Design Of The Indoor Lighting Control System Based On Neural Network

With the rapid development of China's economy,the electricity consumption of the society has been continuously increasing.In 2018,the electricity consumption of the society was 6844.9 billion kWh,which is up by 8.5% compared with last year.In China,indeed,the lighting need accounts for 10% of household electricity consumption and more than 30% of commercial building electricity consumption.All those results demonstrate that the power consumption of architectural lighting constitutes a significant part of China's energy consumption.And in order to improve the utilization ratio of power resources,naturally,then,a more environmentally friendly lighting solutions is needed.The environmental friendliness of the lighting scheme has three dimensions,namely,the requirements of energy-saving,of low-budget and of comfortableness.Concretely speaking,energy-saving means less power resources consumption and a reduction in the energy consumption of building lighting compared to the traditional scheme;low-budget means that the lighting scheme needs to save manpower and material resources;comfortableness means to improve the lighting visual comfort,increasing indoor illumination uniformity and reducing glare.Based on the neural network prediction function to achieve interior luminous environment control,the solution proposed by the subject aims to discover the best equilibrium pointbetween those three requirements of energy-saving,economy and comfort.Previously,a main part of existing louver control schemes relies on established models of louver angles and illumination distributions.But we could commonly discover that such models have been simplified in the process of abstracting optical phenomena into mathematical descriptions,which makes it difficult to accurately describe the indoor illumination distribution,and thus the natural illumination distribution does not meet the design requirements.In addition,existing LED lighting control solutions tend to rely on multiple illuminance sensors in the room which will make interferences with the activity of people.Fisrt,in order to solve the problem of louver control,this paper adopts the analysis of the undone equivalent transmission model and the Dialux simulation data,to establish the louver control scheme which adoptsa rule control based on the illuminance feedback.This control scheme can make use of the louvers to keep the indoor illuminance distribution within the target interval.Second,to solve the problem of indoor illuminance distribution,this paper uses illuminance simulation and data analysis,combined with the parameters trained by NN Toolbox,to establish the BP neural network algorithm for indoor illuminance estimation,and to achieve the goal of estimating the illuminance value of multi-points.And third,to settle the problem of calculation of LEDs,this thesis tries to make the improvement of the transfer function model,and then establish the superposition method,which achieves the fast and accurate calculation of the luminous flux.Finally,the indoor light environment control system was designed and built on the experimental platform.Comparing the theoretical data with the measured data,the results show that the design can effectively control the natural illumination in real time and achieve accurate LED fill light,which achieves the design goal.