首页> 外国专利> AUTOMATED KNOWLEDGE EXTRACTION AND REPRESENTATION FOR COMPLEX ENGINEERING SYSTEMS

AUTOMATED KNOWLEDGE EXTRACTION AND REPRESENTATION FOR COMPLEX ENGINEERING SYSTEMS

机译：复杂工程系统的自动知识提取和代表

页面导航

摘要
著录项
相似文献

摘要

Engineering design is a complex and time-consuming process that can be characterized by a series of decisions. An engineering computing system that includes a design application can train or otherwise help engineers in place of an expert. The systems described herein can improve design cycle times, among other technical improvements. In particular, schemas for systems states are dynamically generated and manifolds defining the designs are adaptively learned through online refinement of state embedding models. Patterns in design decisions can be extracted to automate design decisions and predict next decisions that an engineer may take. Such predictions can be achieved by generating feature-based vectorization of designs and processes, which makes the gathered knowledge utilizable in imitation learning. Furthermore, the learning process can be contextualized by encoding the requirements associated with the engineering process, thereby enabling the generation of real-time in-product contextual decision recommendations.

机译：工程设计是一种复杂且耗时的过程，可以通过一系列决策来表征。包括设计应用程序的工程计算系统可以培训或以其他方式帮助工程师代替专家。这里描述的系统可以改善设计循环时间，以及其他技术改进。特别地，系统状态的模式是通过在线细化的状态嵌入模型的在线细化自适应地学习的动态生成和系统状态的模式。设计决策中的模式可以提取以自动化设计决策并预测工程师可能采用的下一个决策。通过产生基于特征的设计和过程的传感化可以实现这种预测，这使得聚集的知识在模仿学习中使用。此外，可以通过编码与工程过程相关联的要求来进行学习过程，从而能够生成实时内部内容决策建议。

著录项

公开/公告号WO2021247831A1

专利类型
公开/公告日2021-12-09

原文格式PDF
申请/专利权人 SIEMENS CORPORATION;GEORGIA TECH RESEARCH CORPORATION;
展开▼

申请/专利号WO2021US35655
发明设计人 RAMAMURTHY ARUN;SRIVASTAVA SANJEEV;MIRABELLA LUCIA;GRUENEWALD THOMAS;JIN HYUNJEE;SUNG WOONGJE;PINON-FISCHER OLIVIA;
展开▼

申请日2021-06-03
分类号G06N5/02;G06N3/04;G06N7;G06N3/08;
国家 US
入库时间 2022-08-24 22:44:53

相似文献

专利
外文文献
中文文献