A Hybrid Model to Simulate Test Cases of Electrical Power Systems

Abstract

The development of solution algorithms for power system problems is based on hypothetical test systems and test cases. These systems are very scarce, and the degree of variability is relatively low. The constant development of the economic analysis in electrical power systems denotes the need to obtain standardized systems and cases. In this study, the creation of standardized test cases based on a hybrid model using Lévy alpha stable distributions and generalized additive models is proposed. The objective of the work is to present a methodological proposal for the creation of test environments for optimization models based on general information about the operation of particular power systems. The simulation of random values based on Lévy alpha stable distributions lets capturing the series impulsivity and demand peaks, and the use of generalized additive models permits capturing non-linearity in the behavior of the demand for electrical energy. The hybrid model will tolerate simulating as many instances as necessary, with a coherent behavior attached to the reality of the operation of the analyzed electrical systems.