项目名称: 大柔性高应变航天结构动态展开过程的数据驱动控制研究
项目编号: No.61503234
项目类型: 青年科学基金项目
立项/批准年度: 2016
项目学科: 自动化技术、计算机技术
项目作者: 解杨敏
作者单位: 上海大学
项目金额: 20万元
中文摘要: 大柔性高应变复合结构因其轻质、大变形、构型多样化及高刚度的特点,成为航天领域公认的在可展开结构研制上具有极高应用价值的发展方向之一,然而高度非线性和难以解析建模的动力学特点使现阶段对其展开过程的主动控制研究相对空白。本项目提出一套实现高应变结构展开过程动态精确控制的动力学分析和控制方法。为减少模型不确定性的影响,创新性地利用非线性曲度对展开过程进行动力学分析和优化分段;并对单一区域采用前馈-反馈并联控制结构保证系统鲁棒稳定性、抗干扰性能及运动轨迹跟踪性能;最后使用信息融合切换控制技术确保系统在不同区域间切换时控制的平滑过渡。控制系统整体设计完全基于实验测试数据而不依赖于解析模型,实现了高应变结构非线性展开过程无模型的展开控制。本项目所有分析、设计及仿真结果均将在实验平台上进行验证,进行高应变复合结构的受控平滑展开,其项目成果将为大尺寸大柔性可展开航天结构的研制提供理论基础及技术支撑。
中文关键词: 数据驱动控制;可展开航天结构;迭代学习控制;展开过程控制;高应变复合结构
英文摘要: Highly flexible high strain composites have been widely recognized as one of the most valuable R&D for the development of space deployable structures due to its advantages in lightweight, large deformation, diverse designs and high stiffness. However, resulted by its highly non-linear dynamics and being lack of analytical model, research on active control of its deployment process is still insufficient. This project aims to accomplish an analysis and control toolset for precise and smooth control of the deployable structure. To reduce the impact from model uncertainty, the project creatively optimizes the segmentation of the deployment process by its nonlinear curvature value, utilizes feedforward-feedback control to ensure robust stability, disturbance rejection performance and tracking ability, and uses signal blending to realize smooth transitions among sections during the deployment process. The design of the control system is purely data-driven and independent of any analytic model, which achieves modeless control for the deployment process of the highly flexible high strain composite. All of the analysis, design and simulation results in this project will be verified on an experimental system to realize controlled smooth deployment process, which will ultimately provide theoretical and technical supports for the development of large flexible deployable space structure.
英文关键词: Data-driven based control;Deployable space structure;Iterative learning control;Deployment process control;High strain composites