Artificial intelligence distribution network scheduling system based on voice recognition

• Main Authors: HE ZHIPEI, CHEN ZHIWEI, ZHU ZHUOWEN, ZHANG HONG, LUO KANGDING, LUO DONGQIAO, QIU ZEJIAN, YU LILI, ZHANG GUANZHOU, LIU SHUXIN, HAN ZHANYUAN, HUANG DAOU, CHEN FENGCHAO
• Format: Patent
• Language: Chinese; English
• Published: 28.06.2024
• Subjects: ACOUSTICS; MUSICAL INSTRUMENTS; PHYSICS; SPEECH ANALYSIS OR SYNTHESIS; SPEECH OR AUDIO CODING OR DECODING; SPEECH OR VOICE PROCESSING; SPEECH RECOGNITION

The invention discloses a voice recognition-based artificial intelligence distribution network scheduling system, which comprises a first voice recognition module and a second voice recognition module, and is characterized in that the first voice recognition module captures a dependency relationship of voice data of a voice acquisition module by adopting a Bi-LSTM model, extracts intra-frame and inter-frame features of voice, and sends the extracted intra-frame and inter-frame features to the voice acquisition module; outputting a fuzzy mandarin scheduling text and a fuzzy dialect scheduling text corresponding to the dispatcher; and the second voice recognition module is used for performing deep processing on word meanings of the fuzzy mandarin scheduling text and the fuzzy dialect scheduling text by adopting a BERT model and performing iterative prediction calculation on an interactive processing result of the two texts by adopting a multi-layer perceptron, so that semantic consistency of the two texts is me