Kentaro Kameyama

Subspace identification for continuous-time stochastic systems via distribution-based approach

This paper presents a novel approach for system identification of continuous-time stochastic state space models from random input-output continuous data. The approach is based on the introduction of random distribution theory in describing (higher) time derivatives of stochastic processes, and the input-output algebraic relationship is derived which is treated in the time-domain. The efficacy of the approach is examined by comparing with other approaches employing the filters.

Subspace-based prediction of linear time-varying stochastic systems

In this paper, a new subspace method for predicting time-varying stochastic systems is proposed. Using the concept of angle between past and present subspaces spanned by the extended observability matrices, the future signal subspace is predicted by rotating the present subspace in the geometrical sense, and time-varying system matrices are derived from the resultant signal subspace. Proposed algorithm is improved for fast-varying systems. Furthermore, recursive implementation of both algorithms is developed.

Recursive subspace identification for continuous-/discrete-time stochastic systems

Abstract A recursive algorithm is derived for subspace system identification which is flexible and applicable to continuous/discrete-time, time-invariant/varying stochastic systems. Efficacy of the algorithm is demonstrated by simulations.

Subspace-based Identification of Stochastic Systems Using Innovation Model

Abstract In this paper a 4SID algorithm is proposed to identify a class of linear stochastic systems from the noisy input-output data sequence. First, the standard linear stochastic models are replaced equivalently by the innovations representation of Kalman filter equation. Then, by introducing a quasi-stationarity assumption for the error covariance matrix associated with the Kalman filter, our 4SID algorithm is derived. Finally, by simulation studies the effectiveness of our algorithm is shown.

Recursive subspace prediction of linear time-varying stochastic systems

Abstract In this paper, a new subspace method for predicting time-invariant/varying stochastic systems is investigated in the 4SID framework. Using the concept of angle between past and current subspaces spanned by the extended observability matrices, the future subspace is predicted by rotating current subspace in the geometrical sense. In order to treat even time-varying system, a recursive algorithm is derived for implementation. The proposed algorithm is tested by simulation experiments.

Subspace-based identification for continuous-time stochastic systems via distribution-theoretic approach

Abstract Subspace-based identification methods have recently been attracted much attention and active researches for the algorithm of discrete-time stochastic systems are done, and our interest turns to identifying the continuous-time systems. In this paper, we propose a new method to handle the continuous-time stochastic systems in the subspace identification algorithm. Our proposed method uses the concept of the random distribution and describes the derivatives of the (non-differentiable) white Gaussian noise process in the sense of distribution. Then, we construct the input-output algebraic relationship by using the differential information date. Furthermore, signal subspace and quadruple system matrices ( A, B, C, D ) were derived by the algorithm based on subspace framework, and an illustrative numerical example is provided.

Subspace identification of the SYSID2006 benchmarks via distribution-based approach

Abstract This paper presents solutions to the IFAC SYSID2006 Benchmark identification problem by applying the subspace identification via distribution-based approach proposed in Ohsumi, Kameyama and Yamaguchi (2002). The application of the algorithm seems to have succeeded in establishing continuous-time models for the benchmark data.