Model Reduction for Control-based Continuation of Complex Nonlinear Structures

不断提高工程结构性能的驱动力正在导致越来越轻和更灵活的新设计。在这种新的设计中，分布式和局部非线性的存在导致了广泛的复杂的动态现象，这是非常难以预测的，并可能导致灾难性的故障。基于控制的连续（CBC）是一种测试技术，它结合了传感器，执行器和算法，以智能地探测物理系统的动态行为。使用CBC收集的数据被证明在理解非线性系统响应和开发更精确的数学模型方面非常有价值。虽然CBC的原理是非常普遍的，但由于目前使用的简单反馈控制算法，该方法仍然限于简单的学术应用。通过该项目，CBC的性能将得到改善，其适用性将扩展到更广泛的系统，包括复杂的工程结构。特别是，该项目将开发适用于CBC的模型预测控制（MPC）算法，并利用最先进的非线性模型简化技术。

The constant drive to improve the performance of engineering structures is leading to new designs which are increasingly lighter and much more flexible. The presence of distributed and localised nonlinearities in such new designs leads to a wide range of complicated dynamic phenomena which are extremely difficult to predict and can lead to catastrophic failures. Control-based Continuation (CBC) is a testing technique that combines sensors, actuators and algorithms to intelligently probe the dynamic behaviour of a physical system. The data collected using CBC proved extremely valuable in understanding nonlinear system responses and developing more accurate mathematical models. While the principles of CBC are extremely general, the method remains limited to simple academic applications due to the simplistic feedback control algorithm currently used. Through this project, the performance of CBC will be improved and its applicability extended to a broader range of systems, including complex engineering structures. In particular, the project will develop model predictive control (MPC) algorithms that are suitable for CBC and exploit state-of-the-art nonlinear model reduction techniques.