大规模复杂工程跨尺度关联及耦合的非线性动力学与控制研究

项目名称： 大规模复杂工程跨尺度关联及耦合的非线性动力学与控制研究

项目编号： No.11272333

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 数理科学和化学

项目作者： 郭大蕾

作者单位： 中国科学院自动化研究所

项目金额： 80万元

中文摘要： 多尺度特征是目前大规模复杂工程与社会系统的主要非线性影响因素。在工程控制与信息安全领域，探索跨尺度关联与多尺度耦合引起的演化、跃迁、破坏与崩溃等现象的动力学过程及其作用规律等问题，正在随着计算机与数据处理技术的发展，显示出具有突破性的发展前景。本项目主要研究大规模复杂工程跨尺度关联及耦合的非线性动力学机制与控制。首先，建立多尺度多分辨率模拟模型及多分变模型生成的支持环境与解析工具，进行多尺度多分辨率建模与跨尺度联合建模。其次，研究跨尺度突变过程的信息聚融与检索，揭示多源、异构及冗余数据之间的内在关联与层次，建立多时空尺度的数据聚融与数据驱动控制。最后，通过计算模拟和涌现观察，产生和分析复杂系统行为并进行计算实验，克服复杂系统难以进行实验和试验的困难。通过对跨尺度关联及多尺度耦合等非线性动力学与控制研究，力争为我国大规模、复杂、多场耦合的复杂工程与安全领域的控制与管理提供关键理论与技术支撑

中文关键词： 多尺度计算；尺度关联；计算实验；复杂工程系统；非线性动力学

英文摘要： Multi-scale features the main character of the large and complex projects and social systems of nonlinear factors in the fields of engineering and information security. Increasing with the development of computer and data processing technology, it is bringing forth a breakthrough to explore the evolution and rules of cross-scale correlation and multi-scale coupling，jump, destruction and collapse of the dynamics of the phenomenon process. This project attempts to study the nonlinear mechanism and dynamic control for the large-scale complex engineering with multi-scale correlation and trans-scale coupling. First, a multi-resolution and trans-scale for multi-scale factors is modeling by the terms of supportive environment and analytical tool under the multi-resolution and multi-scale simulation model. Then, for the typical multi-source, heterogeneous and redundant data of nonlinear dynamics systems, this stydy adapts data fusion and retrieval to reveal the intrinsic association rules, so as to establish a method combining of data fusion and data-driven control for multi-spatial and temporal scales. Finally, computational experiments are designed to validate the control performance by the means of emergence simulation and oberservation, which is also to overcome the difficulities that the complex systems isn't easy

英文关键词： Multi-scale computational；Scale association；Computational experiment；Complex engineering systems；Nonlinear dynamics