Due to the large size of electric power systems, there is a very high computational burden when obtaining the optimum network by using classical optimization techniques. Several authors have used heuristics and/or sensitivities in order to guide the search of optimal network investments. This paper proposes an Automatic Learning approach in order to decide whether a network change will improve the overall costs or not. More specifically, Decision Trees methods are used to identify a set of simple and reliable rules which combine criteria based on both heuristics and sensitivities. These decision trees are integrated in a subtransmission planning tool, improving dramatically both the "optimality" of the resultant network and the computational time.
The security criteria of a power system require that branch power flows and bus voltages are within their limits, not only in normal operating conditions but also when any credible contingency occurs. In the Spanish electricity market, voltage constraints are solved by connecting a set of offline generators located in the areas where they occur. Thus, for a market participant it is necessary to predict approximately when its generating units are connected in order to prepare the annual budget and/or decide the time and location of new plants. This paper proposes a methodology based on decision trees to estimate the daily load pattern of units that have not been cleared in the daily energy market and can be connected to alleviate the network constraints in the Spanish power system. The method explains the behavior of the daily load patterns of a nonconnected unit (obtained through clustering techniques) with a set of explanatory variables. The explanatory variables consist of the demand-generation imbalance in the electrical area of the generating unit and the maintenance scheduling of the transmission lines that feed the area. The method proposed is illustrated with a case study.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.