Adaptive control of unknown fuzzy disturbance-based uncertain nonlinear systems: application to hypersonic flight dynamics

Keeping into view the human-brain-like capabilities, a novel adaptive control scheme utilizing a defuzzification module has been designed in an effort to endow the control systems with human-like capabilities of learning and processing human-understandable information. It’s been noted that within real-time systems, uncertain and undefined external disruptions can manifest in various ways due to their ambiguous and unpredictable nature. Consequently, attempting to encapsulate these disturbances within precise closed-form mathematical expressions is often impractical. Therefore, opting to describe such ambiguous and incomplete information using linguistic variables, as opposed to rigid mathematical formulations, is a more suitable approach. In this regard, a novel class of uncertain nonlinear systems in non-strict feedback involving fuzzy variables is introduced in which the fuzzy variables are used to express unknown disturbances. As a result, such class is systems become more general and human-interactable but alongside the control of such class of systems becomes more complicated and difficult. Therefore, the goal of this work is to provide an efficient intelligent control technique for the proposed class of systems which can successfully handle these uncertainties and attributes of fuzziness. The proposed control scheme is therefore consists of (a) a defuzzification module to handle fuzzy variables and (b) radial basis function neural network (RBFNN) to approximate the unknown functions. The existing category of uncertain systems is managed by converting the system into an n-step ahead predictor and applying the suggested control approach. By leveraging Lyapunov theory, it can be demonstrated that the comprehensive closed-loop system achieves semi-global uniform ultimate boundedness (SGUUB), and it is established that the error tends to diminish towards zero. Additionally, the computational complexity of the overall control scheme is also analyzed. The control scheme is validated through numerical simulations for its effectiveness and real-time applicability. A simulation example based on hypersonic flight dynamics model developed by NASA-Langley Research Centre for longitudinal dynamics of a hypersonic vehicle is used to demonstrate the scheme’s reliability and real-time applicability.