The Fletcher-Powell approach for large power system optimization

The application of non-linear programming techniques to optimal power system planning and operation problems has been demonstrated in the recent literature. The Fletcher-Powell technique appears to be more efficient with regard to the computer time requirement compared to the other available approaches. The technique, however, fails to provide convergence for optimization studies of large power systems. This paper presents a decomposition approach for the extension of the Fletcher-Powell technique to large power system studies. The application of the technique has been demonstrated by a sample system study. The suggested decomposition approach also requires less computer memory and time and provides a better stability of convergence than the earlier techniques.