Neuroadaptive Fault-tolerant PI Control of Nonlinear Systems with Unknown Control Direction

In this paper, we propose a low-cost and effective neuroadaptive PI control for MIMO nonlinear systems with actuation failures as well as unknown control direction. In addressing both square and nonsquare systems with unknown control direction, we make use of Nussbaum-type function and the matrix decomposition technique to build a generalized PI control with adaptively adjusting gains, which do not require the time-consuming “trial and error” process for determining the gains as in traditional PI control; Furthermore, the neural network unit is constructed with the help of barrier Lyapunov function to guarantee the crucial compact set precondition for neural network training signals. Both theoretical analysis and numerical simulation on 3D trajectory tracking of unmanned vehicle authenticate the effectiveness of the proposed method.