ABSTRACT
The so-called smart cities are among the main aspirations of modern society. Among the premises of the “smart cities” are the interconnection of communication systems, the monitoring of services, and the rational and optimized use of energy. The introduction of the Smart Grid concept produces a convergence between the energy generation, transmission and distribution infrastructure, and the digital communications and data processing infrastructure. This chapter aims to create a current signal database of domestic loads and proposes a technique for identifying such loads; for this, a new algorithm is proposed for the disaggregation of cartoons based on the current measurement of electrical equipment in a home, a necessary step for the disaggregation of loads in the Smart Grid context. At the same time that was created, a method that uses wavelet transform coefficients and neural networks allows the load to be distinguished. The disaggregation of the proposed technique is based on the use of neural networks and wavelet transform. The identification of electrical loads aims to discover what equipment is connected to utility power. The task of load disaggregation is complex and is aggravated in the presence of nonlinear loads, where the harmonics generated by one equipment modify the behavior of the signals measured in others. Thus, with this proposed methodology, it is possible to calculate separately for each device the electricity consumed. The algorithm processing and load identification by wavelet and neural networks were developed using the MATLAB environment. The results provide the efficiency of the proposed technique.
