Research on Key Technologies of Big Data Platform for Operation and Maintenance of Cockpit Based on Neural Network

The operation and maintenance cockpit is becoming increasingly complex with the development of information technology, resulting in a greater amount of operation and maintenance data and higher demands for operation and maintenance. The design, configuration, operation, and monitoring of the operation and maintenance cockpit, as well as its usage and impact on the environment, lead to data imbalance within the operation and maintenance cockpit. O&m cockpit data imbalance refers to the uneven distribution of data of different categories or different labels in O&M data, resulting in too much or too little data of certain categories or labels, which affects the effect and accuracy of O&M data analysis and mining. Despite the significant success achieved by existing graph neural network classification models in the classification task, they can only demonstrate satisfactory classification performance when the number of samples in each category is relatively balanced. Therefore, it is crucial to investigate the application of graph neural networks for unbalanced node classification. In order to address this issue, an integrated graph neural network model is proposed in this paper for the purpose of unbalanced node classification. The graph convolutional neural network based on spectral domain was used as the base classifier to train multiple base classifiers in parallel, and then the Packing ensemble learning method was used to aggregate the prediction results of the multiple base classifiers. Finally, the majority voting method was used to determine the final classification results. The difference of base classifiers mainly comes from the difference of training sets. Each training set is randomly sampled from the original data set. Samples of various types are sampled according to the experimental setting, and then samples of most classes are randomly downsampled according to the imbalance ratio, which is combined with a few classes to form a balanced training set.