Deadbeat Flux Vector Control with Continuous Transition from Linear to Overmodulation including Six-Step Operation Considering the Voltage and Current Limits by Applying One Single Control Law

Six-step operation of interior permanent magnet synchronous machines (IPMSMs) has many advantages such as the full utilization of the DC-link voltage and the minimization of the switching losses. Deadbeat Flux Control (DBFC) is a control strategy designed to achieve overmodulation and six-step operation with competent dynamic performance. This paper presents an upgraded implementation of the DBFC controller. One unique control law can operate in the entire torque-speed plane with a simple methodology for continuous and smooth transition between the four different operating regions (i.e., linear, overmodulation I, overmodulation II, and six-step), and under voltage and current limitations. Combining multiple controllers and switching between different control laws is completely avoided. This paper presents an approach for closed-loop torque regulation that has high dynamic performance, high accuracy, and is robust against parameter variations. The proposed controller is evaluated in simulation and experimentally. It is implemented on an automotive microcontroller for a 300 kW IPMSM traction drive.

I. Introduction

Interior permanent magnet synchronous machines (IPMSMs) have been widely used in automotive applications due to their high power density, high efficiency, and wide operating speed range [1]. The flux weakening (FW) control is a prerequisite for traction drives due to the limited DC-link voltage [2]. The flux weakening mode can be divided into two regions: FW I and FW II (also known as the maximum torque per flux/voltage region (MTPF/MTPV)). By operating in the pulse width modulation (PWM) mode, FW I is entered when the voltage reaches the linear limit . For infinite speed drives operating with Deadbeat-direct torque and flux control (DB-DTFC), the square-root-condition (SRC) becomes a favorable control method in FW II, since it is simple to implement, and it has a minimum sensitivity to machine parameter estimation errors [3].