The introduction of disturbance models into model predictive controllers using harmonic inputs for energy regenerative vibration control
The introduction of disturbance models into model predictive controllers using harmonic inputs for energy regenerative vibration control
复制标题
DOI：
10.1299/transjsme.14-00545
复制
发表时间：
2015
期刊：
影响因子：
--
通讯作者：
T. Nakahara;T. Fujimoto
中科院分区：
文献类型：
--
作者：
T. Nakahara;T. Fujimoto
提出了一种能量再生型主动振动控制系统的模型预测控制方法。该方法的控制输入由谐波函数的线性和组成，减少了计算量，并为预测时间平均再生功率提供了一种简单的方法。然而，由于在预测过程中忽略了干扰输入，预测的平均功率往往小于实际的平均功率。本文提出了一种改进的预测控制器，该控制器采用谐波干扰模型和最小阶观测器来估计和考虑预测过程中的干扰输入。用该控制器对单自由度系统进行了数值仿真，结果表明，在以下情况下，改进后的控制器可以提高时间平均再生功率的预测质量。这些情况下，干扰的类型是正弦输入，其频率是SDOF系统的固有频率或随机输入，其幅度足够小，以满足控制输入约束。
The authors proposed a model predictive control method for energy regenerative active vibration control systems. The control input of this method is composed of a linear sum of harmonic functions to reduce computational costs and this approach gives a simple way to predict the time averaged regenerated power. However, the predicted averaged power tends to be smaller than the actual averaged power because the disturbance input is ignored in the prediction process. In this paper, a modified predictive controller that uses a harmonic disturbance model and a minimal order observer to estimate and consider the disturbance input in the prediction process is proposed. Numerical simulations of a single degree of freedom (SDOF) system with the proposed controller show that the modified controller can improve the prediction quality of the time averaged regenerated power in the following cases. These are the cases that the type of the disturbance is sinusoidal input whose frequency is the natural frequency of the SDOF system or random input whose amplitude is small enough to satisfy control input constraint.