Modelagem e análise de um Smart Condominium Controller utilizando inteligência artificial

Abstract

The expansion of an emerging society in the age of artificial intelligence has led to technological advances and economic growth. However, it has also created challenges in terms of financial savings and residential comfort, both at the individual and condominium levels. As comfort expectations evolve with technological advancements, there is increased financial pressure to maintain or enhance residential comfort. This can create a conflict between the pursuit of comfort and the economic ability to achieve it, particularly if the energy chain fails to integrate and align with the supply-demand relationship in the energy matrix. In this context, it should be noted that it is possible to achieve a balance between energy consumption and residential comfort by operating at the level of the individual residence and at the condominium level. At the first level it is proposed the use of a Smart Home Controller (SHC) that uses Particle Swarm Optimization (PSO) to shift the charges residential schedules that can be activated at a lower cost and with maximum comfort. In addition, an algorithm with fuzzy logic is used to define the comfort levels of residential loads, taking into account the interaction with the user in relation to climate, temperature and humidity. In this way, it is possible to integrate user evidence into the SHC comfort model. At the second level, it is proposed the use of an Smart Condominium Controller (SCC) that uses a Long Short Term Memory (LSTM) to forecast the distributed generation (DG) capacity of electricity for the next period based on climate data. In addition, the SCC distributes the credits resulting from this generated energy among the residences in the condominium, following the terms of the Electric Energy Compensation System (EECS) in Brazil. To this end, studies are carried out on residential comfort models, residential load models, SHCs, SCCs, as well as techniques and interaction models between smart homes, DG and credit sharing between homes. The SCC/SHC integrated model includes a new mathematical model of comfort and ability to perform the dynamic management of electricity in a condominium, considering a macro and micro perspective of consumption and compensation of electricity in the EECS. As a result, comparisons between SHC residential scenarios with different comfort functions demonstrate gains in the consumer-comfort ratio for the user, both at the residential and condominium level, in addition to provide greater predictability of demand for the electricity supplier.