Network construction new energy cluster wide area state perception and data fusion method and system

Abstract

The invention discloses a networking new energy cluster wide area state perception and data fusion method and system, and the method comprises the steps: carrying out the multi-level data collection of all stations of a new energy cluster, and obtaining the multi-level multi-source heterogeneous data; performing multi-time-scale data preprocessing on the multi-level multi-source heterogeneous data to obtain fused data; constructing a multi-level adaptive state sensing model based on the fused data, and performing state sensing based on the multi-level adaptive state sensing model; and constructing a multi-domain associated data mapping mechanism based on the fused data and the state sensing result, so as to realize a closed loop from sensing to evaluation and from evaluation to optimization based on the multi-domain associated data mapping mechanism. According to the method, a technical system of'multi-level acquisition, multi-time-space fusion, adaptive perception and multi-domain mapping 'is constructed, an end-to-end closed loop of a new energy cluster from perception to optimization is realized, and deep collaboration and accurate interaction between new energy and a power grid are realized through technical innovation and fusion application.

Description

Technical Field

[0001] The present invention relates to the technical field of new energy power generation, and more specifically, to a method and system for wide-area state perception and data fusion of a network-forming new energy cluster. Background Art

[0002] In recent years, the global energy structure has been accelerating its transformation towards low-carbon and clean directions, and new energy power generation technologies such as wind energy and solar energy have been rapidly developed and widely applied. However, new energy power generation has characteristics such as intermittency, volatility, and uncertainty. After large-scale new energy is connected to the power grid, it has brought new challenges to the safe and stable operation of the power system. In order to address these challenges, it has become the industry's consensus and development direction to build a regional cluster composed of multiple new energy power stations and improve the new energy consumption capacity through collaborative optimization within the cluster.

[0003] However, the operation and management of the network-forming new energy cluster face the following technical problems:

[0004] (1) The cluster covers a wide area and involves multiple types of power stations such as wind farms and photovoltaic power stations. Each power station contains multiple types of power generation equipment such as wind turbines and inverters inside. The operating environment is complex and changeable, and the types and formats of the collected data vary greatly. There is a lack of a unified perception architecture and fusion method, making it difficult to form a global state awareness.

[0005] (2) There is a large amount of heterogeneous data inside the cluster, including multi-source data such as meteorology, geography, and electrical parameters. The data has a high dimension and strong correlation. Relying on traditional simple aggregation analysis is difficult to mine the intrinsic value of the data and difficult to support the refined management and optimal decision-making of the cluster.

[0006] (3) The operation of the cluster involves multiple levels from power generation equipment, power stations to the cluster. There is a lack of horizontal connection in state perception and operation optimization at each level, and the internal relationship between equipment status and cluster capabilities is not yet clear, resulting in insufficient overall effectiveness of the cluster and insufficient power grid support capabilities.

[0007] (4) In the past, the state perception and capacity assessment of new energy clusters mostly relied on static capacity indicators such as total installed capacity and total connected capacity, which are difficult to reflect the real-time dynamic characteristics of the cluster. The actual output, available capacity, regulation capacity, etc. of the cluster are dynamically affected by equipment status and resource conditions, and there are often large deviations between the assessment results and the actual operation performance, affecting the effect of power grid dispatching.

[0008] In response to the above problems, the existing technologies mainly adopt the following solutions:

[0009] The first is to establish a centralized data collection and monitoring system, connect the operation data of each station to the centralized control center, manually check various data indicators, and evaluate the operation status of the cluster. However, this method is difficult to handle massive heterogeneous data, relies on manual experience, lacks intelligent analysis capabilities, and the evaluation process is delayed and lacks timeliness.

[0010] The second is to establish data analysis models within a single wind farm or photovoltaic power station, such as wind power prediction models and photovoltaic power generation efficiency models, to achieve status assessment of a single scenario. However, this method is limited to the internal site, does not consider data association and status fusion at the cluster level, and lacks understanding of the overall characteristics of the cluster.

[0011] The third is to apply emerging information technologies such as big data analysis and artificial intelligence to improve the processing and analysis capabilities of renewable energy power generation data and tap into the value of data. However, most existing data applications remain at the shallow level, such as anomaly detection and trend prediction, and have not yet reached the deep level of correlation mining from data to status and then to capabilities.

[0012] In summary, it can be seen that the existing technology still has shortcomings in coping with the perception and fusion of multi-source heterogeneous data in network-building new energy clusters, multi-level state penetration assessment, and accurate characterization of dynamic capabilities. A new perception method is urgently needed to establish a three-dimensional state perception system covering the entire cluster, the entire process, and all data to support the intelligent management and control of new energy clusters and enhance the power grid's ability to absorb and regulate new energy.

[0013] Therefore, a method for wide-area state perception and data fusion of networked new energy clusters is needed. Summary of the invention

[0014] The present invention proposes a method and system for wide-area state perception and data fusion of a networked new energy cluster to solve the problem of how to establish a three-dimensional state perception system covering the entire cluster, the entire process, and all data.

[0015] In order to solve the above problem, according to one aspect of the present invention, a method for wide-area state perception and data fusion of a new energy cluster network is provided, the method comprising:

[0016] Carry out multi-level data collection for each station in the new energy cluster to obtain multi-level, multi-source, heterogeneous data;

[0017] Performing multi-time scale data preprocessing on the multi-level, multi-source, heterogeneous data to obtain fused data;

[0018] Building a multi-level adaptive state perception model based on the fused data, so as to perform state perception based on the multi-level adaptive state perception model;

[0019] Construct a multi-domain associated data mapping mechanism based on the fused data and the state perception results, so as to realize a closed loop from perception to evaluation and from evaluation to optimization based on the multi-domain associated data mapping mechanism.

[0020] Preferably, multi-level data collection is performed on each station in the new energy cluster to obtain multi-level multi-source heterogeneous data, including:

[0021] At the level of a single power generation equipment, collect the real-time working condition data of key equipment; wherein, the key equipment includes: wind turbines, photovoltaic modules and converters;

[0022] At the level of a single station, collect environmental data and power data reflecting the overall characteristics of the station; wherein, the environmental data includes: wind resource distribution data of a wind farm, solar radiation data of a photovoltaic power station; the power data includes: operation data of the reactive power compensation device of the station, power flow data of the booster station and in-site power grid topology data;

[0023] At the cluster level, collect spatial association data reflecting the geographical distribution and internal association of the cluster.

[0024] Preferably, multi-time scale data preprocessing is performed on the multi-level multi-source heterogeneous data to obtain fused data, including:

[0025] For equipment-side data, adopt a second-level processing granularity, and perform outlier removal, data smoothing and fast interpolation operations on the data;

[0026] For station-side data, adopt a minute-level processing granularity, and perform missing value filling, consistency verification and association fusion processing on the data;

[0027] For spatial association data, adopt an hour-level processing granularity, and perform topological association and geographical correction processing on the data to construct a spatial association model inside the cluster.

[0028] Preferably, construct a multi-level adaptive state perception model based on the fused data, so as to perform state perception based on the multi-level adaptive state perception model, including:

[0029] Construct an equipment-level adaptive state perception model, and output equipment state indicators based on the equipment-level adaptive state perception model; wherein, the equipment state indicators include: equipment health and remaining life;

[0030] Construct a station-level adaptive state perception model, and output station state indicators based on the station-level adaptive state perception model; wherein, the station state indicators include: available capacity of the station, power generation efficiency and power quality;

[0031] Build a cluster-level adaptive state awareness model, and output cluster state indicators based on the cluster-level adaptive state awareness model; wherein, the cluster state indicators include: real-time standby capacity of the cluster, ramp-up ability, and fault ride-through ability.

[0032] Preferably, building a multi-domain associated data mapping mechanism based on the fusion data and state awareness results includes:

[0033] Establish an association model between equipment state indicators and station yard capacity indicators to construct a mapping from equipment state to station yard capacity;

[0034] Establish a dynamic association model between station yard capacity and cluster capacity to construct a mapping from station yard capacity to cluster capacity;

[0035] Establish a dynamic adaptation model between cluster support capacity and power grid regulation requirements to construct a mapping from station yard capacity to cluster capacity.

[0036] Preferably, the method further includes:

[0037] Visualize the cluster wide-area awareness results; wherein, the cluster wide-area awareness results include: equipment state, station yard capacity, and cluster capacity.

[0038] According to another aspect of the present invention, there is provided a wide-area state awareness and data fusion system for a new energy cluster for power grid construction, and the system includes:

[0039] A data acquisition module, configured to perform multi-level data acquisition on each station yard of the new energy cluster to obtain multi-level multi-source heterogeneous data;

[0040] A data fusion module, configured to perform multi-time scale data preprocessing on the multi-level multi-source heterogeneous data to obtain fusion data;

[0041] A state awareness module, configured to build a multi-level adaptive state awareness model based on the fusion data, and perform state awareness based on the multi-level adaptive state awareness model;

[0042] A multi-domain associated data mapping module, configured to build a multi-domain associated data mapping mechanism based on the fusion data and state awareness results, and implement a closed loop from perception to evaluation and from evaluation to optimization based on the multi-domain associated data mapping mechanism.

[0043] Preferably, the data acquisition module performs multi-level data acquisition on each station yard of the new energy cluster to obtain multi-level multi-source heterogeneous data, including:

[0044] At the level of a single power generation equipment, collect real-time operating condition data of key equipment; wherein, the key equipment includes: wind turbines, photovoltaic modules, and inverters;

[0045] At the single station level, environmental data and power data reflecting the overall characteristics of the station are collected; among them, the environmental data includes: wind resource distribution data of a wind farm, and solar radiation data of a photovoltaic power station; the power data includes: operation data of the station's reactive power compensation device, power flow data of the booster station, and in-site power grid topology data;

[0046] At the cluster level, spatial association data reflecting the geographical distribution and internal association of the cluster is collected.

[0047] Preferably, the data fusion module performs data preprocessing on the multi-level and multi-source heterogeneous data at multiple time scales to obtain fusion data, including:

[0048] For equipment-side data, with a second-level processing granularity, outlier rejection, data smoothing, and fast interpolation operations are performed on the data;

[0049] For station-side data, with a minute-level processing granularity, missing value filling, consistency verification, and association fusion processing are performed on the data;

[0050] For spatial association data, with an hour-level processing granularity, topological association and geographical correction processing are performed on the data to construct a spatial association model within the cluster.

[0051] Preferably, the state perception module constructs a multi-level adaptive state perception model based on the fusion data to perform state perception based on the multi-level adaptive state perception model, including:

[0052] Construct an equipment-level adaptive state perception model, and output equipment state indicators based on the equipment-level adaptive state perception model; among them, the equipment state indicators include: equipment health and remaining life;

[0053] Construct a station-level adaptive state perception model, and output station state indicators based on the station-level adaptive state perception model; among them, the station state indicators include: available capacity of the station, power generation efficiency, and power quality;

[0054] Construct a cluster-level adaptive state perception model, and output cluster state indicators based on the cluster-level adaptive state perception model; among them, the cluster state indicators include: real-time reserve capacity of the cluster, ramping ability, and fault ride-through ability.

[0055] Preferably, the multi-domain associated data mapping module constructs a multi-domain associated data mapping mechanism based on the fusion data and the state perception result, including:

[0056] Establish an association model between equipment state indicators and station capacity indicators to construct a mapping from equipment state to station capacity;

[0057] Establish a dynamic association model between the substation capabilities and the cluster capabilities to construct a mapping from substation capabilities to cluster capabilities;

[0058] Establish a dynamic adaptation model between the cluster support capabilities and the power grid regulation requirements to construct a mapping from substation capabilities to cluster capabilities.

[0059] Preferably, the system further includes:

[0060] A visualization module for visualizing the results of wide-area cluster perception; wherein, the results of wide-area cluster perception include: equipment status, substation capabilities, and cluster capabilities.

[0061] The present invention provides a method and system for wide-area state perception and data fusion of a network-forming new energy cluster, including: performing multi-level data collection on each substation of the new energy cluster to obtain multi-level multi-source heterogeneous data; performing multi-time-scale data preprocessing on the multi-level multi-source heterogeneous data to obtain fused data; constructing a multi-level adaptive state perception model based on the fused data to perform state perception based on the multi-level adaptive state perception model; constructing a multi-domain associated data mapping mechanism based on the fused data and the state perception results to achieve a closed-loop from perception to evaluation and from evaluation to optimization based on the multi-domain associated data mapping mechanism. The present invention realizes an end-to-end closed-loop from perception to optimization of the new energy cluster by constructing a technical system of "multi-level collection, multi-time-space fusion, adaptive perception, and multi-domain mapping". Through technological innovation and integrated application, it breaks through the perception blind spots, evaluation bottlenecks, and regulation barriers of the new energy cluster, and realizes the deep coordination and precise interaction between the new energy and the power grid. This set of technical systems can significantly improve the new energy penetration rate, optimize the power grid operation efficiency, promote the digital transformation of the energy and power system, and has important value and broad application prospects for accelerating the construction of a clean, low-carbon, safe, and efficient modern energy system and achieving the goals of carbon peak and carbon neutrality. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] The exemplary embodiments of the present invention can be more fully understood by referring to the following drawings:

[0063] Figure 1 FIG. 100 is a flowchart of a method for wide-area state perception and data fusion of a network-forming new energy cluster according to an embodiment of the present invention;

[0064] Figure 2 FIG. 101 is an overall architecture diagram of a method for wide-area state perception and data fusion of a network-forming new energy cluster according to an embodiment of the present invention;

[0065] Figure 3 FIG. 102 is a flowchart of three-level adaptive state perception modeling of equipment-substation-cluster according to an embodiment of the present invention;

[0066] Figure 4 A schematic diagram of a multi-domain association mapping mechanism of equipment status, station capability and cluster support capability according to an embodiment of the present invention;

[0067] Figure 5 Schematic diagram of the structure of a wide-area state perception and data fusion system 500 for networking new energy clusters according to an embodiment of the present invention. DETAILED DESCRIPTION

[0068] Now, exemplary embodiments of the present invention are described with reference to the accompanying drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. These embodiments are provided to disclose the present invention in detail and completely and to fully convey the scope of the present invention to those skilled in the art. The terms used in the exemplary embodiments shown in the accompanying drawings are not intended to limit the present invention. In the accompanying drawings, the same units / elements are marked with the same reference numerals.

[0069] Unless otherwise specified, the terms (including technical terms) used herein have the commonly understood meanings to those skilled in the art. In addition, it is understood that the terms defined in commonly used dictionaries should be understood to have the same meanings as those in the context of the relevant fields, and should not be understood as idealized or overly formal meanings.

[0070] The purpose of the present invention is to address the deficiencies of the prior art and propose a method for wide-area state perception and data fusion of a networked new energy cluster, in order to achieve the following technical effects: (1) Design a unified multi-source data acquisition and fusion framework for multiple heterogeneous sites such as wind and solar power to eliminate source data barriers and lay the foundation for data fusion; (2) Construct a vertically connected three-layer state perception model of equipment-site-cluster to explore the state association between layers and achieve accurate characterization of the global state; (3) Establish a horizontally associated feature-state-capability mapping mechanism to reveal the inherent laws between data, state, and capability, and achieve data-driven capability evaluation; (4) Integrate static capacity data and dynamic operation data, propose multi-time scale power grid support capability indicators for actual regulation, and improve the credibility and effectiveness of the evaluation results.

[0071] The core of this invention is to propose a multi-level heterogeneous data fusion, state perception and correlation analysis method, build an end-to-end mapping of "data-state-capability", realize panoramic and three-dimensional state perception of networked new energy clusters, and provide key technical support for the intelligent regulation and high-quality consumption of new energy.

[0072] Figure 1 FIG. 1 is a flow chart of a method 100 for wide-area state perception and data fusion of a new energy cluster network according to an embodiment of the present invention. Figure 1As shown in the figure, the wide-area state perception and data fusion method for a new energy cluster that constructs a power grid provided by the embodiment of the present invention realizes an end-to-end closed-loop from perception to optimization of the new energy cluster by constructing a technical system of "multi-level acquisition, multi-time and space fusion, adaptive perception, and multi-domain mapping". Through technological innovation and integrated applications, it breaks through the perception blind spots, evaluation bottlenecks, and regulation barriers of the new energy cluster, and realizes the deep coordination and precise interaction between the new energy and the power grid. This technical system can significantly improve the new energy penetration rate, optimize the power grid operation efficiency, promote the digital transformation of the energy and power system, and has important value and broad application prospects for accelerating the construction of a clean, low-carbon, safe, and efficient modern energy system and achieving the goals of carbon peak and carbon neutrality. The wide-area state perception and data fusion method 100 for a new energy cluster that constructs a power grid provided by the embodiment of the present invention starts from step 101. In step 101, multi-level data acquisition is performed on each station of the new energy cluster to obtain multi-level multi-source heterogeneous data.

[0073] Preferably, the multi-level data acquisition of each station of the new energy cluster to obtain multi-level multi-source heterogeneous data includes:

[0074] At the level of a single power generation equipment, real-time working condition data of key equipment is collected; wherein, the key equipment includes: wind turbines, photovoltaic modules, and inverters;

[0075] At the level of a single station, environmental data and power data reflecting the overall characteristics of the station are collected; wherein, the environmental data includes: wind resource distribution data of a wind farm, solar radiation data of a photovoltaic power station; the power data includes: operation data of the reactive power compensation device of the station, power flow data of the booster station, and in-site power grid topology data;

[0076] At the cluster level, spatial association data reflecting the geographical distribution and internal association of the cluster is collected.

[0077] Figure 2 The overall architecture of the present invention is shown, which is composed of modules such as data acquisition, multi-time and space scale fusion calculation, adaptive state perception, multi-domain association mapping, and visualization of perception results. Through the collaborative work of these modules, the full-process intelligence of the new energy cluster that constructs a power grid from perception, fusion to mapping is realized. The perception results can be further applied to the digital twin system to support equipment operation and maintenance decision-making; they can also be applied to the energy management system to optimize the power grid dispatching operation.

[0078] Combined with Figure 2 As shown in the figure, in the present invention, multi-level multi-source heterogeneous data acquisition is first performed. That is, multi-type and multi-scene data acquisition devices are deployed on each station of the new energy cluster and its key power generation equipment to realize all-round state data perception at the power generation equipment end and the station end. Specifically, it includes:

[0079] 1. At the level of a single power generation equipment, real-time operating condition data of key equipment such as wind turbines, photovoltaic modules, and converters are collected. The wind turbine collects multi-dimensional data such as wind speed, wind direction, blade inclination, nacelle vibration, and generator current; the photovoltaic module collects multi-parameter data such as light intensity, photovoltaic panel temperature, and string voltage and current; the converter collects multi-type data such as active power, reactive power, harmonic content, and insulation impedance. Through the multi-source heterogeneous data collection of a single piece of equipment, the real-time operating state of the equipment is comprehensively characterized.

[0080] 2. At the level of a single site, environmental and power data reflecting the overall characteristics of the site are collected. The environmental data includes wind resource distribution data of a wind farm, solar radiation data of a photovoltaic power station, etc., which are collected by means of distributed weather stations, satellite images, etc.; the power data includes the operating data of the reactive power compensation device of the site, the power flow data of the booster station, the topology data of the internal power grid of the site, etc., which are collected by means of intelligent substations, distribution automation terminals, etc. Through the associated collection of multi-source data at the site level, a macroscopic perception data set at the site level is formed.

[0081] 3. At the cluster level, spatial data reflecting the geographical distribution and internal association of the cluster are collected. The GIS system is used to collect geographical information data such as the longitude, latitude coordinates and altitude of each site in the cluster, and the power grid dispatching system is used to collect associated data such as the upper and lower layer power grid parameters of the injection point of the cluster and the power flow of the transmission channel. Through the extended collection of associated data at the cluster level, a spatial perception database covering the entire cluster is constructed.

[0082] In step 102, multi-time scale data preprocessing is performed on the multi-level multi-source heterogeneous data to obtain fused data.

[0083] Preferably, performing multi-time scale data preprocessing on the multi-level multi-source heterogeneous data to obtain fused data includes:

[0084] For equipment-side data, a processing granularity of seconds is adopted, and outlier removal, data smoothing and fast interpolation operations are performed on the data;

[0085] For site-side data, a processing granularity of minutes is adopted, and missing value filling, consistency verification and associated fusion processing are performed on the data;

[0086] For spatially associated data, a processing granularity of hours is adopted, and topological association and geographical correction processing are performed on the data to construct a spatial association model within the cluster.

[0087] In the present invention, after data collection is completed, data preprocessing is performed on a multi-time scale. A multi-time scale data preprocessing process is designed according to the different characteristics of the collected multi-source heterogeneous data to improve data quality and strengthen the spatio-temporal correlation of the data. Specifically, it includes:

[0088] 1. For equipment - end data such as fans and photovoltaic modules, with a processing granularity of seconds, perform operations such as outlier removal, data smoothing, and fast interpolation on the data to ensure the integrity and continuity of the data and meet the requirements of millisecond - level control.

[0089] 2. For site - end data such as wind farms and photovoltaic power stations, with a processing granularity of minutes, perform processing such as missing value filling, consistency verification, and correlation fusion on the data to improve the spatial correlation of site data and support minute - scale site power prediction and optimization.

[0090] 3. For spatially correlated data such as cluster coordinates and channel parameters, with a processing granularity of hours, perform processing such as topological correlation and geographical correction on the data to construct a spatial correlation model within the cluster and support the operation optimization of the cluster at the hour - scale and even longer periods.

[0091] In step 103, construct a multi - level adaptive state perception model based on the fusion data, and perform state perception based on the multi - level adaptive state perception model.

[0092] Preferably, constructing a multi - level adaptive state perception model based on the fusion data and performing state perception based on the multi - level adaptive state perception model includes:

[0093] Construct an equipment - level adaptive state perception model, and output equipment state indicators based on the equipment - level adaptive state perception model; wherein, the equipment state indicators include: equipment health and remaining life;

[0094] Construct a site - level adaptive state perception model, and output site state indicators based on the site - level adaptive state perception model; wherein, the site state indicators include: site available capacity, power generation efficiency, and power quality;

[0095] Construct a cluster - level adaptive state perception model, and output cluster state indicators based on the cluster - level adaptive state perception model; wherein, the cluster state indicators include: cluster real - time reserve capacity, ramp - up ability, and fault - ride - through ability.

[0096] Combined Figure 2 As shown, in the present invention, perform adaptive multi - level state perception. For different decision - making requirements, construct an adaptive multi - level state perception model to achieve three - dimensional state evaluation from local to global and from equipment to cluster.

[0097] Figure 3It shows the process of the self - adaptive state perception modeling of the present invention at three levels: the equipment level, the station level, and the cluster level. At the equipment level, through transfer learning and fine - tuning optimization, a state evaluation model for a single device is constructed; at the station level, the internal data of the station is fused and external environmental factors are embedded to construct a station - level state evaluation model; at the cluster level, the spatial topology data of the cluster is associated, and a graph neural network is used to realize the collaborative modeling of the states of the stations within the cluster. The models at the three levels support each other and enhance progressively, ultimately realizing the global three - dimensional state perception.

[0098] Specifically, the process of performing self - adaptive multi - level state perception includes:

[0099] 1. Equipment - level self - adaptive state perception. According to the characteristics of a single power generation equipment, an adaptive equipment state evaluation model is constructed. A transfer learning mechanism is introduced, and the operation big data of similar equipment is used to train a general state evaluation model, and this model is fine - tuned with personalized data to form a single - machine state perception model with self - adaptive ability. The model takes the multi - source heterogeneous data of the equipment as input, outputs state indicators such as equipment health and remaining life, and realizes functions such as fault warning and state prediction.

[0100] 2. Station - level self - adaptive state perception. According to the characteristics of a single new - energy power station, an adaptive station state evaluation model is constructed. Using the multi - machine collaborative data within the station, through spatio - temporal correlation analysis, the power and voltage distribution within the station are characterized, and the power generation performance of the station is evaluated. At the same time, the external meteorological and geographical data of the station are embedded, and an adaptive mechanism is introduced to enable the model to dynamically respond to external environmental changes. The model takes the multi - source heterogeneous data of the station as input, and outputs state indicators such as the available capacity, power generation efficiency, and power quality of the station.

[0101] 3. Cluster - level self - adaptive state perception. According to the wide - area distribution characteristics of the cluster, an adaptive cluster state evaluation model is constructed. Using the geographical proximity relationship between the stations within the cluster, a state association model between the stations is constructed through algorithms such as graph neural networks to realize state transfer learning from local to global. Further, external spatial association data such as the power grid are introduced, and an integrated state perception model of the cluster and the external power grid is constructed through technologies such as knowledge graphs. The model takes multi - level data inside and outside the cluster as input, and outputs a series of support - ability indicators such as the real - time reserve capacity, ramping ability, and fault - ride - through ability of the cluster.

[0102] In step 104, a multi - domain associated data mapping mechanism is constructed based on the fusion data and the state perception result, so as to realize a closed - loop from perception to evaluation and from evaluation to optimization based on the multi - domain associated data mapping mechanism.

[0103] Preferably, constructing the multi - domain associated data mapping mechanism based on the fusion data and the state perception result includes:

[0104] Establish an association model between equipment status indicators and station yard capacity indicators to construct a mapping from equipment status to station yard capacity;

[0105] Establish a dynamic association model between station yard capacity and cluster capacity to construct a mapping from station yard capacity to cluster capacity;

[0106] Establish a dynamic adaptation model between cluster support capacity and power grid regulation requirements to construct a mapping from station yard capacity to cluster capacity.

[0107] Combine Figure 2 As shown, in the present invention, multi-domain associated data mapping is performed. Taking equipment status and cluster capacity as the starting points, a multi-domain data association mapping mechanism is constructed to realize a closed loop from perception to evaluation and from evaluation to optimization.

[0108] Figure 4 The present invention demonstrates the construction of a mapping mechanism among equipment status, station yard capacity, and cluster support capacity by using multi-domain association analysis. First, methods such as correlation analysis and causal reasoning are adopted to construct a mapping model between equipment characteristics and status attributes to realize local status perception; then, technologies such as causal reasoning and knowledge graph are used to construct an association model between station yard status attributes and cluster support capacity indicators to realize global capacity evaluation; finally, the evaluated cluster capacity is matched and optimized with the power grid regulation requirements to form a perception-driven intelligent regulation closed loop.

[0109] In the present invention, multi-domain associated data mapping is performed, specifically including:

[0110] 1. Mapping from equipment status to station yard capacity. Through methods such as correlation analysis and causal reasoning, an association model between equipment status indicators and station yard capacity indicators is established. Analyze the influence laws of status quantities such as equipment failure frequency and loss rate on capacity indicators such as station yard power generation efficiency and power quality, and form a quantitative relationship from local equipment status to overall station yard capacity.

[0111] 2. Mapping from station yard capacity to cluster capacity. Through methods such as network analysis and multi-agent learning, a dynamic association model between station yard capacity and cluster capacity is established. Consider the influence of the spatial distribution and geographical proximity relationship of new energy stations on the comprehensive cluster capacity, analyze characteristics such as output complementarity and frequency response synergy among stations, and form an emergence law from individual station yard capacity to group cluster capacity.

[0112] 3. Mapping from cluster capacity to power grid regulation. Through technologies such as knowledge mining and reinforcement learning, a dynamic adaptation model between cluster support capacity and power grid regulation requirements is established. Combining external knowledge such as power grid load characteristics and unit combination methods, analyze the optimal power grid dispatching strategies under different cluster support capacity states, and realize a closed loop from cluster status perception to power grid intelligent regulation.

[0113] Preferably, the method further comprises:

[0114] Visualize the cluster wide-area perception results; the cluster wide-area perception results include: equipment status, site capabilities and cluster capabilities.

[0115] Combination Figure 2 As shown, in the present invention, cluster wide-area perception results can also be visualized. A visualization system for cluster wide-area perception is designed to achieve intuitive presentation of perception results and human-computer interaction. Specifically, it includes:

[0116] 1. Visualization of equipment status. Based on the three-dimensional equipment model, the real-time working condition data of the equipment is integrated, and VR / AR and other technologies are used to realize the visualization and virtual-real integration of the equipment's operating status.

[0117] 2. Visualization of station capacity. Based on the three-dimensional layout of the plant area, combined with the geographic information system, the real-time status data of the station is connected to realize the visualization of the power generation capacity and output distribution within the station.

[0118] 3. Visualization of cluster capabilities. Based on the cluster geographic wiring diagram, large-screen splicing, data visualization and other technologies are used to integrate related data inside and outside the cluster to achieve a panoramic and three-dimensional display of the cluster's support capabilities.

[0119] Through the above technical solution, the present invention realizes the full process intelligence of building a new energy cluster from perception, fusion to mapping, forming a technical closed loop of "transparent perception, adaptive evaluation, correlation analysis, and closed-loop optimization". Compared with the existing technology, the present invention has the following advantages:

[0120] (1) From the perspective of perception, a multi-level data collection solution is proposed to connect the heterogeneous data in the entire chain from equipment to clusters, and realize complete perception of global information.

[0121] (2) From the perspective of fusion, a multi-scale temporal and spatial data processing mechanism is constructed to improve the quality and relevance of multi-source heterogeneous data, laying the foundation for subsequent precise modeling.

[0122] (3) From the mapping dimension, innovative methods such as multi-level adaptive evaluation and multi-domain correlation analysis are used to achieve three-dimensional status perception from local to global, across levels and across fields.

[0123] (4) From the application dimension, construct new cluster capability indicators for power grid dispatching, provide precise input for real-time optimization and regulation of new energy, and promote the digital transformation of the energy system.

[0124] In summary, the present invention provides an effective technical means for the wide-area perception and global optimization of new energy clusters, which is of great significance for promoting the large-scale and friendly access of new energy and enhancing the power grid's ability to absorb new energy, and is an important technical support for achieving the goals of carbon peak and carbon neutrality.

[0125] The objectives of the present invention are as follows:

[0126] 1. Construct a multi-level and heterogeneous data perception and fusion system. In view of the characteristics of the new energy cluster with grid-forming involving multiple types of stations such as wind farms and photovoltaic power stations, and each station covering multiple types of equipment such as wind turbines and inverters, a hierarchical data acquisition and fusion architecture from power generation equipment, stations to the cluster is designed. At the equipment layer, the operating condition data of a single device is collected to perceive the health status of the device; at the station layer, the status data of each equipment within the station is aggregated to evaluate the power generation performance of the station; at the cluster layer, the characteristic data of each station is fused to analyze the complementary characteristics and operating capabilities of the cluster. By constructing a multi-level data perception system of "equipment - station - cluster", the heterogeneity problems brought about by the diverse new energy types and wide cluster scope are eliminated, and the collaborative perception and fusion calculation of global information are realized.

[0127] 2. Achieve a vertical through-assessment from local status to global capabilities. In view of the characteristics of the multi-level operation of the new energy cluster, a vertical assessment mechanism from local to global and from status to capabilities is established. Starting from the perception of equipment status, through data mining at the equipment layer, the status attributes such as the health level and failure probability of a single device are evaluated; at the station level, the status attributes of each equipment within the station are fused to form macroscopic attributes such as the power generation capacity and available capacity of the station; at the cluster level, the attributes of each station are superimposed to extract the support capacity indicators of the cluster in aspects such as power grid frequency regulation and active and reactive power balance. Through the vertical correlation analysis and fusion calculation of multi-level status-capabilities, it reveals how the microscopic changes in equipment status are transmitted and affect the macroscopic performance of the cluster's capabilities, and realizes the vertical coordination and status through of production factors within the cluster.

[0128] 3. Mine cross-domain data correlations and establish a horizontal mapping between status and capabilities. In view of the characteristics of the new energy cluster operation involving data in multiple fields such as meteorology, geography, and electricity, cross-domain data correlation analysis is carried out to reveal the internal connections between data, status, and capabilities. Through feature engineering methods, key features reflecting equipment status are extracted from multi-source heterogeneous data, and a feature-status mapping model is constructed to realize the correlation perception from data to status; through knowledge mining techniques, the correlation rules between equipment status attributes and cluster support capabilities are analyzed, and a status-capabilities mapping model is constructed to realize the correlation assessment from status to capabilities. Through the multi-domain collaborative analysis of data, status, and capabilities, using data as a link, an end-to-end horizontal mapping of equipment operation characteristics, health status, and cluster power grid support capabilities is established, breaking through the limitations of previous single data analysis and providing a new data-driven idea for cluster capability assessment.

[0129] 4. Propose new capability indicators for actual grid control. In view of the problem that most of the previous new energy cluster capability assessments were based on static capacity indicators and were difficult to support real-time dispatching and control, the present invention proposes a new cluster capability indicator that integrates static and dynamic characteristics. First, medium- and long-term support capability indicators based on capacity data are proposed, such as the cluster's grid-connected capacity, cluster power generation utilization hours, etc., to characterize the support capability of new energy clusters for grid scale expansion in the planning stage; second, short-term dynamic capability indicators based on operating data are proposed, such as cluster real-time standby, cluster climbing rate, cluster unit start-up and shutdown times, etc., to characterize the support capability of new energy clusters for grid frequency regulation and even transient stability in the real-time operation stage. By constructing a three-dimensional cluster capability indicator with full-time scale characteristics, a basis is provided for power grid dispatching organizations to make scientific decisions on the allocation of new energy output.

[0130] In summary, the core purpose of the present invention is to build a technical system of "data collection and fusion, state perception and association, capability assessment and optimization", to open up the technical chain from local to global, from perception to decision-making, and from offline to online, to achieve holographic intelligent perception of the networked new energy cluster with "one map, one model, and one set of indicators", and to achieve "integrated, refined, and real-time" intelligent control of cluster power generation based on perception. The present invention is of great significance for improving the penetration rate of new energy, promoting the digital and intelligent development of the energy and power field, and supporting the implementation of the carbon peak and carbon neutrality strategy.

[0131] The wide-area state perception and data fusion method for networking new energy clusters proposed in this invention realizes an end-to-end closed loop from perception to optimization of new energy clusters by building a technical system of "multi-level collection, multi-time and space fusion, adaptive perception, and multi-domain mapping". Compared with the existing technology, it has the following significant advantages and innovative effects:

[0132] 1. Increase the scale and proportion of new energy access. By sensing multi-source heterogeneous data on the equipment side, adaptively evaluating the real-time operating conditions of single-machine equipment, effectively guiding intelligent operation and maintenance, reducing equipment failure rates, and improving power generation efficiency, the power generation at the level of a single new energy station can be increased by more than 1.5%. By integrating the spatial correlation data of multiple stations within the cluster, optimizing the output coordination between stations, smoothing the fluctuations in the total output of the cluster, and increasing the proportion of new energy consumption at the cluster level by more than 3 percentage points. By associating the cluster support capabilities with the grid control needs, accurately formulate new energy access plans, and increase the proportion of new energy at the grid level by more than 5%.

[0133] 2. Reduce the negative impacts of new energy grid connection. By adaptively integrating the data of the substation environment, dynamically evaluating the support capabilities such as reactive power compensation and active power frequency modulation of the substation, and optimizing the voltage and current distribution inside the substation, the qualified rate of the voltage at the grid connection point of the substation is increased to over 99%. Through multi-level state mapping, quantify the capacity indicators such as cluster reserve, inertia, and ramp rate, guide the optimal dispatching of the power grid, smooth the impacts of the intermittency and volatility of new energy, and make the qualified rate of the frequency at the grid connection point reach over 99.9%. Through the method of expanding transfer learning, predict extreme working conditions such as cluster fault ride-through and island operation, and formulate emergency load shedding strategies for new energy, so that the stability accidents at the grid connection point are reduced by over 80%.

[0134] 3. Promote the multi-energy interaction of power sources, grids, loads, and energy storage. Through the correlation analysis of heterogeneous data, reveal the internal mapping between the output characteristics of new energy and the load electricity consumption law, realize the precise matching of new energy and load demand, and improve the operation efficiency of the power grid by more than 2 percentage points. Through hierarchical perception and collaborative optimization, guide the complementary operation of new energy, renewable energy, and conventional energy, build an integrated operation control mode of power sources, grids, loads, and energy storage, and increase the consumption ratio of renewable energy by more than 10% in typical scenarios. Through the quantification of capacity indicators and demand mapping, support new energy to participate in the electricity spot market, develop the energy and power derivatives market, and cultivate the endogenous motivation for new energy to participate in power grid regulation.

[0135] 4. Promote the structural reform of the new energy supply side. Through the data perception across substations and across energies, depict the dynamic characteristics of different types of new energy, support the precise investment decision-making for the multi-energy complementarity of wind, light, water, and fire, and optimize the spatial distribution of new energy. Through the evaluation of the capabilities of new energy facing the power grid, guide the transformation of new energy construction from "scale and speed type" to "quality and benefit type", and promote the transformation of the new energy development mode from extensive to intensive. Through integrating government affairs and economic data, judge the social and economic effects of new energy development, provide a basis for implementing differentiated new energy planning policies in various regions, and promote the high-quality development of the new energy industry.

[0136] 5. Consolidate the digital foundation of energy and power modernization. By building a heterogeneous data perception network covering the whole process, all equipment, and all scenarios, create an industrial big data resource library in the field of new energy, and lay a data foundation for the digital twin of the energy and power system. Through the research and development of hierarchical intelligent algorithms from equipment to system, realize the distributed intelligence from local edge to cloud center, and form an intelligent technology middle platform for the energy system. By quantifying the mapping of capacity status and regulation requirements, establish a digital closed-loop for energy production, market trading, and policies and regulations, and create a new situation for the digital development of energy and power modernization.

[0137] In summary, through technological innovation and integrated applications, the present invention has achieved deep coordination and precise interaction between new energy and the power grid at the three levels of perception, evaluation, and regulation, providing important fundamental support for building a new power system with new energy as the main body. On the one hand, the present invention enhances the power grid's ability to absorb new energy, reduces the negative impacts of new energy grid connection on power grid stability and power quality, and breaks through the technical bottlenecks restricting the large-scale development of new energy, which is of great significance for promoting the low-carbon transformation of energy and achieving the "dual-carbon" goal. On the other hand, through global perception and multi-domain coordination, the present invention optimizes the matching degree of each link in the "source-network-load-storage" system, improves the overall efficiency of the power grid, and plays a leading role in promoting the construction of the ubiquitous power Internet of Things and realizing energy and power modernization. The popularization and application of the present invention will accelerate the digital, networked, and intelligent development process in the energy and power field, providing solid support for building a clean, low-carbon, safe, and efficient energy system.

[0138] The following describes the best embodiments of the present invention in detail in combination with specific application scenarios.

[0139] Embodiment: Wide-area state perception and intelligent regulation of a new energy cluster in a certain region

[0140] Background: A large-scale new energy cluster has been planned and constructed in a certain regional power grid, including 10 million-kilowatt-level wind power bases and 5 million-kilowatt-level photovoltaic power stations. To achieve the efficient absorption and optimized operation of the new energy cluster, it is necessary to build a wide-area state perception system covering the entire process of wind and solar power generation, multiple power stations, and multiple devices, and make intelligent regulation decisions based on the perception results.

[0141] Implementation steps:

[0142] 1. Multi-level heterogeneous data collection

[0143] At the level of individual power generation devices such as wind turbines and photovoltaic inverters, through Internet of Things gateways, intelligent sensors, etc., real-time operating condition data of the devices are collected, including multi-dimensional time-series data such as wind speed, light intensity, temperature, current, voltage, and power, and the data granularity reaches the second level.

[0144] At the level of each wind farm, photovoltaic power station and other power stations, through meteorological stations, energy management systems, etc., power station environmental data and power operation data are collected, including wind resource distribution, irradiance distribution, power station power flow, power quality and other data, and the data granularity reaches the minute level.

[0145] At the level of the new energy cluster, through geographic information systems, power grid dispatching control systems, etc., cluster spatial topology data and power grid operation data are collected, including power station geographical coordinates, upstream and downstream power flow of the cluster, load curves and other data, and the data granularity reaches the hour level.

[0146] 2. Multi-temporal and multi-scale data fusion

[0147] For the second-level data of devices such as wind turbines and inverters collected, data cleaning, feature extraction and other means are adopted to achieve data noise reduction and key information condensation, and online processing is carried out through real-time stream computing to meet the millisecond-level control requirements.

[0148] For the minute-level data of wind farms and photovoltaic power stations collected, methods such as multivariate time series anomaly detection and missing value filling are adopted to improve data quality, and through spatio-temporal correlation analysis, consistency verification and fusion calculation of multi-device data within the station are realized.

[0149] For the hourly-level data of new energy clusters collected, technologies such as geospatial interpolation and topological association mapping are adopted to achieve complementary fusion of data from different stations at the geographical and electrical levels, and a spatio-temporal integrated data environment within the cluster is constructed.

[0150] 3. Adaptive multi-level state modeling

[0151] At the device level of wind turbines, inverters and other devices, a device health state evaluation model is constructed based on the long short-term memory neural network (LSTM). The general model is trained using the large historical operation data of the device, and the model parameters are fine-tuned using the data of new devices to form a state perception model with the ability of transfer learning between devices, and indicators such as device health and failure probability are evaluated in real time.

[0152] At the station level of wind farms, photovoltaic power stations and other stations, a station power generation capacity evaluation model is constructed based on the graph attention network (GAT). On the one hand, the status quantities of each device within the station are fused, and the topological association between devices is considered to evaluate the overall power generation efficiency of the station; on the other hand, the external environmental meteorological data of the station is fused, and the dynamic impact of environmental factors is considered to evaluate the fluctuation range of the power generation capacity of the station, providing a basis for the power prediction of the station.

[0153] At the new energy cluster level, a cluster power grid support capacity evaluation model is constructed based on the spatio-temporal convolutional network (STCNet). Considering factors such as the power generation capacity of each station within the cluster, the load demand of the upstream and downstream of the cluster, and the grid frequency stability margin, grid support indicators such as the reserve capacity, ramping rate, and frequency modulation ability of the new energy cluster are evaluated, and the support ability of the new energy cluster for the grid security and stability is characterized.

[0154] 4. Multi-domain knowledge association mapping

[0155] By constructing an association rule library for the device status and the station KPIs, the influence law of device failures and abnormal operating conditions on the station production operation KPIs is mined to form a device-station status quantitative association model.

[0156] By constructing an association knowledge graph of the operation of the power station and cluster indicators, characterizing the influence mechanism of different power stations, different capacities, and different types of renewable energy on the power grid frequency, voltage, and power balance, and forming a quantitative association model of power station-cluster capabilities.

[0157] By constructing an association optimization algorithm between cluster capabilities and power grid dispatching, solving the optimal power grid dispatching strategy under the multi-time scale support capabilities of the new energy cluster, guiding the deep peak shaving of conventional power sources, maximizing the consumption of new energy, and forming a matching model of capabilities-regulation requirements.

[0158] 5. Visualization of multi-scenario perception results

[0159] Build a cluster digital twin visualization system, conduct three-dimensional modeling of the meteorological environment, geographical distribution, and electrical topology of the new energy cluster, and map the perceived device status data, power station operation data, cluster capability indicators, etc. to the three-dimensional scene in real time. Through technologies such as AR / VR, achieve rapid positioning of abnormal devices and immersive deduction of fault scenarios.

[0160] Build a large screen for the situation awareness of the new energy cluster, visually display the health status of equipment, power generation efficiency of power stations, and cluster support capabilities in the form of curves, bar charts, radar charts, etc., supplemented by algorithm model traceability explanations to enhance the interpretability of perception results; for the evaluated abnormal scenarios and extreme working conditions, timely push warning information to facilitate rapid response and handling by operation and maintenance personnel.

[0161] Build a "source-network-load-storage" collaborative optimization cockpit, integrate multi-source data such as conventional power sources, new energy, load, and energy storage, comprehensively display the power grid macro situation KPIs, and connect the perceived new energy cluster support capabilities to the power grid AGC system to provide a dispatching basis for new energy to participate in power grid frequency modulation control and provide spinning reserve, realizing the friendly consumption of new energy.

[0162] Application effect: By adopting the method of the present invention, the health status and power generation capabilities of each wind farm and photovoltaic power station in the new energy cluster in this area are accurately evaluated, effectively guiding field-level optimizations such as equipment maintenance and power prediction control. The average power generation efficiency is increased by 2.5% and the failure rate is decreased by 15%; the power grid support capabilities of the entire new energy cluster are dynamically evaluated at the minute level, and the accuracy rates of key indicators such as reserve capacity and ramp rate reach more than 90%, providing a reliable basis for new energy to participate in power grid frequency modulation control. The new energy consumption ratio is increased by 8 percentage points, promoting the coordinated complementarity between new energy and conventional energy; the multi-dimensional visualization of perception results for auxiliary decision-making shortens the identification time of new energy operation anomalies by 80% and saves more than 60% of the manual analysis time.

[0163] This embodiment fully demonstrates the remarkable application effects of the present invention in aspects such as improving the scale of new energy access, ensuring the stable operation of the power grid, and optimizing the new energy consumption level. Through innovative technical means such as collaborative perception of multi-level heterogeneous data, fusion calculation of multi-temporal and multi-spatial scales, adaptive hierarchical modeling, and multi-domain data association mapping, the present invention has better solved the technical challenges in the wide-area state perception and data fusion of new energy clusters, realized the complete acquisition of global information, the efficient processing of data, the accurate assessment of states, and the three-dimensional characterization of capabilities, providing an important theoretical basis and key technical support for the intelligent control of network-forming new energy clusters. The popularization and application of the present invention are of great significance for improving the new energy consumption level, supporting the clean and low-carbon transformation of energy, and ensuring the safe and economic operation of the power grid.

[0164] Figure 5 FIG. 500 is a schematic structural diagram of a wide-area state perception and data fusion system for a network-forming new energy cluster according to an embodiment of the present invention. As Figure 5 shown, the wide-area state perception and data fusion system 500 for a network-forming new energy cluster provided by the embodiment of the present invention includes: a data acquisition module 501, a data fusion module 502, a state perception module 503, and a multi-domain associated data mapping module 504.

[0165] Preferably, the data acquisition module 501 is configured to perform multi-level data acquisition on each station of the new energy cluster to obtain multi-level multi-source heterogeneous data.

[0166] Preferably, the data acquisition module 501 performs multi-level data acquisition on each station of the new energy cluster to obtain multi-level multi-source heterogeneous data, including:

[0167] At the level of a single power generation equipment, real-time working condition data of key equipment is collected; wherein, the key equipment includes: wind turbines, photovoltaic modules, and inverters;

[0168] At the level of a single station, environmental data and power data reflecting the overall characteristics of the station are collected; wherein, the environmental data includes: wind resource distribution data of a wind farm, solar radiation data of a photovoltaic power station; the power data includes: operation data of a reactive power compensation device of the station, power flow data of a booster station, and in-site power grid topology data;

[0169] At the cluster level, spatial association data reflecting the geographical distribution and internal association of the cluster is collected.

[0170] Preferably, the data fusion module 502 is configured to perform multi-time scale data preprocessing on the multi-level multi-source heterogeneous data to obtain fusion data.

[0171] Preferably, the data fusion module 502 performs multi-time scale data preprocessing on the multi-level and multi-source heterogeneous data to obtain fusion data, including:

[0172] For equipment-side data, a second-level processing granularity is adopted to perform outlier removal, data smoothing, and fast interpolation operations on the data;

[0173] For station-side data, a minute-level processing granularity is adopted to perform missing value filling, consistency verification, and association fusion processing on the data;

[0174] For spatially associated data, an hour-level processing granularity is adopted to perform topological association and geocorrection processing on the data to construct a spatial association model within the cluster.

[0175] Preferably, the state perception module 503 is used to construct a multi-level adaptive state perception model based on the fusion data to perform state perception based on the multi-level adaptive state perception model.

[0176] Preferably, the state perception module 503 constructs a multi-level adaptive state perception model based on the fusion data to perform state perception based on the multi-level adaptive state perception model, including:

[0177] Construct an equipment-level adaptive state perception model, and output equipment state indicators based on the equipment-level adaptive state perception model; wherein, the equipment state indicators include: equipment health and remaining life;

[0178] Construct a station-level adaptive state perception model, and output station state indicators based on the station-level adaptive state perception model; wherein, the station state indicators include: station available capacity, power generation efficiency, and power quality;

[0179] Construct a cluster-level adaptive state perception model, and output cluster state indicators based on the cluster-level adaptive state perception model; wherein, the cluster state indicators include: cluster real-time reserve capacity, ramping ability, and fault ride-through ability.

[0180] Preferably, the multi-domain associated data mapping module 504 is used to construct a multi-domain associated data mapping mechanism based on the fusion data and the state perception result to achieve a closed loop from perception to evaluation and from evaluation to optimization based on the multi-domain associated data mapping mechanism.

[0181] Preferably, the multi-domain associated data mapping module 504 constructs a multi-domain associated data mapping mechanism based on the fusion data and the state perception result, including:

[0182] Establish an association model between equipment state indicators and station capacity indicators to construct a mapping from equipment state to station capacity;

[0183] Establish a dynamic association model between the substation capacity and the cluster capacity to construct the mapping from the substation capacity to the cluster capacity;

[0184] Establish a dynamic adaptation model between the cluster support capacity and the power grid regulation demand to construct the mapping from the substation capacity to the cluster capacity.

[0185] Preferably, the system further includes:

[0186] A visualization module for visualizing the results of the cluster wide-area perception; wherein, the results of the cluster wide-area perception include: equipment status, substation capacity, and cluster capacity.

[0187] The wide-area state perception and data fusion system 500 of the network-forming new energy cluster according to the embodiment of the present invention corresponds to the wide-area state perception and data fusion method 100 of the network-forming new energy cluster according to another embodiment of the present invention, which will not be elaborated here.

[0188] The present invention has been described by referring to a few embodiments. However, as is well known to those skilled in the art, other embodiments equivalent to those disclosed above of the present invention equally fall within the scope of the present invention.

[0189] Generally, all terms used in the present invention are interpreted according to their ordinary meanings in the technical field, unless otherwise explicitly defined therein. All references to "a / the [device, component, etc.]" are open to interpretation as at least one instance of the device, component, etc., unless otherwise explicitly stated. The steps of any method disclosed herein do not necessarily have to be run in the exact order disclosed, unless explicitly stated.

[0190] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0191] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processors of general-purpose computers, special-purpose computers, embedded processors, or other programmable data processing devices to generate a machine, such that the instructions executed by the processors of the computer or other programmable data processing devices produce means for implementing the functions specified in one or more flows and / or blocks Figure 1 in one or more flows and / or blocks Figure 1 or in one or more blocks.

[0192] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including instruction means that implement the functions specified in one or more flows and / or blocks Figure 1 in one or more flows and / or blocks Figure 1 or in one or more blocks.

[0193] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operational steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more flows and / or blocks Figure 1 in one or more flows and / or blocks Figure 1 or in one or more blocks.

[0194] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications or equivalent replacements can still be made to the specific embodiments of the present invention, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered by the protection scope of the present invention.

Claims

1. A wide-area state perception and data fusion method for a new energy cluster forming a power grid, characterized in that, The method includes: Performing multi-level data collection on each power station in the new energy cluster to obtain multi-level multi-source heterogeneous data; Performing multi-time scale data preprocessing on the multi-level multi-source heterogeneous data to obtain fused data; Constructing a multi-level adaptive state perception model based on the fused data to perform state perception based on the multi-level adaptive state perception model; Constructing a multi-domain associated data mapping mechanism based on the fused data and the state perception result to achieve a closed loop from perception to evaluation and from evaluation to optimization based on the multi-domain associated data mapping mechanism.

2. The method according to claim 1, characterized in that, Performing multi-level data collection on each power station in the new energy cluster to obtain multi-level multi-source heterogeneous data, including: At the level of a single power generation equipment, collecting real-time operating condition data of key equipment; wherein, the key equipment includes: wind turbines, photovoltaic modules, and inverters; At the level of a single power station, collecting environmental data and power data reflecting the overall characteristics of the power station; wherein, the environmental data includes: wind resource distribution data of a wind farm, solar radiation data of a photovoltaic power station; the power data includes: operating data of the reactive power compensation device of the power station, power flow data of the booster station, and in-site power grid topology data; At the cluster level, collecting spatial association data reflecting the geographical distribution and internal association of the cluster.

3. The method according to claim 1, characterized in that, Performing multi-time scale data preprocessing on the multi-level multi-source heterogeneous data to obtain fused data, including: For equipment-side data, adopting a processing granularity of seconds, performing outlier removal, data smoothing, and fast interpolation operations on the data; For power station-side data, adopting a processing granularity of minutes, performing missing value filling, consistency verification, and association fusion processing on the data; For spatial association data, adopting a processing granularity of hours, performing topological association and geographical correction processing on the data to construct a spatial association model within the cluster.

4. The method according to claim 1, wherein Constructing a multi-level adaptive state perception model based on the fused data to perform state perception based on the multi-level adaptive state perception model, including: Constructing an equipment-level adaptive state perception model and outputting equipment state indicators based on the equipment-level adaptive state perception model; wherein, the equipment state indicators include: equipment health and remaining life; Constructing a power station-level adaptive state perception model and outputting power station state indicators based on the power station-level adaptive state perception model; wherein, the power station state indicators include: available capacity of the power station, power generation efficiency, and power quality; Constructing a cluster-level adaptive state perception model and outputting cluster state indicators based on the cluster-level adaptive state perception model; wherein, the cluster state indicators include: real-time reserve capacity of the cluster, ramp-up capacity, and fault ride-through capacity.

5. The method according to claim 1, characterized in that Constructing a multi-domain associated data mapping mechanism based on the fused data and the state perception result, including: Establishing an association model between equipment state indicators and power station capacity indicators to construct a mapping from equipment state to power station capacity; Establishing a dynamic association model between power station capacity and cluster capacity to construct a mapping from power station capacity to cluster capacity; Establishing a dynamic adaptation model between cluster support capacity and grid regulation requirements to construct a mapping from power station capacity to cluster capacity.

6. The method according to claim 1, characterized in that, The method further includes: Visualize the wide-area perception results of the cluster; among them, the wide-area perception results of the cluster include: equipment status, station capabilities, and cluster capabilities.

7. A wide-area state perception and data fusion system for a new energy cluster forming a power grid, characterized in that, The system includes: A data acquisition module for performing multi-level data acquisition on each station of the new energy cluster to obtain multi-level multi-source heterogeneous data; A data fusion module for performing multi-time scale data preprocessing on the multi-level multi-source heterogeneous data to obtain fusion data; A status perception module for constructing a multi-level adaptive status perception model based on the fusion data and performing status perception based on the multi-level adaptive status perception model; A multi-domain associated data mapping module for constructing a multi-domain associated data mapping mechanism based on the fusion data and status perception results, and realizing a closed loop from perception to evaluation and from evaluation to optimization based on the multi-domain associated data mapping mechanism.

8. The system according to claim 7, wherein The data acquisition module performs multi-level data acquisition on each station of the new energy cluster to obtain multi-level multi-source heterogeneous data, including: At the level of a single power generation equipment, collect real-time operating condition data of key equipment; among them, the key equipment includes: wind turbines, photovoltaic modules, and inverters; At the level of a single station, collect environmental data and power data reflecting the overall characteristics of the station; among them, the environmental data includes: wind resource distribution data of the wind farm, solar radiation data of the photovoltaic power station; the power data includes: operating data of the station's reactive power compensation device, power flow data of the booster station, and in-site power grid topology data; At the cluster level, collect spatial association data reflecting the geographical distribution and internal association of the cluster.

9. The system according to claim 7, wherein The data fusion module performs multi-time scale data preprocessing on the multi-level multi-source heterogeneous data to obtain fusion data, including: For equipment-side data, use a second-level processing granularity to perform outlier removal, data smoothing, and fast interpolation operations on the data; For station-side data, use a minute-level processing granularity to perform missing value filling, consistency verification, and association fusion processing on the data; For spatial association data, use an hourly processing granularity to perform topological association and geographical correction processing on the data to construct a spatial association model within the cluster.

10. The system according to claim 7, wherein The status perception module constructs a multi-level adaptive status perception model based on the fusion data and performs status perception based on the multi-level adaptive status perception model, including: Construct an equipment-level adaptive status perception model and output equipment status indicators based on the equipment-level adaptive status perception model; among them, the equipment status indicators include: equipment health and remaining life; Construct a station-level adaptive status perception model and output station status indicators based on the station-level adaptive status perception model; among them, the station status indicators include: station available capacity, power generation efficiency, and power quality; Construct a cluster-level adaptive status perception model and output cluster status indicators based on the cluster-level adaptive status perception model; among them, the cluster status indicators include: cluster real-time reserve capacity, ramp-up capacity, and fault ride-through capacity.

11. The system according to claim 7, wherein The multi-domain associated data mapping module constructs a multi-domain associated data mapping mechanism based on the fusion data and status perception results, including: Establish an association model between equipment status indicators and station capabilities indicators to construct the mapping from equipment status to station capabilities; Establish a dynamic association model between station capabilities and cluster capabilities to construct the mapping from station capabilities to cluster capabilities; Establish a dynamic adaptation model between cluster support capabilities and grid regulation requirements to construct the mapping from station capabilities to cluster capabilities.

12. The system according to claim 7, wherein The system further includes: A visualization module for visualizing the results of cluster wide-area perception; among them, the results of cluster wide-area perception include: equipment status, station capabilities, and cluster capabilities.

Images