A novel optimization method of asymmetric traffic distribution for motor oil spray cooling system based on surrogate model

The oil spray system can directly cool the end winding and improve cooling capacity of the permanent magnet synchronous motor (PMSM) with high power density. In this manuscript, a novel method is proposed to optimize the traffic distribution in different oil spray nozzles, which is important to suppress the maximum temperature of the end winding. Firstly, the irrelevance of the oil spray nozzle size on the heat transfer efficiency is validated. Thus, the optimization of the oil spray nozzle size can be simplified to the optimization of flow distributions with the same oil spray nozzle size. Subsequently, an adaptive Latin Hypercube Sampling (LHS) method is employed to collect 120 sets of sample datas, and an esponse surface surrogate model is obtained. Combined with the NSGA-II multi-objective genetic algorithm, the flow distribution is optimized. Finally, this method is verified based on the CFD calculations. Compared with the global CFD optimization and traditional Taguchi method, the proposed method has high optimization efficiency and solution coverage rate, respectively.