Font Size: a A A

The Method Research On Adaptive Neruo-fuzzy Inference System For Assessing Urban Sustainability Performance

Posted on:2018-01-27Degree:MasterType:Thesis
Country:ChinaCandidate:C Y ShuaiFull Text:PDF
GTID:2359330536468931Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
Urbanization as the worldwide development engine has made unparalleled achievements.Urbanization in China is considered as an inexorable trend of the economic development becasuse of the largest amount of population.However,the urbanization practice in China has also been faced with many unsustainable challenges.It is therefore significant to assess the urban sustainability performance,and evaluate the urbanization quality and staus for the urbanization process in China.This research reviews the existing sustainability performance assessment methods,and combines the features of the sustainability performance.With aim of overcoming shortage of the subjectivity(expert knowledge)in the traditional sustainability performance assessment,this research thus tentatively introduces the adaptive neuro-fuzzy inference system(ANFIS)assessment method.First,an assessment framework of urban sustainability performance with ANFIS by reviewing the fuzzy-set,neural-network algorithm,and ANFIS theories is proposed.Then,the reports from Urban China Initiative(UCI)are employed as the training,checking and comparison samples of ANFIS to examine the feasibility of ANFIS in urban sustainability performance assessment.Last,nine frequently adopted including linear triangular,trapezoid,and non-linear Gaussian,joint-Gaussian,PI type,sigmoidal,differential sigmoidal,product sigmoidal,and bell-shaped membership functions(MFs)are collected for the membership function optimization(MFO).Three error criteria including general error,average error,and fluctuate error are established for the MF membership function optimization.The main conclusions are: 1)ANFIS is validated as a feasible approach in urban sustainability performance assessment.2)The Gaussian membership function is the most optimal membership function in urban sustainability performance assessment in China.3)The non-linear MFs fit better than the linear MFs in urban sustainability performance assessment in China.The ANFIS method introduced in this research not only can be used in urban sustainability assessment,but also can be applied in nation,province and project sustainability performance assessment,which enrich the sustainability assessment method greatly.Meanwhile,the assessment results from ANFIS can help to guide the sustainable urbanization practice in China.
Keywords/Search Tags:Urban sustainability performance, neuro-fuzzy network, fuzzy-set theory, membership function optimization, sustainable decision-making
PDF Full Text Request
Related items