Research On Method Of Sustainable Supply Chain Risk Resilience Measures Prediction And Configuration Optimization

Due to the depletion of global resources,sustainable supply chain management become an essential strategy in the global economic environment.Although sustainable supply chain has been studied a lot in the academic world,there is a lack of guiding methods to integrate sustainability into supply chain risk management.Sustainable supply chain risk resilience measures(RMs)can reduce the impact of risks and improve the stability of sustainable supply chain,which plays an important role in the study of sustainable supply chain risk management.Therefore,this thesis carries out systematic research on risk factor assessment,prediction and configuration optimization of RMs centre on sustainable supply chain management,mainly including:(1)The ranking model of crucial risk factors(RFs)driven by customer requirements(CRs).Listening to sustainable CRs and integrating them into the supply chain’s RFs can achieve higher customer satisfaction in a customer-oriented market.Therefore,this thesis proposes a ranking method based on fuzzy kano model(FKM),quality function deployment(QFD),single-valued neutrosophic sets(SVNS)and cross-entropy,to determine the importance of sustainable supply chain RFs.And the relationship between CRs and RFs was determined by using SVNS-QFD to deal with the uncertainty and fuzziness in expert evaluation.Finally,the importance of RFs is obtained through the SVNS cross-entropy.It provides important reference for identifying key RMs and formulating risk mitigation strategy configuration schemes.(2)Prediction model of the importance of risk resilience measures based on AFOA-BP neural network.Due to the complicated nonlinear relationship between the RFs and RMs,this thesis built the relationship between RFs and RMs using the weights of RFs and RMs’ historical data.It can obtain the importance of RMs and reduce the subjectivity of traditional analysis methods.Therefore,in order to overcome the problems of slow convergence and easy to fall into the local minimum in the learning process of BP neural network and the characteristics of importance of sustainable supply chain RMs in prediction,this paper propose a RMs importance prediction model based on adaptive fruit fly algorithm(AFOA)optimized BP neural network,referred to as AFOA-BP neural network.Combining the advantages of fruit fly optimization algorithm,this model used AFOA to optimize the initial weight and threshold value of BP neural network.(3)Sustainable supply chain risk resilience measures fuzzy configuration optimization.According to the purpose and characteristics of sustainable supply chain risk mitigation in fuzzy environment,a fuzzy multi-objective hybrid allocation optimization model of sustainable supply chain RMs was constructed.In this model,the uncertain configuration information is expressed by the SVNS,the goal is to maximize risk mitigation and minimize carbon emissions and energy consumption,the constraints are cost,strategy choice and boundary conditions,and the mixed numbers of continuous variables and discrete variables are used to express the decision variables.Considering the problems of mixed variables,fuzzy information and constraints in the optimization model,an adaptive fuzzy genetic algorithm based on dynamic stochastic ranking was designed to solve the model,and the optimal design scheme of sustainable supply chain RMs was obtained.Finally,the supply chain of automobile manufacturing enterprises is taken as an example to verify the feasibility and effectiveness of the research scheme.It provides method guidance for sustainable supply chain risk management.