基于分段参考模型的大行程双边直线开关磁阻电机模型参考自适应速度控制OA

Linear switched reluctance motors(LSRMs)exhibit broad application potential in industrial automation,wave energy conversion,and rail transportation,due to the absence of permanent magnets,robust mechanical structure,low heat consumption,high reliability,and low manufacturing cost.To mitigate disturbance forces,suppress electromagnetic force ripple,and enhance motor efficiency,double-sided LSRMs(DLSRMs)have been developed and have become a key focus of research.However,the dynamic model of DLSRMs is subject to parameter mismatches due to strong nonlinearity,time-varying parameters,and uncertain disturbances,significantly limiting high-performance motion control of DLSRMs.Model reference adaptive control(MRAC)has proven effective at handling uncertainties and improving system robustness.Therefore,this paper proposes a model-reference adaptive speed control method for long-stroke DLSRMs to improve their motion performance during long-stroke operation. A composite strategy combining speed feedforward control and proportional-integral(PI)speed feedback control is adopted to construct the adjustable-speed control system for the long-stroke DLSRM.Differential equation models for both the adjustable motor system and its reference model are established.Then,a generalized position error is defined as the difference between the output positions of the adjustable motor system and those of the reference model.Using this error,a Lyapunov function is formulated.Based on Lyapunov stability theory,the speed-tracking problem is transformed into an asymptotic-convergence problem for a generalized state vector,and an adaptive control law for the long-stroke DLSRM is derived.Considering that the practical operation of the DLSRM involves acceleration,constant speed,and deceleration phases,each with different performance requirements,piecewise reference models are designed for the constant-speed and acceleration/deceleration phases.The stable parameter ranges for the piecewise reference models are determined.Furthermore,the initial values of the adaptive control parameters are set according to the matching criterion between the adjustable motor system and its reference model. Experimental results show that under a loading condition equivalent to 1.5 times the mass of the moving platform,the speed ripples achieved by the proposed method are 0.82%at 300 mm/s and 3.09%at 100 mm/s.Compared with a non-adjustable-parameter control method that combines speed feedforward and PI feedback,the speed ripples during constant-speed operation are reduced by 3.53%and 2.83%,respectively.Compared with an adjustable-parameter fuzzy PI control method,the reductions are 38.35%and 42.13%,respectively.Moreover,the adjustable-parameter fuzzy PI method exhibits smaller maximum absolute speed errors during acceleration and deceleration phases under various load and speed conditions.The proposed method achieves the lowest average maximum absolute speed error during the steady-state constant-speed phase.Thus,the proposed method effectively reduces overall speed ripple and demonstrates strong robustness against load disturbances in long-stroke DLSRMs.Since speed ripple is the critical performance indicator for constant-speed operation in linear motors,rather than maximum absolute speed error,the proposed method is well-suited for long-stroke DLSRM speed-control applications.