Improved Fuzzy Load Models by Clustering Techniques in Distribution Network Control

In operation and planning of the power systems, the analysis of the consumption trends depends on the evolution of economic activities and competition among several sources of energy, which affect forecasts. The loads estimation represents the basis of the system state estimation and influences various aspects of power system planning such as: transformer and conductor sizing, capacitor bank placement and so on. The main difficulties in modeling of the nodal loads result from the random nature of loads, the deficiency of measured data and the fragmentary and uncertain character of information on loads and customers. Thus, a modern method for expressing the uncertainty in load models is fuzzy technique. A fuzzy set is a set containing elements that have varying degrees of membership in the set. There are different ways to derive membership functions. Subjective judgment, intuition and expert knowledge are commonly used in constructing membership function. Because in many situations the choices of the membership functions are subjective, in the paper the clustering techniques are proposed for the improved of the defining of membership functions corresponding to the load profiles and customers consumption categories. Results obtained demonstrate the ability of the fuzzy load models to overcome difficult aspects encountered in process control and operation problems.