Artificial intelligence applied to on-line transient stability assessment of electric power systems

A novel approach to fast transient stability assessment of power systems is presented. It consists of inductive inference which allows the construction of decision rules derived from off-line simulation data. The proposed procedure derives from ID3 by Quinlan, appropriately modified so as to handle a large number of continuous variables. The decision rules are expressed in terms of static, controllable variables. This makes the method able to solve both on-line transient stability analysis and preventive control problems.