Radial basis function-based state-dependent autoregressive (RBF-AR) models are a class of nonlinear combined models which can effectively describe some nonlinear dynamics. This paper focuses on the parameter estimation for the RBF-AR models. To overcome the estimation difficulty due to the highly nonlinear relations between the parameters and the model output, the separated idea is used to transform the original optimization problem into a quadratic and a nonlinear optimization problems. Applying the hierarchical identification principle and the multi-innovation theory, two interactive algorithms are proposed for the RBF-AR models. In addition, an approach based on data weighting is proposed to overcome the data saturation in the algorithms. The simulation results verify the effectiveness of the proposed algorithms from the aspects of parameter estimation accuracy and prediction performance.
In the radial basis function-based state-dependent autoregressive (RBF-AR) models with regression weights, the local linear models are included between the hidden layers and the output layers of the networks. The parameter estimation for the RBF-AR models with regression weights is studied in this brief. Considering the separable feature of the models, two criterion functions based on the increasing data length are defined to fit the observation data of the whole dynamical process. Two sub-algorithms are proposed by minimizing the criterion functions. Aiming to overcome the existence of the singular matrix during the Newton search and to make the algorithm more stable, a positive definite diagonal matrix is introduced to the algorithm. Based on the hierarchical principle, a hierarchical Newton recursive algorithm is proposed, which can realize the on-line parameter estimation. Simulation results verify the validity.
28 January 2024 in International Journal of Adaptive Control and Signal Processing
Most satellite attitude control models are built analytically, which requires a clear understanding of the kinematic equations of the satellites and the various disturbance models to deal with the interferences and uncertainties in the space environment. This paper studies a system identification method to build a system model for satellite attitude control. The satellite model is based on a neural network model, i.e., the RBF-ARX model, which requires only observation data. Unlike the existing off-line identification methods for the RBF-ARX model, this paper proposes an on-line identification method that can constantly correct the model according to the newest data. To reduce the computational complexity of identifying the satellite model, a decomposition scheme is also developed. The computational efficiency of the proposed method is analyzed. Finally, the efficacy of the proposed identification method is demonstrated by applying it to two satellite attitude control systems.