5 results about "Process graph" patented technology

The application discloses a zero sample graph learning method based on a large language model, comprising: neighborhood subgraph sampling on to-be-processed graph structure data to generate graph query text; based on the graph query text, a structured prompt template containing multi-stage reasoning instructions is constructed, and the prompt template instructs the large language model to sequentially perform topological structure analysis, semantic attribute interpretation, multi-candidate answer enumeration and deep reevaluation before generating a final answer; a graph reasoning training data set is constructed; the basic large language model is fine-tuned by full parameters by using the graph reasoning training data set; the large language model is optimized by using a reinforcement learning algorithm based on group relative advantage; for a new graph task to be reasoned, the new graph task is input into the optimized large language model, and a final prediction answer containing a reasoning process is directly output. The application discards the dependence of a graph neural network, and only uses the text reasoning capability of the large language model to complete a graph task, so that the deployment complexity and the calculation cost are greatly reduced.

The embodiment of the invention provides a sample data construction method, a graph model training method and a graph data processing method. The sample data construction method comprises the steps that graph data of a target item and description texts of multiple graph elements in the graph data are acquired; inputting the graph data into a pre-training graph model of the target item to obtain an initial classification result of the plurality of graph elements; the description text of the first graph element and the initial classification result of the first graph element are input into a large language model, a label classification result of the first graph element is obtained, and the first graph element is any one of multiple graph elements; and based on the graph data and the tag classification result of the plurality of graph elements, constructing sample graph data of the target item, the sample graph data being used for training a graph model of the target item. According to the method, the structure information between the graph elements is considered, the tag classification result for the graph elements is generated, limitation on graph model training is avoided, and the accuracy and efficiency of processing graph data under the target item are improved.

The embodiment of the invention discloses a method, a device and equipment for training a graph size model. In the embodiment of the invention, a graph data set corresponding to a set task is obtained, and the graph data set comprises multiple pieces of graph data; respectively carrying out lossless conversion on the plurality of pieces of graph data to generate a plurality of lossless graph sequences; training a generative graph size model corresponding to the set task according to the plurality of lossless graph sequences; performing fine adjustment on the plurality of lossless graph sequences according to the subtask type of the set task to generate a plurality of target lossless graph sequences; and according to a plurality of target lossless graph sequences, carrying out fine tuning on the generative graph size model to generate a target generative graph size model. By means of the method, the target generative graph size model can be trained, and the target generative graph size model has a good effect when processing graph data.

This application discloses a method and system for identifying connection relationships in pipe and instrument diagrams based on graph neural networks. The method includes: acquiring a pipe and instrument flow diagram to be identified; detecting components and connecting lines in the pipe and instrument flow diagram to construct graph structure data; encoding the features of nodes and edges in the graph structure data respectively, and mapping the features of nodes and edges to a high-dimensional latent space to obtain initial feature vectors; inputting the initial feature vectors into a preset graph neural network, using multiple graph convolutional layers for information propagation and aggregation, learning and updating the latent space features of topological relationships between components; based on the updated topological representation features, identifying whether the connection relationships corresponding to each edge in the graph are valid through an edge classifier, and outputting the connection relationship identification result. This invention transforms the connection relationship identification problem into a graph connectivity judgment task, significantly improving the accuracy and robustness of pipe and instrument diagram connection relationship identification in complex scenarios.

The invention relates to a multi-machine-tool thermal error prediction space-time diagram modeling method based on meta-learning, and the method comprises the following steps: data collection: collecting the thermal deformation data of a main shaft and the temperature data of each part of a machine tool at a corresponding moment within a certain time from the starting to the operation of the machine tool at an interval time; then arranging temperature and thermal deformation data into a graph format required by SGAT, dividing a data set into a meta-training set, a meta-test set and a meta-task, and then dividing a training structure into internal circulation and external training by a meta-learning method; the processed graph data is put into an SGAT-Transform, training is carried out in combination with the divided meta training set, meta test set and subtasks, and a complete MAML-SGAT-Transform model capable of adapting to unknown working conditions is obtained; and after the MAML-SGAT-Transform model is trained, rapid adaptation is carried out on a support set of a corresponding meta test set, and the robust performance of modeling under multiple working conditions of a single machine tool and different working conditions of multiple machine tools can be enhanced.