Search

This paper presents a model predictive control for a six-phase permanent magnet synchronous machine (PMSM) with a reduced-dimension cost function. Only the variables in the harmonic subspace are included into the cost function, while the energy conversion related variables are excluded. This is achieved by using the deadbeat direct torque and flux control method to obtain a reference voltage vector and then to track the torque and stator flux properly in the energy conversion related subspace. Then, according to the position of the reference vector, the appropriate prediction vectors can be determined. Subsequently, a reduced-dimension cost function including only the harmonic constraints is defined to evaluate the feasible prediction vectors. In this way, the predictive model and the cost function are simplified without redundant constraints. Meanwhile, the torque and flux can be regulated satisfactorily, and the harmonic currents are suppressed effectively. Finally, experimental results of the proposed method and the conventional model predictive torque control are presented in this paper to verify the validity of the proposed method. [ABSTRACT FROM AUTHOR]

This paper proposes a low-complexity model predictive flux control (MPFC) with current harmonics and torque ripple reduced for a six-phase permanent magnet synchronous machine (PMSM) motor. First, the virtual vectors in two different magnitudes are adopted for the sake of harmonic currents and torque ripple reduction, as well as simplifying the predictive model by eliminating the z1–z2 variables from the cost function. Then, a look-up table is developed to exclude the useless voltage vectors in advance. In this way, the number of prediction vectors is reduced and heavy computation burden is alleviated. Moreover, to avoid the tedious tuning work of the weighting factor in the cost function, the torque and flux amplitude are transformed into an equivalent reference flux vector. Thus, the complexity of the cost function is significantly reduced. Meanwhile, the current and torque performance are highly improved using the proposed method. Also, the fast dynamic response of predictive control is retained. Finally, simulation and experimental results are offered to verify the effectiveness of the proposed MPFC method. [ABSTRACT FROM AUTHOR]

This paper proposes a novel deadbeat current control (DBCC)-based model predictive control for an asymmetrical six-phase permanent magnet synchronous machine. First, the solution of DBCC is adopted to obtain the expected reference voltage vector (RVV). Then, two groups of virtual vectors, in the total number of 24 with different magnitudes, are defined for the sake of current harmonics suppression. Subsequently, two in-phase virtual vectors which are closest to the RVV are selected as the prediction vectors. The next step is to define a cost function which is composed of the error between the RVV and the available prediction vectors. Then, the selected two virtual vectors are evaluated and the one that minimizes the cost function will be applied in the next instant. In this way, only two prediction vectors need to be evaluated and the computation burden is highly alleviated. In the meantime, the weighting factor involved in predictive torque control is avoided. In addition, to achieve the readily implementation with standard pulsewidth modulation switching sequence, 18 virtual vectors are artfully replaced by their corresponding equivalent virtual vectors. Finally, the proposed method is comparatively studied and compared with other benchmark methods. Simulation and experimental results are offered to confirm the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

This paper proposes a model predictive control scheme based remedial control for the single-phase open-circuit fault of an asymmetrical six-phase PMSM machine. The proposed strategy adopts the normal vector space decomposition transformation matrix without reconfiguring the controller topology. The key is to derive the perturbation term of the prediction vectors in α-β and x-y subspace, which is based on the difference of the faulty phase voltage under healthy and faulty condition. Through the compensation of the prediction vectors, the predictive model can be accurately described in the faulty condition. Thus, the current variables in the next instant is accurately predicted and the torque fluctuation is suppressed. Finally, the experimental results are provided to verify the validity of the proposed method on pre- and post-fault tolerant operation. [ABSTRACT FROM AUTHOR]