Search

This paper presents an online maximum torque per ampere (MTPA) control that considers both magnetic saturation and cross-magnetization effects of interior permanent-magnet synchronous machines (IPMSMs). For IPMSM drives, especially in torque-controlled applications, torque accuracy and high-efficiency operations are important issues. They can be dealt as a constrained optimization problem to satisfy both torque reference tracking and loss-minimizing operation. In this paper, nonlinear simultaneous equations are derived from Lagrange multiplier method, which could be solved by numerical algorithms. Among them, Levenberg–Marquardt algorithm (LMA) is employed to guarantee a robust calculation of optimal current references. It is optimized to alleviate calculation burden while maintaining the stability of the proposed algorithm. In addition, a torque reference limiter is implemented to satisfy a current limit in real time. The feasibility of the proposed method is verified under various operating conditions by simulation and experimental results. Through the proposed algorithms, accurate MTPA control is achieved under not only unsaturated but also highly saturated operating conditions. [ABSTRACT FROM AUTHOR]

State-space control could provide both high dynamic performance and sufficient stability margin to voltage source converters with LC filter. However, an inherent digital delay induced by a digital control system deteriorates performance and even induces instability. In this paper, a discrete-time voltage controller is proposed based on a discrete state-space model. First, a state feedback control with a reference feedforward path is designed in the discrete-time domain with consideration for the digital delay. Second, an output current decoupling path is augmented to minimize the effects of an output current disturbance. Controller gains are derived as the functions of system parameters and design specifications, which is based on a direct pole placement and pole-zero cancellation method in a rotating reference frame. Moreover, a parameter sensitivity and digital implementation are discussed to improve the performance and stability of the proposed controller. The effectiveness of the proposed controller is verified with various experimental results. It shows that a filter resonance is well damped out while maintaining wide voltage control bandwidth. [ABSTRACT FROM AUTHOR]

In this paper, the characteristics of ripple current of a multimegawatt drive system fed by a pulse-width modulation (PWM) inverter are investigated. Based on the investigation, a novel PWM scheme is devised to reduce the switching loss of the inverter under an identical amount of torque ripple of a motor. The proposed PWM is composed of two parts. The first is minimum torque ripple PWM (MTRPWM), which is optimized in terms of the minimum torque ripple in high modulation index (MI). The second determines the minimum switching frequency while holding the average torque ripple less than the allowable value under various operating conditions. It is termed reduced switching frequency PWM (RSFPWM). Consequently, by simply changing the PWM method from conventional fixed switching frequency continuous PWM to the proposed PWM, MTRPWM and RSFPWM, simulation results show that the switching loss of the inverter driving a 3.5 MW induction motor is reduced to 51.6% while keeping the torque ripple in the allowable boundary, where MI is 1.12. This approach is applied to small-scale experimental sets consisting of an interior permanent magnet synchronous machine (IPMSM) motor-generator set, and an induction machine MG set in an effort to verify the validity and feasibility of the proposed method. Based on these experimental results, the cost indexes of the IPMSM and the induction machine, as defined in this paper, decrease by up to 66% and 77%, respectively, when the proposed method is applied when MI is 1.12 compared with the outcome with the conventional PWM. [ABSTRACT FROM AUTHOR]

This paper presents a carrier-based pulsewidth modulation (PWM) method that reduces the common-mode voltage (CMV) of a three-level four-leg converter. Based on an analysis of space vector PWM (SVPWM) and sinusoidal-PWM switching patterns, the fourth-leg pole voltage of a three-phase converter, known as the “f pole voltage,” is manipulated to reduce the CMV. To synthesize the f pole voltage for the suppression of the CMV, positive and negative pole voltage references of the f leg are calculated. In addition, the offset voltage to prevent distortion of the a, b, and c phase voltages regarding the neutral point is deduced. The proposed PWM strategy can be easily implemented in the software of a DSP-based converter control. The three-level four-leg converter with the proposed PWM algorithm results in a remarkable reduction in the peak-to-peak value of the CMV. From the simulation and the experimental results, the peak-to-peak value of the CMV when using the proposed PWM method is 33% compared to that when using the SVPWM method, while the number of CMV transitions during the switching period in the proposed PWM method is only 25% of that when using the SVPWM method. [ABSTRACT FROM PUBLISHER]