The drive to save energy has become a global issue and people strive to make this a reality in all walks of life. Lightwave Technologies has researched, developed and released a product that saves energy in buildings using hybrid intelligent computing techniques. The product called the Intelligent Control of Energy (ICE) can potentially interface with any Building Management System (BMS) in the world over the Internet. Through the BMS, ICE directly affects the activation and setpoint control of the heating, ventilation and air conditioning equipment (HVAC). Energy savings are made by learning a building's thermodynamics and its response to the HVAC, training various computationally intelligent algorithms and sending control decisions back via the BMS that optimise energy usage whilst maintaining comfort conditions inside the building.

ICE has been designed around a service orientated model. This model leverages aspects of SOA (service orientated architecture) and alleviates the customer's burden of software maintenance, ongoing operation, and support. ICE's model consists of a centralised set of analysis and data handling servers which communicate to remote agents using protocols such as HTTP and XML. ICE's uses both predictive and data analysis algorithms, coupled with the weather forecasts for the region, to build a set of information models which define the building and learn its thermodynamic footprint. This information allows ICE to predict how a building will react to external and internal stimulus. It also allows ICE to spot inefficiencies and problems with the HVAC heating and cooling cycles.

ICE Optimisation Cycle

ICE uses iterative techniques to learn the building's reactions to various conditions. ICE will use both past historical data from the building and the current building state to generate a set of predictive models on how the building will react. These models coupled with adaptive algorithms allow ICE to make HVAC optimisations, which in turn reduce energy consumption during heating and cooling cycles.

While ICE is running, it continues to gather information about the building and its thermal interactions. This give ICE the ability to better understand the building producing more accurate optimisations and control decisions. ICE is a continuous optimisation system, always on and changing with the seasons. Figure 1 (to the left) show how a typical ICE optimisation cycle takes place. Both weather data and building information is loaded into the ICE predictive models. These models can accurately predict the future state of the building and how it will react to the conditions inside and out. The model is used by the decision subsystems during an analysis phase. The decision subsystem then produces a set of control decisions for the BMS of the modeled building. These decisions are relayed back to the remote ICE agents and put into effect. The decision and the operation feedback from the BMS are fed back into the predictive models and allow for subtle changes and optimisations for the next iteration.

ICE Communication

ICE is built on a standardised set of Internet protocols that have been proven to be robust while remaining flexible. An important part of the ICE system is the interaction with the building's BMS system. This is achieved using a small java enabled remote appliance. The appliance contains a number of BMS protocols, allowing ICE to interface with virtually any pre-existing BMS system without any retrofitting or engineering on the BMS side. The appliance contains a java enabled framework, which has allowed Lightwave to create a set of ICE specific libraries that allow for secure communication between ICE's servers and the client BMS. This effectively allows for the ICE servers to request information from any point in the BMS and also allows for complete control of these points. All communications is made using XML over HTTP and is designed to provided a standard interface into the BMS, providing transparent access irrespective of the BMS used. Comms can be wrapped in an SSL layer providing for enterprise levels of encryption and ensuring that the BMS data is completely secure.

Web Enabled BMS

Using ICE's BMS control portal - ICECube - it is possible to control and monitor an ICE enabled BMS from any web browser. This flexibility allows the user to monitor their building easily and gives the user the ability to monitor several remote building from one location. The web interface can provide real-time energy saving statistics, scheduling, setpoint control, building statistics, Alarming and reporting.

Reporting

The ICE system provides a number of details reports on the operation of a building and its HVAC equipment. ICE reporting can provide real-time building and energy statistics, long term energy usage information, trend analysis, detailed energy costs and savings reports.