Research On Performance Apprasial Of Complex Product Suppliers Based On Heterogeneous Information Fusion

Complex product,such as aircrafts,warship rockets,spacecraft,high-end CN C equipment,and high-precision industrial robots,are large-scale industrial products or complicated systematic projects with complex product structure,high component integration and advanced engineering technology.O wing to the huge complexities of product structure,management process,long development cycle and high technical difficulty,the complex product adopt ”main manufacture-suppliers” production system.As a production organizer and system integrator,the main manufacturer of complex products play an important role in the global collaborative supply chain,which responsible for designing and maintaining the proper functioning of the global supply chain.And the performance appraisal of the system supplier in the development process contribute to the final assembly progress and quality of the final products.From the characteristics of “main manufacture-suppliers” prod uction system,this thesis contributes to supplier management and performance appraisal by designing a set of supplier performance appraisal indicator and indicator weight measurement model.Base on t he extens ive liter ature ana lysis and actual r esearch,a supplie r performance a ppraisal ind icator of co mplex product is construc ted from five perspective of suppliers' schedule management,quality assurance,cost control,comprehensive protection and sustainable development.This thesis analyzes expert's explicit and implicit of appraisal(monotonic hesitant fuzzy language),according to the Weibull distribution functions with three parameters and ABC classification method,a novel computational method for monotonic hesitant fuzzy linguistic terms is proposed to transfer experts' uncertain appraisal information to decision-making data,which can systematically consider and mine expert's obvious explicit and hidden implicit appraisal information.Moreover,feasibility analysis is explored to show the influence of parameters on decision-making precision.A hesitant fuzzy AHP analysis method is constructed based on the relevant theory of AHP analysis,furthermore,the performance weights of the alternative suppliers are determined.Heterogeneous information fusion mechanisms are explored according to the decision information and expert preference information intrinsic connection characteristics.Additionally,the attribute weight measurement model is constructed with the minimum degree of difference between the two types of information to achieve the optimal selection of candidate suppliers.A case study of by C919 suppliers' performance appraisal is used to show the feasibility and effectiveness of the research.A set of performance appraisal indicator is structured to help C919 determined suppliers' performance appraisal content,analyzing the different performance information of each supplier under different indicators and the relative performance information between two suppliers given by experts.Expert's implicit appraisal information is mined to effectively improve the precision of computational based on the understanding experts' appraisal information and decision preference information.An attribute weight measurement model is constructed to ranking the comprehensive performance of each suppler.Finally,performance improvement measures of complex product supplier are proposed to revealing actual contributions.According to the characteristics of ABC classification and Weibull distribution function,focusing on the research on the quantitative characterization method of expert appraisal opinions when evaluating the performance of complex product suppliers,the explicit meaning of expert appraisal opinions are revealed,specially,the comprehensive meanings of monotone decreasing and increasing hesitant fuzzy linguistic terms are investigated and a method of characterizing the hesitant fuzzy linguistic term set of monotonic interval experts is proposed to avoid information loss and improve the decision-making efficiency.Based on the long development time of complex products,facing multiple types of information,different structural characteristics the dual language information fusion method is designed to measure attribute weights and the weight of the stage milestones,the method proposed to further improve the scientific and rationality of the decision-making work of the main manufacturer.