基于改进自抗扰的永磁电磁混合悬浮型磁浮球控制方法

Permanent magnet electromagnetic hybrid levitation magnetic levitation ball control method based on improved auto disturbance rejection

永磁电磁混合悬浮(PEMS)系统在降低能耗和增大悬浮气隙方面优势明显,但其非线性特征和控制难度也相对增加。针对PEMS系统中未知干扰和输入信号噪声引起的控制性能下降问题,提出一种基于改进型Levent微分器(ILevent)的自抗扰控制(ADRC)方法。首先,结合电磁悬浮(EMS)型磁浮球系统的控制结构,搭建PEMS型磁浮球控制系统试验台,并对PEMS型磁浮球系统进行动力学建模,判断系统的稳定性和能观能控特性;其次,采用双曲正切函数对Levent微分器进行改进,降低输出信号抖振,分析改进前后Levent微分器以及ADRC中TD微分器的输出性能,充分融合Levent微分器对信号噪声的鲁棒性和ADRC对干扰的估计与补偿优势,设计了ILevent-ADRC控制器;并利用带约束因子的模拟退火-粒子群(SA-CFPSO)优化算法有效的解决了ILevent-ADRC参数多、关联性强等制约问题;最后,分别对传统ADRC、PID控制和ILevent-ADRC方法在PEMS型磁浮球控制试验台上进行对比实验,验证所提方法的有效性和优越性。实验结果表明:相比于传统ADRC和PID控制方法,所提ILevent-ADRC方法有效提升了PEMS型磁浮球系统的控制性能,不仅能够有效地抑制测量信号噪声对系统的影响,而且还对系统悬浮气隙大范围变化和运行过程中受到的未知扰动具有更强的适应性和鲁棒性。研究成果为磁悬浮智能控制技术的实际应用提供了理论参考,同时对于永磁电磁混合悬浮这一类复杂的非线性系统的控制,也具有较好的参考价值。

Permanent magnet electromagnetic hybrid suspension(PEMS)system has obvious advantages in reducing energy consumption and increasing suspension air gap,but its nonlinear characteristics and control difficulty are relatively increased.Aiming at the problem of control performance degradation caused by unknown interference and input signal noise in the PEMS system,an auto disturbance rejection control(ADRC)method based on improved Levent differentiator(ILevent)was proposed.Firstly,combining the control structure of the electromagnetic levitation(EMS)type magnetic levitation ball system,the PEMS type magnetic levitation ball control system test bed was built,and the dynamic modeling of the PEMS type magnetic levitation ball system was carried out to judge the stability and the characteristics of the system.Secondly,the hyperbolic tangent function was used to improve the Levent differentiator,reduce the output signal chattering,analyze the output performance of the Levent differentiator before and after the improvement and the TD differentiator in ADRC,fully integrate the robustness of the Levent differentiator to signal noise and the advantages of ADRC to interference estimation and compensation,and design the ILevent-ADRC controller.The simulated annealing particle swarm optimization(SA-CFPSO)algorithm with constraints was used to effectively solve the constraints of ILevent-ADRC,such as multiple parameters and strong correlation.Finally,the traditional ADRC,PID control and ILevent-ADRC methods were compared on the PEMS-type magnetic levitation ball control test-bed to verify the effectiveness and superiority of the proposed methods.The experimental results show that compared with the traditional ADRC and PID control methods,the proposed ILevent-ADRC method effectively improves the control performance of the PEMS-based magnetic levitation system.It can not only effectively suppress the impact of measurement signal noise on the system,but also has stronger adaptability and robustness to the large range change of the levitation air gap of the system and the unknown disturbance in the operation processb.The research results can provide a theoretical reference for the practical application of magnetic suspension intelligent control technology,and also have a good reference value for the control of complex nonlinear systems such as permanent magnet electromagnetic hybrid suspension.