This paper presents an online data-driven diagnosis method for multiple IGBTs open-circuit faults and current sensor faults in the three-phase PWM inverter. The Fast Fourier Transform (FFT) algorithm is used to extract the fault frequency spectrum features of the three-phase currents. Then, a feature selection method named ReliefF is introduced to select the most critical features by removing redundant and irrelevant features. In addition, as a novel fast learning technology, Random Vector Functional Link (RVFL) network is applied to learn the faulty knowledge from the historical dataset. Based on the well-learned model, the fault type and location of the converter can be accurately identified as long as the three-phase current signals are measured. Offline test results verify that the proposed method can identify the IGBT and sensor faults with an accuracy of 98.83% and outperforms state-of-the-art learning algorithms. Moreover, the real-time hardware-in-the-loop test results show that the proposed method can successfully identify the IGBT faults and current sensor faults within 22 milliseconds. It is robust to the DC-link voltage fluctuations, model parameters, and speed or load variations. The extensibility of the proposed method is also validated based on the test results in terms of other fault modes and drive systems.

This paper presents a speed estimation scheme by combing second-order sliding-mode observer (SMO) the model reference adaptive system (MRAS) in the sensorless-vector-controlled linear induction motor (LIM) drives for medium-low speed maglev applications. A state space-vector model of the LIM considering the dynamic end effects is rearranged in order to be represented by the form of the super-twisting algorithm (STA). Then, a STA-SMO based stator current observer of the LIM with Popov's hyperstability theory is designed, which is used to replace the reference model of the MRAS-based scheme. Correspondingly, the actual stator current model is regarded as the adaptive model of the MRAS for speed estimation. Meanwhile, because the variation of stator resistance associated with stator temperature may lead to large estimated error and even system instability, a parallel stator resistance on-line identification scheme is adopted in the proposed speed estimation scheme to reduce the speed error and improve the stability of the sensorless control system. Compared to the speed estimation schemes based on Luerberger observer with MRAS and the single-manifold SMO, the proposed speed estimation scheme exhibits better estimated performance. The effectiveness and feasibility of the proposed speed estimation scheme have been verified by simulation and hardware-in-the-loop (HIL) tests.

This paper presents a model-based open-circuit fault diagnosis approach for single-phase three-level neutral-point-clamped (3LNPC) converters in electric railway application. The diagnosis algorithm, which only requires the signals existing in the control system, not only detects open-circuit faults but also identifies the faulty device among the transistors and clamping diodes. The mixed logical dynamic (MLD) model of the converter is built to estimate the grid current. The residual generated from the measured current subtracting the estimated one is analyzed under different open-circuit faults. According to the characteristics of the residual changing rate, the proposed approach allows fault localization. The proposed method is effective both in traction and regenerative braking operation and has fast diagnosis speed. Hardware-in-the-loop (HIL) experiments are carried out to verify the effectiveness of the proposed fault diagnosis algorithm.

Phase-locked loops (PLLs) and frequency-locked loops (FLLs) are of importance in power and energy applications. Both technologies have been introduced to speed-sensorless- controlled motor drives, and increasing applications of PLLs and FLLs for speed estimation are foreseen. To enable a proper and good design, a thorough review of the PLL- and FLL-based speed estimation schemes is then provided in this article. It is revealed through the review that many PLL- and FLL-based estimation schemes fail to accurately track a frequency ramp (i.e., obvious estimation errors appear), which may lead to a compromised estimation accuracy when these schemes that are applied in induction motor drives operating during acceleration and deceleration processes. To address this, the proven speed estimation schemes together with new attempts are also presented in this article. Moreover, various challenges to the PLL- and FLL-based speed estimation schemes, including harmonics, dc offsets, and parameter variations, are considered when evaluating these schemes. Solutions to tackle these disturbances are accordingly presented. In addition, two representative estimation schemes are exemplified through experimental tests. Finally, further challenges in using the PLL- and FLL-based schemes for speed estimation are discussed.

A proper thermal network model (TNM) facilitates the thermal analysis of multichip power modules and is beneficial to the reliability evaluation of power converters. However, many TNMs are struggling to provide the accurate thermal analysis while maintaining acceptable computational efficiency. To address this, a TNM based on the Cauer model is proposed in this article, where the Cauer model is obtained with the assistance of the structure function. Then, the thermal coupling effects in multichip power modules are fully characterized. More specifically, the thermal coupling in the proposed TNM is equivalent to the injected power losses from the specific nodes, and an enumeration method is accordingly used to identify the injected nodes and the injected power losses. Moreover, a discretization approach based on the forward Euler method is implemented to enhance the computational efficiency. Subsequently, the junction temperature of the multichip power module is estimated through the signal-flow graph method. Finally, the proposed TNM uses the typical insulated-gate bipolar transistor modules as a case study, and extensive simulations and experimental tests are conducted to validate the effectiveness of the proposed TNM.

Speed-sensorless control of induction motor drives based on the extended state observer (ESO) attracts much popularity due to its key characteristics, e.g., satisfactory estimation performance, and high robustness against disturbances. However, the conventional ESO generally uses fixed high gains to achieve fast convergence. This may bring about the concern of noise sensitivity. Moreover, a proportional-integral (PI)-type speed controller is mostly adopted in speed-sensorless control of induction motor drives, which may cause unsatisfactory system dynamics. To address these, an adaptive active disturbance rejection control (ADRC) strategy is proposed for speed-sensorless induction motor drives in this article. In this scheme, a third-order adaptive ESO (AESO) is first used to estimate the speed, the phase, and the overall disturbance. Then, a speed controller based on the state error feedback control rate (SEFCR) is designed to improve system dynamics. In practice, disturbances like dc offsets may further challenge estimation performance, and hence, a closed-loop flux observer (CLFO) is employed to deal with this issue. Additionally, a parameter sensitivity analysis of the CLFO is also provided to evaluate the performance of the proposed scheme. Finally, extensive experimental tests have been carried out to validate the effectiveness of the proposed.