Thermal power generating unit oil abnormity monitoring method and system based on knowledge graph

Abstract

The invention relates to the field of intelligent monitoring of power equipment, and discloses a thermal power generating unit oil abnormity monitoring method and system based on a knowledge graph, and the method comprises the steps: integrating a system operation log and historical oil sample data of an oil system in a thermal power generating unit to generate joint operation data; then, a unit knowledge graph covering oil indexes, unit component states and operation associated events is constructed based on the data, an operation cycle change curve is generated by recognizing the oil entity state and combining the component states, and an abnormal curve segment deviating from the benchmark is positioned; and further determining a cooperative part of the oil abnormal event and the associated event, extracting oil degradation characteristics, and finally formulating maintenance countermeasures for specific degradation factors to realize accurate monitoring and maintenance of the oil system. According to the invention, the accuracy of oil abnormity monitoring of the thermal power generating unit can be improved.

Description

Technical Field

[0001] This invention relates to a knowledge graph-based method and system for monitoring oil anomalies in thermal power units, belonging to the field of intelligent monitoring of power equipment. Background Technology

[0002] A thermal power unit is a complete set of power equipment that uses the heat energy generated by the combustion of fuels such as coal, oil or natural gas, and outputs electrical energy through a series of energy conversions. The oil system in a thermal power unit undertakes crucial functions such as lubrication, cooling, sealing and transmission of control signals, and its condition is directly related to the safety and efficiency of the entire unit.

[0003] Currently, oil monitoring in thermal power units mainly relies on periodic sampling and laboratory analysis. By detecting key parameters such as oil viscosity, moisture content, and particulate contamination, and based on preset thresholds, anomaly judgments and alarms are made. However, these existing technologies have a fundamental flaw: the monitoring process is isolated from the overall operating environment of the unit, focusing only on the static comparison of oil indicators, lacking a dynamic understanding of the correlation between oil conditions, and failing to identify early deterioration trends under complex operating conditions in a timely manner, thus making it difficult to achieve accurate early warning and proactive maintenance. Summary of the Invention

[0004] This invention provides a knowledge graph-based method and system for monitoring oil anomalies in thermal power units, with the main objective of improving the accuracy of oil anomaly monitoring in thermal power units.

[0005] To achieve the above objectives, this invention provides a knowledge graph-based method for monitoring abnormal oil levels in thermal power units, comprising: After obtaining the system operation logs corresponding to the oil system in the thermal power unit, the joint operation data corresponding to the oil system is generated based on the system operation logs and historical oil sample data. Based on the joint operation data, a unit knowledge graph corresponding to the thermal power unit is constructed. The unit knowledge graph includes: oil index, unit component status and operation-related events. After identifying the physical state of the oil index in the oil system, and combining it with the state of the unit components, a change curve of the oil system during the operating cycle is generated, and abnormal segments in the change curve that deviate from the historical benchmark are located. Identify the unit cooperating components corresponding to the abnormal oil events and the operation-related events in the abnormal curve segment, and extract the oil deterioration characteristics that appear in the oil system when the unit cooperating components are running; Develop maintenance and response measures corresponding to specific deterioration elements in the oil degradation characteristics to achieve oil monitoring and maintenance of the oil system in the thermal power unit.

[0006] Optionally, constructing a unit knowledge graph corresponding to the thermal power unit based on the joint operation data includes: The oil index items and unit component descriptions in the joint operation data are analyzed, and the joint operation events recorded in the joint operation data are extracted. Categorize the knowledge entity relationships between the oil index items, the unit component descriptions, and the operational linkage events; Based on the knowledge entity relationships, a unit knowledge graph corresponding to the thermal power unit is constructed.

[0007] Optionally, the classification of the knowledge entity relationships between the oil index items, the unit component descriptions, and operational linkage events includes: Identify the entity identifiers in the oil index items, unit component descriptions, and operational linkage events, and determine the function corresponding to the entity identifiers; After defining the corresponding entity triggering condition for the action, the knowledge entity relationship under the entity triggering condition is generated.

[0008] Optionally, generating joint operating data for the oil system based on the system operation log and historical oil sample data includes: Read the sequence of running events corresponding to key log entries in the system operation log, and extract the same period oil sample indicators from the historical oil sample data; Align the sequence of running events with the key timestamps in the oil sample indicators of the same period to obtain the sequence-indicator correspondence; Integrate the joint operation data corresponding to the joint sub-relationships in the sequence-index correspondence.

[0009] Optionally, generating the change curve of the oil system during the operating cycle by combining the status of the unit components includes: Analyze the periodic operation data corresponding to the oil-hydraulic coordination components related to the status of the unit components; The time sequence of oil changes at mid-segment nodes during the operational cycle is collected; Generate the change curves corresponding to the change time points in the oil change time series.

[0010] Optionally, locating the abnormal segments in the change curve that deviate from the historical baseline includes: After analyzing the deviation value of the curve deviation segment between the change curve and the historical baseline curve, the deviation segment of the change curve that exceeds the deviation value is marked as an abnormal candidate segment; Verify that the segment with the largest deviation from the historical benchmark among the candidate abnormal segments is the abnormal curve segment.

[0011] Optionally, the unit coordination component that determines the oil abnormality event and the operation-related event corresponding to the abnormal curve segment includes: Identify the abnormal oil identifiers corresponding to abnormal oil events in the abnormal curve segments; Based on the abnormal oil identification, the event association links of the operation-related events in the unit knowledge graph are traced. Identify the unit coordination component corresponding to the coordination component identifier in the event association link.

[0012] Optionally, the step of extracting the oil degradation characteristics that occur in the oil system during the operation of the unit's cooperating components includes: Define the overlapping time window for the operation of the unit's collaborative components, and retrieve the collaborative status records of the unit's collaborative components within the overlapping time window; Identify the oil state sequence recorded under multiple detections in the collaborative state record, and screen out the oil deterioration characteristics that appear in the oil state sequence when components are working together.

[0013] Optionally, the step of formulating maintenance measures corresponding to specific deterioration elements in the oil deterioration characteristics includes: After analyzing the degree of deterioration corresponding to the specific deterioration elements in the oil deterioration characteristics, the maintenance-related items in the unit knowledge graph are queried. After identifying the maintenance conditions corresponding to the maintenance-related items, formulate maintenance response measures for the oil system under the maintenance conditions.

[0014] To address the aforementioned problems, this invention also provides a knowledge graph-based oil anomaly monitoring system for thermal power units, the system comprising: The joint operation module is used to obtain the system operation logs corresponding to the oil system in the thermal power unit, and then generate the joint operation data corresponding to the oil system based on the system operation logs and historical oil sample data. The knowledge graph construction module is used to construct the unit knowledge graph corresponding to the thermal power unit based on the joint operation data. The unit knowledge graph includes: oil index, unit component status and operation-related events. The abnormal curve module is used to identify the physical state of the oil index in the oil system, and then, in combination with the state of the unit components, generate the change curve of the oil system within the operating cycle, and locate the abnormal curve that deviates from the historical benchmark in the change curve. The feature extraction module is used to determine the unit cooperating component corresponding to the oil abnormal event and the operation-related event in the abnormal curve segment, and to extract the oil deterioration characteristics that appear in the oil system when the unit cooperating component is running. The monitoring and maintenance module is used to formulate maintenance measures corresponding to specific deterioration elements in the oil deterioration characteristics, so as to realize the monitoring and maintenance of the oil system in the thermal power unit.

[0015] Compared to the problems described in the background art, the embodiments of the present invention can comprehensively cover the operating scenarios and oil status information of the oil system, forming more complete data support. This provides basic data that aligns with the actual operating conditions of the unit for accurate judgment of oil anomalies, effectively avoiding the limitations of a single data dimension. Furthermore, the embodiments of the present invention can transform scattered oil-related information and unit operating information into a structured relational network, clearly presenting the intrinsic connection between oil status and unit operation, providing systematic information support for in-depth analysis of oil anomalies, and avoiding the one-sidedness of interpreting single data. Finally, the embodiments of the present invention can achieve dynamic tracking of the operating status of the oil system and transform scattered status information into time-series change curves, fully presenting the state evolution within the operating cycle. By changing the trajectory, the invention can accurately capture abnormal changes that deviate from historical benchmarks, avoiding the lag in static threshold judgment. Furthermore, the embodiments of the invention can break the isolated analysis mode of anomalies and component operation, and by extracting the oil deterioration characteristics of unit cooperating components during operation, it can accurately capture the core impact information behind the anomalies, avoiding judgments based solely on surface phenomena. This provides clear feature indications for tracing the root cause of anomalies. Finally, the embodiments of the invention can achieve precise matching between deterioration problems and solutions, and allow maintenance actions to focus on the core causes of deterioration, improving the targeting of maintenance. This can directly support the targeted monitoring and maintenance of the oil system, ensuring the stable operation of related systems in thermal power units. Therefore, the invention can improve the accuracy of oil anomaly monitoring in thermal power units. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a knowledge graph-based method for monitoring oil anomalies in thermal power units, as provided in an embodiment of the present invention. Figure 2 A schematic diagram of a module for implementing a knowledge graph-based oil anomaly monitoring system for thermal power units, provided as an embodiment of the present invention; Figure 3 A schematic diagram of a computer device for a knowledge graph-based method for monitoring oil anomalies in thermal power units, provided in an embodiment of the present invention; The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0017] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0018] This application provides a knowledge graph-based method for monitoring oil anomalies in thermal power units. The execution entity of this knowledge graph-based method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the knowledge graph-based method for monitoring oil anomalies in thermal power units can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster.

[0019] Reference Figure 1 The diagram shown is a flowchart illustrating a knowledge graph-based method for monitoring oil anomalies in thermal power units according to an embodiment of the present invention. In this embodiment, the knowledge graph-based method for monitoring oil anomalies in thermal power units includes: S1. After obtaining the system operation log corresponding to the oil system in the thermal power unit, generate the joint operation data corresponding to the oil system based on the system operation log and historical oil sample data.

[0020] The embodiments of the present invention can comprehensively cover the operating scenarios and oil status information of the oil system, forming a more complete data support. It provides basic data that fits the actual operating conditions of the unit for accurate judgment of oil abnormalities, effectively avoiding the limitations of a single data dimension.

[0021] The thermal power unit refers to a complete set of power equipment that uses coal, oil, or natural gas as fuel, generates heat energy through fuel combustion, and then converts the heat energy into mechanical energy and then into electrical energy through equipment such as boilers, steam turbines, and generators, and can continuously output electrical energy. Its operation relies on the coordinated work of multiple subsystems to ensure stability and safety. The oil system refers to the subsystem in the thermal power unit that is specifically responsible for providing lubrication, cooling, and sealing services for key moving parts and transmitting control signals. It includes core components such as oil storage devices, oil pipelines, filter components, and detection interfaces, as well as various functional oils circulating within the system, and is an important support for the normal operation of the unit. The system operation log refers to a structured data document that records the relevant operating status of the oil system in real time during the operation of the thermal power unit, including operating time, unit load, etc. Key information such as oil circulation flow rate, system operating temperature, equipment start-up and shutdown frequency, and operation command execution status are continuously stored in a time series. The historical oil sample data refers to the raw data set obtained by laboratory testing or online testing after collecting oil samples from the oil system of the thermal power unit under different operating cycles and conditions in the past. It covers key indicators such as oil viscosity, water content, particulate contamination degree, and chemical composition, and includes related information such as sampling time and operating condition background. The joint operation data refers to the comprehensive dataset formed by associating, matching, and integrating the system operation log of the thermal power unit's oil system with the historical oil sample data according to the time dimension and operating condition dimension. This dataset contains the correspondence between the operating status parameters of the oil system and the performance indicators of the oil itself, and fully covers the operating scenarios and oil status information of the oil system.

[0022] Optionally, the acquisition of the system operation log corresponding to the oil system in the thermal power unit can be achieved through industrial data acquisition tools, such as: by connecting with the monitoring interface of the oil storage device, oil pipeline, circulating pump and other core components of the oil system, key parameters such as running time, unit load, oil circulation flow rate, system operating temperature, and equipment start-up and shutdown status are collected in real time, automatically recorded and stored at preset time intervals, and finally formed a structured system operation log.

[0023] In this embodiment of the invention, generating joint operation data corresponding to the oil system based on the system operation log and historical oil sample data includes: reading the operation event sequence corresponding to key log entries in the system operation log and extracting the contemporaneous oil sample indicators from the historical oil sample data; aligning the key timestamps in the operation event sequence and the contemporaneous oil sample indicators to obtain the sequence-indicator correspondence; and integrating the joint operation data corresponding to the joint sub-relationships in the sequence-indicator correspondence.

[0024] The key log entries refer to the set of entries in the system operation log that carry core operational information of the oil system. These include key parameters directly related to the oil state, such as unit load, oil circulation flow rate, system operating temperature, and equipment start-up / shutdown status. Each entry is time-stamped and serves as the core data unit for extracting the operational characteristics of the oil system, possessing structured and highly correlated attributes. The operational event sequence refers to the set of oil system operational state changes extracted from the key log entries and arranged chronologically. Each sequence element corresponds to a specific operational state adjustment or stable operation phase, covering information such as event occurrence time, duration, and key operational parameter values, fully presenting the dynamic operation process of the oil system. The concurrent oil sample indicators refer to oil performance test data from historical oil sample data whose collection time falls within the same time interval as the occurrence time of each event in the operational event sequence. These indicators include core parameters such as viscosity, moisture content, particulate contamination, and chemical composition, reflecting the corresponding... The data directly reflects the actual performance of the oil under operating conditions. The key timestamps are standardized time markers used to mark the start / end time of operating events and the oil sample collection time. Recorded using a unified time format, they accurately pinpoint the time node of each operating event and its corresponding oil sample collection, providing a clear basis for time alignment between the operating event sequence and concurrent oil sample indicators. The sequence-indicator correspondence refers to the mapping between each operating event and its corresponding concurrent oil sample indicator in the operating event sequence formed after key timestamp matching. Each associated item contains complete parameters for a single operating event and a set of concurrent oil sample indicator data, clearly defining the correspondence between a specific operating state and oil performance. The joint sub-relationship refers to a local data unit in the sequence-indicator correspondence that directly reflects the association between a specific operating scenario of the oil system and its corresponding oil sample performance indicator. Each unit contains the core operating parameters and matching oil sample indicator combination for that scenario, forming the basic association unit for joint operating data.

[0025] S2. Based on the joint operation data, construct the unit knowledge graph corresponding to the thermal power unit. The unit knowledge graph includes: oil index, unit component status and operation-related events.

[0026] The embodiments of the present invention can transform scattered oil-related information and unit operation information into a structured relational network, clearly presenting the intrinsic relationship between oil status and unit operation, providing systematic information support for in-depth analysis of oil anomalies, and avoiding the one-sidedness of interpreting single data.

[0027] The unit knowledge graph refers to a structured semantic network built on joint operation data, with oil indicators, unit component status, and operation-related events of thermal power units as core nodes, and the inherent relationships between nodes, such as the impact of operation events on oil indicators and the correspondence between component status and oil performance, as edges. Through standardized knowledge representation, it systematically integrates information related to unit operation and oil status, and clearly presents the logical relationships between various elements.

[0028] In this embodiment of the invention, based on the joint operation data, a knowledge graph of the thermal power unit is constructed, including: parsing the oil and fluid index items and unit component descriptions in the joint operation data, and extracting the joint operation events recorded in the joint operation data; classifying the knowledge entity relationships between the oil and fluid index items, the unit component descriptions, and the joint operation events; and constructing the knowledge graph of the thermal power unit based on the knowledge entity relationships.

[0029] The oil performance indicators refer to specific data entries extracted from the joint operation data to characterize oil performance. Each entry includes information such as indicator name, measured value, unit of measurement, measurement time, and corresponding operating conditions, covering core aspects such as viscosity, moisture content, particulate contamination, and chemical composition. These are the basic data units reflecting the state of the oil. The unit component description refers to a collection of information about relevant components of the thermal power unit, extracted from the joint operation data. This includes component name, model specifications, installation location, subsystem to which it belongs, and operating parameters such as temperature, pressure, and vibration. It also covers the connection relationship between the component and the oil system, describing the component's attributes. The key information includes: Operational collaboration events refer to event records extracted from joint operation data that involve the coordinated action of the hydraulic system and other subsystems during the operation of the thermal power unit. These records include the event occurrence time, triggering conditions, participating unit components, and parameter changes in the hydraulic system and related subsystems, providing a complete picture of the multi-system collaborative operation process. Knowledge entity relationships refer to the clearly defined interrelationships among the three categories of knowledge entities: hydraulic index items, unit component descriptions, and operational collaboration events. These relationships include the attribution relationship between hydraulic index items and corresponding monitoring components, the impact of operational collaboration events on hydraulic index items, and the participation relationship between unit component descriptions and operational collaboration events.

[0030] Furthermore, in this embodiment of the invention, classifying the knowledge entity relationships among the oil index items, the unit component descriptions, and the operational linkage events includes: identifying entity identifiers in the oil index items, unit component descriptions, and operational linkage events, and determining the action direction corresponding to the entity identifier; after defining the entity triggering conditions corresponding to the action direction, generating the knowledge entity relationships under the entity triggering conditions.

[0031] The entity identifier refers to the feature information extracted from three types of knowledge entities—oil index items, unit component descriptions, and operational linkage events—used to uniquely distinguish each entity. This includes the entity name, unique number, and core attribute keywords. Each identifier corresponds to a single entity and includes a basic attribute description of that entity, enabling precise location and differentiation of different types of knowledge entities. The action direction refers to the direction and object of mutual influence and association between different entities determined after identifying the entity identifier. This clarifies the specific direction in which a certain entity, such as an operational linkage event, affects another entity, such as an oil index item, or whether an entity, such as a unit component description, has a specific attributive association with another entity, such as an oil index item, clearly defining the action logic between entities. The entity triggering condition refers to the specific preconditions defined that can cause different entities to form knowledge entity relationships, including the operating condition range of the entity, parameter value range, and event occurrence sequence. When entities meet these preset conditions, a clear association relationship will be formed between the corresponding entities.

[0032] In detail, the unit knowledge graph includes: oil fluid indicators, unit component status, and operational related events. The oil fluid indicators refer to a specific set of parameters characterizing the performance status of the oil in the hydraulic system of a thermal power unit, covering core parameters such as oil viscosity, moisture content, particulate contamination, and chemical composition content. Each indicator includes the measured value, unit of measurement, measurement time, and corresponding unit operating conditions, serving as a fundamental characterizing element reflecting whether the oil is functioning normally. The unit component status refers to the real-time or historical status of core components associated with the hydraulic system in the thermal power unit during operation. The information includes parameters such as component name, model and specifications, installation location, operating temperature, working pressure, vibration amplitude and sealing integrity, which directly reflect the operating conditions and health status of the components and are directly related to the oil condition. The operation-related events refer to the records of unit operation events that can affect the oil system status or are related to changes in oil indicators during the operation of the thermal power unit. These include information such as the time of event occurrence, triggering cause, participating unit subsystems and components, and the adjustment range of operating parameters, covering key operating scenarios related to the oil condition, such as equipment start-up and shutdown, load changes, and operating condition switching.

[0033] S3. After identifying the physical state of the oil index in the oil system, and combining it with the state of the unit components, generate the change curve of the oil system during the operating cycle, and locate the abnormal segments in the change curve that deviate from the historical benchmark.

[0034] The embodiments of the present invention can realize dynamic tracking of the operating status of the oil system and transform the dispersed status information into a time-series change curve, fully presenting the status evolution trajectory within the operating cycle. It can accurately capture abnormal changes that deviate from historical benchmarks and avoid the lag of static threshold judgment.

[0035] The "oil physical state" refers to the current actual performance state of the oil, determined comprehensively based on real-time monitoring data of oil indicators in the oil system and the state of unit components under corresponding operating conditions. This includes the oil's lubrication capacity, cleanliness, chemical stability, and other actual performance characteristics. It is a concrete description of the oil indicator values, directly reflecting whether the oil currently meets the unit's operating requirements. The "change curve" refers to a time-series curve formed by data fitting, with time as the horizontal axis and the core parameters corresponding to the oil physical state and the associated unit component state parameters as the vertical axis. This curve fully presents the state evolution trajectory of the oil system throughout its entire operating cycle, clearly demonstrating the dynamic changes of each parameter over time. The historical baseline refers to the parameter variation range or benchmark curve determined after statistical analysis and data calibration based on the joint operation data of the oil system of thermal power unit over a long period of normal operation. It covers the normal state range under different operating conditions and different operating stages and serves as a reference standard for judging whether the current oil system state is abnormal. The abnormal curve segment refers to the part of the curve in which the relevant parameters of the oil system exceed the normal range set by the historical baseline, or the variation trend deviates significantly from the historical baseline curve. This curve segment corresponds to the abnormal state in the operation of the oil system and clearly marks the time range and degree of deviation of the abnormality.

[0036] Optionally, the identification of the oil entity state of the oil index in the oil system can be achieved by a fuzzy comprehensive evaluation algorithm. For example, first, core oil indicators such as oil viscosity, water content, and particulate contamination degree are selected, and the corresponding related parameters of the unit components are combined as evaluation factors. The weight of each factor is determined by the analytic hierarchy process, and a fuzzy evaluation matrix is ​​constructed. Then, based on the preset oil state classification standard, fuzzy reasoning and comprehensive calculation are performed on each evaluation factor, and finally the current oil entity state of the oil system is output.

[0037] In this embodiment of the invention, generating the change curve of the oil system within the operating cycle by combining the status of the unit components includes: parsing the periodic operating data corresponding to the oil-coordinated components related to the status of the unit components; collecting the oil change time sequence of the mid-segment nodes of the periodic operating data within the operating cycle; and generating the change curve corresponding to the change time sequence points in the oil change time sequence.

[0038] The oil-hydraulic cooperating components refer to the core components in a thermal power unit that are directly related to and work in conjunction with the oil system. These include oil storage devices, oil pipelines, filter components, circulating pumps, and moving parts lubricated and cooled by the oil. Their operating status is interdependent with the oil performance, making them crucial components for the oil system to achieve lubrication and cooling functions, and directly related to changes in the oil state. The periodic operating data refers to all operating status data of the oil-hydraulic cooperating components within a complete operating cycle of the thermal power unit. This data covers key information such as operating time, operating temperature, operating pressure, vibration amplitude, number of start-stop cycles, and flow parameters, and is continuously recorded in a time sequence, fully covering the component's operating status under different operating conditions within the cycle. The period nodes refer to key time nodes within the thermal power unit's operating cycle, divided according to preset rules, including weekly... The system includes start points, operating condition switching points, periodic inspection points, and load adjustment nodes. Each node corresponds to a specific operating scenario and is used to accurately mark the time of oil condition data collection. The oil condition change time series refers to the records of oil condition changes arranged in chronological order with the period nodes as the time axis. It includes core indicator data such as oil viscosity, moisture content, and particulate contamination for each period node, as well as the associated information of the unit components at the corresponding nodes, fully presenting the dynamic evolution of the oil condition within the operating cycle. The change time series point refers to the specific data unit corresponding to each period node in the oil condition change time series. It includes the time identifier of the node, the specific value of the core oil indicator, and the corresponding unit component status parameters. It is the basic data point that constitutes the oil system change curve and can accurately reflect the oil condition at a specific moment.

[0039] In this embodiment of the invention, locating abnormal segments in the change curve that deviate from the historical benchmark includes: analyzing the deviation value of the curve deviation segment between the change curve and the historical benchmark curve, marking the deviation segment of the change curve that exceeds the deviation value as an abnormal candidate segment; and verifying that the segment with the largest deviation from the historical benchmark among the abnormal candidate segments is an abnormal segment.

[0040] The historical baseline curve refers to a time-series baseline curve formed after data cleaning, statistical calibration, and curve fitting, based on the joint operation data of the hydraulic system of thermal power units over a long period of normal operation. Its horizontal axis represents the operating time, and the vertical axis represents the core hydraulic indicators and related unit component status parameters, covering the normal parameter change trends under different operating conditions and different operating stages. The deviation value is a numerical indicator used to quantify the degree of difference between the current change curve and the historical baseline curve. It is obtained by calculating the parameter difference, trend slope deviation, or curve similarity at corresponding time points of the two curves, with a preset reasonable numerical threshold. This indicator directly reflects the degree of deviation of the current curve segment from the baseline curve and is the core quantitative standard for judging whether the curve segment is abnormal. The abnormal candidate segment refers to the curve segment in the current change curve whose deviation value exceeds the preset threshold after deviation analysis. This segment has significant differences from the historical baseline curve in terms of hydraulic indicator parameter values ​​or change trends. It has not yet been finally verified to be a real abnormality and is only a suspected abnormal curve segment for further verification.

[0041] S4. Determine the unit cooperating component corresponding to the oil abnormality event and the operation-related event in the abnormal curve segment, and extract the oil deterioration characteristics that appear in the oil system when the unit cooperating component is running.

[0042] The embodiments of the present invention can break the isolated analysis mode of anomalies and component operation, and by extracting the oil deterioration characteristics of unit cooperating components during operation, accurately capture the core impact information behind the anomaly, avoiding judgment based solely on surface phenomena, and providing clear feature indicators for tracing the root cause of the anomaly.

[0043] The oil abnormality event refers to a specific event during the operation of the oil system where oil indicators exceed the normal range set by historical benchmarks, or the physical state of the oil deviates from the requirements of normal operation. This includes the time interval of the event, the abnormal core oil indicators, the corresponding unit operating conditions, and the status information of related unit components; it is the specific manifestation of an abnormality in the oil system. The unit cooperating components refer to the set of components in a thermal power unit that are directly related to the operation of the oil system and have a synergistic effect with the oil abnormality event and operation-related events. This includes the core components of the oil system itself, such as oil pipelines and filter components, and related unit components affected by or influencing the state of the oil, such as turbine moving parts and circulating pumps. The oil degradation characteristics refer to the specific manifestations of oil performance degradation that occur in the oil system during the operation of the unit cooperating components. This includes quantifiable characteristic parameters such as abnormal fluctuations in oil viscosity, excessive moisture content, increased particulate contamination, and changes in chemical composition, as well as the trends of these parameters over time, directly reflecting the specific type and degree of oil degradation.

[0044] In this embodiment of the invention, determining the unit collaborative component corresponding to the oil abnormality event in the abnormal curve segment and the operation-related event includes: identifying the abnormal oil identifier corresponding to the oil abnormality event in the abnormal curve segment; tracing the event association link of the operation-related event in the unit knowledge graph based on the abnormal oil identifier; and determining the unit collaborative component corresponding to the collaborative component identifier in the event association link.

[0045] The oil abnormality event refers to the abnormal operation of the oil system corresponding to the abnormal curve segment. Specifically, it refers to a specific event where the oil index exceeds the normal range of historical benchmarks or the physical state of the oil deviates from the operating requirements. It includes the time interval of the event, the type of abnormal core oil index, the corresponding unit operating condition, and the status information of related components. It is a specific representation of the oil system abnormality. The abnormal oil identifier refers to a set of feature information used to uniquely identify the oil abnormality event. It includes the name of the abnormal oil index, a unique code, the timestamp of the abnormality, and the corresponding operating condition label. Each identifier corresponds one-to-one with a single oil abnormality event, which can accurately locate the associated node of the event in the unit knowledge graph. The operation-related event refers to the unit operation event in the unit knowledge graph that is directly or indirectly related to the oil abnormality event, including the event occurrence time. Information such as triggering causes, participating subsystems and components, and operational parameter adjustments covers key operational scenarios that may trigger or be associated with oil abnormalities, including equipment start-up and shutdown, load changes, and operating condition switching. The event association link refers to a structured association path in the unit knowledge graph, starting from the abnormal oil identifier, connecting oil abnormality events with corresponding operational related events and unit components. It consists of multiple knowledge entity nodes, such as abnormal events, operational events, components, and the relationships between nodes, clearly presenting the association logic between various elements. The collaborative component identifier refers to the characteristic information used to uniquely identify collaborative components of the unit, including core identifier elements such as the component's unique number, name, installation location, and subsystem to which it belongs. This identifier is stored in the component nodes of the unit knowledge graph and can be directly extracted through the event association link, serving as a key basis for identifying the corresponding collaborative component of the unit.

[0046] Furthermore, in this embodiment of the invention, the step of extracting the oil deterioration characteristics that appear in the oil system during the operation of the unit's cooperating components includes: defining an overlapping time window during the operation of the unit's cooperating components, and retrieving the cooperating state records of the unit's cooperating components within the overlapping time window; identifying the oil state sequence of the cooperating state records under multiple detections, and screening out the oil deterioration characteristics that appear in the oil state sequence during component cooperation.

[0047] The overlapping time window refers to the overlapping time interval defined after time dimension calibration, based on the actual operating period of the unit's collaborative components and the time interval of the occurrence of abnormal oil events. This window accurately covers the time range in which the operation of the unit's collaborative components and abnormal oil events occur simultaneously, providing a clear time boundary for focusing on the analysis of the correlation between component operation and oil deterioration. The start and end times of the window are marked with standardized timestamps. The collaborative status record refers to the original data record of the operating status of the unit's collaborative components collected by the unit's monitoring equipment within the overlapping time window. It includes core operating information such as component operating temperature, working pressure, vibration amplitude, start-stop status, and flow parameters, as well as interactive correlation data between the component and the oil system. It is stored in a time-series structure to fully reflect the collaborative operating status of the components within this period. The oil state sequence refers to the oil state data chain formed by integrating the results of multiple oil tests, such as viscosity test, moisture test, and particulate contamination test, based on the collaborative status record within the overlapping time window and arranged in chronological order. Each data node corresponds to the core oil indicator test value at a specific time point, fully presenting the continuous changes in the oil state during the collaborative operation of the components.

[0048] S5. Formulate maintenance and response measures corresponding to the specific deterioration elements in the oil deterioration characteristics to achieve oil monitoring and maintenance of the oil system in the thermal power unit.

[0049] The embodiments of the present invention can achieve precise matching of deterioration problems and solutions, and allow maintenance actions to focus on the core causes of deterioration, thereby improving the pertinence of maintenance. It can directly support the targeted monitoring and maintenance of the oil system and ensure the stable operation of related systems of thermal power units.

[0050] The specific deterioration elements refer to the core components that are quantifiable and clearly point to the cause of deterioration, extracted from the characteristics of oil deterioration. These include the numerical range of abnormal fluctuations in oil viscosity, the specific magnitude of water content exceeding the standard, the particle size distribution and the magnitude of exceeding the standard for particulate contamination, and the specific types of changes in chemical composition. Each element has clear detection standards and judgment thresholds and is the basic unit constituting the characteristics of oil deterioration. The maintenance response measures refer to the targeted oil system maintenance operation plan formulated based on the type, degree, and potential impact of the specific deterioration elements. These include specific execution methods such as oil filtration and purification, replacement of deteriorated oil, repair of cooperating components, and adjustment of operating parameters. The operation process, tools used, technical parameters, and timing of execution are clearly defined, making them actionable solutions to address the corresponding oil deterioration problems.

[0051] In this embodiment of the invention, the step of formulating maintenance response measures corresponding to specific deterioration elements in the oil deterioration characteristics includes: analyzing the degree of deterioration corresponding to specific deterioration elements in the oil deterioration characteristics, querying maintenance-related items in the unit knowledge graph; identifying the maintenance conditions corresponding to the maintenance-related items, and formulating maintenance response measures for the oil system under the maintenance conditions.

[0052] The degree of degradation refers to the severity of deviation of specific degradation elements from normal standards in the oil degradation characteristics. It is determined based on quantitative indicators such as the difference between the detected value of the element and the historical benchmark, the duration of exceeding the standard, and the rate of change, and is divided into grades such as mild, moderate, and severe, with each grade corresponding to a specific numerical threshold or characteristic manifestation. The maintenance-related items refer to the set of maintenance-related information directly related to specific degradation elements stored in the unit's knowledge graph, including the treatment schemes corresponding to similar degradation elements in the past, the appropriate maintenance technologies, the maintenance standards of related components, and the maintenance effect verification data, which are composed of the correlation between degradation type and maintenance operation. The maintenance conditions refer to the specific operating scenario conditions of the thermal power unit and oil system when formulating maintenance response measures, covering key factors such as the current load level of the unit, the operating pressure / temperature parameters of the oil system, the health status of cooperating components, and external environmental conditions, which clarifies the execution prerequisites and constraints of maintenance operations.

[0053] Optionally, the oil system monitoring and maintenance in the thermal power unit can be achieved through a closed-loop monitoring and maintenance method. For example, the oil system indicators and unit cooperating component status data are collected in real time by sensors and transmitted to the data processing module. After the module analyzes the data to obtain specific deterioration factors and deterioration degree, it calls the unit knowledge graph to query matching maintenance related items, generates maintenance response measures in combination with the current maintenance conditions, and after the measures are implemented, the oil and component data are collected again to verify the effect. If the standard is not met, the measures are iteratively adjusted to form a closed loop, and finally the monitoring and maintenance of the oil system is achieved.

[0054] Compared to the problems described in the background art, the embodiments of the present invention can comprehensively cover the operating scenarios and oil status information of the oil system, forming more complete data support. This provides basic data that aligns with the actual operating conditions of the unit for accurate judgment of oil anomalies, effectively avoiding the limitations of a single data dimension. Furthermore, the embodiments of the present invention can transform scattered oil-related information and unit operating information into a structured relational network, clearly presenting the intrinsic connection between oil status and unit operation, providing systematic information support for in-depth analysis of oil anomalies, and avoiding the one-sidedness of interpreting single data. Finally, the embodiments of the present invention can achieve dynamic tracking of the operating status of the oil system and transform scattered status information into time-series change curves, fully presenting the state evolution within the operating cycle. By changing the trajectory, the invention can accurately capture abnormal changes that deviate from historical benchmarks, avoiding the lag in static threshold judgment. Furthermore, the embodiments of the invention can break the isolated analysis mode of anomalies and component operation, and by extracting the oil deterioration characteristics of unit cooperating components during operation, it can accurately capture the core impact information behind the anomalies, avoiding judgments based solely on surface phenomena. This provides clear feature indications for tracing the root cause of anomalies. Finally, the embodiments of the invention can achieve precise matching between deterioration problems and solutions, and allow maintenance actions to focus on the core causes of deterioration, improving the targeting of maintenance. This can directly support the targeted monitoring and maintenance of the oil system, ensuring the stable operation of related systems in thermal power units. Therefore, the invention can improve the accuracy of oil anomaly monitoring in thermal power units.

[0055] like Figure 2 The diagram shown is a functional module diagram of a knowledge graph-based oil anomaly monitoring system for thermal power units according to the present invention.

[0056] The knowledge graph-based oil and fluid anomaly monitoring system 300 for thermal power units described in this invention can be installed in an electronic device. Depending on the functions implemented, the knowledge graph-based oil and fluid anomaly monitoring system for thermal power units can be integrated with a module 301, a graph construction module 302, an anomaly segment module 303, a feature extraction module 304, and a monitoring and maintenance module 305. The module described in this invention can also be called a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, stored in the memory of the electronic device.

[0057] In this embodiment of the invention, the functions of each module / unit are as follows: The joint operation module 301 is used to obtain the system operation log corresponding to the oil system in the thermal power unit, and then generate the joint operation data corresponding to the oil system based on the system operation log and historical oil sample data. The graph construction module 302 is used to construct a unit knowledge graph corresponding to the thermal power unit based on the joint operation data. The unit knowledge graph includes: oil index, unit component status and operation-related events. The abnormal curve module 303 is used to identify the physical state of the oil index in the oil system, and then, in combination with the state of the unit components, generate a change curve of the oil system within the operating cycle, and locate abnormal curves in the change curve that deviate from the historical benchmark. The feature extraction module 304 is used to determine the unit cooperating component corresponding to the oil abnormal event and the operation-related event in the abnormal curve segment, and to extract the oil deterioration characteristics that appear in the oil system when the unit cooperating component is running. The monitoring and maintenance module 305 is used to formulate maintenance measures corresponding to specific deterioration elements in the oil deterioration characteristics, so as to realize the oil monitoring and maintenance of the oil system in the thermal power unit.

[0058] In detail, the modules in the knowledge graph-based thermal power unit oil abnormality monitoring system 300 described in this embodiment of the invention employ the same methods as described above. Figure 1 The method used is the same as the knowledge graph-based oil anomaly monitoring method for thermal power units described above, and can produce the same technical effect, so it will not be repeated here.

[0059] In one embodiment, a computer device is provided, which may be a server or a client, and its internal structure diagram may be as follows: Figure 3 As shown. The computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used for communication with external clients via a network connection. When the computer program is executed by the processor, it implements the functions or steps of a knowledge graph-based method for monitoring oil anomalies in thermal power units on the server or client side.

[0060] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps: After obtaining the system operation logs corresponding to the oil system in the thermal power unit, the joint operation data corresponding to the oil system is generated based on the system operation logs and historical oil sample data. Based on the joint operation data, a unit knowledge graph corresponding to the thermal power unit is constructed. The unit knowledge graph includes: oil index, unit component status and operation-related events. After identifying the physical state of the oil index in the oil system, and combining it with the state of the unit components, a change curve of the oil system during the operating cycle is generated, and abnormal segments in the change curve that deviate from the historical benchmark are located. Identify the unit cooperating components corresponding to the abnormal oil events and the operation-related events in the abnormal curve segment, and extract the oil deterioration characteristics that appear in the oil system when the unit cooperating components are running; Develop maintenance and response measures corresponding to specific deterioration elements in the oil degradation characteristics to achieve oil monitoring and maintenance of the oil system in the thermal power unit.

[0061] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor: After obtaining the system operation logs corresponding to the oil system in the thermal power unit, the joint operation data corresponding to the oil system is generated based on the system operation logs and historical oil sample data. Based on the joint operation data, a unit knowledge graph corresponding to the thermal power unit is constructed. The unit knowledge graph includes: oil index, unit component status and operation-related events. After identifying the physical state of the oil index in the oil system, and combining it with the state of the unit components, a change curve of the oil system during the operating cycle is generated, and abnormal segments in the change curve that deviate from the historical benchmark are located. Identify the unit cooperating components corresponding to the abnormal oil events and the operation-related events in the abnormal curve segment, and extract the oil deterioration characteristics that appear in the oil system when the unit cooperating components are running; Develop maintenance and response measures corresponding to specific deterioration elements in the oil degradation characteristics to achieve oil monitoring and maintenance of the oil system in the thermal power unit.

[0062] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or computer device described above can be referred to the relevant descriptions on the server side and client side in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0063] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0064] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0065] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Finally, it should be noted that in the above embodiments, each embodiment can be combined with another or independent of the others; deleting any one of them does not affect the technical implementation of the other embodiments. The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention.

Claims

1. A method for monitoring abnormal oil levels in thermal power units based on knowledge graphs, characterized in that, The system includes: After obtaining the system operation logs corresponding to the oil system in the thermal power unit, the joint operation data corresponding to the oil system is generated based on the system operation logs and historical oil sample data. Based on the joint operation data, a unit knowledge graph corresponding to the thermal power unit is constructed. The unit knowledge graph includes: oil index, unit component status and operation-related events. After identifying the physical state of the oil index in the oil system, and combining it with the state of the unit components, a change curve of the oil system during the operating cycle is generated, and abnormal segments in the change curve that deviate from the historical benchmark are located. Identify the unit cooperating components corresponding to the abnormal oil events and the operation-related events in the abnormal curve segment, and extract the oil deterioration characteristics that appear in the oil system when the unit cooperating components are running; Develop maintenance and response measures corresponding to specific deterioration elements in the oil degradation characteristics to achieve oil monitoring and maintenance of the oil system in the thermal power unit.

2. The method for monitoring abnormal oil levels in thermal power units based on knowledge graphs as described in claim 1, characterized in that, The construction of the unit knowledge graph corresponding to the thermal power unit based on the joint operation data includes: The oil index items and unit component descriptions in the joint operation data are analyzed, and the joint operation events recorded in the joint operation data are extracted. Categorize the knowledge entity relationships between the oil index items, the unit component descriptions, and the operational linkage events; Based on the knowledge entity relationships, a unit knowledge graph corresponding to the thermal power unit is constructed.

3. The method for monitoring oil anomalies in thermal power units based on knowledge graphs as described in claim 2, characterized in that, The classification of the knowledge entity relationships between the oil index items, the unit component descriptions, and operational linkage events includes: Identify the entity identifiers in the oil index items, unit component descriptions, and operational linkage events, and determine the function corresponding to the entity identifiers; After defining the corresponding entity triggering condition for the action, the knowledge entity relationship under the entity triggering condition is generated.

4. The method for monitoring oil anomalies in thermal power units based on knowledge graphs as described in claim 1, characterized in that, The step of generating joint operation data for the oil system based on the system operation log and historical oil sample data includes: Read the sequence of running events corresponding to key log entries in the system operation log, and extract the same period oil sample indicators from the historical oil sample data; Align the sequence of running events with the key timestamps in the oil sample indicators of the same period to obtain the sequence-indicator correspondence; Integrate the joint operation data corresponding to the joint sub-relationships in the sequence-index correspondence.

5. The method for monitoring abnormal oil levels in thermal power units based on knowledge graphs as described in claim 1, characterized in that, The step of generating the change curve of the oil system during the operating cycle by combining the status of the unit components includes: Analyze the periodic operation data corresponding to the oil-hydraulic coordination components related to the status of the unit components; The time sequence of oil changes at mid-segment nodes during the operational cycle is collected; Generate the change curves corresponding to the change time points in the oil change time series.

6. The method for monitoring abnormal oil levels in thermal power units based on knowledge graphs as described in claim 1, characterized in that, Locating the abnormal segments in the change curve that deviate from the historical baseline includes: After analyzing the deviation value of the curve deviation segment between the change curve and the historical baseline curve, the deviation segment of the change curve that exceeds the deviation value is marked as an abnormal candidate segment; Verify that the segment with the largest deviation from the historical benchmark among the candidate abnormal segments is the abnormal curve segment.

7. The method for monitoring oil anomalies in thermal power units based on knowledge graphs as described in claim 1, characterized in that, The unit coordination component that determines the oil abnormality event and the operation-related event corresponding to the abnormal curve segment includes: Identify the abnormal oil identifiers corresponding to abnormal oil events in the abnormal curve segments; Based on the abnormal oil identification, the event association links of the operation-related events in the unit knowledge graph are traced. Identify the unit coordination component corresponding to the coordination component identifier in the event association link.

8. The method for monitoring abnormal oil levels in thermal power units based on knowledge graphs as described in claim 1, characterized in that, The extraction of oil degradation characteristics that occur in the oil system during the operation of the unit's cooperative components includes: Define the overlapping time window for the operation of the unit's collaborative components, and retrieve the collaborative status records of the unit's collaborative components within the overlapping time window; Identify the oil state sequence recorded under multiple detections in the collaborative state record, and screen out the oil deterioration characteristics that appear in the oil state sequence when components are working together.

9. The method for monitoring abnormal oil levels in thermal power units based on knowledge graphs as described in claim 1, characterized in that, The formulation of maintenance measures corresponding to specific deterioration elements in the oil degradation characteristics includes: After analyzing the degree of deterioration corresponding to the specific deterioration elements in the oil deterioration characteristics, the maintenance-related items in the unit knowledge graph are queried. After identifying the maintenance conditions corresponding to the maintenance-related items, formulate maintenance response measures for the oil system under the maintenance conditions.

10. A method for monitoring abnormal oil levels in thermal power units based on knowledge graphs, characterized in that, The system includes: The joint operation module is used to obtain the system operation logs corresponding to the oil system in the thermal power unit, and then generate the joint operation data corresponding to the oil system based on the system operation logs and historical oil sample data. The knowledge graph construction module is used to construct the unit knowledge graph corresponding to the thermal power unit based on the joint operation data. The unit knowledge graph includes: oil index, unit component status and operation-related events. The abnormal curve module is used to identify the physical state of the oil index in the oil system, and then, in combination with the state of the unit components, generate the change curve of the oil system within the operating cycle, and locate the abnormal curve that deviates from the historical benchmark in the change curve. The feature extraction module is used to determine the unit cooperating component corresponding to the oil abnormal event and the operation-related event in the abnormal curve segment, and to extract the oil deterioration characteristics that appear in the oil system when the unit cooperating component is running. The monitoring and maintenance module is used to formulate maintenance measures corresponding to specific deterioration elements in the oil deterioration characteristics, so as to realize the monitoring and maintenance of the oil system in the thermal power unit.

Images