Automatic transmission network capability assessment scheme in daily scheduling using neural networks

The paper presents a scheme for power system daily scheduling of network capability that uses a neural network (NN). For a given set of power transactions, the scheme detects overloaded lines taking into account their apparent power flow security limits. The NN identifies the overloaded lines for each time interval of the power system daily schedule as composed by the system operator and takes into account changes in transmission network topology. The optimal structure of the NN is obtained using the test of hypothesis procedure. As demonstrated on a 14-bus test system, the operation of the proposed scheme is quick and reliable.