Emergency state voltage control using a tree search method

The problem of coordinated emergency state voltage control is addressed, with emphasis on discrete control actions such as capacitor bank switching, load shedding and tap blocking. The control problem is formulated as a combinatorial optimization problem and solved using a tree search approach similar to those used in chess computers. Two different evaluation functions based on a linearized system model are considered: (i) real part of the dominant eigenvalue of the dynamic Jacobian matrix; and (ii) linear prediction of voltages (on a time scale of seconds to minutes). The eigenvalue based evaluation function is shown to be unsuitable in presence of unsettled dynamics at the linearization point. Using two simulated test cases the tree search method with the model predictive evaluation function is shown to perform better than a rule-based strategy based on local measurements in terms of voltage response and the amount of load shedding necessary to stabilize the system.