| Summary: | 碩士 === 國立中興大學 === 電機工程學系 === 91 === This thesis develops methodologies for positioning of a single-sided induction motor (SLIM). Although the PID controller have been extensively used in industry. Such a controller would not given good tracking performance, when the transfer functions or the parameters of plants vary with time, or unpredictable disturbances occur. To deal with these shortcomings, we propose a new type of control structure, which combines a simple recurrent neural-networks identifier (SRNNI), and a simple recurrent neural-networks compensator (SRNNC). The SRNNI and SRNNC make use of the dynamic back-propagation (DBP) method for updating their weightings and ensure that the output of the plant is able to follow the desired output generated by a reference model.
Computer simulation results have used to verify the effectiveness and feasibility of the proposed control method. Through experimental results, the proposed positioning control for the SLIM has been show to outperform the traditional PID control, and to given an excellent control efforts.
|
|---|
