Performance Enhancement of Switched Reluctance Motor Drive Using Direct Torque Control Strategies

Abstract

Due to rare earth permanent magnets (PM) price volatility, there is a significant interest in developing magnet free motor drive configurations that are more efficient, highly dense and ruggedness to the harsh environment in electrified vehicles. Switched reluctance motor (SRM) is considered to be one of the promising alternatives to the traditional AC motors due to distinctive features of the robust structure, high starting torque, wide constant power region, no shoot-through current in the inverter, low cost and low maintenance. However, it suffers from severe torque ripple, acoustic noise and the requirement of a position sensor for control, which limits the wide usage of SRM in vehicular applications. Moreover, smooth torque control is difficult to achieve due to inherent torque pulsations generated due to the doubly salient structure. Over the past years, significant research is carried out to minimize the aforementioned concerns in SRM by improving the design aspects of the motor and/or employing appropriate control strategies. Most of the work related to the motor design is devoted to optimizing machine parameters and developing the multi-phase machines. Later one is achieved by developing advanced optimized control techniques such as current profiling techniques (CPT), and torque sharing functions (TSF), where the torque is an indirect control variable. However, these approaches require intensive computation to store machine non-linear characteristics, long settling time and require optimal tuning of control parameters. Alternatively, controlling torque as a direct control variable led to overcome some of the aforementioned shortcomings through direct instantaneous torque control (DITC), advanced DITC, direct torque and flux control (DTFC) and model predictive control strategies. The performance comparison of different control strategies such as CPT, TSF, DITC, and DTFC is studied and the merits and demerits of each method are well established in the literature. Among them, the DTFC scheme which utilizes the philosophy of direct torque control (DTC) of conventional AC machines draws more attention due to automatic control, reduced acoustic noise, fast dynamic response, insensitive to motor parameters & does not require rotor position information. In this scheme, the phase torque is directly controlled by accelerating or decelerating the flux ilinkage vector, where the magnitude of flux linkage is kept constant within a hysteresis band. However, the active phase has to produce higher positive torque during commutation of outgoing phase, thereby draws higher source current. As a result, there is a reduction in overall torque per ampere, thereby reducing the efficiency of the traction drives. In the present thesis, two new DTC algorithms are proposed with optimized voltage vector selection and appropriate sector partition based on the inductance profile to enhance torque/ampere ratio (T/A) and minimize torque ripple in four-phase (8/6) SRM. A 16 sixteen sector approach is developed to select the vectors more precisely, thereby effectively eliminates the existence of phase current in the zero torque region which leads to increase in T/A and torque ripple minimization. However, negative phase torque is still generated in this method due to existence of phase current in negative inductance slope region. To overcome this issue, a 16 sector 16 vector DTC algorithm is proposed. MATLAB simulations and real time simulations with OPAL-RT are performed to verify the effectiveness of the proposed DTC algorithms in terms of torque ripple, T/A and source current ripple. Moreover, copper losses, switching losses and number of commutations required for one cycle are calculated to show the efficacy of the proposed algorithm over conventional DTC algorithm. In general, four phase SRM requires four current sensors to measure phase currents that increases the cost of the drive. To overcome this issue, a new phase current reconstruction algorithm is developed to detect the four phase currents more accurately using only three senors under proposed DTC algorithm.