Identification Modeling Based on RBFNN for an Aerial Inertially Stabilized Platform*

Abstract

Aiming at the serious influence of multi-source disturbances on the control precision of inertially stabilized platform (ISP), an accurate identification modeling method based on RBFNN for the ISP system is proposed. Since the ISP control system under the multi-source disturbances is a nonlinear, parameters uncertain, and time-varying system, the conventional modeling method cannot accurately describe the system characteristics. Therefore, more accurate modeling should be conducted. In the proposed modeling method, an off-line/on-line composite identification method is proposed to ensure the real-time performance in the dynamic adjustment process of the model, and the basis for the design of adaptive controller is designed. Besides, the simulation analysis and the experimental validation are performed and consistent conclusions are gotten.