基于 DAG 的画布与表达式双向转写研究

打开文本图片集

DOI：10.19981/j.CN23-1581/G3.2026.12.040

中图分类号：TP311.13

文献标志码：A

文章编号：2095-2945（2026）12-0160-05

Abstract： In the process of building mechanism models such as power generation equipment status monitoring and hidden danger warning， in order to improve the real-time visibility of the model modeling and calculation process， class configuration building tools such as operators， canvases， and connections are often used， but they are complex formulas. There is a sense of fragmentation in the logical presentation. This research proposes a method for identifying and verifying canvas elements based on directed acyclic graphs （DAG） to realize the forward transfer from the visualization mechanism model to mathematical expressions; in the reverse transfer process， the three-layer logical relationship of Token-AST-DAG is constructed. Finally， operators and operation relationships are restored in the canvas. Experimental verification shows that this method supports bidirectional transcriptions of most mechanism models， and the error rate is less than 2%.

Keywords： DAG; directed acyclic graph; canvas model to mathematical expression; mathematical expression to canvas model; bidirectional transcription; power generation equipment monitoring and early warning

发电设备因其固有的监测逻辑与故障隐患预警经验，在实时生产监测环节往往采用机理建模，因传统依赖数学表达式的建模方法学习成本高、建模难度大、维护成本高、可视化程度低等劣势，近年来在html5、VUE、canvas等技术加持下，涌现了众多可视化机理建模工具，不同厂商虽有不同侧重点，但基本都是采用画布+算子+有向关联线的方式构建计算逻辑[1]。（剩余6607字）