Comparative studies of adaptive algorithms

This paper presents a generalized truncated least squares (GTLS) adaptive algorithm and comparative studies of adaptive algorithms. The proposed algorithm is directly derived from the normal equation of the generalized truncated least squares method (GTLSM). The special case of the GTLSM, the truncated least squares (TLS) adaptive algorithm, has some distinct features which include the case of a minimum steps estimator. This algorithm appears to be best in the deterministic case. For applications in the presence of disturbances, the GTLS adaptive algorithm is more effective. Since the GTLS adaptive algorithm includes many types of algorithms, such TLS and LS adaptive algorithms, as special cases by setting parameters included in the GTLS algorithms appropriately. This algorithm is also able to realize the case of model matching control when the algorithm is used as an adaptive law in adaptive control. In these cases, a two stage design method is used which combines the design of the adaptive control system with that of conventional control, where each can be treated independently. Using this method, the validity of the presented algorithms is examined by the simulation studies of an indirect adaptive control. [In Japanese.]