Abstract: Identification excitation signal design method is improved for a coarse-fine motion platform. A coarse-fine platform structure is widely used in long stroke, high precision motion platform systems, such as scan lithography tool motion platform and hard disk track-seeking platform. Obtaining an accurate model of course-fine system is an important basis for improving the control of the system, while system identification is an effective way to obtain the system model, in which the design of input signal is an important part of the test of system identification. The system model of coarse-fine motion platform is simplified to a multivariable finite dimensional linear time invariant discrete-time model. Multiple-input multiple-output (MIMO) identification method is introduced into coarse-fine system identification. An optimization design algorithm of excitation signal of multiple-input multiple-output model system identification is proposed. The algorithm parameterizes the multi-sine curve time domain signal and its power spectrum density. Then the parameterized excitation signal is optimized under limited power output. The results of theoretical analysis and simulation show that this method can effectively improve the performance of asymptotic covariance of the estimation model.

Key words: Coarse-fine, Excitation signal, Optimization design, System identification