Muñoz Jorge, Monje Concepción A, Nagua Luis F, Balaguer Carlos
University Carlos III of Madrid, RoboticsLab, Avenida de la Universidad 30, 28911 Leganés, Madrid, Spain.
University Carlos III of Madrid, RoboticsLab, Avenida de la Universidad 30, 28911 Leganés, Madrid, Spain.
ISA Trans. 2020 Oct;105:296-307. doi: 10.1016/j.isatra.2020.05.045. Epub 2020 May 30.
Fractional order controllers are widely used in the robust control field. As a generalization of the ubiquitous PID controllers, fractional order controllers are able to reach design specifications their integer counterparts cannot, and as a result they outperform them at particular situations. Their main drawback is that generalization of the design tools is not always evident, and therefore tuning this kind of controller is always a new and different challenge. Existing methods often use numerical computation to find the controller parameters that fit the specifications. This paper describes a graphical solution for fractional order controllers, which avoids the solution by nonlinear equations and helps designer to solve the control problem in a very intuitive way. This approach is tested in the servomotors of a real bio-inspired soft neck and results are compared with those obtained from other control strategies. The experiments show that the controller tuned by this method works as expected from a robust controller and that this approach is very competitive compared to other state of the art methods, while offering a more simplified and direct tuning process.
分数阶控制器在鲁棒控制领域中得到了广泛应用。作为无处不在的PID控制器的推广，分数阶控制器能够达到其整数对应控制器无法达到的设计规格，因此在特定情况下表现优于它们。其主要缺点是设计工具的通用性并不总是很明显，因此调整这种控制器始终是一个全新的、不同的挑战。现有方法通常使用数值计算来找到符合规格的控制器参数。本文描述了一种分数阶控制器的图形化解决方案，该方案避免了通过非线性方程求解，并帮助设计者以非常直观的方式解决控制问题。这种方法在一个真实的仿生软颈的伺服电机中进行了测试，并将结果与其他控制策略获得的结果进行了比较。实验表明，用这种方法调整的控制器如预期的那样作为一个鲁棒控制器工作，并且与其他现有技术方法相比，这种方法具有很强的竞争力，同时提供了一个更简化和直接的调整过程。
