CN-122020957-A - Method for constructing full life cycle digital twin body of pumped storage power station

CN122020957ACN 122020957 ACN122020957 ACN 122020957ACN-122020957-A

Abstract

The invention relates to the technical field of life cycle construction, in particular to a method for constructing a full life cycle digital twin body of a pumped storage power station, which comprises the steps of obtaining multisource heterogeneous sensing data in a construction period and an operation period, carrying out standardized cleaning and format unified processing, constructing a two-stage space grid coding system comprising a first-stage global grid and a second-stage local fine grid, carrying out space discretization mapping, establishing full life cycle time axis coding, correlating space grid data, carrying out space-time dimension alignment and multi-precision data fusion calculation, and carrying out digital twin geometric modeling, physical attribute mapping and full life cycle state evolution simulation based on space-time fusion data, effectively reducing risks of model distortion and decision bias, improving accuracy of full life cycle state simulation, further avoiding design defects and operation faults, effectively eliminating long-term negative influence on safety and reliability of a power station, and guaranteeing accuracy and scientificity of full life cycle management.

Inventors

WANG BAOGEN
XU YANAN
GE YULIN
LI LINYANG
WANG ZHUOYU
LIU HAITAO
ZHU XI
WANG LUYAO
CHANG YUHONG
YAN JUN
GUO YANG
NIU XIANGYU
LIANG TINGTING
JIANG FAN
LI GENGZE
CAO XINLI

Assignees

国网新源控股有限公司

Dates

Publication Date: 20260512
Application Date: 20251219

Claims (10)

1. The method for constructing the full life cycle digital twin body of the pumped storage power station is characterized by comprising the following steps of: Step S1, multi-source heterogeneous sensing data of a pumped storage power station in a construction period and an operation period are obtained, and standardized cleaning and format unified processing are carried out on the multi-source heterogeneous sensing data to generate multi-source sensing basic data; s2, constructing a two-stage space grid coding system comprising a first-stage global grid and a second-stage local fine grid based on multi-source perception basic data, and performing space discretization mapping to obtain space grid coding data; Step S3, establishing a full life cycle time axis code, and carrying out space-time dimension alignment association on the space grid code data and the full life cycle time axis code to obtain space-time dimension associated data; And S4, carrying out geometric modeling and physical attribute mapping of the digital twin body according to the full life cycle space-time fusion data to obtain a digital twin body basic model, carrying out full life cycle state evolution simulation based on the digital twin body basic model to generate a digital twin body model of the pumped storage power station, and sending the digital twin body model of the pumped storage power station to a management and control terminal.
2. The pumped storage power station full life cycle digital twin body construction method as defined in claim 1, wherein step S1 comprises the steps of: s11, acquiring original sensing signals of a pumped storage power station in a construction period and an operation period by using a laser radar, oblique photography and an Internet of things sensor; step S12, denoising and outlier removing processing are carried out on the original perception signals, and original perception cleaning data are generated; S13, performing semantic annotation and coordinate system conversion on the original perception cleaning data to obtain semantic annotation perception data; and S14, carrying out data structure standardized recombination on the semantic annotation sensing data so as to generate multi-source sensing basic data.
3. The pumped storage power station full life cycle digital twin body construction method as defined in claim 1, wherein step S2 comprises the steps of: S21, carrying out power station space topology structure analysis based on multisource perception basic data to obtain power station space topology data; step S22, dividing a primary macroscopic geographic grid according to the space topology data of the power station to generate a primary global grid index; Step S23, performing high-precision grid subdivision on key equipment and a structural area based on the primary global grid index to obtain a secondary local fine grid index; and step S24, performing two-stage spatial discretization mapping on the multi-source perception basic data according to the first-stage global grid index and the second-stage local fine grid index, so as to obtain spatial grid coding data.
4. A pumped storage power station full life cycle digital twin body construction method as defined in claim 3, wherein step S23 comprises the steps of: Step S231, obtaining the density distribution characteristics of the perceived data in the first-level global grid index, and analyzing the geometric complexity of the key component of the power station space topology data to obtain the geometric complexity data of the key component; Step S232, performing self-adaptive calculation of mesh subdivision granularity according to the geometric complexity data of the key component and the density distribution characteristics of the perception data, and generating mesh subdivision granularity parameters; And S233, performing recursion quadtree subdivision on a specific area in the primary global grid index based on the grid subdivision granularity parameter, thereby obtaining a secondary local fine grid index.
5. A pumped storage power station full life cycle digital twin body construction method as defined in claim 3, wherein step S24 comprises the steps of: s241, extracting space position coordinates of the multisource perception basic data to obtain space position coordinate data; Step S242, performing rough positioning matching on the space position coordinate data based on the first-level global grid index to generate rough positioning matching data; Step S243, performing high-precision position correction and voxelization coding on the rough positioning matching data according to the secondary local fine grid index to obtain high-precision voxelization coding data; step S244, performing hierarchical association combination on the rough positioning matching data and the high-precision voxelized coding data so as to obtain space grid coding data.
6. The pumped storage power station full life cycle digital twin body construction method as defined in claim 1, wherein step S3 comprises the steps of: step S31, performing time stamp analysis and serialization sequencing on the space grid coding data to obtain time serialization grid data; Step S32, performing time slicing division on the time-series grid data based on full life cycle time axis coding to generate time slice grid data; s33, carrying out space-time dimension alignment association on the time slice grid data and full life cycle time axis codes to obtain space-time dimension association data; And step S34, carrying out weighted fusion on the data confidence degrees of the different precision perception sources according to the space-time dimension associated data, thereby obtaining the space-time fusion data of the whole life cycle.
7. The pumped storage power station full life cycle digital twin body construction method as defined in claim 6, wherein step S34 comprises the steps of: Step S341, carrying out spatial overlapping region identification of different perception sources on the time-space dimension related data to obtain spatial overlapping region data; step S342, performing perception source precision grade and data integrity evaluation based on the spatial overlapping region data to generate precision grade and integrity evaluation data; s343, constructing a weighted fusion function matrix according to the precision grade and the integrity evaluation data to obtain weighted fusion function matrix data; and step 344, performing multi-source data complementary fusion operation on the space-time dimension associated data by using the weighted fusion function matrix data, thereby obtaining the space-time fusion data in the whole life cycle.
8. The pumped storage power station full life cycle digital twin body construction method as defined in claim 1, wherein step S4 comprises the steps of: S41, reconstructing a three-dimensional geometric structure based on the full life cycle space-time fusion data to obtain three-dimensional geometric structure data; step S42, extracting texture and physical parameters of the full life cycle space-time fusion data to generate texture and physical parameter data; s43, performing attribute mapping and rendering on the three-dimensional geometric structure data according to the texture and physical parameter data to obtain a digital twin basic model; And S44, performing historical state backtracking and future trend prediction evolution simulation based on the digital twin basic model to generate a digital twin model of the pumped storage power station.
9. The pumped storage power station full life cycle digital twin body construction method as defined in claim 8, wherein step S44 comprises the steps of: Step S441, acquiring current operation state parameters of a digital twin basic model, mining historical operation rules of the whole life cycle space-time fusion data to obtain historical operation rule data; Step S442, performing time-series evolution trend prediction on the current running state parameters based on the historical running rule data to generate time-series evolution trend prediction data; s443, carrying out dynamic behavior driving updating on the digital twin basic model according to the time sequence evolution trend prediction data to obtain a dynamic behavior driving model; and step 444, taking the dynamic behavior driving model as a pumped storage power station digital twin body model to execute the visual monitoring and simulation analysis of the whole life cycle of the power station.
10. A pumped storage power station full life cycle digital twin body construction method as defined in claim 3, further comprising, after obtaining the secondary local fine grid index: Step S25, performing aggregation analysis and visual rendering on the multi-source perception basic data in the corresponding fine grid area based on the secondary local fine grid index to generate a local fine view; and S26, carrying out linkage mapping and comparison analysis according to the local refined view and the global view generated by the primary global grid index to realize anomaly detection and state evaluation of the power station structure.

Description

Method for constructing full life cycle digital twin body of pumped storage power station Technical Field The invention belongs to the technical field of life cycle construction, and relates to a method for constructing a full life cycle digital twin body of a pumped storage power station. Background The pumped storage power station is used as a key peak shaving, energy storage and standby power supply facility in a power system, has the characteristics of large engineering scale, complex structure, long life cycle and changeable running environment, and is a comprehensive engineering involving multiple professions and multiple stages. In the construction and operation process of the pumped storage power station, various sensing data such as geology, hydrology, structure, equipment and the like need to be integrated, the data are derived from different sensing technologies (such as remote sensing, sensors, monitoring systems and the like), the accuracy is obviously different, and once the digital model is distorted due to inaccurate data fusion, design defects, operation faults or safety accidents can be caused, so that the importance of the full life cycle digital twin body construction is highlighted. There are also some areas where optimization is needed for the current construction of digital twin bodies of pumped storage power stations, specifically in the following aspects: 1. The current digital twin body construction method does not perform unified data processing from the perspective of multi-level space coding, is easy to cause fusion inconsistency of sensing data with different accuracies, improves risks of model errors and decision errors, cannot improve accuracy of full life cycle state depiction, causes resource waste and low operation efficiency, and brings long-term hidden danger to power station safety and reliability. 2. At present, space-time integrated coding integration is not carried out on multisource perception data, data continuity and dynamic updating of a digital twin body in construction and operation stages cannot be guaranteed, model faults and prediction deviation caused by data splitting cannot be avoided, meanwhile, the possibility of economic cost increase caused by data precision difference cannot be effectively reduced, and accuracy and scientificity of full life cycle management cannot be guaranteed. Disclosure of Invention In view of the problems of the prior art, the invention provides a method for constructing a full life cycle digital twin body of a pumped storage power station, which is used for solving the technical problems. In order to achieve the above and other objects, the present invention adopts the following technical scheme: The invention provides a method for constructing a full life cycle digital twin body of a pumped storage power station, which comprises the following steps: Step S1, multi-source heterogeneous sensing data of a pumped storage power station in a construction period and an operation period are obtained, and standardized cleaning and format unified processing are carried out on the multi-source heterogeneous sensing data to generate multi-source sensing basic data; s2, constructing a two-stage space grid coding system comprising a first-stage global grid and a second-stage local fine grid based on multi-source perception basic data, and performing space discretization mapping to obtain space grid coding data; Step S3, establishing a full life cycle time axis code, and carrying out space-time dimension alignment association on the space grid code data and the full life cycle time axis code to obtain space-time dimension associated data; And S4, carrying out geometric modeling and physical attribute mapping of the digital twin body according to the full life cycle space-time fusion data to obtain a digital twin body basic model, carrying out full life cycle state evolution simulation based on the digital twin body basic model to generate a digital twin body model of the pumped storage power station, and sending the digital twin body model of the pumped storage power station to a management and control terminal. Step S1 comprises the steps of: s11, acquiring original sensing signals of a pumped storage power station in a construction period and an operation period by using a laser radar, oblique photography and an Internet of things sensor; step S12, denoising and outlier removing processing are carried out on the original perception signals, and original perception cleaning data are generated; S13, performing semantic annotation and coordinate system conversion on the original perception cleaning data to obtain semantic annotation perception data; and S14, carrying out data structure standardized recombination on the semantic annotation sensing data so as to generate multi-source sensing basic data. Step S2 comprises the steps of: S21, carrying out power station space topology structure analysis based on multisource perception basic data to