Small smart grid for the power utilization system
University of New Brunswick
The wide proliferation of renewable energy and deregulation of power grid systems require small power utilization systems to deploy intelligent methods of adjustment to the user power demand. In order to accomplish this goal, the smart power demand forecasting and power consumption optimization methods and algorithms need to be developed. For this purpose, small power utilization systems can benefit from the techniques developed for the smart grid in general. This thesis is devoted to the development of the forecasting model based on the Long Short-Term Memory (LSTM) method and the optimization model based on Genetic Algorithm (GA) adopted for the use in home energy management systems (HEMS). The thesis work presents the small smart grid architecture, which describes the roles of the LSTM and GA in small smart grid. The system also includes the load identification subsystem based on the Newton-phaselet algorithm. The LSTM method and GA are tested using Matlab simulation, as well as implemented using Python programming/scripting language with the PyTorch library. The forecasting LSTM algorithm is tested with a number of interruptible appliances to predict power demand over 24 hour segments. The experiments demonstrate that the developed algorithm generates a stable pattern of daily power demand; therefore, it has a high predictive power. The scheduling algorithm based on the developed optimization model is tested using photovoltaic power and stored battery power. The use of the developed algorithm allows automated shifting of power to achieve the lowest price without sacrificing user’s comfort.