Robust DSP-based efficiency optimization of a variable speed induction motor drive

This paper proposes a practical digital-signal-processor (DSP)-based algorithm for the online efficiency optimization of induction motor drives. The algorithm combines good features of the search control and loss-model-based control, while simultaneously eliminating their major drawbacks. It ensures very fast convergence without the need for any a priori knowledge of the motor loss-related parameters. The minimum loss operating point is decided from a functional approximation of the motor and the power converter losses, in the form of a suitably defined loss function. The loss function parameters are obtained online from input power measurement and a dedicated identification routine acting in conjunction with the common drive control functions. Proposed online identification of the loss function parameters relies upon the signals and states readily available within the drive controller memory and does not involve the user entry nor the initialization procedure. The proposed controller is coded into a conventional low-cost 16-bit DSP and verified on a 2.2-kW induction motor drive prototype.