基于改进Mask R-CNN的建筑屋面光伏利用潜力评估研究——以长春市工业厂房为例  

Study on the Evaluation of Photovoltaic Utilization Potential of Building Roofs Based on an Improved Mask R-CNN Algorithm:A Case Study of Industrial Plants in Changchun City

作　　者：周春艳[1] 路少石 

机构地区：[1]吉林建筑大学建筑与规划学院

出　　处：《建筑与文化》2025年第3期244-247,共4页Architecture & Culture

基　　金：吉林省教育厅科技项目(项目编号:JJKH20240367KJ);吉林省教育科学规划项目(项目编号:GH23596)。

摘　　要：近年来,中国的能源需求随着经济的发展而快速增长。在建筑屋面上利用太阳能资源是实现我国“碳达峰、碳中和”目标的重要途径。文章提出了一种改进后的Mask R-CNN深度学习算法,通过将原模型中的FPN网络改进为PAN网络来提升模型对于图像特征的提取能力,从而提高光伏潜力的评估效率。文章以长春市中心城市区的工业厂房为研究对象并评估其屋面的光伏利用潜力,最终计算得到长春市中心城区的工业厂房屋面面积为82.48×10^(6)m^(2),光伏利用潜力为144.4375×10^(8)kWh/年,可为长春市城市工业厂房屋顶光伏发展规划提供依据。In recent years,China's energy demand has grown rapidly along with its economy.Utilizing solar energy resources on building roofs is an important way to achieve China's carbon peak and carbon neutrality goals.This study proposes an improved Mask R-CNN deep learning algorithm to enhance the model's feature extraction capability by upgrading the FPN network in the original model to a PAN network,thereby improving the efficiency of photovoltaic potential assessment.The study focuses on industrial plants in the central urban area of Changchun City,evaluating the PV potential of their roofs.It is calculated that the total roof area of industrial plants in the central urban area of Changchun City is 82.48×10^(6)m^(2),and the PV potential is 144.4375×10^(8)kWh/year,providing a basis for the development planning of rooftop PV systems for industrial plants in Changchun City.

关 键 词：Mask R-CNN 建筑屋面 光伏利用潜力 长春市工业厂房 

分 类 号：F42[经济管理—产业经济]