International Journal of Computer Applications (0975 – 8887) Volume 184– No.25, August 2022 46 Smart Grid Management Modeling using Blockchain and Machine Learning Technologies Roberto Alexandre Dias Federal Institute of Santa Catarina (IFSC) 950, Av. Mauro Ramos Florianópolis – SC, Brazil Rafaela Oliveira de Azevedo Federal Institute of Santa Catarina (IFSC) 950, Av. Mauro Ramos Florianópolis – SC, Brazil Lucas Moino Armada Federal Institute of Santa Catarina (IFSC) 608, José Lino Kretzer Street São José – SC, Brazil ABSTRACT The present work consists of modeling a system for the maintenance of an infrastructure focused on the generation of electricity integrated into a management system for the sale of energy on the free market. Through this system, companies that provide micro grid generation infrastructure management services will be able to implement virtual power plants by aggregating microgrids implemented in their own or third-party physical spaces. In this way, the service provider will be able to manage the remote maintenance of its assets, aiming at the maintenance of the project specifications, the predictive maintenance of failures and verification of the loss of performance in components of the generation system. The system provides intelligence on contract management in a dynamic way, from data collected as well as computational cloud from consumers and aggregate generating units. Environmental parameters such as insulation, atmospheric and climatological conditions from weather forecast services, available in an open database, can be crossed with information from microgrids for capacity planning, in order to subsidize the sales contract management system of energy on the free market. By implementing the proposed system, it will be possible to define business models to commercially enable the adoption of the system. An example would be the model in which energy consumers act as service subscribers. In this way, the remuneration to the service provider can be made through a monthly fee or a portion of the energy generated in surplus. Acting in an aggregated way, the service provider will be able to carry out the best negotiation possible on the free market. Another example of a business model could be the remuneration of the owner of leased areas for the installation of energy generation microgrids, or power plants on land owned by the service provider. General Terms Smart grids, Block chain, Machine Learning Keywords Smart networks, Demand side management, Virtual Power Plant, Energy Market, Internet of Things, Predictive Maintenance 1. INTRODUCTION The growth in electricity production from alternative sources, due to the reduction in the cost of its assets along with the prospect of expanding the opening of the free market in Brazil, creates a need for development of intelligent solutions for the remote management of these assets. The possibility of aggregating distributed power generation systems constituting virtual power plants still has several open research challenges. Based on this scenario, the present work consists of the research and development of an infrastructure management system for electric energy generation microgrids through alternative sources, with emphasis, initially, on photovoltaic generation. According to the UBS Group (Swiss Investment Bank), the cost of solar and wind energy could be zero by 2030 in Europe. In the state of California (USA), a law already requires the installation of photovoltaic panels in new properties. In addition, many companies that need carbon credits are installing free solar panels in the homes of low- income people (2). A literature review (5) on Virtual Power Plants (Virtual Power Plants – VPP) is presented, in this work the authors present fundamental concepts about the technology of virtual power plants and its taxonomy. Among the types presented in the article, Commercial VPP is described as an aggregation of geographically dispersed power generation units that can be implemented and managed by an energy trading service provider. In accordance with this literature, the challenges from the point of view of communication infrastructure for the implementation of this service are emphasized. In (6) the need to use cloud computing solutions and large- scale data analysis to manage smart power grids of continental scope is discussed. In this model, the management of energy sale contracts is presented as a software service, where the infrastructure used in the management is presented as a commodity through the Internet. This model not only suggests the establishment of secure channels of communication but also describes the infrastructure components of cloud computing services necessary for the development of applications in a smart grid environment. Reference (7) shows a series of characteristics of a smart grid, among which two stand out: (a), a large number of sensors and monitors and (b), a vast amount of data. The need for this comes from the potential that information about such systems has on improving their operational characteristics. On the other hand, when discussing the future of electrical energy, one repeatedly reads that in a few years electricity will predominantly be processed by static converters in its final use, that is, the vast majority of electrical equipment will have either switched sources in its connection with the distribution networks, or rectifiers and inverters to control electric motors. This is already the case with distributed generation technologies, where photovoltaic solar generation is processed by static converters (DC-DC converters and inverters), wind