On-line security optimisation of large power systems using artificial intelligence techniques

This paper presents a method for the robust on-line optimisation of power system transient stability and economy. It is intended that such a tool will be able to continually adapt to on-line changes in demand and availability in a power system and provide advice to power system operators to allow substantial cost savings and reduce the need for additional power system reinforcement. A detailed description of the method, which utilises the power of three different artificial intelligence techniques, is presented. This is accompanied by the results of a series of tests which demonstrate both the overall suitability of the method and clearly show a substantial improvement over traditional optimisation techniques.