Search

Deadbeat predictive current control (DPCC) can predict motor behavior based on SPMSM model. However, during the operation of motor system, motor parameters (such as stator inductance and flux linkage) vary frequently according to different working conditions, which may lead to controller parameter mismatch, causing current harmonic content to increase and efficiency to decrease. In order to solve these problems caused by parameter variation, first, this paper proposes a current prediction error model by considering uncertainties of model parameters. Second, stator inductance and flux linkage are decoupled based on current prediction error model, which can reduce the interaction between parameters. Finally, the Kalman Filter (KF) algorithm is presented to filter the decoupled parameters. It is shown that the stator inductance and flux linkage can be identified accurately and the complexity of computation can be simplified. The traditional DPCC method, Extended Kalman Filter (EKF) based DPCC method and the proposed DPCC method are comparatively analyzed in this paper. Simulation and experiment indicate that the proposed parameter decoupling identification method can effectively reduce current harmonic content, current fluctuation and current tracking errors caused by parameter mismatch. [ABSTRACT FROM AUTHOR]

Speed control during startup is one of the most critical tasks of gas turbine power plant operation. This paper introduces the PI fuzzy gain-scheduling (PI-FGS) controller to solve the foremost speed-control problems, including tracking of the acceleration pattern and rejection of disturbances caused by operation events throughout startup. Fundamentally, the PI-FGS synthesizes a GS controller from multiple locally tuned generalized proportional integral (PI) algorithms by means of a fuzzy system. The PI-FGS structure permits independent tuning of the tracking and rejection responses at any operating point. In addition, tuning can be carried out on-site based on the operator experience and inspection of the current plant response. Simulation experiments show that the PI-FGS may improve speed-control performance well over that provided by commercially available speed controllers, which are based on a single conventional PI algorithm. [ABSTRACT FROM PUBLISHER]