欢迎光临《计算力学学报》官方网站!
 
基于知识图谱和GPT模型的可靠性代码自动生成方法
Automatic Code Generation Method for Structural Reliability Analysis Based on Knowledge Graphs and GPT Models
投稿时间:2023-12-28  修订日期:2023-12-28
DOI:
中文关键词:  知识图谱  结构可靠性  GPT  Transformer  代码生成
英文关键词:Knowledge Graph  Structural Reliability  GPT  Transformer  Code Generation
基金项目:
作者单位邮编
向历霓 大连理工大学 116024
李刚* 大连理工大学 116024
李海江 卡迪夫大学 
摘要点击次数: 110
全文下载次数: 0
中文摘要:
      工程结构服役中广泛使用可靠性分析进行结构安全评估,但可靠性分析方法种类多、分析程序代码自动化程度低、复用难,需要研究可靠性代码自动生成方法。生成式预训练转换器(Generative Pre-trained Transformer,GPT)模型已经在大量替代编程手工作业,进行代码自动生成。但是,它在工程领域中的应用受限于可学习数据量小、问题匹配度不高。本文提出了一种结合多种类可靠性知识图谱、基于GPT的代码自动完成模型进行Matlab可靠性代码预测的方法,使用精心设计的源代码预处理降噪策略,以及知识图谱传播模拟密集型任务解释意图;采用条件代码生成训练,有效提升了小数据样本量的学习性能,实现高准确率、问题匹配的可靠性代码自动生成。最后通过三个可靠性知识图谱案例验证了所提方法的有效性。
英文摘要:
      Reliability analysis is widely used in engineering structures for safety assessment, but the variety of reliability methods, low automation of the analysis programming, and the hard reuse require reliability code generation methods. Generative Pre-Trained Transformer (GPT) models have been generously replacing manual programming work by automatic code generation. However, its application in engineering is limited by the small amount of learnable data and the difficulty of problem matching. In this paper, we propose a code prediction method for Matlab reliability analysis by combining multi-category reliability knowledge graphs and a GPT-based code autocompletion model. We used a well-designed source code preprocessing and denoising strategy, and knowledge graphs to transfer simulation intentions. We also employed conditional code generation training. These efforts drastically increase the learning effect of small data size, and enable automatic code generation with high accuracy and problem matching. Finally, the proposed method is verified by three reliability knowledge graph cases.
  查看/发表评论  下载PDF阅读器
您是第12052208位访问者
版权所有:《计算力学学报》编辑部
本系统由 北京勤云科技发展有限公司设计