State Key Lab of Control and Simulation of Power Systems and Generation Equipments, Department of Electrical Engineering, Tsinghua University, Beijing, China
Auto-generated profile, based on publicly available information
Jiawei Sun
State Key Lab of Control and Simulation of Power Systems and Generation Equipments, Department of Electrical Engineering, Tsinghua University, Beijing, China
The high fault-tolerant ability of multiphase drives is favored in safety-critical applications. Under open-phase faults (OPFs), to guarantee ripple-free torque, stator current references should be revised. In this paper, a global fault-tolerant control strategy based on an online current optimization algorithm (OCOA) is proposed for symmetrical multiphase machines (SMMs) to achieve both the maximal torque production range (TPR) and minimal stator winding losses under all possible OPFs. The OCOA can calculate the minimum losses current references covering the full TPR online, regardless of the number and locations of faulty phases. With the online-optimized references, ripple-free torque and minimum losses in full TPR can be achieved under faulty conditions. Additionally, the proposed method can be adopted to SMMs with arbitrary phase numbers. Simulation and experiments demonstrate the online optimization can be completed in a short time, meanwhile, achieves the same TPR and loss reduction compared with existing offline strategy based on pre-stored look-up tables. Due to the flexible online calculation ability and excellent extensibility, the proposed method is especially favored in modular multiphase converters and machines with high phase numbers, where converters can adapt to machines with different phase numbers or the number of possible faulty conditions can be extremely large.
14 December 2021 in IEEE Transactions on Power Electronics
In recent years, fault-tolerant control of multiphase drives against open-phase faults has been an interesting research topic. However, in contrast, few research efforts have been devoted to fault-tolerant control against open-switch faults in which only one controllable power semiconductor device in a converter leg is open, but its antiparallel diode operates normally. In this article, a fault-tolerant control strategy against open-phase/open-switch faults based on dc current injection is proposed. In open-switch phases, dc currents, whose amplitudes are equal to the amplitudes of the corresponding fundamental phase currents, are injected to eliminate the nonlinearity brought by faults. Therefore, the fundamental currents in open-switch phases can be precisely regulated. To achieve torque-ripple-free postfault operation, minimize the stator winding losses and maximize the postfault torque production, stator winding losses associated with fundamental currents and injected dc currents are optimized over the full torque production range on the premise that no dc current is mapped in the torque-producing subspace. Simulation results on different kinds of multiphase drives and experimental results on a nine-phase induction drive proved the reduced stator winding losses, extended torque operation range, and reduced torque ripple compared with the existing methods.
14 February 2022 in IEEE Transactions on Energy Conversion
Multiphase drives with fault-tolerant capability are favored in high-reliability applications. In most fault-tolerant multiphase drives, the fault types and locations must be known prior to the fault-tolerant control. Therefore, fault diagnosis is an indispensable procedure. This paper proposes a generalized method to detect and identify the open-circuit faults (including open-switch and open-phase faults) in star-connected symmetrical multiphase drives with different phase numbers. The selected fault diagnosis signals are calculated solely from simple arithmetical operations of measured phase currents, resulting in easy real-time implementation. As fault detection signals are derived from the physical constraint imposed by the neutral point, the proposed method is naturally robust and not sensitive to operating points, machine transients and harmonics in stator currents. Therefore, a relatively small fault detection threshold can be used and fast detection can also be achieved. After the fault has been detected, the fault type is further identified according to the polarity of the integration of phase currents. Experimental results on nine-phase and five-phase induction machine drives verify the generality, fast diagnosis speed and robustness of the proposed method.
23 March 2021 in IEEE Transactions on Industrial Electronics
Multiphase drives with high reliability are receiving increasing attention by the research community in safety-critical applications. Condition monitoring is an effective way to improve drive reliability, because it can detect the development of faults and offer the possibility of predictive maintenance. Among the different condition monitoring methods, online estimation of stator resistance is a promising technique for its ability to detect high-resistance connections, overheating of stator windings and other defects. In this article, an online estimation method of stator resistance based on dc-signal injection is proposed for multiphase drives. The dc currents are injected in the subspace with the smallest main inductance to minimize the induced torque ripple. Phase voltages are measured and the dc components are extracted to calculate the resistance in all the phases. Compared with existing methods, the proposed method can estimate the per-phase stator resistance with higher accuracy and higher robustness, which provides additional benefits in condition monitoring applications. In addition, the proposed method is derived in a general way, which makes it applicable to most kinds multiphase drive systems. Experimental results on a nine-phase induction machine and a five-phase induction machine validate the high accuracy, high robustness, and high generality of the proposed technique.
Despite the huge progress of model predictive control (MPC) for electrical drives that has taken place over the last few years, there are still some issues that require further study, such as robustness, steady-state performance, and tracking error reduction. This paper proposed a robust offset-free model predictive current control (ROF-MPCC) for induction motor. The proposed ROF-MPCC applies incremental model and incremental current observer to achieve offset-free prediction and time-delay compensation. ROF-MPCC has good robustness to the rotor flux and model parameters, and can eliminate the steady-state error in control. Experimental results demonstrate that ROF-MPCC has good stead-state and dynamic performance.