This paper presents a novel error compensation method for resolvers to improve motor drive performance. It is shown that the nature of the error terms of the position information provided by the resolver is deterministic to a large extent, as they mainly stem from the physical imperfections regarding the resolver and its mounting. This motivates a model-based approach of extracting the error terms and modelling them as the sum of sinusoidal waveforms. Therefore, the resolver output signal can be corrected by the recorded error model without any delay. In the proposed method, the memory requirement and computation burden to the CPU is negligible. Thus, the method is very convenient for practical applications. To determine the error model, hence the compensation parameters, an offline calibration procedure is run once and the extracted parameters can be used for the entire life cycle of the motor as long as the mounted resolver is not removed. Enhanced performance parameters of the motor driver system are verified on a 100kW Internal Permanent Magnet Synchronous Motor (IPMSM) test setup and the results are presented. It is experimentally shown that the proposed compensation algorithm is able to decrease the peak-to-peak position error magnitude from 3.834 to 0.322 degrees (91.6% reduction).

Early detection of an inter-turn short circuit fault (ISCF) can reduce repair costs and downtime of an electrical machine. In an induction machine (IM) driven by an inverter with a model predictive control (MPC) algorithm, the controller outputs are influenced by a fault due to the fault-controller interaction. Based on this observation, this study developed neural network models using inverter switching statistics to detect the ISCF of an IM. The method was non-invasive, and it did not require any additional sensors. In the fault detection task, an area under receiver operating characteristics curve value of 0.9950 (95% Confidence IntervaI: 0.9949 - 0.9951) was obtained. At the rated operating conditions, the neural network model detected and located an ISCF of 2-turns (out of 104 turns per phase) under 0.1 seconds, a speedup of more than two times compared to the thresholding-based method. Moreover, we published the switching vector data collected at various load torque and shaft speed values for healthy and faulty states of the IM, becoming the first publicly available ISCF detection dataset. Together with the dataset, we provided performance baselines for three main neural network architectures, namely, multi-layer perceptron, convolutional neural network, and recurrent neural network.

In this paper, a novel and non-invasive stator inter-turn short circuit (ITSC) online detection method is presented for an induction machine (IM), driven by a two-level voltage source inverter (2L-VSI) via finite control set model predictive control (FCS-MPC). The main idea of the proposed method is to utilize the controller itself as an observer: under the presence of a fault, the distribution of inverter switching states significantly deviates from the original balanced case. Therefore, by inspecting the inverter switching vectors, which are the outcomes of the FCS-MPC routine’s online optimization procedure, a stator fault can be detected efficiently. It is observed that both the zero-vector allocation over the complex plane and the allocation among the active vectors are influenced by the presence of a stator short-circuit fault. The proposed fault detection strategy introduces little to no extra burden for processor and memory. Experimental results verify the proposed method, and inter-turn short circuits of two and three turns are confidently detected and located for a 500 W, two-pole IM.