Identification of continuous-time nonlinear systems via local Gaussian process models

被引：0

作者：

Hachino, Tomohiro ^{[1
]}

Matsushita, Kazuhiro ^{[1
]}

Takata, Hitoshi ^{[2
]}

Fukushima, Seiji ^{[1
]}

Igarashi, Yasutaka ^{[1
]}

机构：

[1] Graduate School of Science and Engineering, Kagoshima University, 1-21-40, Korimoto, Kagoshima

[2] Kagoshima University, 1-21-40, Korimoto, Kagoshima

来源：

IEEJ Transactions on Electronics, Information and Systems | 2014年 / 134卷 / 11期

关键词：

Continuous-time system; Gaussian process model; Local model; Nonlinear system; System identification;

D O I：

10.1541/ieejeiss.134.1708

中图分类号：

学科分类号：

摘要：

This paper deals with a nonparametric identification of continuous-time nonlinear systems using multiple local Gaussian process (GP) models. Multiple sets of training input and output data are collected to train the local GP prior models. Each local GP prior model is trained by minimizing the negative log marginal likelihood of each set of the training data. The final nonlinear function with confidence measure is estimated by weighted mean of the local estimated nonlinear functions using the predictive variances of local GP posterior distributions. Compared to the standard GP-based identification method, the proposed method can reduce the computational cost and improve the accuracy of identification. Simulation results are shown to illustrate the effectiveness of the proposed identification method. © 2014 The Institute of Electrical Engineers of Japan.

引用

页码：1708 / 1715

页数：7

共 19 条

[1]

Unbehauen H., Rao G.P., Continuous-time approaches to system identification - A survey, Automatica, 26, 1, pp. 23-35, (1990)

[2]

Sagara S., Yang Z.J., Wada K., Identification of continuous-time systems, Systems, Control and Information, 37, 5, pp. 284-290, (1993)

[3]

Jin C.Z., Wada K., Hirasawa K., Murata J., Sagara S., Identification of nonlinear continuous systems by using neural network compensator, IEEJ Trans. EIS, 114, 5, pp. 595-602, (1994)

[4]

Tsang K.M., Billings S.A., Identification of continuous time nonlinear systems using delayed state variable filters, Int. J. Control, 60, 2, pp. 159-180, (1994)

[5]

Liu G.P., Kadirkamanathan V., Stable sequential identification of continuous nonlinear dynamical systems by growing radial basis function networks, Int. J. Control, 65, 1, pp. 53-69, (1996)

[6]

Hachino T., Karube I., Minari Y., Takata H., Continuous-time identification of nonlinear systems using radial basis function network model and genetic algorithm, Proc. of the 12th IFAC Symposium on System Identification, 2, pp. 787-792, (2001)

[7]

Hachino T., Takata H., Identification of continuous-time nonlinear systems using a model expanded by a basis of automatic choosing functions and genetic algorithm, IEEJ Trans. EIS, 119, 7, pp. 848-857, (1999)

[8]

Hachino T., Takata H., Identification of continuous-time nonlinear systems by using a gaussian process model, IEEJ Trans. EEE, 3, 6, pp. 620-628, (2008)

[9]

O'hagan A., Curve fitting and optimal design for prediction (with Discussion), Journal of the Royal Statistical Society B, 40, pp. 1-42, (1978)

[10]

Williams C.K.I., Rasmussen C.E., Gaussian processes for regression, Advances in Neural Information Processing Systems, 8, pp. 514-520, (1996)

← 1 2 →