Studies on nonlinear optimal control for power systems: a constructive method for choosing weighting matrices

The interconnection of regional power systems is developing rapidly in China driven by probable economical benefits; however, the interconnection makes the system more complex. It has been proved that nonlinear optimal controllers can improve system damping and increase stability limits remarkably. One of the key steps in the nonlinear optimal control law design is to choose the weight matrix Q and R for the optimal performance index. In this step, trial and error method and eigenstructure assignment based method are usually used but time consuming. In this paper, a new method for choosing the weighting matrices is proposed, in which the concept of generalized energy is applied and the feasible region of system state and control variables are considered. Comparing with commonly used methods, this method has a lighter calculation burden and is more intuitive and practical. Several cases are used to illustrate the rationality of this method.