Control of bilinear plants with actuator constraints; an l1 methodology

We propose a control scheme for l¹ type optimal control problems of bilinear plants with actuator constraints. The scheme involves an inner control loop which implements a feedback linearization, and an outer linear control loop that guarantees that actuator and state constraints are not violated. The hard-limit actuator constraints become linear constraints on the combined state and auxiliary control vector after feedback linearization. The linear controller is then designed to minimize an l¹ performance index while guaranteeing that the state control trajectories remain in the feasible set for all time, thus guaranteeing the stability and performance of the nonlinear feedback system. The design of the controller is equivalent to a certain linear programming problem