Line start permanent magnet synchronous motors (LS-PMSM) are preferred more and more in industrial applications, because they can start on their own and because of their high efficiency. In this study, a new LS-PMSM rotor typology is suggested, which is modelled using surface mount permanent magnets, in which two different slot types have been used together. The rotor of an asynchronous motor on the industrial market in the IE2 efficiency segment has been remodeled in the study, resulting in an increase in motor efficiency from 85% to 91.8%. A finite elements software was used for determining motor design and performance, in addition to analytical methods.
Starting from the first years of the proposed period right up to the 2000s, there were commercialization concerns of line start permanent magnet synchronous motors. This situation, which developed as a result of not developing methods to adequately explain the motor operating characteristics, and also because of the limitations associated with the non-high-performance manufacturing materials, took place widely until the 1970s. In subsequent years, the development of increasingly efficient materials, as well as the use of innovative methods, such as advanced magnetic circuit models and FEM analysis, which better reflect the performance of the motor, these motors became available for an increasing number of industries. In the past 10 years, research has paid off, and finally, line start permanent magnet synchronous motors have begun to be offered on the market as an alternative to AC induction motors. These motors, which are offered on the market with an IE4 efficiency class limits level that is currently tricky to achieve with traditional AC induction motors, are already included in the catalog of many major motor manufacturers.With this study, a detailed literature search was conducted and filtered on the development and improvement of line start permanent magnet synchronous motors. As a result of the study of numerous researches and more than 50 rotor topologies, design trends were identified that are important in the development of such motors. The development steps from the proposed years to the present have been discussed in terms of manufacturability, efficiency and performance.
5 November 2021 in Journal of Mechatronics and Artificial Intelligence in Engineering
In this study, 5.5-kW 1500 rpm IEC-60034-30-1:2014-IE4 super premium efficiency class Line Start Permanent Magnet Synchronous Motor (LS-PMSM) was presented with an innovative rotor design utilizing two different slot types have been used. The slots offered in two different geometries make a significant contribution to the motor in the form of a high starting torque thus increasing the synchronization ability. The prototype was manufactured for the optimal model obtained via analytical methods and Finite Element Software. Torque ripples that pose an important issue for permanent magnet motors were tried to be reduced by way of a skewed stator structure. In addition, motor frame and stator main dimensions were acquired from a standard IE2 induction motor design during the manufacturing of the prototype. Experimental studies for the suggested LS-PMSM topology put forth a 6 % higher efficiency in comparison with low efficiency IE2 standard induction motors with the same shaft power. Results for the starting performance, torque ripples and back-EMF were presented for the prototype motor under a constant load of 0-1.25 p.u. It was verified as a result of the study that LS-PMSMs can be used as an alternative to induction motors in pump and fan industrial applications.
This study proposes a sequential Hybrid GP-ANN architecture based on the concept of autonomous feature engineering for multi-output performance prediction of single-phase induction motors, serving as a high-fidelity surrogate model ready for digital twin integration. The required high-resolution motor dataset (88,375 samples) was generated using a parametric sweep approach with Ansys RMxprt. In the proposed architecture, Genetic Programming (GP) is not positioned as a final predictor but as an analytical filter that discovers hidden physical relationships in raw input data and converts them into super features, thereby structurally eliminating the polynomial explosion risk and scale sensitivity that arise when raw data are modelled directly. The nonlinear physical relationships discovered autonomously by GP are added to the network input matrix as mathematical vectors, breaking the black-box structure of standard artificial neural networks and transforming it into a physics-inspired grey-box model. The nonlinear terms discovered by GP are fed into the network after independent Z-score normalisation to preserve gradient stability. To comprehensively evaluate this framework, its predictive performance is benchmarked against industry-standard machine learning algorithms, including Random Forest (RF), Support Vector Regression (SVR), and Extreme Gradient Boosting (XGBoost). Results on variation-based unseen test data show that the hybrid model achieves competitive prediction accuracy relative to XGBoost—matching or exceeding it for Output Torque and Input Current, and remaining broadly comparable for Output Power, though XGBoost achieves substantially lower RMSE for Efficiency and Power Factor—while additionally offering transparent mathematical traceability, reducing error rates (RMSE) by 42.1% to 66.1% per parameter compared to the standard ANN. The model achieves an score of 0.9879 for the power factor parameter, where conventional approaches struggle. To rigorously validate the interpretability of this grey-box architecture, a global Permutation Feature Importance (PFI) analysis was conducted, showing that the GP-derived super features collectively account for 61.46% of the decision logic, outweighing the combined contribution of the raw inputs (38.54%). Furthermore, the autonomous feature engineering layer is found to reduce the learning burden on the ANN, allowing it to converge to a substantially lower error floor within the same fixed epoch budget. With an online inference time below 0.02 ms, the proposed architecture offers a robust methodological infrastructure for sustainable motor digital twins through a balanced trade-off between prediction accuracy and computational efficiency.