一种新型模糊神经网络滑模控制器设计

Design of a New Fuzzy Neural Network Sliding Mode Controller

针对基于PID算法的永磁同步电机矢量控制系统易受系统不确定性影响(负载扰动、模型参数变化)的问题,提出一种新型模糊神经网络滑模控制器。首先,在传统滑模控制的基础上,利用模糊神经网络控制器(FNN)构成滑模控制器(SMC)的切换控制项来完成对系统不确定性因素进行逼近以及对滑模控制切换增益的调节;其次,利用麻雀搜索算法来实时更新模糊神经网络滑模控制器参数,为了加快SSA的收敛速度以及防止其陷入局部最优,利用分数阶微积分对传统麻雀算法进行改进;最后,进行了仿真验证,表明所设计的新型模糊神经网络滑模控制器具有较强的鲁棒性和动态性能。

Aiming at the problem that permanent magnet synchronous motor vector control system based on PID algorithm is vulnerable to system uncertainties(load disturbance,model parameter change),a new fuzzy neural network sliding mode controller is proposed.Firstly,on the basis of traditional sliding mode control,fuzzy neural network controller(FNN)is used to form the switching control term of sliding mode controller(SMC)to complete the approximation of system uncertainty factors and the adjustment of switching gain of sliding mode control.Secondly,the sparrow search algorithm is used to update the parameters of the fuzzy neural network sliding mode controller in real time.In order to accelebrate the convergence speed of SSA and prevent it form falling into local optimum,the traditional sparrow algorithm is improved by fractional calculus.Finally,the simulation results show that the designed new fuzzy neural network sliding mode controller has strong robustness and dynamic performance.