There exists an inaccurate measurement problem in permanent magnet synchronous motors (PMSMs) due to low motor speed operation, high temperatures and humid environments, which will degrade the motion performance and stability of PMSMs. In this study, a model reference adaptive system without position measurement is presented in a PMSM to improve the output performance with external disturbance suppression caused by an environmental change. Firstly, a PI adaptive estimation law is designed to identify the motor speed. Then, a optimization method based on the sliding mode variable structure technique is proposed to realize the stability augmentation of the speed loop by using the parametric fuzzy logic design. To reject the current loop noise, an extended Kalman filter (EKF) is adopted to compensate the input signal in the current loop. The effectiveness of this proposed method is verified via a numerical simulation in the case of different speeds and external loads.
Transcritical carbon dioxide waste heat recovery systems and the construction of scroll expanders have recently been hot topics. The flank clearance, located between the orbiting and fixed scroll, has a vital impact on the scroll expander performance. This paper estimates the effect of the flank clearance on the expander’s thermodynamic performance (first-law efficiency) based on computational fluid dynamics (CFD) simulations. The manufacturing cost of different flank clearances is also considered to enhance the feasibility of the machinery design. The computational cost for different flank clearance cases is significantly reduced with a surrogate-assisted multi-objective optimisation algorithm (SAMOA), which also supports modelling the trade-off relationship between manufacturing cost and machinery efficiency. The results indicated that an increasing flank clearance negatively affects the first-law thermal efficiency. The efficiency decreased from 87.41% to 44.83% moving from 20 to 200 μm flank clearances. The SAMOA successfully reduced the computational cost of the dynamic mesh CFD model from 90 h to 15 s with 0.6% discrepancy. The final Pareto solutions presented a clear trade-off relationship between the first-law efficiency and manufacturing cost and promised a diversity of optimum solutions. The “knee points” for the relationship were 25, 55, and 127 μm, which provided flexible clearance choices based on the importance of either machinery efficiency or manufacturing cost.
7 February 2025 in Journal of Statistical Theory and Practice