This paper presents an innovative model for efficient home energy management through a Home Energy Management System (HEMS) based on Deep Learning and Reinforcement Learning. Disaggregated consumption meters have been installed in five household devices (refrigerator, air conditioner, fan, personal computer and lights) and a meter that calculates total consumption. Consumption data have been collected for 12 weeks, with different time periods for each device. Average consumption data per appliance has been obtained and electricity consumption curves have been generated for each appliance. The developed model shows significant potential for optimizing household energy consumption, improving energy efficiency and promoting more sustainable practices.