磁力轴承的智能控制研究

发布时间：2018-04-21 13:56

本文选题：磁力轴承 + PID控制　；参考：《武汉理工大学》2012年硕士论文

【摘要】：磁力轴承是一种新型的支承部件,它具有传统的滚动轴承和滑动轴承轴承无法比拟的优点,如无机械接触、无磨损、无需润滑等,一直以来都受到国内外学术界和工业领域的广泛关注。磁力轴承本质上是非线性系统,具有开环不稳定的特点,需要设计合理的控制器进行闭环控制。常用的控制器设计方法是在小范围内将磁力轴承非线性模型线性化,然后根据线性理论设计经典PID控制器。研究表明,经典PID控制不能达到理想的控制效果,需要研究和设计更优良的控制器。本文针对单自由度磁力轴承系统进行了分析,介绍了其工作原理,并根据电磁理论建立了其线性和非线性模型,并采用经典PID控制对两种模型进行了仿真研究。实验结果表明在小范围内将磁力轴承线性化是合理可行的,但经典PID控制的超调量大,调节时间长。针对经典PID控制效果不理想的情况,本文引入了智能控制方法,将模糊控制和神经网络控制应用于磁力轴承系统中。本文进行了模糊控制器和模糊查找表的设计,采用查表法对磁力轴承系统进行了基本模糊控制和参数自调整模糊PID控制的仿真研究。仿真结果表明,基本模糊控制和参数自调整模糊PID控制都能改善磁力轴承的性能,但前者存在静态误差,后者对性能的提高不明显。利用神经网络的非线性映射能力,本文设计了磁力轴承的PID神经网络单变量控制器,讨论了其控制算法,分析了其稳定性。仿真结果表明,PID神经网络单变量控制的效果优良,与经典PID控制和参数自调整模糊PID控制相比,其超调量小,调节时间短；与基本模糊控制相比,其无静态误差。
[Abstract]:Magnetic bearing is a new type of supporting parts, it has the advantages that traditional rolling bearing and sliding bearing can not compare, such as no mechanical contact, no wear, no lubrication, etc. All along has been the domestic and foreign academic circles and the industrial domain widespread concern. Magnetic bearing is essentially a nonlinear system with the characteristics of open-loop instability, so it is necessary to design a reasonable controller for closed-loop control. The commonly used controller design method is to linearize the nonlinear model of magnetic bearing in a small range, and then to design the classical PID controller according to the linear theory. The research shows that the classical PID control can not achieve the ideal control effect, so it is necessary to study and design a better controller. In this paper, the single-degree-of-freedom magnetic bearing system is analyzed, its working principle is introduced, its linear and nonlinear models are established according to electromagnetic theory, and the two models are simulated by classical PID control. The experimental results show that it is reasonable and feasible to linearize the magnetic bearing in a small range, but the classical PID control overshoot is large and the adjustment time is long. In view of the unsatisfactory control effect of classical PID, the intelligent control method is introduced in this paper. Fuzzy control and neural network control are applied to the magnetic bearing system. In this paper, the design of fuzzy controller and fuzzy lookup table is carried out. The basic fuzzy control and parameter self-adjusting fuzzy PID control of magnetic bearing system are simulated by the look-up table method. The simulation results show that both the basic fuzzy control and the parameter self-adjusting fuzzy PID control can improve the performance of the magnetic bearing, but the former has static error, but the latter has not obviously improved the performance. Based on the nonlinear mapping ability of neural network, the PID neural network single variable controller of magnetic bearing is designed, its control algorithm is discussed and its stability is analyzed. The simulation results show that the single variable control of pid neural network is effective, compared with the classical PID control and parameter self-adjusting fuzzy PID control, its overshoot is small and its adjusting time is short, and compared with the basic fuzzy control, it has no static error.
【学位授予单位】：武汉理工大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TH133.3;TP273

【相似文献】