Intelligent sensing system and method based on multi-dimensional state monitoring

Abstract

The invention belongs to the technical field of thermal power auxiliary equipment maintenance, and particularly relates to an intelligent sensing system and method based on multi-dimensional state monitoring, and the system comprises a state monitoring module, an intelligent maintenance decision module, a maintenance operation standardization module, and a real-time collection equipment physical signal. When an abnormal condition or an early fault symptom occurs, the edge end can carry out abnormal judgment and prompt and alarm; and performing modeling analysis on the data at the cloud, acquiring real-time operation data of the equipment on line to construct a knowledge graph model, performing reasoning and decision analysis on the operation state of the equipment, and identifying a fault risk influencing the reliability of the equipment by combining reasoning and decision analysis results. According to the method, equipment signals are grabbed to analyze the overall reliability, risk and fault of the equipment, standardized and normalized operation guidance is given, and a maintenance strategy is perfected and reconstructed so as to continuously update and iterate.

Description

Technical Field

[0001] The present invention belongs to the technical field of maintenance of thermal power auxiliary equipment, and particularly relates to an intelligent perception system and method based on multi-dimensional condition monitoring. Background Art

[0002] Thermal power auxiliary equipment generally includes induced draft fans, forced draft fans, primary fans, steam-driven feed water pumps, condensate pumps, etc. During daily operation, regular maintenance is required or timely repair is needed when a failure occurs. Therefore, regular inspections by staff are very necessary. For example, signals such as the temperature and sound of each machine are checked to detect abnormalities in a timely manner.

[0003] During daily maintenance, maintenance personnel can improve the maintenance efficiency by obtaining signals such as the temperature and sound of each machine based on Internet of Things technology. For example, the Chinese invention with publication number CN115062653A, an analysis and maintenance system based on a thermal power plant steam turbine, can issue early warnings at different levels through macroscopic and microscopic analysis of the steam turbine, facilitating maintenance personnel to carry out maintenance work of different intensities.

[0004] Regarding the maintenance process based on Internet of Things technology, although early warnings at different levels are divided, there is a lack of guiding opinions for maintenance. Therefore, the invention provides an intelligent perception system and method based on multi-dimensional condition monitoring, which effectively utilizes fault data and archived maintenance data to continuously update the equipment fault model, provides data support for maintenance decision-making, and makes the maintenance work more targeted. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent perception system and method based on multi-dimensional condition monitoring, which captures equipment signals to analyze the overall reliability, risks, and faults of the equipment, gives standardized and normalized operation guidance, and improves and reconstructs the maintenance strategy for continuous update and iteration.

[0006] The technical solution adopted by the present invention is specifically as follows: An intelligent perception system and method based on multi-dimensional condition monitoring, including: A condition monitoring module that continuously collects physical signals of the equipment. When an abnormal situation or early fault symptom occurs, the edge side can make an abnormal judgment and give an alarm; and model and analyze the data in the cloud, and at the same time obtain the real-time operation data of the equipment online to construct a knowledge graph model for reasoning and decision-making analysis of the equipment operation status; A smart maintenance decision-making module that combines the results of reasoning and decision-making analysis to identify the fault risks affecting equipment reliability, and establish fault and maintenance reliability data standards, a digital maintenance data collection and exchange system, and an equipment reliability model; and then optimize the maintenance tasks and plans; The maintenance operation standardization module establishes an exchange standard for maintenance content and data, conducts standardized design and rational reconstruction of maintenance operations, and uses fault data and archived maintenance data to continuously update equipment faults, providing data support for the intelligent maintenance decision-making module.

[0007] As an alternative, the condition monitoring module includes: The software basic unit is used for the configuration of the intelligent maintenance decision-making module, the maintenance operation standardization module, and user management permissions, realizing the flow of data streams among the condition monitoring module, the intelligent maintenance decision-making module, and the maintenance operation standardization module; The condition perception and early warning unit collects physical signals and working condition data through sensors, conducts real-time monitoring and analysis of the operation status of equipment, and generates early warning information when abnormalities are found to prompt users to pay attention to fault risks; The asset structure tree is used for equipment master data management, spare part resource data management, and human resource data management; The fault knowledge base management unit includes a standard fault knowledge base, an equipment structure decomposition library, a fault mode library, a maintenance strategy library, and a health index library.

[0008] As an alternative, the intelligent maintenance decision-making module includes: The FMEA unit sorts out and manages fault factors, discovers existing faults and impacts to help evaluate the reliability and availability of units, systems, or equipment; The risk dynamic analysis unit evaluates the dynamic risk of equipment by constructing a visual risk matrix based on the impact of equipment fault consequences on the environment, cost, and safety dimensions and using the equipment fault law formed by the regression of equipment reliability data, assisting equipment management personnel in making auxiliary decisions; The equipment defect management unit is used for data collection and defect information improvement. Among them, data collection includes defect registration, defect confirmation, operation information, and defect elimination processing, and defect information improvement automatically collects, fills in, and improves defect content through each page field and defect category; The data regression analysis unit regresses historical defect data and maintenance record data, and the parameters of the distribution curve are obtained through regression of historical data or user definition to form a reliability database for comprehensively evaluating the reliability, availability, and maintainability of units; The simulation and optimization unit, based on the reliability database, determines the magnitude of economic risks and failure losses, predicts future production capacity and benefits, and completes the modeling and simulation of the failure causes, failure modes, and inspection and maintenance strategies of the unit and equipment on the basis of simulating the maintenance strategy based on equipment reliability and mean time between failures; among them, considering the impact of maintenance strategies, technological transformation plans, and spare parts strategies on asset performance, by adjusting the cycle and content of the maintenance strategy and comparing the performance differences, a more optimal strategy is formulated.

[0009] As an alternative, the FMEA unit includes: The equipment structure decomposition layer decomposes and sorts out the enterprise assets, decomposes the equipment into repairable parts, and at the same time associates the equipment failure modes with the parts of the equipment. The function analysis layer describes the functions of the equipment or parts and the reference requirements for realizing the functions. The failure mode analysis layer sorts out and defines all possible failure modes of the equipment or parts according to the functions, and at the same time corresponds the equipment failure modes with the failure phenomena of the equipment. The failure impact analysis layer defines the consequence impacts of the failure modes and comprehensively evaluates the multi-dimensional consequence impacts caused by equipment failures. The failure cause analysis layer defines the failure causes corresponding to the failure modes, combines the actual situation on site, comprehensively evaluates the equipment consequence impacts and the occurrence probabilities of the failure causes, and thus analyzes and classifies the hazards of the equipment or parts. The maintenance strategy analysis layer determines the maintenance strategy based on the failure mode, failure impact, failure probability, maintenance time, and the impact of the equipment on the operation objectives, and realizes the management and maintenance of the maintenance content.

[0010] As an alternative, the management of the maintenance content includes maintenance strategy category, maintenance strategy name, maintenance strategy content, and required resources.

[0011] As an alternative, the risk dynamic analysis unit includes: The risk matrix customization layer defines the risk dimensions, risk grading, risk matrix, and its display color of the enterprise equipment. The dynamic risk assessment layer analyzes the failure consequences of the equipment, and based on the historical data of equipment defects, uses the data regression algorithm to evaluate the dynamic risk of the equipment, and calculates the number of failure causes of each risk level in the risk matrix.

[0012] As an alternative, the overhaul operation standardization module includes: The maintenance strategy push and supervision unit, through the interface with the external maintenance execution system, realizes the timely push and execution after the expiration of the preventive maintenance plan, and the tracking closed-loop of timely updating the monitoring status after the execution is completed, serving as the exchange standard for the establishment of inspection and maintenance content and data; The overhaul operation standardization unit is used to standardize the content of the overhaul document packages for power industry major and minor repairs, guide the execution process of major and minor repairs, and comprehensively record the results of overhaul work; The equipment health management unit establishes a standard health model based on business data, constructs an equipment health evaluation standard, evaluates the comprehensive health status of the equipment, and through the relevant business data synchronized by the interface, forms an asset health assessment after calculation by the standard health model, and displays the health status and change trend in real time, so as to comprehensively determine the equipment health status, provide the main cause and secondary cause, support the decision-making of the unit overhaul plan, and assist in rationally reconstructing the overhaul operation.

[0013] As an alternative solution, the maintenance strategy push and supervision unit includes: The preventive maintenance strategy layer is used to display the maintenance strategy content included in the equipment after failure mode and effects analysis, or directly add the maintenance strategy of the equipment manually; The preventive maintenance plan and monitoring layer combines and packages the corresponding maintenance strategies of the equipment according to the preventive maintenance plan requirements, formulates the spare parts list required for maintenance, generates the corresponding preventive maintenance plan for the equipment, and configures the maintenance plan monitoring according to the cycle of the maintenance plan, and centrally displays the preventive maintenance plans in the monitoring state.

[0014] As an alternative solution, the overhaul operation standardization unit includes: The overhaul time plan details the overhaul work to be carried out, the time plan and the arrangement of relevant resources, including the list of equipment, the specific steps and requirements of the overhaul, and the division of labor and responsibilities of personnel; The overhaul operation instruction manual provides specific guidance and requirements, including the steps of the overhaul operation, operating procedures, safety measures, etc.; The overhaul record and report record the key information during the overhaul process, including the status of the equipment, the specific content of the overhaul, the working time and personnel; The overhaul tools and spare parts list the tools and equipment to be used, as well as the list of spare parts and consumables that may need to be replaced or used.

[0015] An intelligent perception method based on multi-dimensional condition monitoring includes the following steps: Step 1. Equipment supervision: Call the condition monitoring module to collect the physical signals of the equipment in real time. When abnormal situations or early fault signs occur, the edge side can make abnormal judgments and give alarms; Step 2. Real-time analysis: Send the collected physical signals to the cloud for modeling and analysis. At the same time, obtain the real-time operation data of the device online to construct a knowledge graph model, and conduct reasoning and decision-making analysis on the operation status of the device; Step 3. Give decisions: Invoke the intelligent maintenance decision-making module, combine the results of reasoning and decision-making analysis, identify the fault risks affecting the reliability of the device, and establish fault and maintenance reliability data standards, a digital inspection and maintenance data collection and exchange system, and a device reliability model; Step 4. Optimization: Integrate reliability data analysis, modeling and simulation, and strategy optimization to optimize the maintenance tasks and plans; Step 5. Maintenance operation guidance: Invoke the maintenance operation standardization module, establish an exchange standard for the inspection and maintenance content and data, and reconstruct the inspection and maintenance operations with standardized design and rationalization; Step 6. Real-time update: Capture the fault data and the archived inspection and maintenance data, send them to the cloud and conduct statistics and archive them into the intelligent maintenance decision-making module to realize the continuous update of device faults and provide data support for the intelligent maintenance decision-making module.

[0016] The technical effects achieved by the present invention are as follows: The present invention utilizes Internet of Things technology to enhance the real-time perception ability of the device state, combines operation and maintenance business data and operation condition data to achieve full-range data sharing and then complete fusion modeling, and realizes intelligent maintenance with the help of artificial intelligence algorithms; through device reliability analysis, overall planning of device status detection and device preventive maintenance is carried out to realize the scientific and reasonable formulation and optimization of maintenance strategies. The overall solution for device reliability and intelligent maintenance will improve and reconstruct the device fault and maintenance records, realize data standardization, and can mine fault rules based on these data to update the device reliability model and fault warning model, realizing the continuous iteration of data and models.

[0017] The present invention utilizes Internet of Things technology to enhance the real-time perception ability of the device state, adopts a cloud-edge collaboration mode to optimize the data processing method and efficiency. By fusing and modeling the perception data and condition data, fully combining and utilizing new technologies and the data results of existing systems, a device anomaly monitoring model is constructed. Scientifically and comprehensively evaluate the device health status, and then realize the support for the key device status maintenance decision-making.

[0018] The present invention analyzes the overall reliability, risks, and faults of the device, plans the overall equipment maintenance management work, and then effectively carries out preventive and predictive maintenance work, and realizes the continuous optimization and improvement of the system. Overall planning and adjustment of the maintenance work throughout the life cycle of the device, and forming a scientific and effective combination of maintenance strategies for preventive maintenance, predictive maintenance, and condition-based maintenance according to local conditions. Through the implementation and continuous improvement of the maintenance strategy, it forms an effective guidance for the maintenance management work.

[0019] By standardizing and normalizing the data during the maintenance execution process, the present invention effectively utilizes the value of business data, restores and updates the equipment reliability model, and thus provides input for the optimization and continuous improvement of strategies. And it uses the execution results to check and improve the plan, forming a complete PDCA (a quality management method, also known as Deming Cycle) cycle in the process of the equipment operation and maintenance management system. Brief Description of the Drawings

[0020] Figure 1 It is the system block diagram of the intelligent perception system based on multi-dimensional state monitoring in Embodiment 1 of the present invention; Figure 2 It is the flowchart of the intelligent perception method based on multi-dimensional state monitoring in Embodiment 2 of the present invention. Detailed Embodiments

[0021] In order to make the purpose and advantages of the present invention clearer, the present invention will be specifically described below in conjunction with embodiments. It should be understood that the following text is only used to describe one or several specific implementation manners of the present invention, and does not strictly limit the scope of protection specifically claimed by the present invention.

[0022] Embodiment 1: The intelligent maintenance decision support system for key auxiliary equipment of thermal power based on multi-dimensional state perception in this embodiment uses Internet of Things technology to strengthen the real-time multi-dimensional state perception of equipment, comprehensively evaluate the equipment health status in combination with equipment reliability, operation and maintenance conditions, etc., and then improve the monitoring and maintenance effects. To further improve the accuracy, timeliness, and integrity of reliability data and lay a solid foundation for the in-depth application of reliability data, reliability analysis technology is used to construct a reliability model in a data-driven manner based on the study of equipment failure laws, determine the optimal combination of various maintenance strategies such as preventive maintenance, condition-based maintenance, and corrective maintenance of equipment, and realize scientific decision-making for maintenance. Establish a data exchange standard for inspection and maintenance, and standardize and reconstruct preventive maintenance work, effectively accumulate reliability data and achieve in-depth application, and realize the improvement of the efficiency of the maintenance work process and the value of archived record data.

[0023] As Figure 1 shown, an intelligent perception system based on multi-dimensional state monitoring includes the following functional modules: A. Software Foundation The software foundation is the basic management function of the software such as function module management configuration, user management permission configuration, etc.; through this module, all function modules of the software are effectively managed, assisting users to complete the basic function configuration for the requirements of each function module, realizing the business flow of data streams within each function module, and meeting the personalized function requirements of user business management through configuration; The system provides independent user management capabilities, supports functions such as user creation, modification, and password reset, and also supports flexible user management functions, including the configuration of user groups, roles, and permissions. Different roles and user groups can be configured with different data or system function permissions; The system can flexibly allocate users' access rights to data, supports a user having multiple roles, and enables convenient management of various types of permissions; This system supports exporting some data into files or importing data. In addition to supporting the project file types of the system itself, the file types for import and export also support the common Excel format; The system provides a to-do task list and in-site message reminders, and can be adapted to support email and SMS notification functions; The system provides a flexible workflow configuration function, supporting users to adjust the approval process of tasks according to management needs; B. Asset Structure Tree (B1) Equipment Master Data Management Build an interface with external systems to obtain equipment master data from them, construct a complete asset digital model, store and manage enterprise equipment master data information, and support manual import of enterprise equipment master data; through sorting out the superior-subordinate relationships of enterprise assets, establish a tree structure of unit-system-equipment to construct the asset structure tree; help the enterprise clarify the relationships among units, systems, and equipment, and provide a data basis for further reliability analysis; The asset structure tree and asset master data can share a set of equipment codes (KKS codes) with the intelligent enterprise control platform and use the same coding system; (B2) Spare Parts Resource Data Management Through associating with the spare parts management function in the intelligent enterprise control platform, realize the unified management and maintenance of spare parts resources, including resource names, resource descriptions, resource codes, procurement price information, spare parts model information, etc.; through keyword query methods, quickly realize the query of spare parts resource information and provide data support for maintenance cost estimation; (B3) Human Resources Data Management Realize the management and maintenance of human resources basic data, including data management in dimensions such as human resources names, resource descriptions, internal personnel numbers, external personnel numbers, fixed costs, time costs, model information, etc.; through keyword query methods, quickly realize the query of human resources information and provide data support for maintenance cost estimation; C. Multidimensional Status Perception and Early Warning Using Internet of Things technology, a large amount of relevant perception data and working condition data are collected through a variety of sensor devices and system integration means. The perception data includes physical signals such as sound, vibration, and temperature captured by sensors, and the working condition data includes current, voltage, rotational speed, and power, etc.; adopting a cloud-edge collaboration mode, the perception data is synchronously processed and matched and diagnosed for typical anomalies at the edge side. The cloud uses low-resource non-cooperative and time series modeling methods to fuse and model the perception data and working condition data, and adopts technologies such as fine feature extraction and knowledge graph reasoning and decision-making to realize abnormal alarm and fault diagnosis of the equipment, laying a foundation for realizing health status assessment and intelligent maintenance; The intelligent sensor hardware device includes: an intelligent sensor with various monitoring and perception capabilities such as sound, vibration, and temperature, supporting wired and wireless data transmission methods to achieve 24-hour uninterrupted real-time monitoring; for the multi-dimensional perception signals such as sound, vibration, and temperature collected by the perception device, using methods such as artificial intelligence and traditional mechanisms to comprehensively, deeply, and real-time monitor the equipment status and conduct intelligent early warning and diagnostic analysis, mainly including the following aspects: (C1) Status monitoring: By obtaining multi-dimensional perception signals such as sound, vibration, and temperature of the equipment, comprehensively monitor the equipment and its main components, and through the fusion analysis of multi-source data, fully understand the operation of the equipment; (C2) Abnormal detection and early warning: Comprehensively utilize technologies such as traditional mechanisms and artificial intelligence models to conduct real-time monitoring and analysis of the operation status of the equipment, and detect abnormal situations early; once an abnormality is found, the system generates corresponding early warning information to prompt the user to pay attention to possible failure risks, so as to take timely maintenance and preventive measures to avoid the occurrence or expansion of failures; (C3) Fault detection: When a fault occurs, the system can assist the user in fault analysis and location. By comparing the equipment operation data and identifying abnormal patterns, the system can help the user identify potential fault hazards and locate the specific cause of the fault, thereby guiding the progress of the maintenance work; Regarding the configuration of the intelligent sensor hardware device, the following arrangements are made in this embodiment: (1) Perception device Support the access of various types of Internet of Things collection devices such as sound, vibration, and temperature, provide the definition of the object model, the service of the object model, event definition, support event subscription, support network protocols such as MQTT and TCP, and support industrial mainstream communication protocols such as Modbus, 104, and 61850. Provide data access capabilities and the parsing and encapsulation of data content standard protocols, provide an SDK method for rapid custom development of protocols, support automatic, continuous or periodic collection, the collection period can be set and adjusted, support the transmission of large amounts of sensor data, and provide a distributed highly reliable data transmission, storage, and fault tolerance mechanism; among them, the main functions include: Physical model definition, using the physical model to uniformly describe sensing devices, including the definition of basic attributes, standard services, and events. The physical model is an abstract representation of a device, using a unified standard data structure and attributes to describe device attributes, commands, events, and other information; Device registration, defining the process and specifications for device registration, including interface and parameter requirements for device registration; Device acquisition parameter configuration, configuring parameters such as the sampling rate and acquisition mechanism (e.g., manual trigger, on-demand acquisition) for various types of data, including automatic, continuous, or periodic acquisition. The device acquisition period is configurable, and data is automatically synchronized; Data access and communication, supporting various communication protocols such as MQTT, TCP, etc. The content of the data protocol is configurable; Device monitoring and maintenance, real-time monitoring of the status and operation of sensing devices, promptly detecting and handling device failures; Device remote control and upgrade: remotely controlling sensing devices, such as data transmission service ports, time synchronization service ports, firmware upgrades, etc., providing remote management and maintenance functions for sensing devices; (2) Data acquisition and storage Sensing device data acquisition and storage, acquiring various physical data generated by sensing devices, including sound, vibration, temperature, ultrasound, angle, image, etc., storing the original data, and performing data processing such as feature extraction, data aggregation, and storage of result data on this basis to meet subsequent application requirements; Data storage, comprehensively using technologies such as time series databases and object databases to store original data and result data; (3) Status monitoring Receiving and processing data transmitted from sensing devices, and real-time monitoring and analyzing the operating status of devices through algorithm models and big data analysis methods (such as RCM analysis, Reliability Centered Maintenance, maintenance analysis centered on reliability, or time series analysis, machine learning, Predictive Analysis, etc.). By using algorithm models, the system can detect outliers, trend changes, etc. in the data and identify potential failure modes and risk factors; In addition, docking with the real-time data PI system to obtain other operating parameters, realizing data interconnection and sharing, and improving overall efficiency and collaboration; through the analysis model, the operating status of the device can be real-time monitored and analyzed, providing users with timely and comprehensive status displays and feedback evaluation results, and providing strong support for the operation and maintenance of the device; the main functions are as follows: The overall display of the device status can provide an overall display of the device operation by integrating the main sensory data, analysis results and other information of the device. Users can view the status of the device at different angles through operations such as rotation and zooming to further understand the working status of the device. Secondly, the device status information can also be displayed in the form of charts, etc. Abnormality and fault alarm: if an early abnormality or fault is found, the user will be reminded by highlighting the display, flashing the alarm light, etc., and the abnormality / fault status of the equipment will be displayed in real time, including the abnormality / fault type, the time when the abnormality / fault occurred, and the cause of the abnormality / fault, so that timely maintenance measures can be taken; Alarm record viewing provides alarm record function, summarizes and displays the historical data of equipment alarms and measuring point alarms, which is convenient for users to query and export; users can choose to configure alarm triggering algorithms, including threshold alarms and trend alarms; Abnormal data viewing, viewing real-time equipment status monitoring data, and supporting data playback function, you can select any time period and data category to play back status monitoring historical data; Combining the above functions, we can fully and intuitively understand the overall situation of the equipment, promptly discover abnormalities in equipment operation, evaluate the health of the equipment, and make corresponding decisions and adjustments to ensure the normal operation and performance optimization of the equipment; (4) Intelligent early warning Based on the continuously accumulated characteristic data of the equipment's entire life cycle, the degradation characteristic parameters are extracted (including the time axis correlation between the early fault sign characteristics and the typical characteristics of mid- and late-stage faults, as well as parameters such as degradation trend and influencing factor coefficient), and the fault development trend is predicted according to the current status of the equipment and the fault level, and the hidden early signs of faults are discovered in advance. The fault trend is predicted using the trend prediction model, the probability of fault occurrence is evaluated, and the possible faults of the equipment are predicted. The operation and maintenance personnel are assisted in formulating the optimal operation and maintenance plan, and the early abnormalities found are provided with early warning and alarm functions to locate specific faults, which are pushed to users through various means; (5) Intelligent fault detection On the basis of conventional data query, analysis and display, this system provides a professional intelligent fault diagnosis toolbox, which mainly covers the following functions: Audio noise reduction analysis and voiceprint comparison function: In addition to being used for abnormality monitoring and fault diagnosis, audio signal processing analysis also provides noise reduction analysis for abnormal sounds. After removing the interference of normal working sounds as much as possible, the abnormal signal part is highlighted and the signal analysis results in the time and frequency domains are given; Model online adaptive update function: As fault data accumulates, users can use the online adaptive model update function to provide a small amount of labeled data to fine-tune the fault diagnosis model and achieve better fault diagnosis results. Time series statistic calculation function. To facilitate experts in analyzing the characteristics of the series and manually adjusting the warning and alarm thresholds, a time series statistic function is provided to calculate statistical information such as the mean, standard deviation, trend, and inflection point of the series; Adaptive estimation of index thresholds. According to the false alarm frequency set by the user and possible fault samples, an adaptive estimation of threshold-type indicators is provided to find the optimal warning threshold; Series correlation analysis function. An arbitrary two-dimensional series correlation analysis tool is provided for users to facilitate the analysis of the relevance of faults and the discovery of changes in the correlation patterns of related series; Signal time-frequency domain analysis function. For time signal series such as sound and vibration, time-frequency domain analysis, periodic analysis, dominant frequency analysis, etc. are provided; including common graphs and analysis methods such as trend graphs, spectrograms, time domain analysis, and frequency domain analysis; Through a series of auxiliary functions provided by the toolbox, expert experience and artificial intelligence are better combined. On the one hand, expert experience provides useful rules and experience for the improvement of the algorithm model. On the other hand, the algorithm model and analysis tools provide more and more convenient means for experts to quickly analyze and locate problems; D. Fault knowledge base management Using enterprise asset data information, an enterprise-owned fault knowledge base is established, which includes a pre-set equipment structure decomposition library, a fault mode library, and a maintenance strategy library, to quickly build a data foundation for the enterprise to construct a reliability-centered equipment management system; at the same time, it can also be combined with the user's past fault and inspection and maintenance records as auxiliary input data for failure mode and effect analysis; Through the invocation of the industry knowledge base in the fault knowledge base, users can provide a reference basis for equipment digital modeling and failure mode and effect analysis. And through the continuous use of the software, users can define their own private fault knowledge bases and accumulate equipment maintenance experience; the content of the equipment type, equipment name, equipment description, and equipment knowledge base is displayed in the fault knowledge base. The content of the fault knowledge base of the equipment is as follows: Equipment structure decomposition library, providing a reference for the structure decomposition of the corresponding equipment in the fault knowledge base; Fault mode library, providing references for the fault causes and fault modes of the corresponding components in the fault knowledge base; Maintenance strategy library, providing references for the fault causes, maintenance measures, and maintenance cycles of the corresponding components in the fault knowledge base; Health index library, providing references for the corresponding associated health indexes of the equipment in the fault knowledge base; Among them, the fault knowledge base is screened according to the knowledge base type, equipment type, and equipment keyword conditions, and the equipment details in the fault knowledge base are selected for viewing; after the target equipment is determined, the equipment structure decomposition library, fault mode library, maintenance strategy library, and health index library are viewed through the call of the equipment knowledge base; in the equipment structure decomposition stage and the initial stage of the failure mode and effects analysis stage, the equipment structure decomposition and failure mode in the initial stage are completed through the call and reference of the knowledge base; E. FMEA (Failure Mode and Effects Analysis) FMEA, by comprehensively sorting out and managing factors such as the failure modes, failure consequences, and failure causes of equipment or components, helps customers understand the weak links in the operation of equipment, discover possible failures and impacts, so as to help evaluate the reliability and availability of units, systems, or equipment, and take corresponding measures specifically, including the following content: E1) Equipment structure decomposition layer: Decompose and sort out enterprise assets, decompose equipment into repairable components according to the ISO14224 standard, and at the same time associate the equipment failure modes with the components of the equipment, clarify the relationship between the reliability of the entire system and individual equipment / components, and prepare the logical relationship of the equipment for subsequent risk analysis. This work is an important basis for failure mode and effects analysis and is finally realized by the asset structure tree; E2) Function analysis layer: Fully and accurately describe the functions of equipment or components, as well as the reference requirements for realizing the functions, so as to lay a foundation for subsequent fault analysis; E3) Failure mode analysis layer: According to the function, sort out and define all possible failure modes of equipment or components, and at the same time correspond the equipment failure modes with the failure phenomena of the equipment to prepare the logical relationship for subsequent risk analysis; E4) Failure impact analysis layer: Define the consequence impacts of each failure mode, and comprehensively evaluate the multi-dimensional consequence impacts caused by equipment failures; E5) Failure cause analysis layer: Define the failure causes corresponding to the failure modes, realize the management of equipment failure causes, including failure causes and failure cause descriptions, and combine the actual situation on site to comprehensively evaluate the equipment consequence impacts and the occurrence probabilities of failure causes, so as to analyze and classify the hazards of equipment or components and provide a basis and decision-making basis for managing equipment or components by grade; E6) Maintenance strategy analysis layer: Define the maintenance strategies for failure causes. Based on factors such as failure modes, failure impacts, failure probabilities, maintenance times, and the impacts of equipment on operation goals, use a risk-based method to objectively determine maintenance strategies, realize the management and maintenance of maintenance contents, including maintenance strategy categories, maintenance strategy names, maintenance strategy contents, required resources, etc. The initial data comes from the existing inspection and maintenance standards of the enterprise. Subsequently, the inspection and maintenance strategies can be updated after simulation and optimization; Failure Mode and Effects Analysis (FMEA) combines tasks such as asset data analysis, equipment structure decomposition, function analysis, failure mode analysis, failure effect analysis, failure cause identification and analysis, and maintenance strategy formulation in the form of analysis projects to complete the entire FMEA business process; F. Dynamic Risk Analysis Based on the impacts of equipment failure consequences on various dimensions such as the environment, cost, and safety, using the equipment failure laws formed by regression of equipment reliability data, and using a visual risk matrix to evaluate the dynamic risks of equipment, to assist equipment managers in making auxiliary decisions using equipment risks in daily equipment management work; The risk matrix is used to evaluate the risks brought to the enterprise due to equipment failures; to make the failure mode and effects analysis more targeted and the analysis results more in line with the actual situation of the enterprise, determine the risk matrix jointly with the enterprise before conducting the failure mode and effects analysis, as follows: F1) Risk Matrix Customization Refer to the power industry standard "DLT302.2-2011 Technical Guide for Equipment Maintenance Analysis in Thermal Power Plants - Part 2: Risk-Based Maintenance Analysis" to define the risk dimensions, risk grading, risk matrix, and its display colors, etc. for the equipment within the implementation scope in the enterprise; Refer to the enterprise standards "Hazard Source, Environmental Factor Identification, Evaluation and Control Procedure" and "Production Accident (Unsafe Event) Reporting and Investigation and Handling Procedure" to formulate the risk matrix evaluation criteria as the risk assessment method for the intelligent maintenance decision support system project of key auxiliary equipment in thermal power plants based on multi-dimensional state perception; construct a risk matrix system for the main systems and equipment objects involved in the project implementation from multiple dimensions such as personnel safety, equipment safety, economy, environmental protection, reputation, and social impact, as an important reference basis for the daily equipment maintenance, inspection and repair plan formulation / adjustment, real-time reliability assessment, and establishment of equipment health indicators in the power plant; Provide a standardized risk matrix for the power industry, analyze the failure consequences of equipment, and use data regression algorithms to evaluate the dynamic risks of equipment through the standardized historical equipment defect data, providing a decision-making basis for intelligent equipment maintenance; F2) Dynamic Risk Provide a standardized risk matrix for the power industry, analyze the failure consequences of equipment, and use data regression algorithms to evaluate the dynamic risks of equipment through the standardized historical equipment defect data, providing a decision-making basis for intelligent equipment maintenance; Dynamic risk provides a visual display that can simultaneously show the inherent risks of equipment analyzed through Failure Mode and Effects Analysis and the dynamic risks of equipment after data regression, as well as the risk distributions of equipment within the unit and the causes of failures within the equipment; it can statistically analyze the risk levels of different failure causes of asset equipment or components, calculate the number of failure causes for each risk level in the risk matrix, and be used for screening key equipment, determining the effectiveness of maintenance strategies, and judging the equipment status, etc. G. Equipment Defect Management The main functions of equipment defect management are as follows: Defect data collection includes the content in business processes such as defect registration, defect confirmation, operation information, and defect elimination processing; defect information improvement automatically collects, fills in, and improves defect content through each page field and defect category, realizing the standardization of all defect data. Defect query can browse and view all defects synchronized from the big data platform, including information such as defect information improvement status, number, defect description, affiliated unit, defect category, etc. Defect information improvement: In the defect information improvement interface, the results of Failure Mode and Effects Analysis or data from the fault knowledge base can be called, and the defect can be improved according to the defect description information, select the component information where the defect occurs, and associate with the failure mode, fault information, etc. H. Reliability / Maintenance Data Regression Analysis Regress the historical defect data and maintenance record data of equipment, and the parameters of the distribution curve can be obtained by regressing historical data or user-defined; through this function, engineers can quickly establish their own reliability database containing equipment reliability models, reliability parameters, etc., for comprehensively evaluating the reliability, availability, and maintainability of the unit. Reflect the failure laws of each system or the entire unit in the process required by the user, and effectively avoid the misleading results generated when using the general database. For all reliability and maintenance data, by creating a new regression task and selecting the data content to be regressed, perform reliability regression calculations on the failure causes corresponding to the components in the reliability data, so as to calculate the new MTBF (Mean Time Between Failures, statistical average) of this equipment, component, and failure cause, and then view the reliability parameters and reliability curve through the regression report. I. Simulation and Optimization Based on the historical failure data of the equipment, after analyzing the probability and consequences of failures to establish a reliability model, determine its economic risks and the magnitude of failure losses, and predict the future production capacity and benefits; analyze, evaluate, and optimize the reliability, availability, and maintainability of the unit or equipment, and evaluate the impacts of different maintenance strategies on availability, economic benefits, etc. On the basis of simulating the maintenance strategies based on the equipment reliability and mean time between failures, complete the modeling and simulation of the failure causes, failure modes, inspection and maintenance strategies of the unit and equipment; Users can flexibly select the objects for evaluation, the length of the equipment operation life cycle for simulation and prediction, the analysis time nodes, and the sample density; Consider the impacts of maintenance strategies, technical renovation plans, and spare parts strategies on asset performance. By adjusting the cycle and content of the maintenance and repair strategies, and comparing the performance differences of different maintenance and repair strategy plans, formulate a better asset management plan; J. Maintenance Strategy Push and Supervision Collect the comprehensive equipment information provided by experienced equipment management personnel, establish equipment maintenance strategies and maintenance plans that are consistent with the company's financial, production, and safety goals, and centrally manage and monitor the important maintenance strategies that need to be monitored, so as to overall arrange the maintenance plan and improve management efficiency; For each maintenance plan, it can be set whether to monitor, monitoring points, early warning period, associated strategies. The management interface can be filtered according to various requirements, such as by equipment classification, specialty, unit, etc., or by early warning, expired, overdue, etc.; In the management interface, the strategies that have been early warned, expired, and overdue can be displayed in different colors, and the corresponding maintenance plans can be output to the corresponding systems. When outputting, the other strategies of this equipment can be selectively merged and output, and it can be selected whether to synchronously output the corresponding maintenance content and spare parts; Based on factors such as failure modes, failure consequences, failure probabilities, repair times, and the impacts of equipment on operation goals, use a risk-based method to objectively determine maintenance strategies to help users achieve maximum return on investment; Equipment risk priorities. At the factory level, determine which equipment problems should be solved first based on the impacts of equipment on the factory operation goals, and give priority to work to achieve the maximum return on investment based on risk; It is required that the calculation of risk must be based on failure consequences, failure probabilities, and repair times; There are mainly the following two aspects: 1) Preventive Maintenance Strategy For all equipment, display the content of the maintenance strategies included in the equipment after failure mode and effects analysis. At the same time, the maintenance strategies of the equipment can also be manually added directly; 2) Preventive Maintenance Plan and Monitoring According to the requirements of the preventive maintenance plan, combine and package the corresponding maintenance strategies for the equipment, formulate the spare parts list required for maintenance, and generate the corresponding preventive maintenance plan for the equipment; according to the cycle of the maintenance plan, configure the maintenance plan monitoring, and centrally display the preventive maintenance plans in the monitoring state; this system will automatically monitor the maintenance records associated with the equipment maintenance plan according to the warning time and allowed overdue time configured in the maintenance plan monitoring, and timely remind the due situation and execution situation of the maintenance plan; Through the interface with the external maintenance execution system, this part can realize the timely push and execution after the preventive maintenance plan expires, and the tracking closed-loop of timely updating the monitoring status after the execution is completed; K. Standardization of overhaul operations The standardization of overhaul operations is used to standardize the content of the major and minor repair document packages in the power industry, guide the execution process of major and minor repairs, and comprehensively record the results of overhaul work. This system will establish standard overhaul operation standardizations for different types of equipment. After the overhaul items are determined, the standard overhaul operation standardizations can be used as the basis to plan the current major and minor repairs. Users can adjust the standard overhaul operation standardizations according to the situation. The standardization of overhaul operations mainly includes the following contents: K1) Overhaul time plan List in detail the overhaul work to be carried out, the time plan and the arrangement of relevant resources, which should include the list of equipment, the specific steps and requirements of the overhaul, as well as the division of labor and responsibilities of personnel; K2) Overhaul operation instruction manual Provide specific guidance and requirements, including the steps of overhaul operations, operating procedures, safety measures, etc.; these instruction manuals should describe in detail the operating requirements of each work step to ensure the correctness and safety of overhaul operations; K3) Overhaul records and reports Record the key information during the overhaul process, including the status of the equipment, the specific content of the overhaul, the time and personnel of the work, etc.; these records can be used to monitor the maintenance situation of the equipment and track problems, and as a reference for future overhauls; K4) Overhaul tools and spare parts List the tools and equipment to be used, as well as the list of spare parts and consumables that may need to be replaced or used; The standardization of overhaul operations supports APP-based (mobile applications), and can push relevant overhaul work content and procedures to mobile terminals to guide on-site operations of overhaul personnel; after the overhaul operations are completed, relevant overhaul situations and inspection results can be directly transmitted back and archived through the mobile terminal; through structured design, the system can convert some content into reliability data, as the data basis for identifying fault laws and optimizing strategies, and further enhance the later value of overhaul data; L. Equipment health management Combined with the characteristics of different types of equipment, comprehensively considering the inherent fault law characteristics and short-cycle performance and other states of the equipment, as well as according to different specialties, equipment types, and equipment classifications, different standard health models can be established based on different business data, construct equipment health evaluation criteria, and evaluate the comprehensive health status of the equipment; through the relevant business data synchronized by the interface, after the calculation of the standard health model, an asset health assessment is formed, and the health status and change trend are displayed in real time, including health score, influencing factor analysis, asset health list, etc.; comprehensively judge the health status of the equipment, provide the main cause and secondary cause, and provide decision-making support for the unit maintenance plan; Equipment health indicators are used for the daily analysis and evaluation of equipment status by equipment maintenance and operation personnel. As an important internal equipment management means, the operation of the equipment is strengthened through the management of key equipment fault characteristic quantities, select the equipment that needs to be concerned about and monitored, and dynamically monitor the current health indicators of the equipment by accessing the real-time status values of the equipment transmitted by the SIS (Safety Instrumented System) system, and display the current health status of the equipment; Equipment health indicator support: Change the indicators manually processed by humans into standardized equipment status indicator management; Display the equipment health status in a unified view; Early judge equipment failures so as to arrange and make proactive maintenance decisions in a timely manner before equipment failures; Classify the equipment warning indicators to facilitate users to take different measures according to different levels of warnings; Build the correlation relationship between equipment parameter anomalies and equipment failure modes, receive real-time data such as equipment SIS / PI (Performance Indicator, interface / protocol), fault information and other data and information affecting the reliability of the equipment, through equipment modeling and simulation evaluation, through combining the calculated value of the MTBF of this system and other information, comprehensively judge the current state of the equipment, realize intelligent analysis and auxiliary decision-making, and help enterprises improve the efficiency of equipment monitoring and status evaluation.

[0024] Among them, the RCM analysis includes the following steps: S1. Determine the key functions of the equipment: clarify the core role of the equipment in the system. For example, the function of a pump is to transport liquid; S2. Identify functional failure modes: analyze the specific ways that may lead to function failure. For example, bearing overheating, seal leakage; S3. Evaluate the consequences of failures: judge the impact of failures on safety, environment, production, and cost. For example, shutdown losses, safety hazards; S4. Analyze the causes of failures: trace the root causes of failures. For example, insufficient lubrication, material fatigue; S5. Develop a preventive maintenance strategy: Select targeted maintenance tasks, such as regular inspections, condition monitoring, and design improvement; S6. Optimize the maintenance plan: Adjust the maintenance frequency and methods by considering risk, cost, and technical feasibility; S7. Continuously monitor and provide feedback: Verify the effectiveness of the strategy through operating data and dynamically optimize the maintenance plan.

[0025] For algorithm models, there are the following examples: I. Time series prediction 1. Statistical models ARIMA / SARIMA: Suitable for sensor data with linear trends and obvious seasonality (such as periodic changes in temperature and humidity); Prophet: Deals with non-stationary data with holiday effects (such as traffic flow prediction); 2. Machine learning LSTM / GRU: Captures long-term and short-term dependencies (such as equipment vibration signal prediction); Transformer: Processes long sequences in parallel (such as power grid load prediction); 3. Ensemble methods LightGBM / XGBoost: Handles medium and short-term predictions with multiple features (such as combining temperature and pressure to predict equipment status); II. Anomaly detection 1. Statistical methods 3σ principle: Outlier detection based on Gaussian distribution (suitable for univariate stationary data); Dynamic threshold: Calculates the mean / variance using a sliding window (real-time detection of mutations); 2. Machine learning Isolation Forest: Efficiently identifies low-dimensional anomalies (such as sensor failures); One-Class SVM: Unsupervised learning of normal patterns (suitable for scenarios with scarce labels); 3. Deep learning Autoencoder: Detects anomalies through reconstruction error (such as voiceprint sensor anomalies); LSTM-Autoencoder: Handles anomalies with time series dependencies (such as industrial equipment vibration anomalies).

[0026] Example 2: To enable those skilled in the art to quickly get started with the intelligent perception system based on multi-dimensional status monitoring in Embodiment 1, this embodiment provides a set of steps for operating the system, which can reduce the training time of technicians, lower the difficulty of getting started, and accurately understand the intelligent maintenance decision-making of key auxiliary equipment in thermal power plants using this system.

[0027] As Figure 2 shown, an intelligent perception method based on multi-dimensional status monitoring includes the following steps: Step 1: Equipment supervision: Invoke status monitoring to collect physical signals of the equipment in real time (such as temperature and humidity, vibration, pressure). When abnormal conditions or early fault signs occur, the edge device can make an abnormal judgment and give an alarm prompt. Step 2: Real-time analysis: Send the collected physical signals to the cloud for modeling and analysis. At the same time, obtain the real-time operation data of the equipment online to build a knowledge graph model, and conduct reasoning and decision-making analysis on the operation status of the equipment. Step 3: Give a decision: Invoke the intelligent maintenance decision-making module, combine the results of reasoning and decision-making analysis, identify the fault risks affecting equipment reliability, and establish fault and maintenance reliability data standards, a digital inspection and maintenance data collection and exchange system, and an equipment reliability model. Step 4: Optimization: Integrate reliability data analysis, modeling and simulation, and strategy optimization to optimize the maintenance tasks and plans. Step 5: Maintenance operation guidance: Invoke the standardization of maintenance operations, establish an exchange standard for inspection and maintenance content and data, and reconstruct the inspection and maintenance operations with standardized design and rationalization. Step 6: Real-time update: Capture the fault data and the archived inspection and maintenance data, send them to the cloud, and statistically archive them into the intelligent maintenance decision-making to achieve continuous update of equipment faults and provide data support for the intelligent maintenance decision-making module.

[0028] Embodiment 3: An intelligent perception device, relying on the existing hyper-converged platform, is deployed through the virtualized server mode, including: Data server, configured with a CPU (Central Processing Unit) with more than 40 cores, more than 256G of memory, dual gigabit network cards, 2 * 800W power supplies, and a hard disk of no less than 960G * 2 SSD (Solid State Disk) + 4TB * 6 SSD RAID (Redundant Arrays of Independent Disks), realizes interface docking and integration with the SIS real-time data monitoring system, and obtains relevant data to implement business functions; all interface development work between the intelligent perception system side and the SIS real-time data monitoring system in the first embodiment fully considers integration and convenience, automatically realizes the unification of business processes and data, and the sharing and commonality of data; Among them, read the device-related measurement point information and measurement point data through the interface, receive the alarm information of the device-related measurement points in the SIS system status monitoring, obtain the characteristic quantities for the device asset health assessment, and regularly obtain and analyze the alarm information pushed by the device comprehensive status monitoring; Application server, configured with a 20-core CPU, 128G of memory, a hard disk of no less than 2TB * 2 SSD RAID, dual gigabit network cards, 2 * 800W power supplies, used to build each functional module, developed based on a modular and componentized architecture, and users can complete the operation of the intelligent perception system in the first embodiment using a Web browser without installing other client software; Among them, the back-end web application of the intelligent perception system in the first embodiment is developed based on SpringBoot (an open-source application framework on the Java platform) and SpringSecurity (a security framework that can provide a descriptive security access control solution for enterprise application systems based on Spring), and at the same time combines gateway layer components such as Nginx to achieve reverse proxy and traffic load balancing; the front-end browser page is developed based on UI frameworks such as Vue, ElementUI, and EChart, and can be adapted to all browsers based on the Chrome kernel; NAS storage, configured with 32G of running memory, 4 cores, 2 * NVME disk positions, and 8 * 18T NAS dedicated 3.5-inch mechanical keyboards.

[0029] Strictly keep confidential the relevant data secrets, document materials, business secrets and other information obtained during the intelligent perception process, and take corresponding confidentiality measures.

[0030] Adopt a strict security system to ensure the security of data throughout the process of processing and transmission. Ensure that the intelligent perception system in the first embodiment can operate normally without being attacked and damaged.

[0031] Ensure that the information in the intelligent perception system in Embodiment 1 is not accessed without authorization, and divide the operation permissions of operators according to the organizational structure. The usage permissions are uniformly configured by the system administrator.

[0032] In terms of data storage, for sensitive information such as password information in the application system, it is encrypted and stored using the Bcrypt method.

[0033] In terms of data transmission, for the application system that can be directly accessed from the external network, all sensitive information is transmitted in an encrypted manner.

[0034] For the access and download behaviors of all business data within the system, especially the core business data, it must be accessed after authentication and authorization, and anonymous access to the business system data is prohibited.

[0035] For the storage and read-write analysis of business data, a relational database such as MySQL is used. The business data of the intelligent perception system in Embodiment 1 is stored using the relational database MySQL 8.0. For other mainstream relational databases such as SQL Server, Oracle, KingbaseES, and DM, the system can also perform database engine adaptation.

[0036] It can meet the use of different Linux and Windows server sides, such as CentOS, Ubuntu, Windows Server, etc., and support the deployment and operation and maintenance of domestic Xinchuang operating systems such as Kylin and OpenAnolis. This system is containerized based on Docker and can automatically adapt to the host operating system.

[0037] This system supports access through browsers based on Chrome and Firefox kernels. The front-end technology stack of this system uses mainstream UI libraries such as Vue and ElementUI, avoiding the introduction of third-party libraries with compatibility problems, and can be compatible with browsers with mainstream Chrome and Firefox kernels on the market.

[0038] It supports third-party systems to call system-related data and functions through REST API (cloud server). This system provides OpenAPI with the HTTP Restful protocol for third-party systems to call, for accessing the business data and related functions of this system.

[0039] Embodiment 4: A computer-readable storage medium including program instructions, which realizes the steps of the intelligent perception method in Embodiment 2 when the program instructions are executed by an application server. For example, this computer-readable storage medium can be the above-mentioned NAS storage including program instructions, and the above program instructions can be executed by the CPU of the application server to complete the above-mentioned fault diagnosis method.

[0040] As an alternative embodiment, a computer program product is further provided, which includes a computer program executable by a programmable device, and the computer program has a code portion for performing the intelligent perception method in the second embodiment when executed by the programmable device.

[0041] The above are only alternative embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention. The structures, devices, and operation methods not specifically described and explained in the present invention are implemented according to the conventional means in the art without special description and limitation.

Claims

1. An intelligent perception system based on multi-dimensional state monitoring, characterized in that, Including: A status monitoring module that collects physical signals of the device in real time. When an abnormal situation or early fault symptom occurs, the edge side can make an abnormal judgment and give an alarm prompt; And perform data modeling and analysis in the cloud. At the same time, obtain the real-time operation data of the device online to build a knowledge graph model, and conduct reasoning and decision-making analysis on the operation status of the device; A smart maintenance decision-making module that combines the results of reasoning and decision-making analysis, identifies the fault risks affecting the device reliability, establishes fault and maintenance reliability data standards, a digital inspection and maintenance data collection and exchange system, and a device reliability model; and then optimizes the inspection and maintenance tasks and plans; An inspection operation standardization module that establishes an exchange standard for inspection and maintenance content and data, conducts standardized design and rational reconstruction of inspection operations, and uses fault data and archived inspection and maintenance data to continuously update device faults, providing data support for the smart maintenance decision-making module.

2. The intelligent perception system based on multi-dimensional state monitoring according to claim 1, wherein: The status monitoring module includes: A software basic unit for configuring the smart maintenance decision-making module, the inspection operation standardization module, and user management permissions, and realizing the flow of data streams among the status monitoring module, the smart maintenance decision-making module, and the inspection operation standardization module; A status perception and early warning unit that collects physical signals and working condition data through sensors, conducts real-time monitoring and analysis of the operation status of the device, and generates early warning information when abnormalities are found to prompt users to pay attention to fault risks; An asset structure tree for device master data management, spare part resource data management, and human resource data management; A fault knowledge base management unit that includes a standard fault knowledge base, a device structure decomposition library, a fault mode library, a maintenance strategy library, and a health index library.

3. An intelligent perception system based on multi-dimensional state monitoring according to claim 1, characterized in that: The smart maintenance decision-making module includes: An FMEA unit that sorts out and manages fault factors, discovers the faults and impacts that occur, to help evaluate the reliability and availability of the unit, system or device; A risk dynamic analysis unit that, based on the impacts of the fault consequences of the device on the environment, cost, and safety dimensions, uses the device fault laws formed by the regression of device reliability data to construct a visual risk matrix to evaluate the dynamic risks of the device, and assist device managers in making auxiliary decisions; A device defect management unit for data collection and defect information improvement. Among them, data collection includes defect registration, defect confirmation, operation information, and defect elimination processing, and defect information improvement automatically collects, fills in, and improves defect content through each page field and defect category; A data regression analysis unit that regresses historical defect data and inspection record data, and the parameters of the distribution curve are obtained by regressing historical data or user-defined, to form a reliability database for comprehensively evaluating the reliability, availability, and maintainability of the unit. The simulation and optimization unit, based on the reliability database, determines the magnitudes of economic risks and failure losses, predicts future production capacity and benefits, and, on the basis of completing the simulation of maintenance strategies based on equipment reliability and mean time between failures, completes the modeling and simulation of the failure causes, failure modes, and inspection and maintenance strategies of the unit and equipment; among them, considering the impacts of maintenance strategies, technological transformation plans, and spare part strategies on asset performance, by adjusting the cycle and content of the maintenance and repair strategies and comparing the performance differences, a more optimal strategy is formulated.

4. An intelligent perception system based on multi-dimensional state monitoring according to claim 3, characterized in that: The FMEA unit includes: The equipment structure decomposition layer decomposes and sorts out the enterprise assets, decomposes the equipment into repairable components, and at the same time associates the equipment failure modes with the components of the equipment. The function analysis layer describes the functions of the equipment or components, as well as the reference requirements for realizing the functions. The failure mode analysis layer sorts out and defines all possible failure modes of the equipment or components according to the functions, and at the same time corresponds the equipment failure modes with the failure phenomena of the equipment. The failure impact analysis layer defines the consequence impacts of the failure modes and comprehensively evaluates the multi-dimensional consequence impacts caused by equipment failures. The failure cause analysis layer defines the failure causes corresponding to the failure modes, combines the actual situation on site, comprehensively evaluates the equipment consequence impacts and the occurrence probabilities of the failure causes, and thus analyzes and classifies the hazards of the equipment or components. The maintenance strategy analysis layer determines the maintenance strategy based on the failure mode, failure impact, failure probability, maintenance time, and the impact of the equipment on the operation objectives, and realizes the management and maintenance of the maintenance content.

5. The intelligent perception system based on multi-dimensional status monitoring according to claim 4, characterized in that: The management of the maintenance content includes the maintenance strategy category, maintenance strategy name, maintenance strategy content, and required resources.

6. An intelligent perception system based on multi-dimensional state monitoring according to claim 3, characterized in that: The risk dynamic analysis unit includes: The risk matrix customization layer defines the risk dimensions, risk classification, risk matrix, and its display color of the enterprise equipment. The dynamic risk assessment layer analyzes the failure consequences of the equipment, and based on the historical data of equipment defects, uses the data regression algorithm to evaluate the dynamic risk of the equipment, and calculates the number of failure causes of each risk level in the risk matrix.

7. An intelligent perception system based on multi-dimensional status monitoring according to claim 1, characterized in that: The overhaul operation standardization module includes: The maintenance strategy push and supervision unit, through the interface with the external maintenance execution system, realizes the timely push and execution after the preventive maintenance plan expires, and the tracking closed-loop of the timely update of the monitoring status after the execution is completed, and serves as the exchange standard for the establishment of inspection and maintenance content and data. The overhaul operation standardization unit is used to standardize the content of the major and minor repair file packages in the power industry, guide the execution process of major and minor repairs, and comprehensively record the results of the overhaul work. The equipment health management unit establishes a standard health model based on business data, constructs an equipment health evaluation standard, evaluates the comprehensive health status of the equipment, and through the relevant business data synchronized by the interface, forms an asset health assessment after the calculation of the standard health model, and real-time displays the health status and change trend, so as to comprehensively determine the equipment health status, provide the main cause and secondary cause, support the decision-making of the unit overhaul plan, and assist in the rational reconstruction of the overhaul operation.

8. An intelligent perception system based on multi-dimensional state monitoring according to claim 1, characterized in that: The maintenance strategy push and supervision unit includes: The preventive maintenance strategy layer is used to display the maintenance strategy content included in the equipment after failure mode and effects analysis, or to manually directly add the maintenance strategy of the equipment; The preventive maintenance plan and monitoring layer combines and packages the corresponding maintenance strategies of the equipment according to the preventive maintenance plan requirements, formulates the spare parts list required for maintenance, generates the corresponding preventive maintenance plan for the equipment, and configures the maintenance plan monitoring according to the cycle of the maintenance plan, and centrally displays the preventive maintenance plans in the monitoring state.

9. An intelligent perception system based on multi-dimensional state monitoring according to claim 1, characterized in that: The overhaul operation standardization unit includes: The overhaul time plan details the overhaul work to be carried out, the time plan and the arrangement of relevant resources, including the list of equipment, the specific steps and requirements of the overhaul, and the division of labor and responsibilities of personnel; The overhaul operation instruction manual provides specific guidance and requirements, including the steps of the overhaul operation, operating procedures, safety measures, etc.; The overhaul records and reports record the key information during the overhaul process, including the status of the equipment, the specific content of the overhaul, the working time and personnel; The overhaul tools and spare parts list the tools and equipment to be used, as well as the list of spare parts and consumables that may need to be replaced or used.

10. An intelligent perception method based on multi-dimensional state monitoring, applied to the intelligent perception system based on multi-dimensional state monitoring according to any one of claims 1-9, characterized in that, It includes the following steps: Step 1, Equipment supervision: Call the status monitoring module to collect the physical signals of the equipment in real time. When abnormal situations or early fault signs occur, the edge side can make abnormal judgments and prompt alarms; Step 2, Real-time analysis: Send the collected physical signals to the cloud for modeling and analysis. At the same time, obtain the real-time operation data of the equipment online to build a knowledge graph model, and conduct reasoning and decision-making analysis on the operation status of the equipment; Step 3, Give decisions: Call the intelligent overhaul decision-making module, combine the reasoning and decision-making analysis results, identify the fault risks affecting the equipment reliability, and establish fault and maintenance reliability data standards, digital inspection and maintenance data collection and exchange systems, and equipment reliability models; Step 4, Optimization: Integrate reliability data analysis, modeling and simulation, and strategy optimization to optimize the overhaul tasks and plans; Step 5, Overhaul operation guidance: Call the overhaul operation standardization module to establish an exchange standard for the inspection and maintenance content and data, and reconstruct the overhaul operation with standardized design and rationalization; Step 6, Real-time update: Capture the fault data and the archived inspection and maintenance data, send them to the cloud and conduct statistical archiving to the intelligent overhaul decision-making module to realize the continuous update of the equipment faults and provide data support for the intelligent overhaul decision-making module.

Images