Customized Normalization Method to Enhance the Clustering Process of Consumption Profiles

Abstract

The restructuring of electricity markets brought many changes to markets operation. To overcome these new challenges, the study of electricity markets operation has been gaining an increasing importance.With the emergence of microgrids and smart grids, new business models able to cope with new opportunities are being developed. New types of players are also emerging, allowing aggregating a diversity of entities, e.g. generation, storage, electric vehicles, and consumers. The virtual power player (VPP) facilitates their participation in the electricity markets and provides a set of new services promoting generation and consumption efficiency, while improving players` benefits. The contribution of this paper is a customized normalization method that supports a clustering methodology for the remuneration and tariffs definition from VPPs. To implement fair and strategic remuneration and tariff methodologies, this model uses a clustering algorithm, applied on normalized load values, which creates sub-groups of data according to their correlations. The clustering process is evaluated so that the number of data sub-groups that brings the most added value for the decision making process is found, according to players characteristics. The proposed clustering methodology has been tested in a real distribution network with 30 bus, including residential and commercial consumers, photovoltaic generation and storage.

Remuneration and Tariffs in the Context of Virtual Power Players

“…This methodology is tested using a real smart grid with 82 consumers, that includes several consumers of different types (residential and commerce). This work proposes a clustering methodology that uses different data normalization methods and a new customized normalization method has been introduced [7]. Preliminary results demonstrated the advantages of data mining methodologies, based on the application of clustering process to group typical load profiles according to their similarity to allow proposing specific consumption tariffs to each group, so that consumers load profile is taken into account to meet the objectives of the SG aggregator.…”

Customized normalization clustering meth-odology for consumers with heterogeneous characteristics

“…Analysing the results of previous works (C. Ribeiro et al, 2013), (C. Ribeiro et al, 2016), is possible to verify that aggregation strategies have very good results and are very useful, because they provide a good separation according to what is intended.…”

“…The standardization factor must be carefully chosen taking into account the type of data available, the analysis that is intended to be carried out, as well as the type of final results desired to be obtained. Based on previous studies concerning the characterization of electric energy consumers (C. Ribeiro et al, 2016), the maximum power value of the representative load diagram of each consumer was selected as a normalization factor. With the application of the normalization factor, all load diagrams assume the same order of magnitude, belonging to the interval [0,1], being able to be used by clustering algorithms, in order to be grouped according to a criterion of similarity, without losing information related to differences between amounts of consumption among consumers.…”

“…These number of clusters were defined not to create a broad portfolio of tariffs. From (C. Ribeiro et al, 2015) it has been concluded that, by analysing MIA and CDI results from the clustering of non-normalized data, the best clustering results are achieved with the use of 3 clusters, as the clustering error is minimal. MIA and CDI results are presented in Figure 3.…”

“…Remuneration and Tariff Mechanism -RemT, (C. Ribeiro et al, 2013)(C. Ribeiro et al, 2015 is a decision support mechanism that is being developed to support the VPP actions in the definition of the best tariff and remuneration to apply to each of the aggregated players, regarding the VPP objectives and the individual goal of each aggregated player. VPPs in the scope of MASCEM use RemT to remunerate aggregated players, according to the results obtained in the electricity market, the penalties for breach of contract, contracts established to guarantee reserve, demand response programs and incomes of aggregated consumers.…”