Design and optimization of a V-type interior permanent magnets motor for EV application

Electric vehicles (EV) gained an increasing popularity over the years, especially in the last decade, with significant growth of the sales by 46\% to 69\%. There are multiple reasons for that: the regulations on C02 emissions pushed the automotive industry to focus more on them than the traditional ICE vehicles since they are the only ones capable of respecting the latest restrictions. Furthermore, they are suitable to reach higher performance, but also to adapt to new technologies. Consequently, all the components of the electric traction line are required to have high efficiency. Among them, a lot of research have been focusing on the electric motor and on how taking full advantage from its features. In particular, looking at the already existing solutions on the market, it seems that the V-type interior permanent magnet motor (IPM) is the most promising and used one thanks to the high efficiency, specific power and constant power speed range, but also to the remarkable and torque density it can achieve. For this reason, the aim of this thesis is to design and optimize a V-type IPM motor that enhances the characteristic required by the EV application. As primary task, the working points of the motor have been obtained through the Worldwide Harmonised Light Vehicle Test Procedure (WLTP). It is a driving cycle, introduced in 2017 and adopted to retrieve the energy consumption of EVs. It is more reliable than the others because it is based on real-driving data. The design part has been carried out through the development of a proper reluctance network of the motor of interest. It represents an equivalent magnet circuit of the electric machine. In other words, the aim is to model the magnetic flux path inside the motor, adopting a simple circuit. The main benefit of this approach is to save computational time, although the accuracy could decrease, depending on how complicate the geometry is. The next step has been to properly define the parameters that affect the rotor geometry, as various properties depend on it. At first stage, we implemented two reluctance networks, for d and q axes, on Simulink. Afterwards, the networks have been analytically solved, including the non-linearity effect of the iron core and the cross coupling between the axes. This last step involved adding an extra path on the q axis, representing the bridges reluctances, present on d-axis. Different geometries have been drawn to test the model and validate it through the finite element software FEMM 4.2. The results have been compared and the error on the torque has been calculated. The model made it possible to retrieve the current-flux maps of the motor, which were used to develop a simple open loop control following the MTPA and MTPV laws. The control gives the possibility to follow properly the working points suggested by the driving cycle. The advantage of the model is to permit a straightforward control, acting directly on the d and q components of the current. Finally, the Particle Swarm Optimization (PSO) has been used to optimise the design in those points.