Neural-network-based near-optimal control for a class of nonlinear descriptor systems with control constraint

The near-optimal control problem for nonlinear constrained descriptor systems is solved by greedy iterative DHP(GI-DHP) algorithm. The descriptor system is first conceptually reduced to a state space form and then a nonquadratic functional is developed in order to deal with the control constraint problem. Then the GI-DHP algorithm is proposed to solve the optimal control problem of the state space system. For facilitating the implementation of the iterative algorithm, two neural networks are utilized to approximate the costate function and compute the optimal control policy respectively. An example is given to demonstrate the validity and feasibility of the proposed optimal control scheme.