Neural Network Trajectory Tracking Control on Electromagnetic Suspension Systems

A new adaptive-like neural control strategy for motion reference trajectory tracking a nonlinear electromagnetic suspension dynamic system is introduced. Artificial networks, differential flatness and sliding modes are strategically integrated in the presented adaptive network design approach. The robustness efficiency of magnetic on desired smooth position profile can be improved this fashion. single levitation parameter tuned on-line from perspective by using information error signal only. mode discontinuous action approximated network-based continuous function. Control firstly developed theoretical modelling physical system. Next, dependency substantially reduced integrating B-spline networks technique. On-line accurate estimation uncertainty, unmeasured external disturbances uncertain nonlinearities conveniently evaded. effective performance robust approach depicted multiple simulation operating scenarios. capability active disturbance suppression furthermore evidenced. based extended to other controllable complex systems where internal represent relevant issue. Computer simulations analytical results demonstrate method.