The closed-loop performance of an inertial stable platform (ISP) affects the operation of navigation in moving objects. In this paper, the issue of high-performance control of ISP is discussed. By proposing a novel backstepping controller, the ISP plant with external stochastic disturbance, unknown dynamics, and actuator saturation is stabilized and regulated to the desired reference. Since unfamiliar terms appear in the practical ISP plant, a novel adaptive neural network model is suggested. To deal with the stochastic disturbance and modeling error, the stochastic bounded stability scheme is considered. Moreover, the practical problem of actuator saturation is involved in the design procedure. The suggested robust controller needs only one scalar adaptation law for all of the neural network gains. The virtual command inputs are propagated into a first-order filter to eliminate the conventional procedure of calculating the time-derivative terms. Eventually, a novel control technique is suggested for a nonlinear three Degrees of Freedom (3-DOF) ISP plant case study. Results illustrate the high performance of the robust controller in the presence of stochastic disturbance and input saturation.
Robust Backstepping Control Based on Neural Network Stochastic Constrained for Three Axes Inertial Stable Platform
Journal of Aerospace Engineering ; 35 , 1
2021-09-30
Article (Journal)
Electronic Resource
Unknown
Three Axes RLG Inertial Sensor Assembly
British Library Conference Proceedings | 1995
|Dynamic Neural Network-Based Robust Backstepping Control approach for Quadrotors
British Library Conference Proceedings | 2008
|AUTO-BALANCING TRANSPORTATION DEVICE WITH STABLE PLATFORM PIVOT AXES
European Patent Office | 2020
|