Nonstationary optimal positioning controller design using GA meta-optimization

For positioning control problems of vibration systems, a nonstationary optimal regulator (NOR) is known as one of the effective control methods. NOR corresponds to an LQ regulator based on a time-varying criterion function including time-varying weights. In conventional studies on NOR, the time-varying weights were determined by iteration of numerical simulation. Taking control constraints into account, this study shows NOR design using GA (Genetic Algorithm) meta-optimization for the determination of the time-varying weights. This method is applied to a single-degree-of-freedom vibration system positioned by a DC servo motor and a ball screw. The effectiveness of this method is verified by numerical calculations.