Recurrent Adaptive NeuroFuzzy Paradigms

Vehicle stability and comfort of passenger is greatly in influenced by the uneven and bumpy roads. When visualizing the conventional semi-active and passive systems, they have fixed parameters and can not adapt with uneven roads. Main objective of this work is to get enhanced active suspension of the full car model. Mathematical modeling of full car model is evaluated and update parameter equations of the active controllers, i.e. Recurrent Fuzzy Wavelet Neural Network-1(a),Recurrent Fuzzy Wavelet Neural Network-1(b), Recurrent Fuzzy Wavelet Neural Network-2(a), Recurrent Fuzzy Wavelet Neural Network-2(b), Recurrent Fuzzy Wavelet Neural Network-3(a) and Recurrent Fuzzy Wavelet Neural Network-2(b) controllers have been derived. Then these controllers are tested using MATLAB/Simulink on full car model. At the end comparison of all the above mentioned controllers with passive suspension and semi-active suspension is done. It is concluded that these active controllers have greatly improved the displacement and acceleration of Heave, Pitch, Roll and Seat of car, as a result comfort of passenger and vehicle is improved.

Majida Nazir is involved in teaching and supervision in a defense organization. Laiq Khan is a professor of Power System Dynamics and Control in the department of electrical engineering, COMSATS IIT Abbottabad, Pakistan. Shahid Qamar is a lecturer in the department of electrical engineering, COMSATS IIT, Abbottabad, Pakistan.