Online search based fuzzy optimum efficiency operation in steady and transient states for DC and AC vector controlled motors

This paper presents an efficiency optimization control system which can be incorporated as an outer loop to a standard indirect vector speed control system for both AC and DC motor drives (i.e. Induction motor, Permanent Magnet Synchronous motor, Synchronous Reluctance motor and DC motor drives), when not operating at full load. The proposed control system adjusts the air-gap current space vector i
_od
(magnetizing current) while meeting the speed and load torque demands. During transient states, the optimization objective is the minimization of the motor power losses by a loss model controller (LMC), based on a generalized d-q loss-minimization model expression for the above specific motors. On the other hand, during steady states, the optimization objective is the minimization of the drive power losses by using a fuzzy logic search control system (FLSC). The FLSC system combines two fuzzy logic search controllers, one introduced as FLSC1 and a second one introduced as FLSC2. The FLSC1 is activated after the LMC and the FLSC2 after the FLSC1, when a higher torque or speed demand occurs. The LMC estimates a sub-optimum magnetizing current command signal, which is used as initial value from the FLSC system to accelerate the on line search process. Subsequent fine adjustments are made by the FLSC1 by decreasing the magnetizing current in steps, or by the FLSC2 by increasing in steps the last magnetizing current command signal at minimum. The torque current command signal is generated by a fuzzy controller introduced as FLC3. The drive's power losses are calculated at every sampling step as the difference between the measured input DC link power and drive's output power. The change of drive power losses are calculated over a constant time interval as it is needed by the efficiency fuzzy control algorithm. When there is no a sufficient knowledge of the motor model and motor's parameters, the proposed control system benefits from the best characteristics of both controllers: high speed of the LMC, and optimum suggested value of magnetizing current of the FLSC system. Due to the use of FLSC2 and FLC3, the drive system is keeping good dynamic response. The LMC and the fuzzy logic controllers prototypes were created and tested using Simulinkreg. Simulations are accomplished by using Simulinkreg in order to validate our method. In order to validate the importance of the efficiency optimization during transient states, the same simulations are done with and without including the computations by LMC.