Operation monitoring and early warning method for intelligent energy-saving control equipment

Abstract

The invention discloses an operation monitoring and early warning method for intelligent energy-saving control equipment, and relates to the technical field of equipment early warning, and the method comprises the steps: collecting the operation data of the equipment in real time; performing data cleaning to remove abnormal and repeated data; establishing an intelligent energy-saving control equipment mathematical model, and predicting system output; analyzing the residual error of the system, and determining the fault occurrence condition of the process; the reasoning ability of process experts on monitoring is simulated in a reasoning deduction mode, and equipment fault early warning is completed; internal information in the process data is mined, a fault early warning mathematical model is established, and a process state is expressed; training the fault early warning mathematical model by using historical data; and optimizing model parameters of the fault early warning mathematical model based on the actual operation effect. The fault occurrence condition of the process is determined by analyzing the residual error of the system, and qualitative and quantitative description is carried out on the connection relation and the fault propagation mode among the units in the process.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment early warning, and in particular to an operation monitoring and early warning method for intelligent energy-saving control equipment. Background Art

[0002] Intelligent energy-saving control equipment has a wide range of application scenarios. In addition to the industrial manufacturing field, this method can also be applied to multiple fields such as building safety monitoring, environmental monitoring, traffic management, and energy management. For example, in building safety monitoring, this method can monitor the structural health and operating status of the building in real time, and promptly discover and deal with potential safety issues; in environmental monitoring, this method can monitor the data of environmental factors such as air quality, water quality, and noise in real time, providing strong support for environmental protection; in traffic management, this method can monitor road conditions and traffic flow in real time, providing data support for traffic planning and optimization.

[0003] In a modern industrial manufacturing environment, the stability and reliability of equipment are key factors in ensuring production efficiency and quality. However, with the increasing complexity of equipment, traditional manual monitoring methods can no longer meet the requirements for accuracy and real-time warning of equipment failures. Summary of the invention

[0004] In order to solve the above technical problems, a method for monitoring and early warning the operation of an intelligent energy-saving control device is provided. This technical solution solves the problems raised in the above background technology.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is: An operation monitoring and early warning method for intelligent energy-saving control equipment comprises: Install sensors on the equipment to collect the operation data of the equipment in real time, and transmit the collected data to the data center wirelessly. The sensors include temperature sensors, humidity sensors, current sensors and voltage sensors; After receiving the raw data from the sensor, the data center cleans the data, removes abnormal and duplicate data, and extracts warning-related features from the cleaned data; Based on the equipment operation mechanism, a mathematical model of intelligent energy-saving control equipment is established to predict the system output and compare it with the actual measurement value to obtain the system residual; Analyze the system residuals, determine the occurrence of process failures, analyze the failure types, and qualitatively and quantitatively describe the connection relationship between each unit in the process and the failure propagation mode based on the experience and knowledge of experts and operators; When abnormal signs appear in the equipment, the reasoning ability of process experts in monitoring is simulated through reasoning and deduction to complete the equipment failure warning; Mining the intrinsic information in process data, establishing a mathematical model for fault warning and expressing process status, and monitoring the implementation process based on the model; Use historical data to train the fault warning mathematical model, adjust the model parameters, and input the newly collected data into the trained model in real time to perform fault warning analysis; Based on the actual operation results, the model parameters of the fault warning mathematical model are optimized.

[0006] Preferably, the analysis of the system residuals, determination of the occurrence of the process fault, analysis of the fault type, and qualitative and quantitative description of the connection relationship and fault propagation mode between the units in the process based on the experience and knowledge of experts and operators specifically include: Use chart tools to draw a trend chart of residuals changing over time, observe the residual trend chart, and analyze the changing pattern of residuals; Based on the characteristics of the system, a residual threshold is set. When the residual exceeds the threshold, the current system is output to have a fault, and the fault warning mechanism is triggered to send warning information to relevant personnel; Combine residual data with system characteristics to analyze the location of the fault, and further confirm the specific location of the fault through on-site inspection and equipment testing; Based on the manifestation, cause and impact of the fault, the fault is divided into sensor fault, controller fault and actuator fault; Extracting typical features of each fault type, wherein the typical features include the frequency, duration and impact range of the fault; Draw the unit connection diagram of the intelligent energy-saving control equipment, and analyze the propagation path and impact range of the current fault in the system; A fault propagation model is established using Bayesian networks to quantitatively describe the fault propagation process. In the Bayesian network, each fault and symptom is defined as a node. The fault node represents the possible type of fault, and the symptom node represents the symptoms or abnormal manifestations related to the fault. Based on the characteristics of the system and the fault propagation mechanism, the connection relationship between the fault node and the symptom node is determined. The connection relationship reflects the propagation mode of the fault in the system. For each fault node, determine the prior probability of its occurrence based on historical data, and for each symptom node, determine its conditional probability under a given fault node state; In Bayesian networks, the propagation of faults is analyzed by simulating fault injection, setting the state of a specific faulty node to “occurred” and observing the state changes of other nodes; After the fault is injected, the posterior probabilities of other nodes are calculated based on the reasoning ability of the Bayesian network, and the output is the probability of other nodes failing or showing specific symptoms under a given fault state; By observing the changes in the posterior probability, we can determine which nodes are most relevant to a specific fault, locate the source of the fault, and take corresponding repair measures.

[0007] Preferably, after the equipment shows abnormal signs, simulating the reasoning ability of process experts in monitoring by way of reasoning and deduction to complete the equipment failure warning specifically includes: Based on the working principle and failure mode of the equipment, a fault tree model is constructed. The fault tree connects basic events and top events with logic gates to reflect the propagation path and logical relationship of the fault. The fault cause is input into the basic event, and the equipment failure is input into the top event. Based on historical data, the probability of occurrence of each basic event is assigned, and based on abnormal symptoms and combined with the fault tree model, a fault hypothesis is proposed; Track the propagation path of the fault in the fault tree, and update the probability of basic events in the fault tree in real time based on new observation data to improve the accuracy of early warning; Based on the result of reasoning and deduction, determine the warning level of the fault, which includes emergency warning, general warning and potential risk prompt; The early warning information is sent to relevant personnel through the automated system, and the early warning information includes key information such as fault type, possible cause, urgency and recommended measures.

[0008] Compared with the prior art, the present invention has the following beneficial effects: It improves the real-time monitoring and can quickly issue early warnings when equipment shows abnormal signs, effectively reducing the impact of faults on production operations. It also provides qualitative and quantitative descriptions of the connection relationships between units in the process and the fault propagation patterns. This integration makes fault warnings more comprehensive and in-depth, helping operators to quickly locate the source of the fault and take appropriate treatment measures. It uses historical data to train the model and adjust parameters, making the model more adaptable to the actual operating status of the equipment. It optimizes the model parameters based on the actual operating results, further improving the accuracy and stability of the early warning system. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 It is a flow chart of the operation monitoring and early warning method of the intelligent energy-saving control device of the present invention; Figure 2 A flow chart of a method for cleaning data, removing abnormal and duplicate data, and extracting warning-related features from the cleaned data according to the present invention; Figure 3 A flow chart of a method for establishing a mathematical model of an intelligent energy-saving control device according to the present invention; Figure 4 A flow chart of the connection relationship between the units and the fault propagation mode method in the qualitative and quantitative description process of the present invention; Figure 5 The present invention simulates the reasoning ability of process experts in monitoring by way of reasoning and deduction, and completes the flow chart of the equipment failure early warning method; Figure 6 To mine the intrinsic information in the process data of the present invention, a mathematical model for fault warning and a flow chart of the method for expressing the process status are established; Figure 7 This is a flow chart of the method for training a fault warning mathematical model using historical data according to the present invention. DETAILED DESCRIPTION

[0010] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0011] Reference Figure 1 As shown, an operation monitoring and early warning method for intelligent energy-saving control equipment includes: Install sensors on the equipment to collect the operation data of the equipment in real time, and transmit the collected data to the data center wirelessly. The sensors include temperature sensors, humidity sensors, current sensors and voltage sensors; After receiving the raw data from the sensor, the data center cleans the data, removes abnormal and duplicate data, and extracts warning-related features from the cleaned data; Based on the equipment operation mechanism, a mathematical model of intelligent energy-saving control equipment is established to predict the system output and compare it with the actual measurement value to obtain the system residual; Analyze the system residuals, determine the occurrence of process failures, analyze the failure types, and qualitatively and quantitatively describe the connection relationship between each unit in the process and the failure propagation mode based on the experience and knowledge of experts and operators; When abnormal signs appear in the equipment, the reasoning ability of process experts in monitoring is simulated through reasoning and deduction to complete the equipment failure warning; Mining the intrinsic information in process data, establishing a mathematical model for fault warning and expressing process status, and monitoring the implementation process based on the model; Use historical data to train the fault warning mathematical model, adjust the model parameters, and input the newly collected data into the trained model in real time to perform fault warning analysis; Based on the actual operation results, the model parameters of the fault warning mathematical model are optimized.

[0012] Reference Figure 2 As shown in the figure, after receiving the raw data from the sensor, the data center performs data cleaning to remove abnormal and duplicate data, and extracts warning-related features from the cleaned data, including: The data center receives the raw data from each sensor and makes a preliminary understanding of the received data, checking the structure, data type, variable meaning and missing values ​​of the raw data; Through statistical description and data visualization, data characteristics can be intuitively displayed to identify outliers and raw data trends; For numerical data, the mean is used to fill missing values, and for categorical data, the most frequent value is used to fill missing values; Identify outliers in the original data by measuring the proportional relationship between the difference between the original data and the overall mean of the data and the standard deviation of the original data; For the identified outliers, the boundary value replacement method is used to update the outliers; Use the timestamp as the unique identifier of the data to determine whether the data is duplicated. If so, delete the duplicate records or merge the duplicate records based on business needs. If not, do not output them. Select relevant features based on early warning needs. When predicting equipment failure, select temperature and vibration frequency features. When predicting environmental changes, select humidity and air pressure features. Extracting numerical features from the data, wherein the numerical features include temperature, humidity, current value, and voltage value; When the data contains text information, use text analysis technology to extract keyword features from the text information of the data; For time series data, the moving average method is used to extract the trend characteristics of the time series data.

[0013] When predicting equipment failures, select temperature and vibration frequency features, which can reflect the operating status and potential failures of the equipment. When predicting environmental changes, select humidity and air pressure features, which can reflect the changes in humidity and air pressure in the environment. Extract numerical features from the data, including temperature, humidity, current value, and voltage value, etc. These features can be directly used for subsequent data analysis and early warning model construction; if the data contains text information (such as equipment name, fault description, etc.), use text analysis technology (such as keyword extraction, sentiment analysis, etc.) to extract keyword features from the text information of the data, which will help data center personnel better understand the data content and discover potential associations and patterns; for time series data, use the moving average method to extract the trend characteristics of time series data. The moving average method can smooth data fluctuations and reveal long-term trends and cyclical changes in the data.

[0014] Reference Figure 3 As shown in the figure, based on the equipment operation mechanism, a mathematical model of intelligent energy-saving control equipment is established to predict the system output and compare it with the actual measured value to obtain the system residual, which includes: Obtain the main functions, performance requirements and energy-saving targets of intelligent energy-saving control equipment; In-depth analysis of the physical principles, working principles, and energy conversion processes of the equipment, and the characteristics and interactions between sensors, controllers, and actuator components; Acquiring operating data of the device under at least one working condition, the operating data including energy consumption, load, temperature and humidity parameters; Based on the equipment operation mechanism, the mathematical model of intelligent energy-saving control equipment is constructed using differential equations. The input of the model is the load and environmental parameters, and the output is the energy consumption and efficiency parameters. Determine the unknown parameters in the mathematical model of intelligent energy-saving control equipment through data fitting and experimental verification, and use known data to verify the mathematical model of intelligent energy-saving control equipment to ensure that the model can reflect the operating characteristics of the equipment; Align the predicted data with the actual measured data in two dimensions in time and space, keeping the timestamps and measurement points consistent; The difference between the predicted value and the actual measured value is calculated to obtain the system residual.

[0015] Select representative operating conditions (such as full load, no load, partial load, etc.), and use measuring instruments or the equipment's own monitoring system to collect operating data. The collected data should include key parameters such as energy consumption (such as electricity consumption), load (such as the amount of material or workload processed by the equipment), temperature, humidity, etc.

[0016] Reference Figure 4 As shown, the system residuals are analyzed to determine the occurrence of process failures, analyze the failure types, and based on the experience and knowledge of experts and operators, qualitatively and quantitatively describe the connection relationship between the units in the process and the fault propagation mode, including: Use chart tools to draw a trend chart of residuals changing over time, observe the residual trend chart, and analyze the changing pattern of residuals; Based on the characteristics of the system, a residual threshold is set. When the residual exceeds the threshold, the current system is output to have a fault, and the fault warning mechanism is triggered to send warning information to relevant personnel; Combine residual data with system characteristics to analyze the location of the fault, and further confirm the specific location of the fault through on-site inspection and equipment testing; Based on the manifestation, cause and impact of the fault, the fault is divided into sensor fault, controller fault and actuator fault; Extracting typical features of each fault type, wherein the typical features include the frequency, duration and impact range of the fault; Draw the unit connection diagram of the intelligent energy-saving control equipment, and analyze the propagation path and impact range of the current fault in the system; A fault propagation model is established using Bayesian networks to quantitatively describe the fault propagation process. In the Bayesian network, each fault and symptom is defined as a node. The fault node represents the possible type of fault, and the symptom node represents the symptoms or abnormal manifestations related to the fault. Based on the characteristics of the system and the fault propagation mechanism, the connection relationship between the fault node and the symptom node is determined. The connection relationship reflects the propagation mode of the fault in the system. For each fault node, determine the prior probability of its occurrence based on historical data, and for each symptom node, determine its conditional probability under a given fault node state; In Bayesian networks, the propagation of faults is analyzed by simulating fault injection, setting the state of a specific faulty node to “occurred” and observing the state changes of other nodes; After the fault is injected, the posterior probabilities of other nodes are calculated based on the reasoning ability of the Bayesian network, and the output is the probability of other nodes failing or showing specific symptoms under a given fault state; By observing the changes in the posterior probability, we can determine which nodes are most relevant to a specific fault, locate the source of the fault, and take corresponding repair measures.

[0017] A fault propagation model is established using a Bayesian network. In the Bayesian network, each fault and symptom is defined as a node. The fault node represents the type of fault that may occur, and the symptom node represents the symptoms or abnormal manifestations related to the fault. Based on the characteristics of the system and the fault propagation mechanism, the connection relationship between the fault node and the symptom node is determined. These connection relationships reflect the way the fault propagates in the system and are the basis of Bayesian network reasoning. For each fault node, its prior probability of occurrence is determined based on historical data. For each symptom node, its conditional probability under a given fault node state is determined. These probability values ​​are used for reasoning calculations in the Bayesian network.

[0018] Reference Figure 5 As shown in the figure, after the equipment shows abnormal signs, the reasoning ability of process experts in monitoring is simulated through reasoning and deduction to complete the equipment failure early warning, which specifically includes: Based on the working principle and failure mode of the equipment, a fault tree model is constructed. The fault tree connects basic events and top events with logic gates to reflect the propagation path and logical relationship of the fault. The fault cause is input into the basic event, and the equipment failure is input into the top event. Based on historical data, the probability of occurrence of each basic event is assigned, and based on abnormal symptoms and combined with the fault tree model, a fault hypothesis is proposed; Track the propagation path of the fault in the fault tree, and update the probability of basic events in the fault tree in real time based on new observation data to improve the accuracy of early warning; Based on the result of reasoning and deduction, determine the warning level of the fault, which includes emergency warning, general warning and potential risk prompt; The early warning information is sent to relevant personnel through the automated system, and the early warning information includes key information such as fault type, possible cause, urgency and recommended measures.

[0019] Emergency warning: indicates that the equipment is about to fail and immediate action is required. General warning: indicates that the equipment has a potential risk of failure, but has not yet reached the level of urgency and requires close attention. Potential risk warning: indicates that the equipment may have certain problems, but has not yet reached the warning level and preventive maintenance is recommended.

[0020] Reference Figure 6 As shown in the figure, the intrinsic information in the process data is mined, a fault warning mathematical model is established and the process status is expressed, and the implementation process is monitored based on the model. Specifically, it includes: Extract the mean, variance, maximum and minimum statistical features of process data. The statistical features reflect the stability of the equipment operation status. The frequency spectrum characteristics of process data are extracted by Fourier transform method to analyze the periodic changes in the operation process of intelligent energy-saving control equipment; Analyze the correlation between the sensor data and extract the composite features that reflect the overall operating status of the intelligent energy-saving control equipment; Combine equipment characteristics and historical data to establish a state space model for intelligent energy-saving control equipment; The operating status of the equipment is represented as a state vector, and the operating status and fault warning information of the equipment are visualized in the form of a chart; The implementation process is monitored based on the state space model of intelligent energy-saving control equipment.

[0021] Perform fast Fourier transform on the collected time series data, convert the time domain signal into frequency domain signal, analyze the spectrum diagram, identify the periodic components in the equipment operation, such as motor rotation frequency, vibration frequency, etc., use correlation coefficient, covariance matrix and other methods to analyze the correlation between the sensor data, and identify the interdependent relationship. For example, the changes in current and voltage may be highly correlated, while the changes in temperature and pressure may reflect the thermal load status of the equipment.

[0022] Reference Figure 7 As shown, the fault warning mathematical model is trained using historical data, the model parameters are adjusted, and the newly collected data is input into the trained model in real time. The fault warning analysis specifically includes: Divide historical data into training set, validation set and test set; The training set is used to train the state space model of intelligent energy-saving control equipment; The validation set is used to adjust the state space model parameters of the intelligent energy-saving control equipment; The test set is used to evaluate the performance of the state space model of intelligent energy-saving control equipment; The newly collected data is input into the trained model in real time for fault warning analysis.

[0023] During the training process, use the validation set data to regularly evaluate the model performance, such as accuracy, recall, F1 score, etc., and adjust the model parameters based on the evaluation results, such as increasing the number of iterations, adjusting the learning rate, changing the model structure, etc., to improve the performance of the model on the validation set.

[0024] Furthermore, the present solution also proposes a computer-readable storage medium on which a computer-readable program is stored. When the computer-readable program is called, the operation monitoring and early warning method of the above-mentioned intelligent energy-saving control device is executed.

[0025] It is understandable that the storage medium may be a magnetic medium, such as a floppy disk, a hard disk, or a magnetic tape; an optical medium, such as a DVD; or a semiconductor medium, such as a solid state drive (SSD).

[0026] To sum up, the advantages of the present invention are: it improves the real-time monitoring, and can quickly issue an early warning when the equipment shows abnormal signs, effectively reducing the impact of faults on production operations, and qualitatively and quantitatively describes the connection relationship between the units in the process and the fault propagation mode. This integration makes the fault warning more comprehensive and in-depth, which helps operators to quickly locate the source of the fault and take corresponding treatment measures. The model is trained and the parameters are adjusted using historical data, so that the model is more adapted to the actual operating status of the equipment, and the model parameters are optimized based on the actual operating results, further improving the accuracy and stability of the early warning system.

[0027] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. An operation monitoring and early warning method for intelligent energy-saving control equipment, characterized in that: include: Install sensors on the equipment to collect the operation data of the equipment in real time, and transmit the collected data to the data center wirelessly. The sensors include temperature sensors, humidity sensors, current sensors and voltage sensors; After receiving the raw data from the sensor, the data center cleans the data, removes abnormal and duplicate data, and extracts warning-related features from the cleaned data; Based on the equipment operation mechanism, a mathematical model of intelligent energy-saving control equipment is established to predict the system output and compare it with the actual measurement value to obtain the system residual; Analyze the system residuals, determine the occurrence of process failures, analyze the failure types, and qualitatively and quantitatively describe the connection relationship between each unit in the process and the failure propagation mode based on the experience and knowledge of experts and operators; When abnormal signs appear in the equipment, the reasoning ability of process experts in monitoring is simulated through reasoning and deduction to complete the equipment failure warning; Mining the intrinsic information in process data, establishing a mathematical model for fault warning and expressing process status, and monitoring the implementation process based on the model; Use historical data to train the fault warning mathematical model, adjust the model parameters, and input the newly collected data into the trained model in real time to perform fault warning analysis; Based on the actual operation results, the model parameters of the fault warning mathematical model are optimized.

2. The operation monitoring and early warning method of an intelligent energy-saving control device according to claim 1 is characterized in that: After receiving the raw data from the sensor, the data center performs data cleaning to remove abnormal and duplicate data, and extracts warning-related features from the cleaned data, including: The data center receives the raw data from each sensor and makes a preliminary understanding of the received data, checking the structure, data type, variable meaning and missing values ​​of the raw data; Through statistical description and data visualization, data characteristics can be intuitively displayed to identify outliers and raw data trends; For numerical data, the mean is used to fill missing values, and for categorical data, the most frequent value is used to fill missing values; Identify outliers in the original data by measuring the proportional relationship between the difference between the original data and the overall mean of the data and the standard deviation of the original data; For the identified outliers, the boundary value replacement method is used to update the outliers; Use the timestamp as the unique identifier of the data to determine whether the data is duplicated. If so, delete the duplicate records or merge the duplicate records based on business needs. If not, do not output them. Select relevant features based on early warning needs. When predicting equipment failure, select temperature and vibration frequency features. When predicting environmental changes, select humidity and air pressure features. Extracting numerical features from the data, wherein the numerical features include temperature, humidity, current value, and voltage value; When the data contains text information, use text analysis technology to extract keyword features from the text information of the data; For time series data, the moving average method is used to extract the trend characteristics of the time series data.

3. The operation monitoring and early warning method of an intelligent energy-saving control device according to claim 2 is characterized in that: Based on the equipment operation mechanism, a mathematical model of intelligent energy-saving control equipment is established to predict the system output, and the system residual is obtained by comparing it with the actual measurement value, which specifically includes: Obtain the main functions, performance requirements and energy-saving targets of intelligent energy-saving control equipment; In-depth analysis of the physical principles, working principles, and energy conversion processes of the equipment, and the characteristics and interactions between sensors, controllers, and actuator components; Acquiring operating data of the device under at least one working condition, the operating data including energy consumption, load, temperature and humidity parameters; Based on the equipment operation mechanism, the mathematical model of intelligent energy-saving control equipment is constructed using differential equations. The input of the model is the load and environmental parameters, and the output is the energy consumption and efficiency parameters. Determine the unknown parameters in the mathematical model of intelligent energy-saving control equipment through data fitting and experimental verification, and use known data to verify the mathematical model of intelligent energy-saving control equipment to ensure that the model can reflect the operating characteristics of the equipment; Align the predicted data with the actual measured data in two dimensions in time and space, keeping the timestamps and measurement points consistent; The difference between the predicted value and the actual measured value is calculated to obtain the system residual.

4. The operation monitoring and early warning method of an intelligent energy-saving control device according to claim 3 is characterized in that: The analysis of the system residuals, determination of the occurrence of process failures, analysis of the failure types, and qualitative and quantitative description of the connection relationship between the units and the failure propagation mode in the process based on the experience and knowledge of experts and operators specifically include: Use chart tools to draw a trend graph of residuals changing over time, observe the residual trend graph, and analyze the changing pattern of residuals; Based on the characteristics of the system, a residual threshold is set. When the residual exceeds the threshold, the current system is output to have a fault, and the fault warning mechanism is triggered to send warning information to relevant personnel; Combine residual data with system characteristics to analyze the location of the fault, and further confirm the specific location of the fault through on-site inspection and equipment testing; Based on the manifestation, cause and impact of the fault, the fault is divided into sensor fault, controller fault and actuator fault; Extracting typical features of each fault type, wherein the typical features include the frequency, duration and impact range of the fault; Draw the unit connection diagram of the intelligent energy-saving control equipment, and analyze the propagation path and impact range of the current fault in the system; A fault propagation model is established using Bayesian networks to quantitatively describe the fault propagation process. In the Bayesian network, each fault and symptom is defined as a node. The fault node represents the possible type of fault, and the symptom node represents the symptoms or abnormal manifestations related to the fault. Based on the characteristics of the system and the fault propagation mechanism, the connection relationship between the fault node and the symptom node is determined. The connection relationship reflects the propagation mode of the fault in the system. For each fault node, determine the prior probability of its occurrence based on historical data, and for each symptom node, determine its conditional probability under a given fault node state; In the Bayesian network, the propagation process of the fault is analyzed by simulating fault injection, setting the state of a specific faulty node to "occurred" and observing the state changes of other nodes; After the fault is injected, the posterior probabilities of other nodes are calculated based on the reasoning ability of the Bayesian network, and the output is the probability of other nodes failing or showing specific symptoms under a given fault state; By observing the changes in the posterior probability, we can determine which nodes are most relevant to a specific fault, locate the source of the fault, and take corresponding repair measures.

5. The operation monitoring and early warning method of an intelligent energy-saving control device according to claim 4 is characterized in that: When abnormal signs appear in the equipment, the reasoning ability of process experts in monitoring is simulated by reasoning and deduction to complete the equipment failure warning, which specifically includes: Based on the working principle and failure mode of the equipment, a fault tree model is constructed. The fault tree connects basic events and top events with logic gates to reflect the propagation path and logical relationship of the fault. The fault cause is input into the basic event, and the equipment failure is input into the top event. Based on historical data, the probability of occurrence of each basic event is assigned, and based on abnormal symptoms and combined with the fault tree model, a fault hypothesis is proposed; Track the propagation path of the fault in the fault tree, and update the probability of basic events in the fault tree in real time based on new observation data to improve the accuracy of early warning; Based on the result of reasoning and deduction, determine the warning level of the fault, which includes emergency warning, general warning and potential risk prompt; The early warning information is sent to relevant personnel through the automated system, and the early warning information includes key information such as fault type, possible cause, urgency and recommended measures.

6. The operation monitoring and early warning method of an intelligent energy-saving control device according to claim 5 is characterized in that: The mining of intrinsic information in process data, establishing a fault warning mathematical model and expressing process status, and monitoring the implementation process based on the model specifically include: Extract the mean, variance, maximum and minimum statistical features of process data. The statistical features reflect the stability of the equipment operation status. The frequency spectrum characteristics of process data are extracted by Fourier transform method to analyze the periodic changes in the operation process of intelligent energy-saving control equipment; Analyze the correlation between the sensor data and extract the composite features that reflect the overall operating status of the intelligent energy-saving control equipment; Combine equipment characteristics and historical data to establish a state space model for intelligent energy-saving control equipment; The operating status of the equipment is represented as a state vector, and the operating status and fault warning information of the equipment are visualized in the form of a chart; The implementation process is monitored based on the state space model of intelligent energy-saving control equipment.

7. The operation monitoring and early warning method of an intelligent energy-saving control device according to claim 6 is characterized in that: The method of using historical data to train the fault warning mathematical model, adjusting model parameters, and inputting newly collected data into the trained model in real time to perform fault warning analysis specifically includes: Divide historical data into training set, validation set and test set; The training set is used to train the state space model of intelligent energy-saving control equipment; The validation set is used to adjust the state space model parameters of the intelligent energy-saving control equipment; The test set is used to evaluate the performance of the state space model of intelligent energy-saving control equipment; The newly collected data is input into the trained model in real time for fault warning analysis.

Images