Multi-Objective Optimal Design Of Interior Permanent Magnet Synchronous Motors Based On A Combined Surrogate Model

Permanent magnet synchronous motors(PMSM)are used in a wide range of applications such as robotics and industrial automation where high performance motors are required due to their high accuracy,efficiency and reliability.With the increase of PMSM applications,the topology of the motor is constantly being improved,but the improvement of the motor structure relies on reliable surrogate models and optimisation techniques,which is one of the difficulties in the motor optimisation process.Therefore,this thesis proposes a strategy based on a combination of a combined surrogate model and an optimisation method to optimise the structure of an interior PMSM(IPMSM).The specific research methodologies are as follows:(1)Firstly,based on the theoretical formulae for motor design,the main dimensions of the PMSM are determined,a suitable pole-slot combinations and winding distribution scheme are selected.Secondly,the stator slot structure parameters and rotor structure parameters are estimated.Then,based on the magnetic circuit method,the structural parameters of the motor are adjusted appropriately to ensure that the output performance of the motor meets the design requirements.Finally,finite element analysis(FEA)is adopted to obtain the characteristics of the PMSM under no-load and load conditions to verify the designed PMSM meets the requirements.(2)Firstly,the target performance to be optimised is determined,the structural parameters impacted on the target performance are analysed.Secondly,the structural parameter combinations are obtained by the latin hypercube design method and the target performance for each parameter combination is calculated using FEA.Then,surrogate models for each target performance are constructed based on data samples and multiple regression models.Finally,the accuracy of the constructed surrogate models is judged according to the evaluation metrics,and the surrogate models with the highest accuracy are selected in turn to be combined.(3)The combination of Non-dominated Sorting Genetic Algorithm-II(NSGA-II)and the Pareto solution set,the appropriate solution and the corresponding finite element model(FEM)are selected to verify the reliability of the optimisation method.Based on the NSGA-II optimised motor model,a secondary local optimisation of the motor is conducted by using the sensitivity analysis method and Taguchi’s ones.(4)The theoretical analysis of the causes of cogging torque is conducted and an auxiliary slotting method is used to reduce the cogging torque of the PMSM.Then,the impact on the motor resulting from the opening of the auxiliary slot was evaluated.