Machine Learning Based Power Regulation for Smart Multi Storage Building

Abstract

The problem of power regulation has been well studied and there exist number of approaches in regulating the power supply to multi storage buildings. Some of the method uses number of supply units present in each floor, and number of control units present in the floors of the building. However, they suffer to achieve higher performance in power regulation. To handle this issue, a Genetic Algorithm Based power regulation model (GAPRM) is presented in this paper. The model considers various factors like number of electric stoves, no of heater, no of iron, no of vacuum cleaner, no of television, no of hair dryer, no of refrigerator, no of electric kettle, no of microwave oven, no of washing machines and Voltage requirement of each towards power regulation. The model fetches the details of each storage level and identifies the set of all electric devices present in the floor. Using these details, the method converts them into feature vector. With the incoming voltage, the method applies genetic algorithm to identify maximum units can be triggered with the supply at each floor. The genetic algorithm generates number of mutations and for each mutation; the method computes the fitness value in the name of Smart Voltage Fitness (SVF) measure. Based on the value of SVF, the method populates a subset of population. For the selected population, the method computes Power Regulation Value (PRV). According to the value of PRV, specific voltage has been regulated to the specific storage. This will be iterated for each storage level of the building. The proposed method improves the performance of power regulation in smart multi storage building.