Efficiency Aspects of Vector Contro l Applied to Synchronous Reluctance Motors John E. Fletcher, Barry W. Williams and T im C. Green Department of Com puting and Electrical Engineering Heriot-Watt University Scotland, UK EH14 4 A S Abstra ct-Cor e losses in a synch ronous reluctance machin e are modelled. The empirical model obtained is used to implement a control scheme to compensate for equivalent core loss currents. This enables accurate control of the magnetising currents, henc e torque . Efficie ncy over the base speed operating range of th e machine is compared for tw o different yector control schemes. Methods of triplen series injection for vector controllers are discussed and a new method proposed. The new technique h as advanta ges in terms of overall trip len content and computational require ments. Inducta nce ripple in the machine is estimated using direct torque ripple measurements and incorporated in a machine model val id for low speed operation. I. INTRODUCTION F o r a given frame size, the maximum power output from a synchronous reluctance machine is similar to that of the induction m ach ine[l]. In addition, the rotor d-q axes are well defined and rotor losses are lower. Tor que control is per formed by regulation of the machin e current vector. Previous publications have directed their studies towards optimising transient torqu e performance[2] [3] and power factor [3]. By virtue of the torque equation of the machine, other than at full load torque, there is a degree of freedom in the choice of current vector applied for a specific output torque. This allows optimisation of a specific performance parameter of the drive This paper investigates issues regarding the steady state efficiency for a particular synch ronous reluctance machine. An empirical model for the core losses of the machine is used to obtain eff iciency curves over the base speed range. A comparison is made for two possible control schemes. These being maximum rate of cha nge of torque and maximum torque per ampere control. The concepts of triplen series injection onto sinusoidal, pulse width modulating signals are considered for use in vector controllers. These are applicable not only to synchronous reluctance machines but are transferable to other three phase machines. Finally, direct torque measurement confirms the effects of inductance ripple on the torque production of the machine, highlighting the need for proper rotor geometry design to reduce ripple components. 11. EQUIV ALENT CIRCUITS 0-7803-3 008-0195$4.000 1995 IEEE 294 I I Fig. 1. d-q axis equivalent circuits. The equivalent d-q axis circuits are shown, Fig. 1in a L, , L,, R S stator winding resistance (Q). R, , R , id m iq m ids, , , zd c , zq c vd s vqs 0, electrical rotor frequency (rad.s-'). P number of pole pairs. @d m = L,,, i Wb). reference frame rotating synchronously with the rotor where d , q axis magnetising inductances (H). equivalent core loss resistances (Q). d, q axis magnetising currents (A). d, q axis stator currents (A). d, q axis core loss currents (A). d, q axis stator voltages (V). . . . I IJqm = L , . i d , (Wb). Leakage effects are neglected. The virtual work princip le gives the torque equation as T = p iqm Ld , - Lq,) Nm (1 )
