TYJL-III Computer Interlocking Equipment Training System Key Technology

This paper carries out key technology research on TYJL-III computer interlocking equipment training system, aiming to improve the training efficiency and quality of railway signal maintenance personnel through augmented reality, natural language modelling and other new technologies. Firstly, the overall scheme of the system based on augmented reality is proposed, which solves the problems of unclear recognition of the equipment, lack of understanding of the internal structure, and incomplete mastery of the fault situation in the traditional practical training. Secondly, the 3D structure recovery and reconstruction technology of the equipment was studied to achieve high-precision modelling of the equipment through multi-view images and depth information, and to provide data support for position tracking. Further, a real-time tracking and registration technique for virtual and real environments is developed to optimise the multimodal sensor data fusion by combining deep learning and attention mechanism, which significantly improves the accuracy and stability of position estimation. In addition, the human-computer intelligent interaction technology based on natural language model is introduced to enhance the naturalness and real-time interaction in the practical training process. The experimental results show that the optimised system performs well in the position estimation accuracy, real-time performance and fault simulation test, and significantly improves the operation efficiency, accuracy and safety compared with the traditional method, while reducing the training cost. Despite the achievements, the system still needs to be further improved in terms of the comprehensiveness of fault simulation scenarios, compatibility with other railway signalling systems and user experience.