主动磁轴承是利用电磁力进行悬浮的一种新型轴承，具有无接触、无润滑、无磨损等特点，正越来越多的应用于各种高精度机械领域和极端环境中。但是由于轴承电磁力与电流、位移的近似平方关系，以及存在漏磁、饱和等现象，主动磁轴承系统存在强烈的非线性特征，所以它的控制器设计问题一直是制约其广泛应用的重要原因之一。本文在对主动磁轴承控制理论及方法调研的基础上，通过仿真和实验研究，总结了PID控制器参数设置与控制特性的基本关系， 及PID控制器应用于非线性特征很强的主动磁轴承系统时，一些性能上的不足。为了解决这个问题，开展了模糊控制器设计的研究工作。将模糊控制方法与常规PID控制相结合，利用模糊推理模块实现控制器参数根据转子不同状态进行平滑切换的变参数控制。当转子居于平衡位置的小邻域内时，系统设置较高刚度以承受载荷，减小振动位移响应；当转子受到扰动离开平衡位置的小邻域范围时，系统设置较小刚度，减小转子振荡，获得较大范围内的稳定性。模糊PID控制器很好的实现了系统在小范围内保证高刚度、在较大范围内保持稳定性的目标，这一特性对于主动磁轴承系统而言非常重要。通过仿真与实验研究，初步归纳总结出一套模糊PID控制器的基本设计流程。起浮仿真与实验结果表明，与常规PID控制器相比，模糊PID控制器在转子起浮过程中可以同时获得更小的过冲以及更小的静差；力冲击响应仿真与实验结果表明，大部分情况下模糊PID控制器的抗干扰能力要优于常规PID控制器，模糊PID控制器在保证小范围内高刚度的同时，还可以实现较大范围内的稳定性；位移跟随响应仿真与实验的结果表明，模糊PID在小范围内与常规PID控制器特性基本一致，而在较大范围内则表现出较小的过冲和更好的稳定性。在仿真与实验中还发现，时间延迟环节会对控制器产生比较严重的影响，一旦延迟时间较长，同时采用e和ec进行模糊推理的模糊PID控制器的性能会明显变差，因此在设计控制器时需要对延迟环节进行充分的估计，尽量减小延迟时间对控制器特性的影响。
An active magnetic bearing (AMB) is a novel technology which can be used to support rotors with electromagnetic forces. It has the following advantages—no contact, no lubrication, no wear etc.. So it is increasingly being used in various high-precision machineries or extreme environments. However, since there are strong nonlinear characteristics in an AMB system, including the square relations of the electromagnetic force formula, magnetic flux leakage and saturation etc, it is a difficult task to design a controller for it, and so wider application of AMBs is restricted. Based on a survey on academic theories about AMB control, simulation and experiment research on AMB controllers were carried out. Basic relations between PID controller parameter settings and its features were obtained. The disadvantage of a PID controller when it was applied to an AMB system with strong nonlinear characteristics was discussed. To solve the problem, fuzzy controller design for an AMB system was studied. Combining traditional PID control with fuzzy control, a fuzzy logic block was used to smoothly change parameter settings of a fuzzy PID controller for a AMB rotor according to the rotor states. When the rotor was in a small displacement range near its equilibrium position, high stiffness parameters were used in the controller to take load and reduce vibration response displacement. When the rotor was out of the range because of disturbance, small stiffness parameters are used in the controller to reduce vabrition and keep the rotor stable in a larger displacement range. When fuzzy PID controller was used, the performance required was achieved. Such performance is very important for the AMB system. Based on the simulations and experiments, a design process for the fuzzy PID controller was summarized. The simulations and experiments results during an AMB levitation process showed that: compared with the traditional PID controller, smaller overshoot and smaller static error could be obtained with the fuzzy PID controller. The simulations and experiments results about the force impulse response of the AMB system showed that: the anti-shock ability of the fuzzy PID controller was better than the traditional PID controller in most cases; With the fuzzy PID controller, high stiffness in a small displacement range nearby the equilibrium position and stability in a big displacement range could be achieved at the same time. The simulations and experiments results about the displacement tracking showed that: there were similar characteristics for both the fuzzy and traditional PID controllers in a small displacement range nearby the equilibrium position, but the fuzzy PID controller performed better with smaller overshoot and better stability in a big displacement range. It was also found that time delay in the close loop AMB system affected the characteristics of the controller seriously. When the time delay was exceed a certain limit, the performance of the fuzzy PID controller with two input signal of e and ec would be deteriorated obviously. So the time delay should be considered and minimized during the controller design.
