摘要
室内照明系统的智能化改造是降低建筑能耗、提升人居环境的关键,本文提出基于图像处理的调控方法:一是构建空间光场模型,动态关闭冗余光源并协同补光;二是结合LSTM行为预测与多目标优化算法实现多用户场景的个性化照明适配。实验表明:办公区该系统的应用降低能耗31.2%的同时将照度达标率提升至96%;会议室场景下,用户行为学习模型能够实现进一步节能,行为学习使日均能耗再降17.5%。
The intelligent transformation of indoor lighting system is the key to reduce building energy consumption and improve living environment.This paper puts forward the control methods based on image processing:first,build a spatial light field model,dynamically turn off redundant light sources and supplement light cooperatively;The second is to combine LSTM behavior prediction and multi-objective optimization algorithm to realize personalized lighting adaptation of multi-user scenes.The experiment shows that the application of the system in office area can reduce energy consumption by 31.2%and improve the illuminance compliance rate to 96%.In the conference room scene,the user behavior learning model can further save energy,and the average daily energy consumption is reduced by 17.5%by behavior learning.
作者
吴文玲
WU Wenling(School of Computer,Sichuan Technology and Business University,Chengdu 611745,China)
出处
《中国照明电器》
2025年第7期81-83,共3页
China Light & Lighting
基金
四川工商学院2024年度教学质量与教学改革工程项目“计算机图形学智慧课程建设”(ZHKC2024001)。
关键词
图像处理
室内照明
自动调节
能效优化
image processing
indoor lighting
automatic adjustment
energy efficiency optimisation
