Backstepping Based Neuroadaptive Control for Uncertain Robot Systems

This paper gives a command filter backstepping method to design an adaptive controller to achieve position tracking for robot systems with uncertain parameters. Command filter is used to deal with computing complex problem of classical backstepping strategy. The neutral network is used to approximate uncertain dynamics. The error compensation signal is used to eliminate the error caused by the filtering. An example is applied to demonstrate effectiveness of control method.