Room illumination model suitable for energy management systems

Controllable shading in living and business spaces can help with reducing the energy costs of heating and cooling while keeping the desired levels of illumination necessary for everyday activities.

As a first step, the illumination model of the Laboratory for Renewable Energy Systems conference room is developed using a combination of development tools such as Rhino and Grasshopper software. Interpretation of the input-output dependence in the illumination model is performed through implementation of a neural network based model. Three room configurations are considered: (i) only with daylight, (ii) with daylight and artificial lighting and (iii) with daylight, artificial lighting and planned controllable window shadings.

Verification and additional calibration of the model using surface illuminating sensors are performed combining it with the Raspberry Pi controller. Additionally, Raspberry Pi controller is used for fetching local measurements of solar irradiance from the laboratory meteorological station and the weather forecast for the location obtained from the Croatian Meteorological and Hydrological Service which will be used for control of the room shading with automated blinds after identification of the illuminance model.

Keywords: Daylight illumination, lighting, model identification, model predictive control, dynamic blinds

Title: Predictive control of natural lighting in an office space

Natural lighting can be used for creating pleasant working atmosphere and saving energy in the office. However it could also reduce the quality of the working atmosphere because of glare occurrence. Side effects when daylight is used for lighting can be reduced by implementing window shadings. When designing control algorithm it is very important to define the requirement that window shadings, which dose the amount of daylight in office, act very slowly and planned so they would not disturb the employees. In that case predictive controls is posed as a logical choice. In this paper, an office surface lighting control system of interest is established, which consists of a controllable shading element mounted on the window opening of the office room, a lighting sensor, an acquisition-control element and a server computer. Illumination data is stored in a database on a server computer and using a weather forecast a predictive daylight control algorithm is designed that can be implemented on the described illumination control system.

Keywords: daylight illumination, lighting, model identification, model predictive control, dynamic blinds