained, evaluated and optimized, and finally an intelligent operation and maintenance prediction model is constructed.

6. The intelligent factory operation and maintenance management system based on multi-dimensional data drive according to claim 5 is characterized by: Through the back propagation algorithm and reinforcement learning algorithm, the preliminary intelligent operation and maintenance prediction model is trained, evaluated and optimized, and finally the intelligent operation and maintenance prediction model is constructed. The specific steps are as follows: The historical plant operation and maintenance knowledge graph converted into time series data format is divided into a training set and a validation set through random partitioning. The prediction performance of the LSTM model is evaluated by minimizing the loss function based on the training set, and the weights and parameters of the LSTM model are optimized through the back-propagation algorithm; Through the prediction value output by the LSTM model, the deep reinforcement learning model uses the policy gradient method to adjust the decision process according to the reward signal, adjust the output of the LSTM model, and optimize the fault risk prediction accuracy; Use the validation set to evaluate the prediction performance of the preliminary intelligent operation and maintenance prediction model by analyzing the accuracy, recall, and F1-score, and fine-tune the reward function, LSTM layer parameters, and reinforcement learning strategy based on the evaluation results; The k-fold cross-validation method is used to verify the initially trained intelligent operation and maintenance prediction model, and finally the intelligent operation and maintenance prediction model is output.

7. The intelligent factory operation and maintenance management system based on multi-dimensional data drive according to claim 1 is characterized by: LSTM model is based on time points To the time point The initial prediction value of the plant operation and maintenance knowledge graph output in the time series data format is converted into the following specific steps: The LSTM model analyzes the plant operation and maintenance knowledge graph converted into a time series data format to predict the plant operation and maintenance status at the current moment, and obtains the preliminary prediction value by weighted averaging the prediction results at historical moments. The expression is: ； in, represents the number of historical moments, An index variable representing a historical moment, Indicates The weight coefficient of the LSTM model at each historical moment, express arrive The time interval is converted into a plant operation and maintenance knowledge graph in the time series data format. Indicates time point Move forward time plus time period , Indicates time point Move forward A moment.

8. The intelligent factory operation and maintenance management system based on multi-dimensional data drive according to claim 1 is characterized by: DRL model is based on time point Factory Operation and Maintenance Knowledge Graph and time point Factory operation and maintenance status The output prediction value, the specific steps are as follows, The deep reinforcement learning model uses the plant operation and maintenance knowledge graph, combined with the complex dynamic relationship between status and behavior in historical data, to continuously improve the prediction ability of plant operation and maintenance behavior through the decision optimization process. It also captures the dynamic relationship between plant operation and maintenance status and behavior by analyzing the multi-dimensional plant operation and maintenance data at past moments, and obtains the prediction value by weighted averaging the prediction results at historical moments. The expression is: ； in, Indicates The weighted coefficient of the DRL model at each historical moment, Indicates time point The knowledge graph of factory operation and maintenance at all times, Indicates time point The factory operation and maintenance status at all times.

9. The intelligent factory operation and maintenance management system based on multi-dimensional data drive according to claim 1 is characterized in that: The specific steps of generating early warning notification and formulating maintenance measures based on the fault risk prediction value are as follows: Based on historical failure risk data, cluster analysis is used to set the upper risk threshold. and the lower risk threshold ; Based on the failure risk prediction value , passing the upper risk threshold and the lower risk threshold , assess the failure risk level of plant operation and maintenance; when When , it indicates that the failure risk of plant operation and maintenance is in the low failure risk zone; when When , it indicates that the failure risk of plant operation and maintenance is in the medium failure risk zone; when When , it indicates that the failure risk of plant operation and maintenance is in the high failure risk zone; Based on the fault risk level of plant operation and maintenance, real-time tracking and decision support of equipment operation status and potential fault risks are carried out, and early warning notifications are automatically generated; Based on the fault risk prediction value and early warning notification, and combined with the equipment operating environment and equipment status, formulate corresponding maintenance measures.

Images