Research on Vehicle Service Simulation Dispatching Telephone System Based on Natural Language Processing

Abstract In view of the current situation that the training process of the existing vehicle simulation training system is not complete and quite different under abnormal circumstances, a vehicle simulation dispatching telephone system based on natural language processing (NLP) is developed. The system uses ASR and TTS, combined with embedded design, to simulate the actual on-site traffic attendant’s voice interaction with each post under abnormal conditions, and secondly recognize the received speech recognition text through NLP which is converted to railway standard terminology for system evaluation. Through the actual case application, it is verified that the system can better identify railway terminology, and ensure that the single person completes the whole process of training and assessment. The system realizes system identification and human-computer interaction, which is important for the efficient and reliable training of the vehicle duty attendant and the expansion of the training system function.