Advanced Control of Indoor Thermal Environments Using Fan Coil Units

The work described in this paper relates to advanced control systems, specifically designed for heating, ventilation and air conditioning in office buildings. This work specifically focuses on the use of state of the art fan coil units with advanced instrumentation and control. The premise of the work is that control systems can be significantly enhanced by using real-time data from a distributed sensor network deployed in the building. Specifically, the performance of control systems can be improved by augmenting predictive (feed-forward) control operations with techniques to improve the accuracy of models. A control algorithm for heating, ventilating and airconditioning systems is described in this paper that integrates an advanced feedforward control algorithm with conventional feedback control. This paper further contains a description of a functional prototype used to demonstrate the proposed control algorithm for indoor thermal environmental control. The test-bed used in this work – the Robert L Preger Intelligent Workplace (IW), at Carnegie Mellon University, involves a large number of variables and hence a complex control task, i.e., the test bed contains multiple sources of thermal energy, and multiple constraints and disturbances – both measurable and immeasurable. The algorithms demonstrated in this test-bed are expected to perform satisfactorily on other environments with smaller number of variables. This paper contains a description of experiments that were performed to validate the comfort and energy benefits of increased sensing using fan coil units that are in installed in two spaces in the IW. INTRODUCTION A driver in the development of building control systems is the possibility of facilitating increased occupant comfort while adhering to constraints for performance and energy usage. This work is a novel sensor-based model reference control scheme to achieve improved thermal performance and energy effectiveness with an emphasis on the predictability of the indoor thermal environment. Thermal performance is addressed as system response time and the ability to track a setpoint, factors that are assumed to affect occupant comfort and productivity. A typical control strategy for HVAC system is based on one element – the comparison of measured space air temperature with the air temperature setpoint from the occupant. In this work, a method to compute "Comfort Temperature" is developed that is used to calculate the setpoint temperature required using a model of the occupant. Goals of this work include the demonstration of the comfort and energy benefits of increased sensing, an increased maintenance of comfort conditions, and a reduced cost of operation through the use of advanced control algorithms. This work included the selection, installation, commissioning and testing of advanced fan coil units along with the necessary sensing and actuation instrumentation and hardware. These were installed in two open plan offices in the Robert L Preger Intelligent Workplace (IW), at Carnegie Mellon University. The control strategy developed in this work improves on existing strategies by using integrated feedforward and feedback loops that effectively use data from a large number of sensors measuring various quantities (air temperature, surface