Internal Project fully funded by CeDInt internal resources.
This project describes the design and development of a HVAC control system for a commercial building. This system is aimed to optimize user comfort and to reduce energy consumption obtained by current commercial control systems.
Most of commercial HVAC control systems work in a fixed and predetermined way. The novelty of the proposed system is that it adapts dynamically to the user and to the building environment.
For this purpose the system has been designed under the concepts and technologies proposed by the paradigm of Ambient Intelligence. The system also takes advantage of new technologies like sensor networks and smartphones.
The main goal of this Project is to develop a HVAC control system for a commercial building that optimizes users’ thermal comfort while reducing energy consumption of HVAC devices.
In order to achieve this goal, the system has to be able of monitoring and controlling indoor and outside ambient parameters and the indoor units associated with each space. The system also has to provide users easy-to-use interfaces that allows them to control and configure devices operation, based on their own preferences.
This goal involves completing the following tasks:
- State-of-the-art of HVAC control system and Ambient Intelligence
- Design the overall solution
- Design and development of a central software control platform
- Design and development of an thermal comfort algorithm
- Development of an user indoor location system
- Implementation and test over a real-world environment
A control algorithm for thermal comfort has been developed aimed to calculate optimal configuration parameters (those providing thermal comfort of occupants) of each indoor unit based on multiple heterogeneous information sources. This is a mixed-AI model algorithm. In it, the problem has been divided into different modules based on the model units of the HVAC system. Each module solves a particular subproblem by employing techniques using Artificial Intelligence (AI) (if necessary).
The test environment has been the south-west wing of CeDInt building, comprising: 5 rooms, 8 Indoor units Daikin VRV and 13 users. From the results it is concluded that thanks to ClimApp system 43% working time of climate equipment has been reduced.
- Advances in new technologies make it possible to develop applications based on AmI in real-world systems:
- Sensor Networks
- Smart Phones
- The incorporation of concepts and technologies of Ambient Intelligence (AmI) systems can improve commercial control systems.
- The developed system implements a system based on thermal comfort using existing equipment and with a minimum investment control.
- Locating system developed allows to detect and identify users without preconfigure or training.
- Database server: MySQL, JDBC
- Sensor Network and HVAC control technology: KNX, Calimero
- Concepts and technologies from Ambient Intelligence (AmI)
- Algorithm development: MATLAB
- Machine learning algorithms: Artificial Neural Networks
- Thermal comfort methodology (Fanger’s equations): Novel development of a thermal comfort control algorithm for multiple users and multiple indoor units in the same space.
- WiFi-based indoor location system: Android, Novel development of a self-learning algorithm for obtaining WiFi signals fingerprints of each space.
- Web server application: ClimApp.ear (Java EE)
- Android indoor location app