标题： 通过最小化无多个闭环测试的时间加权绝对误差积分的分数阶控制器整定 显示英文标题

标题： Fractional-order controller tuning via minimization of integral of time-weighted absolute error without multiple closed-loop tests

摘要： 本研究提出了一种非迭代的线性分数阶（FO）控制器整定技术，该技术基于时间加权绝对误差积分（ITAE）准则。 最小化ITAE是整定FO控制器的传统方法。 该技术减少了超调和欠调，并抑制了稳态误差。 与基于ITAE的常规控制器整定方法不同，所提出的方法不需要进行多次闭环实验或基于模型的仿真来评估ITAE。 一次性输入/输出数据是从被控对象中收集的。 在所收集的输入和输出信号的基础上定义了一个虚拟参考信号，这使我们能够评估任意控制器参数提供的闭环响应。 为了避免常规方法中必要的重复实验，我们使用虚拟参考信号重新表述了ITAE最小化问题。 通过求解基于虚拟参考信号的优化问题，获得使ITAE最小化的期望FO控制器参数。 数值研究证明了所提出方法的有效性。 避免重复实验显著降低了线性FO控制器的开发成本，从而促进了它们的实际应用。 显示英文摘要

摘要： This study presents a non-iterative tuning technique for a linear fractional-order (FO) controller, based on the integral of the time-weighted absolute error (ITAE) criterion. Minimizing the ITAE is a traditional approach for tuning FO controllers. This technique reduces the over/undershoot and suppresses the steady-state error. In contrast to conventional approaches of ITAE-based controller tuning, the proposed approach does not require multiple closed-loop experiments or model-based simulations to evaluate the ITAE. The one-shot input/output data is collected from the controlled plant. A fictitious reference signal is defined on the basis of the collected input and output signal, which enables us to evaluate the closed-loop response provided by the arbitrary controller parameters. To avoid repeated experiments that are necessary in the conventional approach, we reformulate the ITAE minimization problem using the fictitious reference signal. The desired FO controller parameters minimizing the ITAE are obtained by solving the optimization problem that is based on the fictitious reference signal. The validity of the proposed approach is demonstrated by a numerical study. The avoidance of repeated experiments significantly reduces the development cost of linear FO controllers, thereby facilitating their practical application.