Power grid equipment BIM model parameter optimization system, electronic equipment and storage medium

Abstract

The invention relates to a power grid equipment BIM model parameter optimization system, electronic equipment and a storage medium, belongs to the technical field of data processing, and solves the problem of long iteration period caused by dependence on manual analysis in existing BIM model optimization. Comprising a component data processing module used for collecting multi-source data of each component and extracting a multi-modal feature vector of each component at an analysis moment according to a neural network model of each modal; the dynamic risk prediction module is used for acquiring the multi-modal feature vector, output by the component data processing module, of each component at the analysis moment, and transmitting the multi-modal feature vector into a risk prediction model to obtain a risk prediction result; the model parameter optimization module is used for identifying the to-be-optimized component and the priority thereof according to the risk prediction result output by the dynamic risk prediction module; and based on the parameter adjustment rule base, obtaining a BIM model parameter adjustment rule of the to-be-optimized component, and forming an optimization scheme according to the priority. And optimization of the dynamic risk-driven BIM model parameters is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and in particular to a power grid equipment BIM model parameter optimization system, an electronic device and a storage medium. BACKGROUND

[0002] In the digital operation of power grid engineering, the system based on building information model (BIM) has become the core carrier, and the model data covers geometry, electricity, operation and maintenance, and environment and other multi-source heterogeneous information. However, the existing BIM platform and related system have obvious limitations in data processing and decision support.

[0003] At the data fusion level, the existing system lacks effective multi-source heterogeneous data correlation analysis capability. The traditional system usually uses rule engines or simple statistical analysis modules to process data, and these modules cannot deeply mine the complex and nonlinear implicit correlations between the BIM geometric structure, the electrical parameter time series curve, the unstructured operation text and the environmental monitoring data. For example, the system cannot automatically establish a deep relationship model between the "regional continuous heavy rain" data and the "foundation settlement risk" in the BIM model of the power transmission tower, resulting in that the data value is buried, forming a "data island", and the overall perception ability of the system is limited.

[0004] At the risk warning level, the built-in warning function of the existing system is mostly based on a static threshold judgment module. Such a module usually presets fixed rules (such as "temperature exceeding limit means alarm"), and cannot reflect the dynamic evolution trend of the risk, so the warning is seriously lagging behind. At the same time, the system lacks core components for time series modeling and analysis of historical and real-time data, and cannot capture the time series dependency between data, resulting in low prediction accuracy and insufficient foresight of the system.

[0005] At the model optimization level, the model updating and maintenance function of the existing BIM system is highly dependent on manual intervention. When receiving the feedback that "some components frequently malfunction and need to be adjusted in design", the engineer must manually find and modify the related component parameters in the BIM platform. The whole system lacks a core algorithm that can automatically mine the model optimization direction from multi-source data, resulting in a long iteration cycle of the BIM model, and the inability to quickly adapt to the dynamic changes of the actual running state, so that the effectiveness of the digital twin model is greatly reduced. SUMMARY

[0006] In view of the above analysis, the embodiments of the present application aim to provide a power grid equipment BIM model parameter optimization system, an electronic device and a storage medium, to solve the problem of long iteration cycle caused by the dependence of the existing BIM model optimization on manual analysis.

[0007] In one aspect, the embodiments of the present application provide a power grid equipment BIM model parameter optimization system, comprising:

[0008] The component data processing module is used to collect multi-source data of each component and extract the multi-modal feature vectors of each component at the analysis time according to the neural network model of each modality. The dynamic risk prediction module is used to obtain the multimodal feature vectors of each component at the analysis time output by the component data processing module, and input them into the risk prediction model to obtain the risk prediction results. The risk prediction results include: component risk probability, component risk level, component failure type, equipment risk index, and equipment risk type. The model parameter optimization module is used to identify components to be optimized and their priorities based on the risk prediction results output by the dynamic risk prediction module; obtain the BIM model parameter adjustment rules for the components to be optimized based on the parameter adjustment rule library, and form an optimization scheme according to the priority of the components to be optimized; identifying the components to be optimized includes: calculating the basic risk value of the component based on the component risk probability and component risk level by weighted summation; calculating the dynamic risk threshold of the component based on the equipment risk index, high risk threshold, and basic risk threshold; selecting components whose basic risk value is greater than their dynamic risk threshold as the first component; selecting neighboring components whose component risk probability is greater than the neighbor screening threshold and do not belong to the first component from the neighboring components of each first component as the second component; the neighbor screening threshold is less than the basic risk threshold; and merging the first and second components as the component to be optimized.

[0009] Based on further improvements to the above system, the parameter adjustment rule base is obtained through the following steps: Natural language processing technology is used to parse power grid engineering design standards, technical specifications, and expert experience documents to construct a knowledge graph; Based on historical operation and maintenance records and knowledge graphs, the mapping relationship between fault types and BIM model parameters is extracted, and a fault-parameter association table is established. Using the fault-parameter association table and knowledge graph as the knowledge context, the large language model is guided to generate preliminary adjustment rules through prompt word templates. After being reviewed and corrected by domain experts, these rules are stored in the parameter adjustment rule library.

[0010] Based on further improvements to the above system, the basic risk value of a component is calculated using the following formula: , in, Representation of components The basic risk value, and Representation of components The component risk probability and component risk level, and These represent the weights of the component risk probability and the component risk level, respectively. This represents the normalization function.

[0011] Based on the further improvement of the above system, the dynamic risk threshold is calculated by the following formula: , Wherein, represents the dynamic risk threshold of the component ; and respectively represent the preset high risk threshold and the basic risk threshold, ; represents the weight coefficient; represents the device risk index of the component and the device to which the component belongs.

[0012] Based on the further improvement of the above system, the priority of the first component in the component to be optimized is obtained by calculating the product of the basic risk value of the first component and the topological centrality weight of the first component; the priority of the second component is obtained by calculating the product of the basic risk value of the second component, the edge weight between the second component and the first component associated with it, and the topological centrality weight of the second component.

[0013] Based on the further improvement of the above system, the topological centrality weight is calculated based on at least one of the degree centrality, the betweenness centrality and the eigenvector centrality of the corresponding component.

[0014] Based on the further improvement of the above system, the optimization scheme is formed according to the priority of the component to be optimized, comprising: for the first component in the component to be optimized, arranging it in descending order of priority to form a main sequence; for each first component in the main sequence, arranging the second components associated with it in descending order of priority to form a subsequence of each first component; According to the order of the main sequence and each subsequence, the corresponding parameter adjustment rule is output in sequence to obtain the optimization scheme.

[0015] Based on the further improvement of the above system, the parameter adjustment rule satisfies three constraints at the same time: adjustment range constraint, coordination constraint and amplitude constraint; wherein, the adjustment range constraint is used to ensure that the adjusted parameter value is within the upper and lower limits specified by the standard; the coordination constraint is used to ensure that the parameter adjustment between the associated components is matched; the amplitude constraint is used to limit the single parameter adjustment amplitude not to exceed the preset safety threshold.

[0016] Based on the further improvement of the above system, the multi-source data of each component includes: BIM geometric data, topological relationship data, electrical parameter data, fault data and environmental parameter data; the multi-modal feature vector of each component at the analysis time is extracted according to the neural network model of each modality, including: For each component, based on the BIM geometric data and topological relationship data, the topological features of the component are extracted using a graph attention network; based on the analysis time and the fault data before the analysis time, the fault text features of the component at the analysis time are extracted using a text embedding model; based on the electrical parameter data, the electrical parameter features of the component at the analysis time are extracted using an electrical time series feature extraction model; based on the environmental parameter data, the environmental parameter features at the analysis time are extracted using an environmental time series feature extraction model, and the components of the power grid equipment in the same region share the same environmental parameter features. The topological features, fault text features, electrical parameter features and environmental parameter features of each component at the analysis time constitute the multi-modal feature vector of each component at the analysis time.

[0017] Based on the further improvement of the above system, the topological features of the component are extracted using a graph attention network, including: The graph structure data is constructed with the components as nodes, the initial features of the BIM geometric data of the components as node features, and the relationships between the components as edges; the relationships include electrical connection relationship, mechanical linkage relationship and spatial proximity relationship; The graph structure data is input into a multi-layer multi-channel graph attention network to output node embedding representation, thereby obtaining the topological features of each component; each channel corresponds to the edges of a kind of relationship, and is used to independently calculate the attention score and normalize the edges of the kind of relationship.

[0018] Based on the further improvement of the above system, after the electrical parameter data and the environmental data are time-aligned according to a unified time axis, the electrical time series data and the environmental time series data of three time scales are extracted from the data that have been time-aligned before the analysis time, wherein different time scales are resampled at different sampling frequencies.

[0019] Based on the further improvement of the above system, the model structures of the electrical time series feature extraction model and the environmental time series feature extraction model are the same, and each includes in turn: a parameter-aware time series encoding module, a gated fusion module and a dynamic attention pooling module; wherein the parameter-aware time series encoding module is used to encode the time series data of each parameter of each time scale to obtain the feature representation of each time scale; the gated fusion module is used to fuse the feature representations of the same parameter at different time scales to obtain the fusion feature representation of the parameter; the dynamic attention pooling module is used to aggregate the fusion feature representations of multiple parameters of the same component to obtain the electrical parameter features or the environmental parameter features corresponding to the component.

[0020] Based on the further improvement of the above system, the parameter-aware time series encoding module includes in turn: A parameter type encoding unit is used to generate an embedding vector for each parameter type; An input mapping unit is configured to map the concatenated parameter-type embedding vectors and the current time position encoding of each time step in different time scales to an input sequence through a linear layer. A plurality of improved LSTM units are configured to encode the input sequence of one time scale respectively, wherein the input gate, the forget gate and the output gate of each improved LSTM unit are fused with the parameter-type embedding vectors, and the attention weights of the hidden states of all time steps are calculated through an attention mechanism to obtain a feature representation of the current time scale.

[0021] Based on the further improvement of the above system, the risk prediction model comprises a multi-modal correlation model and a hierarchical prediction model in sequence; the multi-modal correlation model is configured to output a global context representation, a correlation analysis result and a multi-modal fusion representation of each component according to the multi-modal feature vector of the component at the analysis moment, and transmit the multi-modal fusion representation to the hierarchical prediction model; the hierarchical prediction model is configured to output a risk prediction result of each component and the device to which the component belongs in a future period.

[0022] Based on the further improvement of the above system, the multi-modal correlation model performs the following steps: The multi-modal feature vectors of the components at the analysis moment are projected to a unified dimension and an input sequence is constructed, and the projected feature vectors of each modality in the input sequence are stacked with respective mixed position encodings through element-wise addition; The input sequence with the mixed position encodings is input into a Transformer encoding layer, and cross-correlation between modalities is performed according to a unified multi-head self-attention layer and a hierarchical attention mask matrix, to output a multi-modal fusion representation of each component in each sample and a global context representation; The correlation analysis result is extracted from the attention weights of the Transformer encoding layer.

[0023] Based on the further improvement of the above system, the positions that are mutually attentive in the hierarchical attention mask matrix are set to 0, and the positions that are prohibited from attention are set to negative infinity, wherein the positions that are mutually attentive include: The position of the global token is mutually attentive with the positions of all other vectors in the input sequence; The positions of the feature vectors of different modalities within the same component are mutually attentive; The positions of the feature vectors of the same modality of different components are mutually attentive only when the corresponding components are neighbors in the topological relationship.

[0024] Based on the further improvement of the above system, the correlation analysis result includes: inter-modal correlation matrix and cross-component modal correlation matrix; the inter-modal correlation matrix is obtained by averaging the attention weights between different modal position sets in all Transformer layers and attention heads; the cross-component modal correlation matrix is obtained by averaging the attention weights of the same modal between neighbor components in the topological relationship in all Transformer layers and attention heads.

[0025] Based on the further improvement of the above system, the hierarchical prediction model successively includes: a component-level risk prediction module, a device-level risk prediction module and a prediction enhancement module; the component-level risk prediction module is used for outputting component risk probability, component risk level and component failure type by adopting a multilayer perceptron according to the multi-modal fusion representation of each component; the device-level risk prediction module is used for outputting device risk index and device risk type according to the global context representation and the multi-modal fusion representation of each component after being aggregated according to devices respectively; and the prediction enhancement module is used for enhancing the component risk probability and the device risk index according to the correlation analysis result, and outputting the final component risk probability and device risk index.

[0026] In another aspect, the embodiment of the present application provides an electronic device, which comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor executes the functions of the power grid equipment BIM model parameter optimization system of any one of the above aspects when executing the computer program.

[0027] The embodiment of the present application also provides a computer readable storage medium, which stores a computer program, and the functions of the power grid equipment BIM model parameter optimization system of any one of the above aspects are realized when the computer program is executed by the processor.

[0028] Compared with the prior art, the present application can at least realize one of the following beneficial effects: 1. From data preprocessing, correlation mining, risk prediction to the formation of a parameter optimization scheme, a complete closed-loop system is realized, effectively solving the problems of data island, early warning lag and long BIM model parameter optimization iteration period in the prior art; realizing autonomous, closed-loop and continuous iteration to promote parameter optimization according to the actual running state and risk prediction result, quickly adapting to the changing operation requirements of the power grid, reducing the dependence on manual analysis, reducing human errors, and having great value for guaranteeing the safe, stable and efficient operation of the power grid.

[0029] 2. Through the unified standardized processing of the component data processing module, the system can map heterogeneous multi-modal data such as BIM geometric data, topological relationship data, electrical parameter data, fault text data, and environmental parameter data to a unified semantic space; instead of treating various types of data in isolation, the system actively discovers and quantifies the complex, nonlinear implicit correlations between them, greatly exploiting the value of multi-source data and providing a comprehensive data view for accurate decision-making.

[0030] 3. The dynamic risk prediction module uses the fused deep features to capture the time evolution law through an improved, parameter-aware LSTM, and combines multi-task learning and context enhancement to perform hierarchical (component level / device level) and multi-dimensional (probability / level / type) risk prediction, making the prediction more consistent with physical reality, with high accuracy and low false alarm rate; it changes the traditional fixed threshold-based lag alarm mode, enabling the operation and maintenance to change from passive response to active intervention, greatly reducing the probability of unplanned shutdown and accident risk.

[0031] In the present application, the above technical solutions can be combined with each other to achieve more preferred combination solutions. Other features and advantages of the present application will be described in the subsequent specification, and some advantages will become apparent from the specification or will be understood by implementing the present application. The purpose and other advantages of the present application can be achieved and obtained from the contents specifically pointed out in the specification and the drawings. BRIEF DESCRIPTION OF DRAWINGS

[0032] The accompanying drawings are included to provide a further understanding of the present application and are incorporated in and constitute a part of this application, illustrate embodiments of the present application and serve to explain the principles of the present application, but are not intended to limit the scope of the present application. It should be understood that the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application. Figure 1 The structure schematic diagram of the power grid equipment BIM model parameter optimization system in embodiment 1 of the present application. DETAILED DESCRIPTION

[0033] The preferred embodiments of the present application will be described in detail below with reference to the accompanying drawings, which form a part of this application, and are used to illustrate the principles of the embodiments of the present application, but are not intended to limit the scope of the present application. It should be understood that the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0034] The skilled in the art should know that the following specific embodiments or specific modes of implementation are a series of optimized settings listed by the present application to further explain the specific inventive content, and these settings can be used in combination or in association with each other, unless the present application explicitly indicates that certain or a specific embodiment or mode of implementation cannot be associated with other embodiments or modes of implementation or cannot be used together. At the same time, the following specific embodiments or modes of implementation are only the most optimized settings, and not as an understanding of the scope of protection of the present application.

[0035] Embodiment 1 One specific embodiment of the present application discloses a power grid equipment BIM model parameter optimization system, which generates optimization decision of BIM model parameter through deep data fusion and intelligent algorithm with dynamic risk perception driving of components in power grid equipment.

[0036] As shown in Figure 1 The system of the present embodiment includes three major functional modules: component data processing module 100, dynamic risk prediction module 200 and model parameter optimization module 300. These modules are connected in turn, iteratively executed, and form a closed-loop system from data perception to decision execution.

[0037] The component data processing module 100 is used to collect multi-source data of each component, and extract multi-modal feature vectors of each component at the analysis time according to the neural network model of each modality; The dynamic risk prediction module 200 is used to obtain the multi-modal feature vectors of each component at the analysis time output by the component data processing module, and input into the risk prediction model to obtain the risk prediction result; The model parameter optimization module 300 is used to identify the components to be optimized and their priorities according to the risk prediction result output by the dynamic risk prediction module; based on the parameter adjustment rule library, obtain the BIM model parameter adjustment rule of the component to be optimized, and form an optimization scheme according to the priority of the component to be optimized.

[0038] Specifically, the component data processing module 100 is the basis of the system, and its goal is to build a unified and standardized "component-level multi-modal feature map". The component data processing module specifically implements steps S101 and S102.

[0039] S101, collect and pre-process multi-source data of each component: It should be noted that the power grid equipment is divided into 5 categories according to application categories: transmission lines, substations (converter stations), communication equipment, monitoring devices and public facilities. Each power grid equipment contains many components, and according to actual needs, the components to be optimized are selected, and multi-source data are collected, including but not limited to: BIM geometric data, topological relationship data, electrical parameter data, fault data and environmental parameter data.

[0040] Wherein, the BIM geometry data and the topological relationship data are obtained by analyzing the BIM model file to obtain the geometry data of each component in the power grid equipment and the relationship between the components; the electrical parameter data and the environmental parameter data are obtained by numerically completing and time sequence aligning the collected electrical parameters of each component and the meteorological data of the region where the power grid equipment is located based on a unified time axis; the fault data are obtained by analyzing the operation and maintenance records to obtain the fault data of each component.

[0041] ① Collection and preprocessing of BIM geometry data.

[0042] It should be noted that the BIM geometry data is derived from the BIM model file, which is parsed by calling the interface provided by the BIM platform to extract the BIM geometry data of each component, including: device type, component type, coordinate, size and component attribute.

[0043] Wherein, the device type and the component type are the classification of the device and the component in the actual application scenario, such as the device type is transformer, and the component type is radiator; the coordinate is used to represent the relative position of the component in the device; the size is the length, width and height of the bounding box of each component; the component attribute includes basic geometric attribute and optimizable parameter attribute.

[0044] Exemplarily, for the transformer radiator, its "radiator area" and "radiator fin density" are the key optimizable parameters affecting the heat dissipation efficiency; for the high-voltage bushing, "insulation material" and "dry arc distance" are the core attributes determining its insulation performance. By targeted extraction and normalization of these key attributes, on the one hand, it is convenient to distinguish components and reflect the unique characteristics of each component, and on the other hand, it is convenient to compare the changes before and after adjusting the parameter attribute value.

[0045] In order to convert the above BIM geometry data into model processable numerical features, the following preprocessing is needed: Category variables such as device type and component type are one-hot encoded to convert them into binary vectors; The coordinates of all components are scaled to the [0, 1] range based on the unified BIM model coordinate system using the min-max normalization method to eliminate the influence of the absolute coordinate value dimension; The size of each component is standardized to eliminate the influence of the dimension; The numerical value of the component attribute is normalized.

[0046] Finally, the processed device type, component type, coordinate, size and component attribute features are spliced to form the geometry feature sequence of each component. This series of processing effectively solves the problem of mixed qualitative and quantitative data and different dimensions in BIM geometry data.

[0047] ②Collection and preprocessing of topological relationship data.

[0048] It should be noted that the relationship information between components is also included in the BIM model file. The analysis of the BIM model file obtains the relationship between the core components, including: electrical connection relationship, mechanical linkage relationship and spatial proximity relationship.

[0049] The collected relationship between components is represented in the form of an adjacency table or an adjacency matrix. If there is a relationship between components, it is set to 1, otherwise it is set to 0. The component association relationship implicitly contained in the BIM model is made explicit and structured, providing key information for understanding the propagation path of risks within the equipment.

[0050] ③Collection and preprocessing of electrical parameter data.

[0051] It should be noted that a component may contain multiple electrical parameters. Electrical parameter data is usually obtained from SCADA systems or other monitoring files, and the format is usually CSV. In this embodiment, Pandas library is used to parse CSV data to obtain the electrical parameter time series data of each component, such as sleeve current, radiator inlet and outlet oil temperature, etc.

[0052] For cases where only device-level data can be obtained (such as total current of transformer), the data is copied to all components as shared electrical parameter data.

[0053] Since electrical parameter data is time series data, after collecting the electrical parameters at a specific analysis time, historical data before the analysis time is still needed to analyze the trend of the data. Therefore, the data at the historical time is also stored for analysis.

[0054] The preprocessing of electrical parameter data includes data cleaning, time series alignment and completion, and standardization processing.

[0055] Among them, data cleaning is to eliminate obvious outliers, such as illegal data with current greater than 500A.

[0056] Preferably, an outlier detection method based on dynamic threshold is adopted. For each parameter of each component, the moving average and standard deviation of its historical data are calculated. Data points that are obviously outside the range (such as ±5 times the standard deviation) are considered outliers and are removed. This method can adapt to the normal fluctuation range of different devices and different parameters, and the cleaning is more accurate.

[0057] Time series alignment and completion is to unify data with different sampling frequencies to the same time granularity (such as 5 minutes). For missing values, linear interpolation method is used for filling in to ensure the continuity of time series.

[0058] The standardization processing is a Z-Score standardization processing on continuous electrical parameters.

[0059] The above preprocessing effectively solves the common noise, missing and dimension problems in real-time data, and lays a solid foundation for accurate time series pattern mining.

[0060] (4) Collection and preprocessing of fault data.

[0061] The fault data is derived from the equipment fault records and inspection reports in recent years, and the format is unstructured Word or PDF document.

[0062] In this embodiment, the text is extracted from the document, and the NLP tool (such as jieba segmentation) and pre-trained model of Python are used for automatic analysis and coding of the text. The preprocessing includes: text cleaning, entity recognition and text vectorization.

[0063] Among them, the text cleaning includes: removing special characters, stop words, and unifying professional terms to reduce text noise.

[0064] Entity recognition is to extract equipment and / or component names, fault types, fault descriptions, severity, and other key information from the text, thereby converting unstructured text into structured knowledge.

[0065] Preferably, a two-level entity linking strategy is adopted during entity recognition: the first level uses a predefined equipment and component dictionary for accurate matching. The second level, for fault descriptions that cannot be accurately matched, uses a semantic similarity model (such as Sentence-BERT) to calculate the similarity between the fault description and all component function descriptions, and associates the fault description to the component with the highest similarity.

[0066] In addition, the fault time is extracted from the text based on the date and time regularization rules, and is bound with the fault record, which is convenient for associating with electrical parameter data and environmental parameter data according to time, so as to accurately label the risk label of the corresponding time.

[0067] It should be noted that if the extracted fault data is specific to a component, a higher association confidence is set; if the extracted fault data is at the device level and cannot be clearly associated with a specific component, the fault data is associated with all core components of the device, and a lower association confidence is set. Through this intelligent association strategy, both the full use of information and the accuracy of association are taken into account.

[0068] Finally, the preprocessed fault data includes: fault time, equipment name and / or component name, fault type, fault description, severity, and association confidence.

[0069] (5) Collection and preprocessing of environmental parameter data.

[0070] The environmental parameter data is derived from meteorological data of the area where the power grid equipment is located, in JSON format. By calling the meteorological data interface, the time series data of the environmental temperature, environmental humidity, and rainfall accumulation is obtained.

[0071] During the preprocessing of the environmental parameter data, the same time axis as the electrical parameter data is adopted, and linear interpolation or spline interpolation is used to fill in the data and align the time series.

[0072] Based on the above preprocessing of multi-source data, the multi-source data of each component at the analysis time is obtained, including BIM geometric data and topological relationship data, different types or quantities of electrical parameter data for each component, different fault data, and the same environmental parameter data for components in the same area.

[0073] S102, respectively according to the neural network model of each modality, the multi-modal feature vector of each component at the analysis time is extracted.

[0074] For each component, the following processing steps are used to obtain the topological feature, fault text feature, electrical parameter feature, and environmental parameter feature at the analysis time, which constitutes the multi-modal feature vector of each component at the analysis time.

[0075] ①Based on the BIM geometric data and topological relationship data, the topological feature of the component is extracted using the graph attention network.

[0076] Specifically, the component is taken as the node, the BIM geometric data of the component is taken as the initial feature of the node, and the relationship between the components is taken as the edge to construct the graph structure data; the relationship includes the electrical connection relationship, the mechanical linkage relationship, and the spatial proximity relationship.

[0077] Preferably, the edge weight is set according to the relationship type of the edge, such as the weight of the electrical connection relationship is 1.0, the weight of the mechanical linkage relationship is 0.8, and the weight of the spatial proximity relationship is 0.5.

[0078] It should be noted that the graph attention network used in this step is a multi-layer multi-channel graph attention network, which adopts a separate pre-training strategy, learns the topological structure representation of the power grid component through a graph reconstruction task, and takes the difference between the reconstructed adjacency matrix of the node embedding and the original adjacency matrix as the loss function. After pre-training, the risk prediction model in the dynamic risk prediction module 200 of the present embodiment is fine-tuned end-to-end.

[0079] In implementation, the graph structure data is input into the trained multi-layer multi-channel graph attention network, and the node embedding representation is output to obtain the topological feature of each component; each channel corresponds to an edge of a certain relationship, which is used to independently calculate the attention score and normalize the edge of this relationship. This multi-channel attention mechanism enables the model to learn the differences in the influence of different types of neighbor relationships on the center node.

[0080] Preferably, a two-layer multi-channel graph attention network is used. First, the initial features of all nodes are linearly transformed and shared across all channels. Then, the features are analyzed using relational methods. Taking the channel as an example, the attention score is calculated using the following formula: , in, Indicates a relationship Next Layer nodes and nodes Unnormalized attention scores between them Representing relations Next The learnable attention vector of the layer, and They represent the first Layer nodes and nodes Features after linear transformation This represents the learnable scaling factor. Representing relations The weights (edge ​​weights). This indicates vector concatenation, and T represents the transpose operation; This represents the activation function.

[0081] For nodes All relations are The neighbors are then subjected to softmax normalization, as shown in the following formula: , in, This represents the normalized attention weights. Represents a node In relationship The set of neighbors below; This represents an exponential function.

[0082] Furthermore, in the relationship Under the channel, aggregate neighbor information to obtain relationships. The aggregation result is shown in the formula below: , in, Indicates a relationship Next Layer nodes The aggregation result.

[0083] The attention weight is calculated independently for each relationship type, and the neighbor information is aggregated to obtain the node representation of each channel; the outputs of all channels are spliced and linearly transformed to generate the node feature of this layer, i.e., the first-order message of the node (i.e., the output of the first-layer multi-channel graph attention network), and the above process is repeated to aggregate the second-order neighbor information, and the final embedding representation of the node is output. The representation not only contains the attributes of the node itself, but also integrates the structural context information in the topological network, providing a feature basis rich in spatial and functional relationships for subsequent multi-modal fusion.

[0084] Through the multi-channel attention mechanism model, it can be learned that when evaluating the risk of a "breaker arc chamber", the load state of its electrical connection components (such as current transformers) should be focused on, and the state of its mechanical linkage components (such as operating mechanisms) should be relatively weakened. The ability of neighbor screening and attention focusing makes the generated topological features no longer a simple average of neighbor features, but a refined structured representation containing network functional logic, laying a solid foundation for risk propagation prediction in topological networks.

[0085] ②Based on the fault data at the analysis time and before the analysis time, the fault text features of the components at the analysis time are extracted by using a text embedding model.

[0086] Specifically, based on a unified time axis, the fault data up to the analysis time is extracted, and the text embedding model is used to convert it into a fixed-dimensional feature vector to obtain the fault text feature representation of each component at the analysis time.

[0087] Exemplarily, the text embedding model uses a Bert model, which is trained by conventional methods, which will not be described here.

[0088] Preferably, the correlation confidence of the fault data is used as a weight, and the fault text feature representation is obtained by weighted averaging of the fault data up to the analysis time.

[0089] Through the text vectorization process, abstract text fault data is converted into numerical vectors with semantic information, so that text information can participate in the fusion processing of multi-modal data together with other numerical modal data.

[0090] ③Based on the electrical parameter data, the electrical time sequence feature extraction model is used to extract the electrical parameter features of the components at the analysis time; based on the environmental parameter data, the environmental time sequence feature extraction model is used to extract the environmental parameter features at the analysis time, and the components of the power grid equipment in the same region share the same environmental parameter features.

[0091] The electrical parameters and environmental parameters of the power grid are typical time series data, which contain evolution rules. The embodiment effectively captures short-term fluctuations, medium-term and long-term trends in these data through three time scales (time windows).

[0092] Specifically, for electrical parameter data and environmental parameter data, three time scale electrical time series data and environmental time series data are extracted from the data that has been time series aligned before the analysis time and before it.

[0093] Exemplarily, based on 7-day time series data before each time, the three time scales are 3 hours, 24 hours and 7 days, respectively, and the resampling frequencies are 5 minutes, 1 hour and 12 hours, respectively, that is, each time corresponds to 3 time scale data sequences. If the electrical parameter data and the environmental parameter data are time series aligned at a frequency of 5 minutes, resampling at a frequency of 1 hour collects all data points at 5-minute intervals within each hour, and aggregates the data points (such as taking the average) to obtain a data point representing 1 hour.

[0094] The model structure of the electrical time series feature extraction model and the environmental time series feature extraction model is the same, and both include a parameter-aware time series encoding module, a gating fusion module and a dynamic attention pooling module in turn; wherein the parameter-aware time series encoding module is used to encode the time series data of each parameter of each time scale to obtain the feature representation of each time scale; the gating fusion module is used to fuse the feature representations of the same parameter at different time scales to obtain the fusion feature representation of the parameter; and the dynamic attention pooling module is used to aggregate the fusion feature representations of multiple parameters of the same component to obtain the electrical parameter feature or the environmental parameter feature corresponding to the component.

[0095] Specifically, the parameter-aware time series encoding module includes in turn: a parameter type encoding unit for generating an embedding vector for each parameter type; an input mapping unit for concatenating the embedding vector of the parameter type and the current time position encoding after encoding the numerical value of each time step in different time scales, and mapping the input sequence through a linear layer; a plurality of improved LSTM units for encoding the input sequence of one time scale, wherein the input gate, the forget gate and the output gate of each improved LSTM unit all fuse the embedding vector of the parameter type, and the attention mechanism is used to calculate the attention weight of all time steps to obtain the feature representation of the current branch corresponding to the time scale.

[0096] It should be noted that the implementation of the parameter type encoding unit is through a trainable embedding layer (EmbeddingLayer) to map discrete parameter types (such as "current", "temperature") into a dense embedding vector.

[0097] At the same time, the standard position encoding (such as sine-cosine encoding) is used to capture the time sequence information, and the current time position encoding is generated for different time scales.

[0098] In the input mapping unit, the embedding vector of the parameter type and the current time position encoding are spliced behind the numerical value of each time step in different time scales, and are mapped into an input sequence of the same dimension through a linear layer. This way makes the model consider the physical nature of the parameter when memorizing or forgetting information. For example, for the "temperature" parameter, the model may learn a slow-changing, inertial memory mode; while for the "current" parameter, it may learn a mode more sensitive to instantaneous spikes.

[0099] Further, the number of improved LSTM units is the same as the number of time scales, and in this embodiment, there are three improved LSTM units to process short-term, medium-term and long-term time series respectively. The improvement is that the input gate, forget gate and output gate of the LSTM unit all fuse the embedding vector of the parameter type, and the formula is as follows: , wherein, , and represent the input gate, forget gate and output gate vectors at time . , and represent the weight matrix of the input gate, forget gate and output gate, respectively, for projecting the spliced vector to the hidden layer dimension; represents the hidden state at the last time, which contains the historical information up to time . represents the Sigmoid activation function, which compresses the output to the range of (0, 1); represents the input data at time . represents the embedding vector of the parameter type; , and represent the bias vectors of the input gate, forget gate and output gate, respectively.

[0100] Further, the temporal attention mechanism is added after the LSTM output to fully capture the temporal patterns. The hidden states of all time steps are taken as the input, and a trainable network is used to calculate the attention scores of each time step, which are then normalized by Softmax to obtain the weights. The final feature vector is the weighted sum of all weights and the corresponding hidden states. This mechanism enables the model to automatically focus on key time segments, such as periods of abnormal fluctuations, rather than treating all time steps equally.

[0101] Specifically, the gating fusion module receives the feature representations of three time scales output by the temporal feature extraction model for each parameter, and concatenates the feature representations of different time scales. A fully connected layer with a Sigmoid activation function is used to calculate the gating weight vector (three-dimensional vector) of different time scales, as shown in the following formula: , wherein, represents the gating weight vector, including the weights of short-term, medium-term, and long-term three time scales 、 and ; and represent the weight matrix and bias vector of the fully connected layer, 、 and represent the feature of short-term, medium-term, and long-term three time scales, respectively.

[0102] Further, the feature representations of the corresponding time scales are weighted and summed according to the gating weight vector of different time scales to obtain the fusion feature representation of the parameter.

[0103] It should be noted that the gating vector of the gating fusion module is dynamic. For predicting short-term sudden failures, the weight may automatically increase; for predicting long-term aging failures, the weight will rise. This dynamic scale selection mechanism enables the model to adapt to different risk patterns.

[0104] Since a component usually has multiple electrical parameters (variable in number), the dynamic attention pooling module is used to aggregate the fusion feature representations of variable parameters into a fixed-dimensional component feature representation, including: a parameter representation enhancement unit for concatenating the fusion feature representation of each parameter with the embedding vector of the parameter type and the importance prior weight to obtain the enhanced feature representation of each parameter; wherein the importance prior weight is preset based on domain knowledge to distinguish the importance of different electrical parameters and environmental parameters for risk prediction; The attention pooling unit calculates the attention weight of each parameter using a learnable global query vector, and aggregates the enhanced feature representation of each parameter according to the attention weight, and outputs the electrical parameter feature or the environmental parameter feature corresponding to the component.

[0105] Specifically, the learnable global query vector Q is a global context shared by all components, and the enhanced feature representation of each parameter is respectively mapped into a key vector K and a value vector V through an MLP; further, the attention score of the query vector with all key vectors is calculated, and after normalization, it is used as the attention weight to perform weighted summation on all value vectors V, to obtain the electrical parameter feature or the environmental parameter feature corresponding to the component.

[0106] It should be noted that the two time series feature extraction models are independently trained in advance using an autoencoder reconstruction loss or a contrastive learning loss, or can be trained end-to-end with the model of the subsequent steps of the embodiment, without the need for an independent loss function, but learning the model parameters from the final risk prediction loss through gradient backpropagation.

[0107] When independently training, if the autoencoder reconstruction loss is used, the time series feature extraction model is used as an encoder, and a reverse LSTM or a fully connected layer is designed as a decoder to reconstruct the original sequence from the output of the encoder, and the loss function is the mean square error of the reconstructed sequence and the original sequence. If the contrastive learning loss is used, two time enhancements are performed on the same time series, and features are extracted using the time series feature extraction model, and the loss function is to make the two features as close as possible in the representation space, and as far away as possible from the features of other time series.

[0108] It should be noted that the two time series feature extraction models have the same structure, but use independent model parameters, that is, the time series of multiple time scales of electrical parameters and the time series of multiple time scales of environmental parameters are used to train the respective time series feature extraction models.

[0109] The electrical time series data of three time scales of each component at the analysis time are transmitted to the trained electrical time series feature extraction model to obtain the electrical parameter feature at the analysis time; the environmental time series data of three time scales of each component at the analysis time are transmitted to the trained environmental time series feature extraction model to obtain the environmental parameter feature at the analysis time.

[0110] It should be noted that the risk prediction model in the dynamic risk prediction model 200 sequentially includes a multi-modal correlation model and a hierarchical prediction model, and is obtained by end-to-end training through construction of a sample set, and the specific implementation steps include S201-S204.

[0111] S201, constructing a sample set.

[0112] The component data processing module 100 is used to collect multi-source data of each component at each historical time, and further obtain a multi-modal feature vector of each component at each historical time.

[0113] Further, the historical time at each time is labeled with a risk label. The risk label includes a component level label and a device level label.

[0114] Specifically, the component level label includes: Component risk probability label, based on historical each time, the component that occurs failure in subsequent T days is labeled as 1, otherwise 0; Exemplarily, T is 7 days.

[0115] Component risk level label, which is set according to the severity in the fault data.

[0116] Exemplarily, 1 to 5 levels are set, including: slight abnormality, risk level label for 1 level that does not affect operation; General defect, risk level label for 2 level that needs planned maintenance; Significant abnormality, risk level label for 3 level that needs to be handled soon; Serious defect, risk level label for 4 level that needs emergency treatment; Immediate failure, risk level label for 5 level that needs to shut down immediately.

[0117] Component failure type label, which is one-hot encoded according to the failure type in the fault data. Exemplarily, the component failure type includes: overheating failure, insulation failure, mechanical failure and electrical failure.

[0118] Specifically, the device level label includes: Device risk index label, which is a comprehensive continuous value calculated according to the component risk level and the device failure coefficient corresponding to each historical time, and the formula is as follows: , Wherein, represents the device risk index label, and respectively represent the preset ratio; represents the function of taking the minimum value; represents the high-risk component ratio, which is based on the component risk level in the fault data to obtain the ratio of the number of high-risk components in the device to the total number of components in the device; represents the historical device failure coefficient, which is calculated according to the number of failures of the device before each historical time, for example, if no failure has occurred, the historical device failure coefficient is 0, if the failure has occurred within 3 times, the historical device failure coefficient is 0.5; If the failure has occurred more than 3 times, the historical device failure coefficient is 1.

[0119] The device risk type label is a one-hot encoding of the risk type of the whole device, including insulation aging, overload risk, environmental impact, and comprehensive factors.

[0120] All components are sorted, and the topological features, fault text features, electrical parameter features, and environmental parameter features of all components at each time are taken as a sample, which is put into the sample set together with the corresponding risk label. That is, each sample is a snapshot data at a time point, including the feature vectors of all components in all modalities.

[0121] S202, construct a multi-modal correlation model.

[0122] The multi-modal correlation model is used to mine and quantify the complex and non-linear implicit correlation between different modal data, thereby breaking the "data island" and providing feature representation rich in context information for accurate risk prediction.

[0123] The multi-modal correlation model is a Transformer model fused with hierarchical attention mask, which realizes deep and controllable multi-modal fusion through fine position encoding, hierarchical attention mask, and cross-attention mechanism, while maintaining physical rationality and interpretability. The following steps are used to output the multi-modal fusion representation, global context representation, and correlation analysis result of each sample: ①Project the feature vector of each modality of each component in each sample to a unified dimension and construct an input sequence. The projected feature vectors of each modality in the input sequence are stacked by element-wise addition.

[0124] Specifically, the feature vector of each modality of each component in each sample is projected to a unified dimension using a linear layer ; according to the component index and fixed modality order, a special CLS token of the same dimension is added at the beginning of the sequence as a global token for aggregating global information, obtaining an input sequence.

[0125] It should be noted that the global token is initially a learnable vector. Each position in the input sequence corresponds to a dimension vector, and each component has 4 modalities, so the total number of positions in the input sequence is , where represents the number of components, and the 4 modalities are arranged in the order of topology, fault text, electrical parameter, and environmental parameter. Position 0 corresponds to the feature vector of the CLS token, positions 1 to correspond to the topological feature vectors of the respective components, positions to correspond to the fault text feature vectors of the respective components, and positions to Corresponding to the electrical parameter feature vector of each component, position to Corresponding to the environmental parameter feature vector of each component.

[0126] Further, the feature vectors of each modality projection in the input sequence are stacked with respective hybrid position encodings by element-wise addition to help the model distinguish different modalities and different components. The dimension of the hybrid position encoding is the same as that of the feature vector after each modality projection, and the dimension of the feature vector of each modality is not changed after stacking. The hybrid position encoding includes: Intra-modality position encoding, which is generated by assigning an index to each component according to its functional importance in the device and using a sine-cosine function dimension vector; Intra-component position encoding, which is generated by assigning a fixed position offset to each modality in the input sequence dimension vector; Inter-modality position encoding, which is obtained according to the embedding vector of the modality type dimension vector.

[0127] It should be noted that the hybrid position encoding is also calculated for the global token, the intra-modality position encoding is calculated by assigning an index of 0, the intra-component position encoding is calculated by assigning a position offset of -1, and the inter-modality position encoding is obtained by obtaining the embedding vector of "CLS".

[0128] Finally, the input sequence with hybrid position encoding is obtained, and the Transformer encoding layer is input.

[0129] In the self-attention mechanism of the Transformer, a pre-defined hierarchical attention mask matrix is applied to inject domain knowledge, control the path of information flow, and enhance the physical rationality and interpretability of the model.

[0130] It should be noted that the dimension of the hierarchical attention mask matrix is the same as the length of the input sequence, which is a dimension matrix, used to control the attention permission between different positions in the input sequence.

[0131] Specifically, the positions that can be mutually observed in the hierarchical attention mask matrix are set to 0, and the positions that are prohibited from being observed are set to negative infinity. The positions that can be mutually observed include: The position of the global token and the positions of all other vectors in the input sequence are mutually observed to converge the global context; The positions of the feature vectors of different modalities within the same component are mutually observed, allowing them to freely calculate attention; The positions of the eigenvectors of the same mode of different components only concern each other when the corresponding components are neighbors in the topological relationship (such as electrical direct connection, belonging to the same device, spatial proximity).

[0132] The input sequence with mixed position encoding is input into the Transformer encoding layer, and cross correlation between modalities is performed according to the unified multi-head self-attention layer and the hierarchical attention mask matrix, to output a multi-modal fusion representation and a global context representation of each sample.

[0133] It should be noted that the Transformer encoding layer includes a plurality of Transformer encoders, and each Transformer encoder includes a multi-head self-attention mechanism, a feedforward network, a residual connection, and layer normalization.

[0134] Specifically, when using the self-attention mechanism of the Transformer, the query vector, the key vector, and the value vector are calculated according to the input sequence with mixed position encoding; then, the original attention score is calculated according to the query vector and the key vector, and the final attention score is obtained by adding the hierarchical attention mask matrix; and the attention weight is obtained by performing softmax normalization. At this time, since the attention score of the position to be prohibited from attention is negative infinity, the value of these positions is still negative infinity after being added to the original attention score, and is 0 after softmax normalization, thereby realizing the set attention mechanism.

[0135] In this embodiment, the query vector, the key vector, and the value vector are calculated according to the input sequence, without explicitly specifying the modal source of the query, the key, and the value. Each attention head projects the input sequence into different feature subspaces through a respective independent parameter matrix, naturally learns different modal cross modes, and can realize rich and diverse multi-modal cross fusion.

[0136] Through this hierarchical and controlled attention mechanism, the model is forced to learn the interaction mode conforming to the physical law during the training process. For example, the electrical parameter feature of a component cannot directly pay attention to the text description of another non-neighbor component, but can indirectly affect it through the shared global token or the full connection within the component. The “operation and maintenance text” feature (such as describing “abnormal noise”) of a component will first be deeply fused with its own “electrical feature” (such as increased current harmonic) and “topological feature” (such as being located at a vibration-sensitive mechanical connection point). Then, this fused information affects the judgment of other components through the CLS token or the topological neighbor relationship. This enables the model to learn complex interactions while avoiding illogical associations, thereby improving the reliability of the results. This process simulates the reasoning chain of human experts when performing fault diagnosis, and realizes guided and interpretable deep fusion.

[0137] After the cross correlation of the Transformer encoding layer, the output vectors of multiple modalities of each component are aggregated to generate a multi-modal fusion representation, and the output of the global token is used as a global context representation.

[0138] ③Extract the correlation analysis result from the attention weight of the Transformer encoding layer.

[0139] The correlation analysis result includes an inter-modal correlation matrix and a cross-component modal correlation matrix. The inter-modal correlation matrix is used to quantify the average correlation strength between different modalities, such as the correlation between the topology modality and the text modality. The cross-component modal correlation matrix is used to quantify the correlation strength between the same modalities of neighboring components, such as the correlation between the topology modality of component A and the topology modality of neighboring component B. These correlation matrices are not only key to explainability, but also provide important context features for subsequent risk prediction.

[0140] It should be noted that the correlation analysis result is extracted and analyzed from the attention weight matrix of each Transformer layer and each attention head, and is obtained by grouping and aggregating the weights by modality and component.

[0141] Specifically, the inter-modal correlation matrix is obtained by averaging the attention weights between different modality position sets in all Transformer layers and attention heads.

[0142] The cross-component modal correlation matrix is obtained by averaging the attention weights of the same modalities between neighboring components in the topology relationship in all Transformer layers and attention heads.

[0143] Compared with the prior art, the multi-modal correlation model guides the model to learn data correlation patterns consistent with the physical reality of the power grid by mixing position encoding and hierarchical attention masks, ensuring sufficient fusion of different modal features within the same component, avoiding direct influence between non-neighbor components that is not logical, and realizing controllable and physically consistent multi-modal information interaction. The inter-modal correlation strength and the cross-component modal correlation strength are extracted from the model attention weight, forming a quantitative correlation analysis result, which is used in the hierarchical prediction model to realize intelligent perception and fusion of context.

[0144] S203, constructing a hierarchical prediction model.

[0145] The hierarchical prediction model of the embodiment simultaneously performs component-level fine-grained prediction and device-level macro-situation assessment. The model sequentially includes a component-level risk prediction module, a device-level risk prediction module, and a prediction enhancement module.

[0146] It should be noted that the component-level risk prediction module is configured to output a component risk probability, a component risk level and a component failure type by using a multi-layer perception according to the multi-modal fusion representation of each component in each sample, so as to realize risk identification at a component level in a fine-grained manner.

[0147] Specifically, the multi-layer perception extracts abstract features highly related to risks from the multi-modal fusion representation through layer-by-layer nonlinear transformation. Subsequently, the branch is divided into three independent outputs, realizing multi-task learning. The component risk probability is a scalar between 0 and 1, representing the probability of failure of the component in a future time period; The component risk level outputs a multi-dimensional vector corresponding to the component risk level label. For example, a 5-dimensional vector is output, representing the probability distribution of the component belonging to “level 1” to “level 5” risks, corresponding to slight, general, significant, severe and immediate failure respectively. This classification output facilitates quick classification management by operation and maintenance personnel. The component failure type outputs a multi-dimensional vector corresponding to the component failure type label, identifying possible failure types. This classification output facilitates the disclosure of main risk inducements.

[0148] It should be noted that the device-level risk prediction module is configured to output a device risk index and a device risk type after aggregating the global context representation of each sample and the multi-modal fusion representation of each component, so as to realize macro-situation assessment at a device level.

[0149] Specifically, the device-level risk prediction module first aggregates the global context representation of each sample and the multi-modal fusion representation of each component according to the device, and then encodes and decodes through a fully connected network to output the device risk index and the device risk type.

[0150] The aggregation of the global context representation of each sample and the multi-modal fusion representation of each component according to the device includes: According to the mapping relationship from the component to the device, the multi-modal fusion representations of all components of the same device are aggregated within the device to obtain a device feature representation. The intra-device aggregation includes maximum pooling, average pooling and risk attention pooling operations. The global context representation and the device feature representation are spliced to obtain a comprehensive feature representation.

[0151] It should be noted that the risk attention pooling is to obtain the weighted sum of the multi-modal fusion representations of each component within the same device by taking the predicted component risk probability as the attention weight.

[0152] The comprehensive feature representation is encoded and decoded by a fully connected network, and finally outputs a device risk index and a device risk type. The device risk index is a scalar between 0 and 1, which is used to quantify the overall risk level of the device; the device risk type outputs a multi-dimensional vector corresponding to the device failure type label, which identifies the possible dominant risk type of the device.

[0153] It should be noted that the prediction enhancement module is used to enhance the component risk probability and the device risk index according to the correlation analysis result, and output the final component risk probability and device risk index.

[0154] Specifically, the inter-modal correlation matrix in the correlation analysis result is flattened, the cross-component modal correlation matrix in the correlation analysis result is maximum-pooled and average-pooled, and the context vectors are obtained by splicing them respectively. The context vectors are mapped to enhancement coefficients by a multilayer perceptron.

[0155] Further, according to the enhancement coefficients, the component risk probability and the device risk index are enhanced respectively by the following formula: , wherein, represents the enhanced component risk probability or device risk index, is a learnable parameter, represents the component risk probability output by the component-level risk prediction module or the device risk index output by the device-level risk prediction module, represents the enhancement coefficient.

[0156] The prediction enhancement module adopts a learnable scaling mechanism. When the correlation context strongly indicates risk (for example, the state of a component itself is acceptable, but its multiple key neighbors and the environment it is in show abnormalities), the model will moderately adjust its risk value to make it more consistent with the actual physical situation.

[0157] S204, end-to-end training using the sample set.

[0158] Using the sample set, the multi-modal correlation model and the hierarchical prediction model are optimized by end-to-end training, and the overall loss function is obtained by calculating the weighted sum of the component-level joint loss , the device-level joint loss and the correlation-driven loss , as shown in the following formula: , wherein, , and represent the weights of the component-level joint loss, the device-level joint loss and the correlation-driven loss, respectively.

[0159] It should be noted that the component-level joint loss is a weighted sum of the component risk probability loss, the component risk level loss, and the component failure type loss. Specifically, the component risk probability loss is obtained by calculating the binary cross-entropy loss between the component-level risk probability prediction and the true label; the component risk level loss is obtained by calculating the cross-entropy loss between the component-level risk level classification prediction and the true label; and the component failure type loss is obtained by calculating the multi-label binary cross-entropy loss between the component failure type multi-label prediction and the true label.

[0160] It should be noted that the equipment-level joint loss is a weighted sum of the equipment risk index loss and the equipment risk type loss; the equipment risk index loss is obtained by calculating the mean squared error loss between the predicted equipment risk index value and the true label; the equipment risk type loss is obtained by calculating the cross-entropy loss between the predicted equipment risk type classification and the true label.

[0161] It should be noted that the correlation-driven loss is calculated based on the predicted value of the equipment risk index and the inter-modal correlation matrix in the correlation analysis results. This means that when the model finds strong correlations between certain modes, its prediction of the equipment risk index should be more reliable (probability value close to 0 or 1). Specifically, the formula is as follows: , in, This indicates the number of samples in the training batch. Indicates the first Number of devices in each sample Indicates the first In the nth sample Predicted risk index value for each device Indicates the first The modal correlation matrix of each sample. This represents the smoothing constant, a very small value used to prevent mathematical errors in numerical calculations.

[0162] During implementation, the component data processing module 100 outputs the multimodal feature vectors of each component at the analysis time and transmits them to the dynamic risk prediction module 200. The trained multimodal association model outputs the global context representation, association analysis results, and multimodal fusion representation of each component based on the multimodal feature vectors of the components at the analysis time, and transmits them to the trained hierarchical prediction model. The hierarchical prediction model outputs the risk prediction results of each component and its associated equipment in the future time period, including: component risk probability, component risk level, component failure type, equipment risk index, and equipment risk type.

[0163] Preferably, the various neural network models involved in each module of the system of the present embodiment are encapsulated as Restful API interfaces and integrated into the background service of a domestic BIM platform (such as Guanglianda GLD-BIM), which is automatically called periodically to obtain risk prediction results.

[0164] The risk prediction results are visually rendered in the three-dimensional model of the BIM platform. The color of the corresponding component changes dynamically according to its risk level. The operation and maintenance personnel click on the highlighted component to view the corresponding multi-source data information and risk prediction results.

[0165] Further, the model parameter optimization module 300 receives the risk prediction results from the dynamic risk prediction module 200, and if it identifies that there are components to be optimized, it intelligently provides adjustment schemes for the BIM model parameters of the components to be optimized based on a pre-constructed parameter adjustment rule library, thereby forming a complete closed loop of "perception-decision-optimization". The specific implementation steps of this module include S301-S303.

[0166] S301, construct a parameter adjustment rule library.

[0167] Parameter adjustment is not arbitrary and must strictly follow grid design standards and engineering experience. For this purpose, natural language processing technology is introduced to assist in constructing an intelligent and structured parameter adjustment rule library.

[0168] ① Use natural language processing technology to analyze grid engineering design standards, technical specifications, and expert experience documents to construct a knowledge graph.

[0169] Collect and input relevant grid engineering design standards, technical specifications, and expert experience documents. These documents include but are not limited to national standards and industry specifications such as "DL / T 1573-2016 Electrical Power Transformer Maintenance Guidelines", "GB 50150-2016 Electrical Equipment Interim Test Standards for Electrical Device Installation Engineering", etc.

[0170] Through natural language processing technology, the input documents are deeply semantically analyzed to extract key technical requirements, parameter constraints, safe operation boundaries, and recommended engineering practices. The analysis process focuses on identifying the following: the allowable range of technical parameters of equipment and components, parameter adjustment suggestions under different working conditions, threshold boundary conditions for safe operation, and specific technical measures for fault prevention.

[0171] Based on the information obtained from the analysis, entities, entity attributes, and relationships between entities are constructed to form a structured knowledge graph.

[0172] ② Based on historical operation and maintenance records and the knowledge graph, the mapping relationship between fault types and BIM model parameters is extracted to establish a fault-parameter association table.

[0173] The fault data in the component data processing module is collected by the method, and each fault type (such as insulation breakdown, overheating aging, mechanical fatigue, etc.) and parameter in the historical operation and maintenance record is obtained; By comparing the similarity of entity vectors in the knowledge graph, the names of the fault type and the parameter are standardized, and the associated parameters are obtained according to the knowledge graph to form a fault-parameter association table.

[0174] For example, "insulation failure" is mapped to "insulation material grade", "insulation thickness", etc. Parameters; "overheating failure" is mapped to "heat dissipation area", "cooling efficiency", etc. Parameters; "mechanical failure" is mapped to "structural strength", "material toughness", etc. Parameters.

[0175] Preferably, by calculating the correlation between the occurrence probability of the fault in the historical operation and maintenance record and the parameter, such as calculating the Pearson correlation coefficient, the sensitivity of different parameters to the fault type is obtained after normalization, which is used to quantify the relative importance of each parameter in risk mitigation.

[0176] ③Take the fault-parameter association table and the knowledge graph as the knowledge context, and guide the large language model to generate the preliminary adjustment rule through the prompt word template. After being reviewed and corrected by the field experts, it is stored in the parameter adjustment rule library.

[0177] It should be noted that the prompt word template instructs the large language model to generate a parameter adjustment rule based on the knowledge context, which includes the fault type, parameter name, adjustment direction, recommended amplitude, basis clause, and expected effect.

[0178] For example, the prompt word template is: "You are a senior power grid engineering expert and BIM model standard consultant, proficient in DL / T 1573-2016 Power Transformer Maintenance Guide and other power grid design and operation standards. Your task is to generate accurate and executable BIM model parameter adjustment rules based on the provided domain knowledge. Please generate rules based on the following two parts of knowledge context: \n 1. Fault-parameter association table\n[Enter the fault-parameter association table here]\n 2. Related standard knowledge\n[Enter the knowledge fragment retrieved from the knowledge graph according to the fault type here]\n. For fault type [Enter the specific fault type, such as: insulation failure], please generate specific parameter adjustment rules, including: 1. Which parameters need to be adjusted; 2. Adjustment direction and recommended amplitude; 3. Parameter constraints and standard clauses relied on; 4. Expected risk reduction effect.".

[0179] The preliminary adjustment rule generated by the large language model is sent to the field experts for manual review, including: verifying the rationality and feasibility of the adjustment amplitude, supplementing special constraint conditions in actual engineering, correcting rule logic that does not conform to engineering reality, and optimizing the application priority of the rule.

[0180] After optimization by domain experts, the final adjustment rules are formed, formatted and stored in the parameter adjustment rule library. The rule library is integrated into the BIM platform, and a version management mechanism is established to support subsequent continuous updating and improvement.

[0181] S302, according to the risk prediction result output by the dynamic risk prediction module, identifying the to-be-optimized component and its priority.

[0182] It should be noted that the to-be-optimized component is identified by the following steps: ① According to the component risk probability and the component risk level, the basic risk value of the component is calculated by weighted summation.

[0183] Specifically, the basic risk value of the component is calculated by the following formula: , Wherein, represents the basic risk value of the component, and represent the component risk probability and the component risk level of the component, and represent the weights of the component risk probability and the component risk level, represents the normalization function, and exemplarily, the minimum-maximum normalization function is adopted. ② According to the equipment risk index, the high risk threshold and the basic risk threshold, the dynamic risk threshold of the component is calculated.

[0184] This embodiment adopts a dynamic threshold mechanism, and combines the equipment risk index for enhanced identification. The risk threshold of the component of the equipment with high overall risk is appropriately relaxed, and the dynamic risk threshold of each component is calculated by the following formula:

[0185] , Wherein, represents the dynamic risk threshold of the component, and respectively represent the preset high risk threshold and the basic risk threshold, ; represents the weight coefficient; represents the equipment risk index of the equipment to which the component belongs.

[0186] ​​​​③ Select components whose basic risk value is greater than their dynamic risk threshold as the first component; select neighbor components whose component risk probability is greater than the neighbor screening threshold and do not belong to the first component from the neighbor components of each first component as the second component; the neighbor screening threshold is less than the basic risk threshold.

[0187] This step includes related components whose propagation risk exceeds a lower neighbor screening threshold as components to be optimized, ensuring synergy in parameter optimization. For example, the neighbor screening threshold is set to 0.3.

[0188] ④ The first and second components are combined as the component to be optimized.

[0189] For example, if the component risk probability is 0.65, the component risk level is 4 (minimum level is 1, maximum level is 5, normalized to 0.75), and the weights are 0.7 and 0.3 respectively, then the basic risk value is 0.68. The high-risk threshold is set to 0.7, the basic risk threshold is set to 0.5, and the weighting coefficients are... With a base risk value of 0.3 and an equipment risk index of 0.8, the calculated dynamic risk threshold is approximately 0.642. Since the base risk value 0.68 > 0.642, this component is classified as a first-tier component. If selection were based solely on a high-risk threshold, it would not be selected as a first-tier component because 0.68 < 0.7. This demonstrates how a dynamic threshold mechanism appropriately relaxes the selection criteria.

[0190] Furthermore, the components to be optimized are processed in descending order of priority to ensure that key risks are resolved first.

[0191] It should be noted that the first component is the direct source of risk and has a higher priority than the second component. Therefore, the priority of the first component is obtained by multiplying its basic risk value by its topological centrality weight; the priority of the second component is obtained by multiplying its basic risk value, the edge weights between the second component and its associated first component, and its topological centrality weight.

[0192] The topological centrality weight is calculated based on at least one of the degree centrality, betweenness centrality and eigenvector centrality of the component, wherein the shortest path used to calculate betweenness centrality is determined based on the electrical function path.

[0193] It should be noted that the topological centrality weight can also be preset according to the importance of the component in the topological network, for example, the weight of a key node is 1.2 and that of an ordinary node is 1.0.

[0194] S303. Based on the parameter adjustment rule library, obtain the BIM model parameter adjustment rules of the component to be optimized, and form an optimization scheme according to the priority of the component to be optimized.

[0195] According to the component failure type of each component to be optimized, the corresponding parameter adjustment rule is queried from the parameter adjustment rule library, and an optimization scheme is formed according to the priority of the component to be optimized.

[0196] Specifically, the optimization scheme is formed according to the priority of the component to be optimized, comprising: The first component in the component to be optimized is arranged in descending order of priority to form a main sequence; The second component associated with each first component in the main sequence is arranged in descending order of priority to form a sub-sequence of each first component; According to the order of the main sequence and each sub-sequence, the corresponding parameter adjustment rule is output in sequence to obtain the optimization scheme.

[0197] This two-level sorting method facilitates simultaneous adjustment of the parameters of the direct risk source and coordinated adjustment of the component associated with the highest risk.

[0198] When applying the adjustment rule to adjust the BIM model parameters of the component, the following three constraints are strictly followed: Adjustment range constraint: Ensure that the new parameter value is within the upper and lower limits specified by the standard.

[0199] Coordination constraint: Check and ensure that the parameter adjustment between associated components is matched (for example, if the sleeve insulation level is improved, the insulation level of the connected conductor also needs to be considered accordingly).

[0200] Amplitude constraint: Limit the adjustment amplitude to no more than 30% to avoid overdesign and severe impact on existing systems.

[0201] Preferably, one or more BIM model parameters are adjusted in the test environment BIM model first, new risk prediction results are obtained based on the latest real-time data, and if the component risk probability is not reduced, the adjustment amplitude is appropriately adjusted or an alternative parameter is selected for a new round of calculation until a scheme that meets all constraints and effectively reduces the risk is found.

[0202] Through the parameter synchronization interface, the finally determined parameter adjustment scheme is automatically written into the parameter database of the localization BIM platform. The BIM platform updates the geometric and attribute information of the three-dimensional model in real time, and realizes the real-time evolution of the digital twin model.

[0203] Preferably, the adjustment log of each parameter is recorded to facilitate comparison and analysis before and after adjustment, and to trace back to any model version in history, providing data trajectory for long-term optimization strategy analysis.

[0204] Compared with the prior art, the power grid equipment BIM model parameter optimization system provided by the embodiment realizes a complete closed loop system from data preprocessing, correlation mining, risk prediction to the formation of a parameter optimization scheme, effectively solves the problems of data island, early warning lag and long BIM model parameter optimization iteration period in the prior art, realizes autonomous, closed loop and continuous iteration promotion of parameter optimization according to the actual operation state and the risk prediction result, quickly adapts to the changing operation demand of the power grid, reduces the dependence on manual analysis and human errors, and has great value for guaranteeing the safe, stable and efficient operation of the power grid. Through unified and standardized processing of the component data processing module, the system can map heterogeneous multi-modal data such as BIM geometric data, topological relationship data, electrical parameter data, fault text data and environmental parameter data to a unified semantic space; instead of considering various types of data in isolation, the system actively discovers and quantifies the complex and nonlinear implicit correlations between them, greatly excavates the value of multi-source data, and provides a comprehensive data view for accurate decision-making. The dynamic risk prediction module uses the fused deep features to capture the time evolution law through an improved parameter-aware LSTM, and combines multi-task learning and context enhancement to perform hierarchical (component level / equipment level) and multi-dimensional (probability / level / type) risk prediction, so that the prediction is more in line with the physical reality, has high accuracy and low false alarm rate; the traditional lag alarm mode based on a fixed threshold is changed, so that the operation and maintenance are changed from passive response to active intervention, greatly reducing the probability of unplanned shutdown and accident risk.

[0205] Embodiment 2 Another embodiment of the present application discloses an electronic device, which comprises a memory, a processor and a computer program stored in the memory and executable on the processor.

[0206] The electronic device of the embodiment can be a server, a terminal or any other electronic device with necessary computing and / or processing capability.

[0207] In a specific embodiment, the electronic device comprises a processor, a memory and a communication interface connected through a system bus. The processor of the electronic device provides necessary computing, processing and / or control capability. The memory of the electronic device comprises a non-volatile storage medium and an internal memory, the non-volatile storage medium stores an operating system, a computer program and serves as a working area during runtime, and temporarily stores intermediate results and final results; the internal memory provides a running environment for the operating system and the computer program in the non-volatile storage medium; the communication interface of the electronic device is used for network connection and communication with external devices; and the computer program is executed by the processor to implement the power grid equipment BIM model parameter optimization system of embodiment 1.

[0208] Embodiment 3 In another embodiment of the present application, a computer readable storage medium is disclosed, and the computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the power grid equipment BIM model parameter optimization system of embodiment 1.

[0209] Those skilled in the art can understand that all or part of the processes of the above-mentioned embodiments can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer readable storage medium. The computer readable storage medium is a disk, an optical disk, a read-only memory, a random access memory, etc.

[0210] The above is only a preferred specific embodiment of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. A BIM model parameter optimization system for power grid equipment, characterized in that, include: The component data processing module is used to collect multi-source data of each component and extract the multi-modal feature vectors of each component at the analysis time according to the neural network model of each modality. The dynamic risk prediction module is used to obtain the multimodal feature vectors of each component at the analysis time output by the component data processing module, and input them into the risk prediction model to obtain the risk prediction results. The risk prediction results include: component risk probability, component risk level, component failure type, equipment risk index, and equipment risk type; The model parameter optimization module is used to identify components to be optimized and their priorities based on the risk prediction results output by the dynamic risk prediction module; obtain BIM model parameter adjustment rules for the components to be optimized based on the parameter adjustment rule library, and form an optimization scheme according to the priority of the components to be optimized; the identification of components to be optimized includes: calculating the basic risk value of the component by weighted summation based on the component risk probability and component risk level; calculating the dynamic risk threshold of the component based on the equipment risk index, high risk threshold, and basic risk threshold; selecting components whose basic risk value is greater than their dynamic risk threshold as first components; selecting neighboring components whose component risk probability is greater than the neighbor screening threshold and do not belong to the first components from the neighboring components of each first component as second components; the neighbor screening threshold is less than the basic risk threshold; and merging the first and second components as the components to be optimized.

2. The power grid equipment BIM model parameter optimization system according to claim 1, characterized in that, The parameter adjustment rule base is obtained through the following steps: Natural language processing technology is used to parse power grid engineering design standards, technical specifications, and expert experience documents to construct a knowledge graph; Based on historical operation and maintenance records and knowledge graphs, the mapping relationship between fault types and BIM model parameters is extracted, and a fault-parameter association table is established. Using the fault-parameter association table and knowledge graph as the knowledge context, the large language model is guided to generate preliminary adjustment rules through prompt word templates. After being reviewed and corrected by domain experts, these rules are stored in the parameter adjustment rule library.

3. The power grid equipment BIM model parameter optimization system according to claim 1, characterized in that, The basic risk value of the component is calculated using the following formula: ， in, Representation of components The basic risk value, and Representation of components The component risk probability and component risk level, and These represent the weights of the component risk probability and the component risk level, respectively. This represents the normalization function.

4. The power grid equipment BIM model parameter optimization system according to claim 1, characterized in that, The dynamic risk threshold is calculated using the following formula: ， in, Representation of components Dynamic risk threshold; and These represent the preset high-risk threshold and basic risk threshold, respectively. ; Indicates the weighting coefficient; Representation of components The equipment risk index of the equipment.

5. The power grid equipment BIM model parameter optimization system according to claim 1, characterized in that, The priority of the first component in the components to be optimized is obtained by calculating the product of the basic risk value of the first component and the topological centrality weight of the first component; The priority of the second component is obtained by calculating the product of the basic risk value of the second component, the edge weight between the second component and its associated first component, and the topological centrality weight of the second component.

6. The power grid equipment BIM model parameter optimization system according to claim 5, characterized in that, The topological centrality weight is calculated based on at least one of the degree centrality, betweenness centrality, and eigenvector centrality of the corresponding component.

7. The power grid equipment BIM model parameter optimization system according to claim 5, characterized in that, The process of forming an optimization scheme based on the priority of the components to be optimized includes: The first component in the optimization components is arranged in descending order of priority to form the main sequence; For each first component in the main sequence, its associated second components are arranged in descending order of priority to form a subsequence for each first component; Based on the order of the main sequence and each subsequence, the corresponding parameter adjustment rules are output sequentially to obtain the optimization scheme.

8. The power grid equipment BIM model parameter optimization system according to claim 7, characterized in that, The parameter adjustment rules simultaneously satisfy three constraints: adjustment range constraint, coordination constraint, and amplitude constraint. The adjustment range constraint ensures that the adjusted parameter value is within the upper and lower limits specified in the standard. The coordination constraint ensures that the parameter adjustments between related components are matched. The amplitude constraint limits the amplitude of a single parameter adjustment to no more than a preset safety threshold.

9. The power grid equipment BIM model parameter optimization system according to claim 1, characterized in that, The multi-source data for each component includes: BIM geometric data, topological relationship data, electrical parameter data, fault data, and environmental parameter data; the extraction of multimodal feature vectors for each component at the analysis time based on the neural network model of each modality includes: For each component, based on BIM geometric data and topology relationship data, graph attention network is used to extract the topological features of the component; based on the fault data at and before the analysis time, text embedding model is used to extract the fault text features of the component at the analysis time; based on electrical parameter data, electrical time-series feature extraction model is used to extract the electrical parameter features of the component at the analysis time; based on environmental parameter data, environmental time-series feature extraction model is used to extract the environmental parameter features at the analysis time, and components of power grid equipment in the same area share the same environmental parameter features. The topological features, fault text features, electrical parameter features, and environmental parameter features of each component at the analysis time constitute the multimodal feature vector of each component at the analysis time.

10. The power grid equipment BIM model parameter optimization system according to claim 9, characterized in that, The extraction of topological features of components using graph attention networks includes: Using components as nodes, the initial features of the nodes are the BIM geometric data of the components, and the relationships between components are the edges to construct the graph structure data; the relationships include: electrical connection relationships, mechanical linkage relationships, and spatial proximity relationships; The graph structure data is input into a multi-layer, multi-channel graph attention network, which outputs a node embedding representation to obtain the topological features of each component. Each channel corresponds to an edge of a certain relationship, which is used to independently calculate and normalize the attention score of the edge of that relationship.

11. The power grid equipment BIM model parameter optimization system according to claim 9, characterized in that, After the electrical parameter data and environmental data are time-aligned according to a unified time axis, electrical time-series data and environmental time-series data of three time scales are extracted from the time-aligned data at and before the analysis time. Different sampling frequencies are used for resampling at different time scales.

12. The power grid equipment BIM model parameter optimization system according to claim 11, characterized in that, The electrical timing feature extraction model and the environmental timing feature extraction model have the same model structure, both including, in sequence: a parameter-aware timing encoding module, a gated fusion module, and a dynamic attention pooling module; wherein, the parameter-aware timing encoding module is used to encode the timing data of each parameter at each time scale to obtain the feature representation of each time scale; the gated fusion module is used to fuse the feature representations of the same parameter at different time scales to obtain the fused feature representation of the parameter; the dynamic attention pooling module is used to aggregate the fused feature representations of multiple parameters of the same component to obtain the electrical parameter features or environmental parameter features corresponding to the component.

13. The power grid equipment BIM model parameter optimization system according to claim 12, characterized in that, The parameter-aware timing coding module includes, in sequence: Parameter type encoding unit, used to generate an embedding vector for each parameter type; The input mapping unit is used to encode the numerical concatenation parameter type embedding vector and the current time position of each time step in different time scales, and then obtain the input sequence through linear layer mapping; Multiple improved LSTM units encode input sequences at different time scales. The input gate, forget gate, and output gate of each improved LSTM unit are all incorporating parameter-type embedding vectors. The attention weights of the hidden states at all time steps are calculated and weighted by an attention mechanism to obtain the feature representation of the current time scale.

14. The power grid equipment BIM model parameter optimization system according to claim 1, characterized in that, The risk prediction model includes a multimodal correlation model and a hierarchical prediction model. The multimodal correlation model is used to output a global context representation, correlation analysis results, and a multimodal fusion representation of each component based on the multimodal feature vector of the component at the analysis time, and then transmits them to the hierarchical prediction model. The hierarchical prediction model is used to output the risk prediction results of each component and its associated equipment in the future time period.

15. The power grid equipment BIM model parameter optimization system according to claim 14, characterized in that, The multimodal association model performs the following steps: The multimodal feature vectors of each component at the analysis time are projected to a unified dimension and an input sequence is constructed. The feature vectors after the modality projection in the input sequence are superimposed with their respective hybrid position codes by adding them element by element. The input sequence with hybrid positional encoding is fed into the Transformer encoding layer, and cross-correlation between modalities is performed based on a unified multi-head self-attention layer and a hierarchical attention mask matrix. The output is a multi-modal fusion representation of each component in each sample and a global context representation. The correlation analysis results are extracted from the attention weights of the Transformer encoding layer.

16. The power grid equipment BIM model parameter optimization system according to claim 15, characterized in that, In the hierarchical attention mask matrix, mutually interested positions are set to 0, and prohibited positions are set to negative infinity. The mutually interested positions include: The position of the global token is correlated with the positions of all other vectors in the input sequence; The positions of the eigenvectors of different modes within the same component are mutually related; The positions of eigenvectors of the same mode in different components are mutually concerned only when the corresponding components are neighbors in the topological relationship.

17. The power grid equipment BIM model parameter optimization system according to claim 15, characterized in that, The correlation analysis results include: inter-modal correlation matrix and cross-component modal correlation matrix; the inter-modal correlation matrix is ​​obtained by averaging the attention weights between different modal position sets in all Transformer layers and attention heads; the cross-component modal correlation matrix is ​​obtained by averaging the attention weights of the same modes between neighboring components in the topological relationship in all Transformer layers and attention heads.

18. The power grid equipment BIM model parameter optimization system according to claim 14, characterized in that, The hierarchical prediction model sequentially includes: a component-level risk prediction module, an equipment-level risk prediction module, and a prediction enhancement module. The component-level risk prediction module is used to output the component risk probability, component risk level, and component failure type using a multilayer perceptron based on the multimodal fusion representation of each component. The equipment-level risk prediction module is used to aggregate the global context representation and the multimodal fusion representation of each component according to the equipment, and output the equipment risk index and equipment risk type. The prediction enhancement module is used to enhance the component risk probability and equipment risk index based on the correlation analysis results, and output the final component risk probability and equipment risk index.

19. An electronic device, characterized in that, The electronic device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the functions of the power grid equipment BIM model parameter optimization system as described in any one of claims 1-18.

20. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the function of the power grid equipment BIM model parameter optimization system as described in any one of claims 1 to 18.

Images