Investigations On Solving the Load Flow Problem By Genetic Algorithms

This paper presents the application of genetic algorithms search procedures based on the mechanics of natural genetics and natural selection to the load flow problem in electrical power systems. Numerical experiments on smallsize networks show that the simple genetic algorithm (SGA) quickly finds an abnormal solution. By specifying an additional term in the objective function, the normal load flow solution can be found. An attempt to search for multiple load flow solutions simultaneously with the help of a sharing function is made and examined. Finally, an approach to improve the precision of genetic algorithm results is presented.