Evaluation Methodologies Of Car-sharing Service Systems Based On Fuzzy Set Theory

The era of “Internet +” is changing people's mobility patterns to a great extent.Following the prevalence of “bicycle sharing”,car-sharing comes into existence in many urban cities.With the combination of mobile internet and traditional mobility patterns,car-sharing makes full advantage of present transport capacity and serves more people by improving the efficiency of car usage.Car-sharing Service System(CSS)is the application of product service system in the field of transportation and mobility sector,with better economic,social and environmental impacts.The operation of car-sharing calls for car purchase or rental,maintenance,and corresponding infrastructures including parking lots and so forth.And it is necessary to operate on a rather considerable basis in order to make a profit.That is why many Chinese car sharing platforms spared no effort to attract users with unreasonable subsidies at the beginning.Comprehensive evaluation of car-sharing service system helps to comprehend different car-sharing options and their economic,social,system performance and environmental impacts,promoting the healthy and orderly development of car-sharing service system.Funded by National Key Technology R&D Program,China(Grant No.2015BAH21F01)entitled “Key technologies on cloud platform for industrial design”,Science & Technology Planning Projects of Xi'an City(Grant No.FW1409)entitled “Cloud platform for cultural innovation and design service” and the 111 Project(Grant No.B13044),comprehensive evaluation research on car-sharing service system is conducted based on the following theories: sharing economy,product service system,comprehensive evaluation theory,and fuzzy set theory.An evaluation model is improved on the basis of fuzzy set theory and verified through a case study.Main contributions of the dissertation are listed as follows:(1)Key elements of car-sharing service system are discussed,which can be classified into two categories: hardware elements and software elements.Hardware elements refer to sharing cars and parking lots;and there are three aspects of software elements in CSS,namely,economic,social,and technical elements.Economic elements include profit modes,operation patterns and trip cost;social elements include user requirements and their willingness to participate in sharing,safety,and policies;and technical ones include interface and cloud platform.On the basis of the above-mentioned elements,a model for constructing car-sharing service system is established and comprehensive evaluation of car-sharing service system is carried out thereafter.(2)An indicator system for comprehensive evaluation of car-sharing service system is established.The principles,methods of such an indicator system,and the testing methods for the necessity,recognition,and feasibility of underlying individual indicators are proposed.Principles of choosing indicators for comprehensive car-sharing system evaluation are discussed,i.e.SCUMN(Specific,Comprehensive,Understandable,Measurable,and Neutral).Moreover,a four-dimentional comprehensive evaluation indicator system is established with 24 indicators from the following dimensions: economy,environment,system performance,and society.And they are classified into three categories: cost indicators,benefit indicators and public welfare ones according to different characteristics of these indicators.(3)Key methodologies of comprehensive evaluation for car-sharing service system are proposed,including weights determination for dimensions,indicators and evaluation members respectively and ranking methods of potential alternatives as well.Fuzzy Analytic Hierarchy Process(F-AHP)is applied to determine the fuzzy weight for each dimension and indicator;ten factors for evaluation members are settled from their professional knowledge,influencing power,and reacting capacities;Fuzzy Technique for Order Preference by Similarity to Ideal Solution(F-TOPSIS)is applied to rank different options according to their distances from positive and negative ideal solutions.(4)On the basis of the above-mentioned key methodologies,a comprehensive evaluation model for car-sharing service system is proposed based on fuzzy set theory.A case study is carried out to demonstrate its application process.Five different car-sharing options are evaluated and ranked,namely,car rental,leasing,driving-for-you,piggy backing,and drive-sharing.Sensitivity analysis experiments are conducted to find out the influence of different indicator weights on the results of evaluation.Principles to select CSS evaluation indicators are proposed and a four-dimentional comprehensive evaluation system for car sharing system is developed.The principles are abbreviated as SCUMN and four dimensions refer to ecomonic,environmental,system,and social dimension.Besides,indicators are classified into three categories: cost,benefit,and public welfare indicators according to different characteristics.F-AHP method is improved to determine weights of dimensions,indicators,and evaluators for comprehensive CSS evaluation.And F-TOPSIS method is improved to rank potential options.A comprehensive evaluation model for car sharing system based on fuzzy set theory is developed.Fuzzy set theory is applied to deal with uncertainty and reduce the influence of subjectivity and vagueness to the final evaluation results and achieve a rather objective evaluation.To conclude,taking car-sharing service system as a research object,the thesis focuses on key methodologies of comprehensive evaluation of car-sharing service system.Key elements of car-sharing system are identified and a CSS model is established accordingly;a four-dimentional comprehensive evaluation system with 24 indicators is established and an evaluation model is improved based on fuzzy set theory.It provides an innovative method and theoretical basis for comprehensive evaluation of car-sharing service system,providing a solid foundation for the widespread and acceptance of car-sharing service system.The application example demonstrates the feasibility and reliability of the proposed evaluation model.