Open AccessResearch on Modelling Method of Natural Illuminance Based on RBFNN
Author(s) -
Yujie Zhang,
Jing Guo
Publication year - 2022
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/2179/1/012006
Subject(s) - illuminance , benchmark (surveying) , computer science , optics , physics , geodesy , geography
Aiming at the problem that the indoor natural illuminance is difficult to calculate accurately due to many complex factors in the environment. This paper proposes a method to establish the natural illuminance model by using radial basis function neural network (RBFNN). The RBFNN was trained by collecting indoor natural light data to obtain the benchmark model for calculating the natural illuminance distribution. The illuminance sensor arranged indoors was used to measure the real-time natural illuminance, and the output of the benchmark model was modified accordingly, so as to obtained a model that could calculate the real-time natural illuminance at any point in the room. The experimental results show that the maximum normalized error between the calculated and measured illuminance values of the model in this paper is no more than 15%, and the error of most test data is less than 8%, which solves the problem of accurate calculation of indoor natural illuminance, so as to provide support for indoor lighting control system.