Innovative Method for Efficiency Improvement in IPMSM Drive through Sensor Reduction

Abstract

An innovative approach to enhance the performance and efficiency of interior permanent magnet synchronous motor (IPMSM) drives is introduced, focusing on two key aspects: maximizing torque per ampere (MTPA) control and addressing sensor reduction challenges in three-phase motor drives. A simple and computationally efficient MTPA control method for IPMSM drives is developed to enhance overall drive efficiency. To calculate MTPA current references accurately, a self-correction of parameters equivalent base current is developed for IPMSMs, which exhibit significant variation in flux linkage and quadrature axis inductance due to temperature and magnetic saturation, but show negligible variation in direct axis inductance during normal operating conditions. In addition, this novel MTPA control law helps to avoid the formation of memory-intensive look-up tables (LUTs), complex computation of actual motor parameters, and several types of MTPA indicators, which are widely used to deal with parameter variations for MTPA operation. Consequently, the proposed method can provide accurate MTPA current references readily and in real time, enabling rapid MTPA control with less computational burden compared to conventional MTPA methods. Simulation and experimental results validate the simplicity, efficacy, and robustness of the proposed MTPA technique which proves to be a great alternative for implementation in low-cost industrial IPMSM drives for numerous applications. Secondly, the study addresses issues inherent in current sensor reduction, particularly in single current sensor (SCS)-based IPMSM drives with MTPA control, focusing on current sampling delay errors that can lead to reduction of efficiency due to speed fluctuations and high torque ripple during IPMSM's wide range of operations. Due to their reduced efficiency, SCS-based three-phase drives are not attracting significant interest despite their potential for addressing sensitivity mismatch issues and reducing costs in various industry applications. In this study, a computationally efficient and accurate model-based current sampling delay error compensation (SEC) scheme based on a new SCS position with a double branch current sampling technique is proposed for an IPMSM drive. In order to improve current reconstruction precision, the compensation strategy includes rotor angle compensation of IPMSM and dead time effect mitigation for the reduction of non-ideal behavior of the inverter, which does not require major computation. Also, dc-link voltage utilization is improved significantly without any external dc-link control. By reducing current reconstruction errors, this scheme enhances drive performance without imposing significant computational burden on the controller. Furthermore, the study explores SCS-based quasi-Z-source inverter (qZSI) drives to remove inherent current measurement dead zones in conjunction with IPMSM instead of putting an additional dc-dc converter. Also, a comprehensive analysis of SCS-based qZSI drive with IPMSM is carried out considering elevated inductor current ripple in conventional qZSI-IPMSM during removal of outer sector boundary dead zone. A new dead zone compensation technique is proposed based on overlapping of active states with ZSVMs pulse width modulation (PWM) strategy. Both current sampling and qZSI operation with ZSVMs are investigated to test the feasibility of three-phase current reconstruction in current measurement dead zones in order to improve the operating range of the IPMSM drive. The practical implementation of ZSVMs-based current sampling also covers single dc-link branch SCS topology, which not only enables variable dc-link voltage control for efficiency enhancement along with MTPA control, but also guarantees inductor size, overall cost, and volume reduction. The proposed SCS-based qZSI-IPMSM drives with MTPA control is verified with experimental results, demonstrating their potential for low-cost, reliable, and efficient motor drives for electric vehicles and industrial applications, while maintaining high computational efficiency.