Smart lighting system using ANN-IMC for personalized lighting control and daylight harvesting

Abstract

Lighting contributes a significant portion to the overall energy consumption in an office building. It is thus important to reduce the energy consumption of lighting systems especially for Net Zero Energy Buildings (NZEB). Maximizing daylight harvesting can significantly increase the energy savings. With increase in demand for satisfying occupant preferences in visual comfort, the need for personalized lighting in the office space is also rising. In this paper, a novel lighting control system for Net Zero Energy Buildings (NZEB) is proposed which models the lighting system using Artificial Neural Network (ANN) and utilizes this model with the Internal Model Control (IMC) principle for controller design. Modeling the lighting system using ANN reduces the challenge of modeling a large and complex system with inherent process variability without the need to analyze extensive data-sets. The proposed ANN-IMC controller uses feedback from sensors on the task table to maintain desired illuminance, is easy to tune with just one parameter and is robust to process variability. The proposed control design is applicable to square systems where the number of lights and number of sensors are equal. However, the proposed architecture can also be extended for controlling other lighting accessories such as roller blinds. The performance of the proposed lighting control system to harvest the daylight effectively is demonstrated using both simulation results and an experimental setup in test-bed environment. The versatility of the proposed system will allow an operator to deploy personalized lighting in an office space.

A Smart Adaptive Lighting System for a Multifunctional Room

“…The study also focused on the implementation on the on-off system to lower the energy usage [22] [23,24] so that the lights and the other related parameter such as blinds can be controlled to support the energy saving. Some approaches are deployed to gain better control for lighting, for example, the use of artificial neural network (ANN) to simplify the model of parameter tuning [25] and the use of reinforcement learning to gain the knowledge on the schedule-based and occupancy-based control scenarios and use it to control lighting and blind [26,27].…”

Consumption Optimization in an Office Building Considering Flexible Loads and User Comfort

“…Recently, a main concern in energy consumption minimization topics is user comfort [8] and keeping a balance between energy consumption minimization and user preference needs that are formulated with adequate constraints in order to observe optimization purposes and user easement at the same time [10]. Mostly, in office buildings, more attention is paid to air conditioners (ACs) [11,12], while according to [13], 29% of the total energy consumption in office buildings occurs due to lighting. The lights in offices are considered as flexible loads if these are controllable by existing equipment [14].…”

“…In [13], the authors proposed a smart lighting control based on the internal mode controller of an artificial neural network which maintained occupant's preferences while natural light was also used. In [15], a real model of an optimization-based SCADA model was presented, which focused on optimizing the energy usage of lights and AC devices of an office building in order to participate in DR events.…”

“…OF (1) is applied to minimize the power consumption of devices without consideration of thermal comfort of users in the equation, however, OF (2) is the complete form of objective function in (1), which aims to minimize the power consumption of devices and discomfort of the users. Equations (13) and (14) show the OF (1) and OF (2), respectively.…”

A Novel Adaptive Lighting System Which Considers Behavioral Adaptation Aspects for Visually Impaired People

“…Some approaches are deployed to gain better control for lighting by using Artificial Intelligence (AI). One of the AI methods is an Artificial Neural Network (ANN) to simplify the model of parameter tuning [29] and the use of reinforcement learning to gain knowledge on the schedule-based and occupancy-based control scenarios and use it to control lighting [30].…”