Hybrid Neural Network Modeling Method for Professional Language Understanding of Electric Power Dispatching
In order to improve the accuracy and generalization of power dispatch professional language understanding,a hybrid neural network modeling method for professional language understanding of electric power dispatching is proposed.First,the dispatch profes-sional language is represented as a low-dimensional feature vector based on the pre-trained BERT model.The characterization ability of dispatch professional language is enhanced by fine-tuning the initial BERT parameters.Then the recognition ability of dispatch profes-sional language slot label is improved from the overall situation to access the CRF layer in the neural network.The mapping relationship between scheduling professional language and scheduling intention is trained based on text convolutional neural network(TextCNN).The dispatch professional language understanding is realized by the joint multi-model recognition results.Finally,through the verification of power dispatch professional language of a control center,it is found that the proposed professional language understanding method has higher understanding accuracy,and the mean F1 values for dispatching intention and slot recognition are 3.41%and 3.99%,respective-ly,higher than those of other methods.
power dispatchhybrid neural networkintention recognitionentity recognitionsemantic understanding
万方数据
维普
