A digital twin factory construction method based on generative AI and causal inference

By using generative AI and causal inference technology, the infrastructure for a digital twin factory is automatically constructed, solving the problems of fragmented design processes and insufficient analysis of potential coupling effects in existing technologies, and achieving efficient and safe factory design and production optimization.

CN122389490APending Publication Date: 2026-07-14SEC ZHILIAN TECH (JIANGSU) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
SEC ZHILIAN TECH (JIANGSU) CO LTD
Filing Date
2026-05-26
Publication Date
2026-07-14

AI Technical Summary

Technical Problem

Existing digital twin factory construction technologies suffer from fragmented design processes, long iteration cycles, low efficiency, difficulty in deeply analyzing potential physical field coupling effects, and a lack of advanced semantic understanding and automated generation capabilities, leading to design defects and production interruption risks.

Method used

By employing generative AI and causal inference technologies, an initial 3D physical layout diagram is generated by analyzing the text of process capacity requirements. This constructs a digital twin factory infrastructure that is spatially compliant and thermodynamically and electrically compliant. Causal models are used to identify and avoid potential coupling effects, thereby achieving automated design optimization.

Benefits of technology

It has improved the intelligence and foresight of factory design, shortened the design cycle, enhanced the robustness and compliance of the design, and ensured high-precision production and safe operation of the factory.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122389490A_ABST
    Figure CN122389490A_ABST
Patent Text Reader

Abstract

The present application relates to the technical field of computer, and relates to a digital twin factory construction method based on generative AI and causal inference; the present application firstly analyzes process capacity demand text through a generative large language model to generate an initial three-dimensional physical layout directed graph; spatial cross conflict testing is performed by using a causal graph conditional independence theorem to generate a three-dimensional decoupling infrastructure model to eliminate physical interference; high-heat and high-precision processing equipment nodes are extracted to construct a structure causal model containing heat conduction paths, and a causal intermediary analysis is used to quantify indirect thermodynamic deformation effects, and a targeted heat insulation facility is generated when the standard is exceeded; a causal dependence network atlas of equipment start-stop timing and bus voltage drop is established by using a causal discovery algorithm, a root cause equipment cluster is identified through counterfactual evaluation, and finally a variable power distribution compensation wiring scheme is generated and fused to form a multi-physical field rule digital twin factory infrastructure base.
Need to check novelty before this filing date? Find Prior Art