CN-122022600-A - Method, device and system for public engineering and energy management of chemical production

CN122022600ACN 122022600 ACN122022600 ACN 122022600ACN-122022600-A

Abstract

Embodiments of the present disclosure provide methods, apparatus, and systems for utility and energy management for chemical production. In response to the presence of data source anomalies, performing a data source anomaly root cause query in an operational knowledge graph to determine first anomaly KPIs of which root causes are anomalous from anomaly KPIs obtained in public engineering and energy management operation processes, and performing a KPI entity root cause query based on an upstream-downstream relationship in the operational knowledge graph for second anomaly KPIs of the anomaly KPIs to determine first KPI entities which are root causes of each second anomaly KPI, wherein the second anomaly KPIs comprise residual KPIs of which the first anomaly KPIs are removed in the anomaly KPIs. By utilizing the method, the root cause of the KPI abnormality problem can be rapidly positioned by constructing the operation knowledge graph for reflecting semantic association between KPIs and business scenes, KPIs and between KPIs and bottom data and analyzing the KPI abnormality root cause based on operation knowledge, and the accuracy of the KPI abnormality root cause analysis is improved.

Inventors

LU XIAOXUAN
ZHU YU

Assignees

巴斯夫一体化基地(广东)有限公司

Dates

Publication Date: 20260512
Application Date: 20260413

Claims (12)

1. A method for utility and energy management for chemical production, the method comprising: after acquiring an abnormal key performance index set in the public engineering and energy management operation process, determining whether a data source is abnormal or not; In response to the presence of a data source anomaly, performing a data source anomaly root cause query in an operational knowledge-graph to determine a first anomaly key performance indicator rooted at the data source anomaly from the anomaly key performance indicator set, and For the second abnormal key performance indicators in the abnormal key performance indicator set, performing a key performance indicator entity root cause query based on an upstream-downstream relationship in the operation knowledge graph to determine a first key performance indicator entity serving as a root cause of each second abnormal key performance indicator, wherein the second abnormal key performance indicators comprise the rest abnormal key performance indicators after the first abnormal key performance indicators are removed in the abnormal key performance indicator set, The operation knowledge graph comprises entity nodes and entity node relations, wherein the entity nodes comprise key performance index entity nodes, data source entity nodes and business scene entity nodes, the entity node relations are used for reflecting logic relations among the entity nodes, and the logic relations comprise the belonging relation of the key performance index entity nodes relative to the business scene entity nodes, the data source relation of the key performance index entity nodes relative to the data source entity nodes and upstream and downstream relations among the key performance index entity nodes.
2. The method of claim 1, wherein said performing a data source anomaly root cause query in an operational knowledge-graph to determine a first anomaly key performance indicator that is root cause anomaly of a data source from the anomaly key performance indicator set comprises: Inquiring whether a third abnormal key performance index with a calculation basis from data generated by an abnormal data source exists in the abnormal key performance index set in the operation knowledge graph; querying the operational knowledge-graph for a fourth abnormal key performance indicator having an upstream relationship with the third abnormal key performance indicator in the presence of a third abnormal key performance indicator whose computational basis is derived from data produced by an abnormal data source, and And determining the third abnormal key performance index and the fourth abnormal key performance index as first abnormal key performance indexes based on the abnormality of the data source.
3. The method of claim 1 or 2, wherein the entity node relationships further comprise causal relationships between key performance indicator entity nodes, the method further comprising: performing causal relation inquiry in the operation knowledge graph to determine second key performance index entities which have causal relation with first key performance index entities serving as root causes of the second abnormal key performance indexes; Performing semantic analysis based on a key performance indicator knowledge base for a first key performance indicator entity for which the corresponding second key performance indicator entity is normal or does not have the corresponding second key performance indicator entity, and And generating root cause analysis results of each second abnormal key performance index based on semantic analysis results of the abnormal second key performance index entity and/or the first key performance index entity which does not have the corresponding second KPI entity and is normal or not.
4. The method of claim 3, wherein the entity node further comprises an entity attribute, the method further comprising: And generating an operation abnormality diagnosis report based on root cause analysis results of the first abnormal key performance index and the second abnormal key performance index, wherein the operation abnormality diagnosis report comprises an abnormality attribution analysis result of the abnormality of the key performance index.
5. The method of claim 4, wherein the operational anomaly diagnostic report further includes countermeasures for a critical performance indicator anomaly.
6. The method of claim 4, wherein the generating an operational anomaly diagnostic report based on root cause analysis results of the first and second anomaly key performance indicators comprises: and providing the root cause analysis results and the root cause analysis processes of the first abnormal key performance index and the second abnormal key performance index to a large language model to generate an operation abnormal diagnosis report with a structured data structure.
7. The method according to claim 1 or 2, wherein the operational knowledge-graph is constructed based on at least one of the following knowledge-graph construction methods: constructing the operation knowledge graph based on the entity nodes, the entity node relations and the entity node attributes which are input by the user according to the appointed format; The operation knowledge map is constructed by carrying out semantic analysis based on a large language model on operation service scene description, data source description and key performance index definition input by a user in natural language to obtain entity nodes, entity node relations and entity node attributes; the operation knowledge graph is constructed by extracting the semantics of the large language model based on the operation specification document with unstructured text format input by the user to obtain entity nodes, entity node relations and entity node attributes, wherein the operation specification document comprises an operation document, an operation rule and a key performance index definition manual, and And constructing the operation knowledge graph based on the historical operation data of the public engineering and energy management system.
8. The method of claim 1, wherein the method further comprises: Acquiring a key performance index query request initiated by a user, wherein the key performance index query request comprises at least one of a historical trend of key performance indexes, influence factors and upstream and downstream key performance indexes; Querying the operation knowledge graph for the query result of the key performance index query request, and And generating a key performance indicator query report based on the query result of the key performance indicator query request to provide to the user.
9. An apparatus for utility and energy management for chemical production, the apparatus comprising: The data source abnormality determining unit is configured to determine whether the data source abnormality exists after acquiring an abnormal key performance index set in the public engineering and energy management operation process; A data source anomaly root cause determination unit configured to perform a data source anomaly root cause query in the operational knowledge graph in response to the presence of a data source anomaly to determine a first anomaly key performance indicator of the root cause of the data source anomaly from the anomaly key performance indicator set, and A key performance indicator entity root cause determination unit configured to perform, for second abnormal key performance indicators in the abnormal key performance indicator set, a key performance indicator entity root cause query based on an upstream-downstream relationship in the operation knowledge graph to determine first key performance indicator entities as root causes of respective second abnormal key performance indicators, the second abnormal key performance indicators including remaining abnormal key performance indicators after the first abnormal key performance indicators are removed in the abnormal key performance indicator set, The operation knowledge graph comprises entity nodes and entity node relations, wherein the entity nodes comprise key performance index entity nodes, data source entity nodes and business scene entity nodes, the entity node relations are used for reflecting logic relations among the entity nodes, and the logic relations comprise the belonging relation of the key performance index entity nodes relative to the business scene entity nodes, the data source relation of the key performance index entity nodes relative to the data source entity nodes and upstream and downstream relations among the key performance index entity nodes.
10. A system for utility and energy management for chemical production, the system comprising: An operation knowledge graph construction device configured to construct an operation knowledge graph of a public engineering and energy management operation process of chemical production; an operation monitoring device configured to perform data source anomaly detection and key performance indicator anomaly detection for the utility and energy management operation process, and The utility and energy management device for chemical production of claim 9.
11. An apparatus for utility and energy management for chemical production, the apparatus comprising: At least one processor; A memory coupled to the at least one processor, and Computer program stored in the memory, the at least one processor executing the computer program to implement the method for utility and energy management for chemical production as claimed in any one of claims 1 to 8.
12. A computer readable storage medium storing executable instructions that when executed cause a processor to perform the method for utility and energy management for chemical production of any one of claims 1 to 8.

Description

Method, device and system for public engineering and energy management of chemical production Technical Field Embodiments of the present disclosure relate generally to the field of chemical production, and more particularly, to methods, apparatus, and systems for utility and energy management for chemical production. Background In the chemical production process, public engineering and energy management (availability AND ENERGY MANAGEMENT) are required. In chemical production, utility is a "basic facility and auxiliary system" that supports the stable operation of the core production equipment of chemical production, which does not directly participate in product synthesis, but determines the continuity, safety, and economy of chemical production. The energy management (ENERGY MANAGEMENT) is the whole-flow management of overall planning, optimal configuration and monitoring accounting for various energy consumed in the public engineering and chemical production process, and is used for realizing energy conservation, cost reduction, compliance, emission reduction and efficient and stable supply. In the operation of utility and energy management, it is necessary to collect a large amount of operation data through operation data collection components (e.g., terminal sensing devices, process control components, metering accounting devices, etc.) deployed at the full-flow nodes of the various systems of the utility (e.g., energy supply, media delivery, consumer terminals, environmental emissions). The collected operation data may include, for example, flow, pressure, temperature, etc., critical parameters of chemical production. These operational data can be translated into various key performance indicators (Key Performance Indicator, KPIs) for helping users to understand the operational status of current utility and energy management and make auxiliary decisions. When KPI abnormal alarm occurs, the potential risk behind the KPI abnormal alarm needs to be judged, and corresponding measures are taken to solve the problem. The existing KPI abnormality alarm processing mechanism is that a digital technician and an operation technician cooperate to construct a static business intelligence (Business Intelligence, BI) report to display the KPI with abnormality alarm and manually analyze the current early warning condition and potential reasons. The existing KPI abnormality alarm processing mechanism needs to perform manual analysis, so that the user expertise is highly relied on, abnormality and trend cannot be automatically identified or root cause analysis cannot be performed, and the auxiliary decision making efficiency is reduced. In addition, the displayed content needs to be predefined, such as a preset chart and a preset data view, so that flexibility is lacking, and dynamic adjustment cannot be performed according to specific problems or application scenes. Moreover, the displayed content presentation lacks context adaptation capability and cannot be automatically adjusted according to the current business context, role or task, resulting in insufficient relevance or insufficient precision of the displayed content. In addition, the information in the displayed content is fragmented, and the user needs to switch among a plurality of dashboards to splice out a complete service scene, so that the cognitive burden and the operation complexity of the user are increased. Disclosure of Invention In view of the foregoing, embodiments of the present disclosure provide methods, apparatus, and systems for utility and energy management for chemical production. By utilizing the method, the root cause of the KPI abnormality problem can be rapidly positioned by constructing the operation knowledge graph for reflecting semantic association between KPIs and business scenes, KPIs and between KPIs and bottom data and analyzing the KPI abnormality root cause based on operation knowledge, and the accuracy of the KPI abnormality root cause analysis is improved. According to one aspect of the embodiments of the present disclosure, a method for public engineering and energy management in chemical production is provided, which includes determining whether there is a data source abnormality after acquiring an abnormal key performance index set in a public engineering and energy management operation process, performing a data source abnormality root cause query in an operation knowledge graph in response to the presence of the data source abnormality to determine a first abnormal key performance index of the root cause abnormality from the abnormal key performance index set, and performing a downstream relationship-based key performance index entity root cause query in the operation knowledge graph for a second abnormal key performance index in the abnormal key performance index set to determine a first key performance index entity serving as a root cause of each second abnormal key performance index, wherein the second abnormal key pe