磁悬浮开关磁阻电机是一个复杂的非线性强耦合系统,且运行过程中容易出现磁饱和现象,增大了数学模型建立及解耦控制的难度。针对上述问题,在利用有限元方法分析其磁场及电磁力特性的基础上,计算了一种对电机磁路线性及饱和状态均适用的新数学模型。分析了系统的可逆性,采用神经网络逆实现了转矩和两自由度径向力的解耦。使用dSPACE系统试验验证了该方法的正确性和有效性,可以弥补现有基于无磁饱和假设的各种建模及相应的解耦控制方法不适用于BSRM磁饱和工况的缺陷,也可以为电机的运行特性分析、本体优化设计以及控制策略研究提供更准确的理论依据。
In view of complicated non-linearity,coupling,and magnetic saturation,it is very difficult to gain an accurate mathematic model and realize decoupling for a bearingless switched reluctance motor(BSRM).So after analyzing the magnetic field and force characteristics with finite element method,a novel mathematical model was computed.This model could be fit for both linear and saturated state,and even meet reversible requirement.Then a neural network inverse model was established to decouple BSRM.Lastly,three closed-loop controllers were designed for pseudo-linear systems.Experimental results based on dSPACE system validated this method.It can remedy the shortcomings of those existing decoupling means based on non-saturation hypothesis,which are not suitable for magnetic-saturated work condition,and can provide more reliable theoretical bases for running state analysis,motor design,and control strategy study.