This paper presents a novel approach to convert a conventional house air conditioning installation into a more efficient system that individually controls the temperature of each zone of the house through Wi-Fi technology. Each zone regulates the air flow depending on the detected temperature, providing energy savings and increasing the machine performance. Therefore, the first step was to examine the communication bus of the air conditioner and obtain the different signal codes. Thus, an alternative Controller module has been designed and developed to control and manage the requests on the communication bus (Bus–Wi-Fi gateway). A specific circuit has been designed to adapt the signal of the serial port of the Controller with the communication bus. For the acquisition of the temperature and humidity data in each zone, a Node module has been developed, which communicates with the Controller through the Wi-Fi interface using the Message Queuing Telemetry Transport (MQTT) protocol with Secure Sockets Layer / Transport Layer Security (SSL/TLS) certificates. It has been equipped with an LCD touch screen as a human-machine interface. The Controller and the Node modules have been developed with the ultra-low power consumption CC3200 microController of Texas Instruments and the code has been implemented under the TI-RTOS real-time operating system. An additional module based on the Raspberry Pi computer has been designed to create the Wi-Fi network and implement the required network functionalities. The developed system not only ensures that the temperature in each zone is the desired one, but also controls the fan velocity of the indoor unit and the opening area of the vent registers, which considerably improves the efficiency of the system. Compared with the single-zone system, the experiments carried out show energy savings between 75% and 94% when only one of the zones is selected, and 44% when the whole house is air-conditioned, in addition to considerably improving user comfort.
A novel approach for ultra low-power WSN node generation