Research on underground engineering collaborative design platform based on “cloud + terminal” architecture and LSTM-MVCNN algorithm

In the field of underground engineering, the construction model of two-dimensional design and decentralized management is no longer able to enhance the efficiency trend of multi-party collaboration and information management, nor can it meet the future demand for refined management and sustainable development of engineering construction. Therefore, this paper first summarizes and analyses the current development status of collaborative design platforms at home and abroad, and proposes a “cloud + terminal” overall technical architecture suitable for underground engineering collaborative design management. This architecture consists of a foundation layer, data layer, service layer, functional application layer, and user interaction layer. The functional modules of the platform also adopt microservice architecture design. In response to the shortcomings of model checking, an IFC component of the model checking module was designed based on an optimized multi view convolutional neural network algorithm to correct errors in the lowest level category information data. At the same time, technical routes were proposed around “three zone independence” and “problem positioning.” Developed a prototype platform with basic functions and applied it to the design process of a certain engineering project. By comparing with traditional design patterns, the results show that the collaborative platform proposed in this paper effectively improves work efficiency and achievement quality. Next, we will continue to expand the functions and application scenarios of the collaborative platform, providing strong support for improving the construction quality and efficiency of underground engineering throughout its entire life cycle.