Techniques for Nonlinear Real-time Hybrid Testing

实时混合测试是一种强大的实验技术，但由于控制回路的延迟补偿问题，它目前还不能用于大多数实际工程应用。这项工作旨在利用前向预测和自适应调整系统机械性能的新组合来解决这一挑战。在商业环境中利用该方法将提高质量并缩短上市时间，在学术环境中，它将为广泛的领域提供有价值的实验工具。实时混合测试（RTHT）是一种在实验室中对大型或复杂结构部件进行实验的技术。该组件实际存在于实验室中，但系统的其余部分仅存在于计算机模拟中。马达、执行器和传感器在两者之间建立了一个虚拟的链接，因此它们作为一个系统运行。这项技术可以大大节省工业发展计划的时间和成本，并提供一种独特的强大的研究工具，但其潜力目前受到连接系统两部分的电机和执行器的延迟的限制。在许多实际工程系统中，如汽车减震器、飞机控制机构和用于设计安全建筑和桥梁的材料，最先进的延迟补偿技术在存在强非线性的情况下仍然无效。该项目解决了对常见非线性部件进行实时混合测试的困难，以实现RTHT目前尚未开发的潜力。采用了一种新颖的方法，从人体用于补偿神经信号传输延迟的机制中获得灵感。采用了两种策略的组合：对所需的力和位置的前向预测，以及对系统的机械特性进行调整以补偿这些力和位置预测中的误差。两个关键的研究问题需要解决，首先是调整系统以匹配所需机械性能的能力，其次是改进RTHT能力的程度。确定了三个目标，围绕该配置的实验调查和其性能的评价：将进行系统配置的分析研究，以确定适当的组件尺寸；构建系统，评估力学性能调整的有效性；最终目标是进行完整的RTHT实验，以确定结果的保真度和稳定性的改进程度。想要进行混合测试的原因有很多。他们已经取得显著成功的一个领域是大型土木结构的地震测试。这里的动机有三个方面：首先，像建筑物或桥梁这样的大型结构不能装进实验室进行整体测试。其次，地震中所经历的条件不容易复制到整个建筑物中。第三，实验性测试可能有意或无意地导致失败，在这种情况下，这将是危险和昂贵的。其他例子包括模拟零重力下的卫星对接，在没有飞行测试的危险或费用的情况下测试航空设备，以及在整车制造之前测试汽车零部件。这些领域的企业和研究人员正准备利用新技术，这将极大地扩展可以从实时混合测试中受益的系统类别。从提高生产力和降低成本方面对企业的好处，以及从更好的研究工具方面对学术界的好处，将转化为最终用户和公众在提高产品质量、提高生活水平和改善安全方面的好处。

Real-time hybrid testing is a powerful experimental technique but it cannot presently be used for a majority of practical engineering applications because of problems with compensating for delays in the control loop. This work seeks to address this challenge using a novel combination of forward prediction and adaptive tuning of the mechanical properties of the system. Exploitation of the method in commercial settings will lead to improved quality and reduced time-to-market, and in an academic context it will offer a valuable experimental tool across a breadth of fields.Real-time hybrid testing (RTHT) is a technique for performing experiments on components of large or complex structures in a laboratory. The component is physically present in the laboratory, but the rest of the system exists only in a computer simulation. Motors, actuators, and sensors create a virtual link between the two so they behave as one system. The technique can offer radical time and cost savings in industrial development programmes and provides a uniquely powerful research tool, but its potential is currently limited by delays in the motors and actuators linking the two parts of the system. State-of-the-art delay compensation techniques remain ineffective in the presence of strong nonlinearities, as found in many practical engineering systems such as automotive shock absorbers, aircraft control mechanisms, and materials used to design safe buildings and bridges. This project addresses the difficulties in performing real-time hybrid tests of common nonlinear components, to allow the realisation of the currently untapped potential of RTHT.A novel approach is taken, drawing inspiration from the mechanisms used in the human body to compensate delays in nerve signal transmissions. A combination of two strategies is employed: forward prediction of the forces and positions required, and tuning of the mechanical properties of the system to compensate for errors in those force and position predictions. Two key research questions are to be addressed, concerning firstly the ability to tune the system to match the desired mechanical properties, and secondly the extent of the improvements this makes to the RTHT capabilities. Three objectives are identified, centring on the experimental investigation of this configuration and the evaluation of its performance: analytical studies of the system configuration will be conducted to determine appropriate component sizing; the system will be built and the effectiveness of tuning of the mechanical properties will be evaluated; the final objective is to conduct full RTHT experiments to determine the extent of improvements to fidelity and stability of the results. There are a variety of reasons for wanting to conduct hybrid tests. One field where they have established marked success is earthquake testing of large civil structures. The motivation here is threefold: firstly, large structures such as buildings or bridges cannot be fitted into a laboratory for testing in their entirety. Secondly, the conditions experienced in an earthquake are not readily reproduced for an entire building. Thirdly, experimental testing may deliberately or inadvertently lead to failure, which in this case would be dangerous and expensive. Other examples include simulations of satellite docking in zero gravity, testing of aeronautical equipment without the danger or expense of flight testing, and testing of automotive components before the full vehicle has been manufactured. Businesses and researchers in these fields are poised to take advantage of new techniques, which will greatly expand the classes of system which can benefit from real-time hybrid testing.The benefits to businesses in terms of increased productivity and reduced costs, and academics in terms of better research tools, will translate to end users and the general public in terms of better quality of products, higher standards of living and improved safety.