Research on Feature Extraction of Power Grid Technical Transformation Project Based on NLP-LSTM

The power grid production technology transformation project is an important guarantee task for ensuring the safe and reliable operation of the power grid. The power grid technical renovation projects under different project types and voltage levels also have their own characteristics. Therefore, the factors related to various power grid technical renovation projects are very complex and diverse, making the design of evaluation index systems for these projects very difficult. The obtained index systems often fail to accurately reflect the characteristics of cost management and evaluation of technical renovation projects, and cannot guarantee the objectivity and accuracy of evaluation results. This paper proposes to use the principle of natural language processing technology, combined with word segmentation technology to classify the existing technical innovation projects, and use LSTM neural network combined with the existing evaluation index set of the whole process cost management to screen the key index features, finally forming a relatively simplified evaluation index system. Through analysis, it is found that the feature extraction method proposed in this paper is highly accurate, and the feature extraction results are reasonable.