智能变体结构作为未来先进无人飞行器等装备研制的关键技术,其分布式主动变形结构可实现光滑连续与多自由度变形,是显著提升结构性能与任务适应性的有效手段。针对这一需求,提出了一种基于形状记忆合金驱动的智能点阵结构的创新设计与控制方案。首先,提出的拟热变形法可用于高效评估SMA 驱动器的变形性能,通过仿真与试验验证该方法对智能点阵结构的变形性能分析具有5% 以内的误差精度,并成功实现了其结构的多模式可控变形。进一步构建了以能耗优化为目标、变形精度为约束的分布式驱动设计模型,在翼型结构应用中仅需16.67%的全局能量即可实现8 个控制点400 mm 的高精度变形(误差1%)。针对大规模结构的实时控制难题,采用BP神经网络实现了多自由度变形的精确预测与控制,该方法具有突出的普适性,可拓展至多种形式的SMA 驱动形式及复合翼面等智能结构设计,为兼具力学性能与智能变形的新一代智能变体结构系统提供了新的解决方案。
Abstract
Smart morphing structures represent a key technology for future advanced equipment such as unmanned aerial vehicles
where distributed active deformation structures enable smooth
continuous
and multi-degree-of-freedom shape changes
serving as an effective mean to significantly enhance structural performance and mission adaptability. Addressing this need
this study proposes an innovative design and control scheme for shape memory alloy (SMA)-driven smart lattice structures. Firstly
the proposed pseudo-thermal deformation method (PTDM) provides an efficient approach to evaluate deformation performance of SMA actuators. Simulation and experimental verification demonstrate that this method achieves deformation analysis accuracy within 5% error for smart lattice structures. It successfully realizes multi-degreeof-freedom controllable shape morphing. Furthermore
an optimized distributed actuation design model was developed with energy consumption minimization as the objective function and deformation precision as constraint condition. In a wing structure application
this model achieved high-precision 400 mm deformation at 8 control points with less than 1% error while consuming only 16.67% of the global energy. To address real-time control challenges in large-scale structures
a BP neural network was employed to achieve precise prediction and control of multiple degrees of freedom deformation. The proposed method exhibits remarkable versatility
being extendable to various SMA actuation configurations and smart structural designs like composite wings
providing a new solution for next-generation smart morphing structural systems that integrate both mechanical performance and smart deformation capabilities.