At present time, research in the field of Electric Vehicles (EV) is significantly intensifying around the world due to the ambitious goals of many countries, including the UK, to prohibit the sale of new gasoline and diesel vehicles, as well as hybrid vehicles, in the near future around 2030-35. The primary goal of this Ph.D. research is to improve the propulsion system of electric vehicles’ powertrains through improvements in the control of Interior Permanent Magnet Synchronous Motors (IPMSM), which are commonly used in EV applications. The proposed approaches are supported by simulations in Matlab, Matlab-Simulink and laboratory-based experiments.
The research initially proposes an analytical solution in implicit view for a combined Maximum Torque per Ampere (MTPA) and Maximum Efficiency (ME) control, allowing to determine the optimal d-axis current, based on the concept of minimisation of the fictitious electric power loss. With the exception of two parameters, the equation is identical to that of the ME control. Therefore, upgrading the ME control to the combined MTPA/ME control is relatively easy and doesn’t require any change in hardware beyond a few minors of controller code in the software. The presented research demonstrates an easy-to-apply combined MTPA/ME control leading to the ‘Transients Optimal and Energy-Efficient IPMSM Drive’ providing smooth transitions to the MTPA control during transients and to the ME control during steady states.
A concept of ‘Nonlinear Optimal Control of IPMSM Drives’ is also introduced in this Ph.D. research. The velocity control loop develops nonlinearities when energy consumption optimisation methods like MTPA, ME, or combined MTPA/ME are added. In addition, the control system’s parameters can be inaccurate and fluctuate depending on the operating point or possible uncertainties in real-time operation. In the proposed method, the control structure is the same as in the Field Oriented II

Control (FOC), with the close velocity and two current loops, but the Proportional-Integral (PI) controllers are replaced by Nonlinear Optimal (NO) Controllers. The linear part of the controller is designed as a Linear Quadratic Regulator (LQR) with integral action for each loop separately. This is, in fact, a PI controller with optimal gain parameters for a specific operating point. The nonlinear part takes the required fluctuations of the control system’s optimal gain parameters in real-time operation as new control actions to improve a robust control structure. The design procedure for the nonlinear part is similar to that of the LQR, but the criterion of A. Krasovsky’s generalised work is used, and the analytical derivations lead to an explicit control solution for the nonlinear optimal part. The nonlinear part emulates the adjustments for updating the linear part’s optimal LQR gains based on operating conditions, instead of employing extensive look-up tables or complicated estimation algorithms. The proposed control is robust in the allowed range of the system’s parameters.

In conclusion, upgrading existing industrial IPMSM drives into a robust and optimal energy-efficient version that can be used for electric vehicle applications is the main advantage of the novel control concept described in this Ph.D. research. For this upgrade, only a small portion of the software that is related to the PI controllers needs to be changed; no new hardware is needed. Therefore, it is cost-effective and simple to transform existing industrial IPMSM drives into a better version with the proposed method. This feature also leads to the design of more adequate IPMSM drives to meet the demands of Electric Vehicle (EV) operating cycles.