CN-121996319-A - IEC 61850-oriented relay protection parameter automatic modeling and consistency verification method

CN121996319ACN 121996319 ACN121996319 ACN 121996319ACN-121996319-A

Abstract

The invention belongs to the technical field of relay protection, and discloses an IEC 61850-oriented relay protection parameter automatic modeling and consistency verification method, which comprises the steps of synchronizing digital twin topology and extreme working condition data through a multi-source data acquisition module, adding a topology adaptation layer in a meta model, embedding dynamic update triggering conditions, setting a new energy special logic node, and refining adaptation parameter configuration; the model can automatically adjust parameters according to topology reconstruction and new energy power fluctuation to avoid protection misoperation caused by adaptation deviation, performs parameter verification by combining a physical mechanism and an algorithm to ensure that core protection parameters meet actual operation requirements, embeds verification logic into a modeling flow in advance, eliminates the error parameters in real time through three-stage prepositive verification, and then relies on an edge and cloud collaborative verification engine and an interpretable traceability mechanism to enable problems found by verification to be directly related to a modeling specific link.

Inventors

GAO WEIQIANG
REN HEHE
XU LEI
HUANG HUANG
LI CAN
SUN HAIBO
SONG YONGLIANG
ZHANG HONGGUANG
LIU YI
WANG LING
HAN XIAOYING
HAN XUEHUA
WANG ANQIN
Du Tingbin

Assignees

兖矿能源集团股份有限公司
兖矿能源集团股份有限公司山东煤炭科技研究院分公司

Dates

Publication Date: 20260508
Application Date: 20251226

Claims (8)

1. An IEC 61850-oriented relay protection parameter automatic modeling and consistency verification method is characterized by comprising the following steps: establishing a multi-source data acquisition module, realizing multi-protocol adaptation, new energy equipment data acquisition, digital twin topology and extreme working condition data synchronization, extracting a fixed value single core parameter to generate a three-dimensional semantic dictionary, establishing a data quality grading evaluation mechanism, and generating a standardized data middleware through missing value filling and outlier removing processing; The meta model construction stage comprises the steps of establishing a five-layer architecture meta model of an equipment layer, a logic node layer, a data object layer, a topology adaptation layer and a collaborative check layer based on standardized data middleware, setting a new energy special logic node and an expansion specification, embedding digital twin topology association rules and collaborative check constraints to form a trinity rule base, generating an resolvable description file, configuring a custom expansion interface and supporting novel scene expansion; The automatic modeling stage comprises developing a meta model drive integrated engine integrating edge calculation, deploying local edge nodes to realize analysis and hierarchical structure generation, synchronously executing three-stage pre-verification to generate an SCD initial draft with a digital twin identifier and a traceability report, and supporting special parameter manual intervention and model self-learning; The verification rule construction stage comprises the steps of establishing a three-dimensional core verification rule system comprising topology adaptation, collaborative verification and extreme working condition adaptation, integrating verification logic, configuring a dynamic self-learning module optimization rule, and establishing a validity evaluation mechanism to simplify a rule base; The automatic verification stage comprises the steps of deploying a parallel computing engine with cooperative edges and cloud, performing full-dimensional automatic verification on a scene by scenes, establishing a modeling link traceability mechanism, generating a visual traceability map, integrating verification results to carry out problem classification, outputting a classification error report and providing real-time monitoring and interactive adjustment functions; The abnormal optimization stage comprises the steps of establishing a modeling, checking and operation and maintenance full-chain linkage traceability map, locating an abnormal source through an algorithm, quantifying influence weight, combining digital twin to realize targeted optimization, generating a corresponding optimization scheme, shortening a checking period by adopting an incremental checking mechanism, and forming an optimization and checking closed loop; the management stage is to build a digital twin integrated full life cycle management platform to realize the full flow intellectualization of the model, and has health evaluation and version management functions and supports dynamic update.
2. The IEC 61850-oriented relay protection parameter automatic modeling and consistency verification method according to claim 1, wherein in the data acquisition preprocessing stage: the multi-source data acquisition module acquires ICD files and operation data of the new energy protection equipment by adapting a basic IEC61850 protocol and a new energy expansion IEC61850 protocol through a preset protocol analysis template, synchronizes power grid topology, new energy access information and extreme working condition topology prejudgement data through a power grid digital twin real-time association interface, and realizes the linkage of physical and virtual models; Integrating a multi-mode parameter extraction engine, integrating a professional knowledge model in the electric power field with a deep learning algorithm to extract a constant value single core parameter, mining a protection logic association relationship, and generating a parameter, topology and logic three-dimensional semantic dictionary; And the high-priority data adopts a physical mechanism type missing value filling based on a power grid tide equation and a short circuit calculation model, and an abnormal value removing strategy of statistical identification and mechanism verification to finally generate an IEC61850 standard data middleware fused with digital twin topology information.
3. The IEC 61850-oriented relay protection parameter automatic modeling and consistency verification method according to claim 2, wherein in the meta model construction stage: the five-layer architecture meta-model adopts a physical mechanism and an AI algorithm hybrid driving mechanism, sets a new energy special logic node under protective logic equipment, and refines parameter items and a value range; The topology adaptation layer is embedded with digital twin topology association identifications, dynamic update triggering conditions and extreme working condition adaptation rules, and scene self-adaptive adjustment is realized by defining association relation between topology change threshold values and modeling adjustment strategies; The custom expansion interface supports the training of the type and verification constraint of the expansion logic node through rule importing and model, and automatically completes rule grammar checking and rationality verification in the expansion process.
4. The IEC61850 oriented relay protection parameter automatic modeling and consistency verification method according to claim 3, wherein in the automatic modeling stage: the meta model driving integrated engine is based on XML format meta model description files and a collaborative rule base, edge computing nodes are deployed in the local of the transformer substation, ICD file characteristics are analyzed, digital twin topology information and standard data middleware are combined, a logic device and logic node hierarchical structure are automatically generated, topology association identifiers are bound, and dynamic update triggering conditions are bound; the three-stage pre-verification sequentially verifies parameter naming and format compliance, association consistency of parameters, topology and logic, physical rationality of parameter value, eliminates error parameters in real time and triggers optimization suggestion; For special parameters not covered by the meta-model, manual configuration rule supplement is supported, and the system automatically learns rule features through an AI algorithm and converts the rule features into standardized meta-model rules to update the rule base.
5. The IEC61850 oriented relay protection parameter automatic modeling and consistency verification method according to claim 4, wherein in the verification rule construction stage: The three-dimensional core verification rule system is constructed based on digital twin power grid data, and the corresponding rule of each dimension is as follows: The topology adaptation dimension comprises topology association consistency verification and dynamic updating adaptation verification rules, and real-time adaptation of parameters and power grid topology is ensured by defining association relations between topology change thresholds and parameter adjustment strategies; the cooperative check dimension comprises a cooperative check rule including grammar, semantics and functions and a new energy access impact check rule, the functions cooperative check and fusion tide analysis model verifies the adaptability of parameters to the operation of a power grid, and the new energy access impact check focuses on the matching of the parameters of a distributed power supply access point, short-circuit current and power fluctuation; The extreme working condition adaptation dimension is based on extreme working condition data generated by the digital twin simulation platform, a parameter adaptation verification rule is constructed, and the reliability of a protection function in an extreme scene is ensured; The dynamic self-learning module optimizes rule parameter threshold values and supplements new rules through a machine learning algorithm by collecting historical verification data, fault cases and operation and maintenance experience, and the validity assessment mechanism periodically screens rule libraries to remove redundancy, failure and conflict rules.
6. The IEC61850 oriented relay protection parameter automatic modeling and consistency verification method according to claim 5, wherein in the automatic verification stage: The parallel computing engine adopts an edge and cloud collaborative architecture, an edge node executes topology adaptation and grammar collaborative verification, and a cloud node executes semantic, function collaboration and extreme working condition adaptation verification based on a digital twin model to verify parameter dynamic suitability and robustness; The verification result is graded according to the four levels of the severity and the influence range, and a classification error report containing problem description, a tracing path, influence evaluation and optimization suggestion is output; The real-time progress monitoring and interactive adjustment functions are provided, and the problem jump tracing, online parameter correction and local re-verification are supported.
7. The IEC61850 oriented relay protection parameter automatic modeling and consistency verification method according to claim 6, wherein in the anomaly optimization stage: constructing a full-chain linkage traceability map based on the traceability identification, associating modeling parameter configuration, verification problem records and operation and maintenance fault data, positioning an abnormal source through a graph neural network algorithm and quantifying influence weights; Generating an optimal parameter adjustment scheme aiming at the topology adaptation abnormality and updating a topology association identifier, generating a parameter correction suggestion aiming at the modeling abnormality, and verifying the validity of the optimization scheme by simulation aiming at the extreme working condition adaptation abnormality; And (3) checking the correction link and the associated topology region by adopting an incremental checking mechanism, reserving a checking log, and updating the model version or returning to the optimization link according to the checking result.
8. The IEC61850 oriented relay protection parameter automatic modeling and consistency verification method according to claim 7, wherein in the management phase: The health assessment function calculates a model health score and predicts failure risk through an AI algorithm based on real-time operation data, a verification result and a fault case; Version management function records model full information, supports version backtracking, comparison and incremental update, and can execute dynamic update and pertinence check aiming at power grid topology reconstruction, new energy source addition or standard update; The system has a cross-system interaction function, and realizes data interaction and collaborative operation and maintenance with a substation monitoring system, a dispatching master station and a digital twin platform through a standardized interface.

Description

IEC 61850-oriented relay protection parameter automatic modeling and consistency verification method Technical Field The invention belongs to the technical field of relay protection, and particularly relates to an IEC 61850-oriented relay protection parameter automatic modeling and consistency verification method. Background The smart power grid construction is deepened continuously, new energy is connected in a large-scale mode, so that the power grid topological structure is more complex, and higher requirements are put on the reliability, compatibility and operation and maintenance efficiency of a relay protection system. The IEC61850 standard is used as an international general specification of a transformer substation communication network and system, and provides technical support for interconnection and interworking of relay protection equipment by virtue of a unified information model and communication protocol, but at present, the following technical problems still exist in relay protection parameter modeling and consistency verification based on IEC 61850: Most of the existing modeling modes are static construction, a real-time linkage mechanism of model and power grid topology change is lacked, when the conditions of new energy unit access, power grid topology reconstruction and the like occur, the model is needed to be manually readjusted or modeling is carried out again, response is lagged, and protection misoperation can be caused due to topology adaptation deviation; In the prior art, verification logic is not integrated in the modeling stage, and verification work is only carried out independently after modeling is completed, so that problems found by verification are difficult to trace back to specific deviation points of modeling links quickly, manual link-by-link investigation is needed, full-flow modeling is needed to be carried out again after correction, and the cooperative efficiency is low. Disclosure of Invention The invention aims to provide an IEC 61850-oriented relay protection parameter automatic modeling and consistency verification method so as to solve the problems in the background technology. In order to achieve the aim, the invention provides the technical scheme that the automatic modeling and consistency verification method for the relay protection parameters facing IEC61850 comprises a data acquisition preprocessing stage, a meta model construction stage, an automatic modeling stage, a verification rule construction stage, an automatic verification stage, an abnormal optimization stage and a management stage: Preferably, the data acquisition preprocessing stage designs a multi-source data acquisition module with protocol adaptation, topological association and digital twin association, can automatically adapt to a basic protocol including IEC61850-8-1 and IEC61850-9-2 through a preset protocol analysis template, is compatible with a new energy expansion protocol of IEC61850-7-420, can acquire a special ICD file of new energy protection equipment and operation data including a real-time action state and a historical fault record, and simultaneously constructs a digital twin real-time association interface of a power grid, synchronously acquires topological structure of the power grid, access position and operation state of a new energy unit and topological prejudgement data under extreme working conditions through a high-speed communication link, and realizes real-time linkage of the physical power grid and a virtual model; Integrating a multi-mode parameter extraction engine, fusing a professional knowledge model and a deep learning algorithm in the electric power field, aiming at fixed value sheets in different formats, automatically identifying a document structure and extracting core parameters, and simultaneously generating a three-dimensional semantic dictionary of parameters, topology and logic by deep mining protection logic association relations corresponding to the parameters; The power field expertise comprises IEC61850 standard constant value item definition, new energy protection equipment parameter constraint, topology-parameter association rule and protection logic matching principle, and is used for carrying out standardized verification, alias unification and association relationship completion on candidate parameters extracted by a deep learning algorithm, so as to ensure that the parameters meet the operation requirements of a power system. The deep learning algorithm adopts a hybrid model of a bidirectional encoder architecture based on a transducer and a convolutional neural network, and a model level comprises an input layer, a text coding layer, a feature fusion layer and a parameter extraction layer: The input layer converts PDF, DOCX, XLSX and other fixed value sheets with different formats into a standardized text sequence, tokenization processing is carried out through a WordPiece word segmentation device, and the input dimension is [ batch_siz