抄録
Magnetic levitation system is one system that floats a metal by power of absorption of magnet without contact. It controls a gap between coil and metal with information of its position. For practical example, we give magnetic bearings. But its non-linearity is big problem for controlling. For controlling this system, we first choose a parameter so that we can describe magnetic levitation system as linear parameter varying (LPV) system. And we try to use gain scheduling theory with online parameter information. LPV system, (A(p). B(p). C(p). D(p)), is depend linearly on parameter p. In gain scheduling theory, parameter gain of linear controller changes for scheduling parameter and it compensates for all range. In recent theory, which was originally proposed by Apkarian, we check stability in vertices for polytope system, then we obtain a stable controller inside of polytope. We additionally try to use Loop Shaping Method to get high performance. The method, which was originally proposed by McFarlane and Glover, demonstrates that if appropriate weighting functions are chosen for a considering model, a high performance controller can be obtained. This method, is compared to other methods, allows more freedom in choosing the weighting function to achieve the desired performance. Our theme is how to combine gain scheduling theory with loop shaping method.
