基于T-S和AIFNN的旧住宅性能评价研究  

Study on the Evaluation of Old Residential Building Performance Based on Adaptive Intuitionistic Fuzzy Neural Network

作　　者：高宇波[1] 韩卫成[1] 胡川晋[1] 

机构地区：[1]太原理工大学建筑与土木工程学院,太原030024

出　　处：《太原理工大学学报》2017年第3期296-302,共7页Journal of Taiyuan University of Technology

基　　金：国家自然科学基金资助项目:旧住宅性能评价及实践效用研究(51378329)

摘　　要：在我国旧住宅资源日益增多的情况下,如何把握其性能的衰退程度和可利用程度,为旧住宅下一步的改造再利用提供科学的指导,是目前旧住宅研究面临的关键问题。基于旧住宅问题现状,从技术、经济、社会3个方面构建旧住宅性能评价指标体系,包含环境与规划、功能与空间、设备于设施、结构与安全等4个方面的内容;利用Takagi-Sugeno(T-S)直觉模糊推理实现不确定信息的量化表达,构建了基于自适应直觉模糊神经网络(adaptive intuitionistic fuzzy neural network,AIFNN)的旧住宅性能评价模型;结合仿真样本与实证案例,分析验证了模型的可行性及有效性。评价结果运用到旧住宅的市场实践中,为下一步的再利用提供科学的量化技术支持。In the case of old residential resources increasing in our country,how to grasp the degree of decline and the degree of utilization,to provide scientific guidance for the reconstruction and reuse of old residential areas is the key problem in the research of the old house.On the basis of the current problem of building performance,the evaluation index system of old residential structure and safety performance was established from three dimensions：technology,economy and society.Through T-S intuitionistic fuzzy reasoning to quantify uncertain information,an evaluation model,which is based on the Adaptive Intuitionistic Fuzzy Neural Network（AIFNN）,was established for old residential building performance.The feasibility and effectiveness of the proposed model were analyzed and verified by simulation examples and case studies.The application of the evaluation results to the old residential market practice,can provide quantitative scientific support for the next reuse.

关 键 词：住宅性能评价 旧住宅 T-S模糊推理 自适应直觉模糊神经网络 

分 类 号：X321[环境科学与工程—环境工程]