Motor Control Algorithms plays an increasingly important role in the control of modern drives. In a cascaded controller structure, a subordinate torque or current controller is required in addition to a speed controller and the superimposed position controller. The increasing utilization of the electric machine leads to saturation of the magnetic circuit, increased cogging torque formation, harmonics in the magnetic field and other secondary effects. The associated non-linear relationship between current amplitude and generated torque constantly increases the demands on the controller design.
In the area of conventional three-phase drives, i.e. the synchronous machine and the asynchronous machine, the control of longitudinal and transverse currents has largely established itself in the rotor-flow-oriented machine model. This allows the flux- and torque-forming current components to be adjusted independently of one another and represented for stationary operating points by constant variables.
The machine types of the transverse flux machine and the switched reluctance machine examined at the IEW do not form a magnetic rotary field. For this reason, a transformation into the rotor-flow-oriented representation in the classical sense is not possible. Therefore, the IEW is working on the creation of suitable models that map the physical properties of the electric machine in different degrees of detail. This allows the design of controllers under different aspects for different applications. For example, taking into account a minimum torque ripple, minimizing losses in special operating points or in a large operating range, for maximum dynamics or for optimizing positioning accuracy in actuators. The effects described above in highly utilized electrical machines place the highest dynamic demands on the subordinate control loop, which excludes the use of conventional PI controllers. The performance of the controller can be further improved by the close connection with machine knowledge from preliminary calculations, measurements or learning drives.
If the later power electronic topology and the used controller structure and strategy are already considered in the design process during the design steps of the electric machine, there is still a very large development potential in the optimization of the entire drive train. The use of a continuous development chain at IEW in MATLAB/Simulink enables controller design without design breaks. The designed controllers can be tested directly on different target platforms (dSpace, µC or FPGA) on prototypes in the laboratory and the simulation and controller models can thus be continuously validated and further developed.