Design and optimisation of electric drive system with high performance and low cost.

As the main component in the global trend of electrification, the electric machine will experience growing demands in many applications such as industrial equipment, domestic appliances, electric vehicles and renewable energy systems. However, nowadays the majority of electric machines are induction machines, which are robust, low-cost, but with relatively low efficiency compared to PM motors, and they face challenges to meet the stringent efficiency mandatory standards. Synchronous machines with rare-earth permanent magnets are the most promising to improve efficiency and fulfil mandatory requirements, but the price of rare-earth materials is every high and volatile. The low-price and abundant ferrite magnet offers a promising solution for non-rare-earth permanent magnet synchronous machines to replace induction machines with similar material cost but higher efficiency. The ideas of high performance and low cost are not only limited on the machine body itself, but also include the considerations of its whole life cycle, such as motor design, motor manufacture, and motor controller. In the thesis, the particle swarm optimisation algorithm is applied in the ferrite IPM machine design, and with the help of VBScript programming language, a time-efficient electric machine design method is proposed to reduce substantially the designer’s designing time and hence the total cost of machine design. The design of experiment (DOE) assisted Six Sigma DMADV method is proposed for the quality control of the electric machine, and the motor quality is improved by realistic adjustment of the tolerances to minimise variations in output performance. The iron loss analysis considering the effects of press fitting stress, shrink fitting, and PWM harmonics results in accurate loss estimation and provides useful advice for implementation of vector control for high performance of the motor. To lower the controller cost, optimised speed and current sensing algorithms are developed based on low-cost Hall Effect speed feedback and single-shunt current sampling. Finally, a prototype machine is manufactured and extensive tests are undertaken to validate the proposed low cost and high performance motor technology.