With the development of the supply chain network, the port supply chain has become a hot concern and research. The port supply chain network is complicated and the members of build form are different from level. Therefore, in order to realize the reasonable allocation of resources, complementary advantages, improving the efficiency of the whole supply chain operation become the biggest goal in the port supply chain management. In addition, in the form of global economic integration, the port supply chain industry is more and more involved; the scope is more and more big. In the process of production, transportation and storage of goods, there are a lot of indefinite factor, if these uncertainties can not be detected early and stop, it will bring the unpredictable consequences for the entire supply chain, so the port supply chain risk assessment is necessary and it is also an important task to ensure the development of the port supply chain coordination.In order to reduce the uncertainty and risk of the port supply chain and strengthen the collaboration among supply chain members, in this paper, the network equilibrium model and risk assessment of port supply chain is studied. In this paper, the main research results are as follows:1. Based on competition relationship with the same layer and mutual cooperation relationship between members of different layers in port supply chain network,a three-layer port supply chain network equilibrium model which consists of supplier, port logistics providers and client is proposed, and on the basis of the model the electronic commerce is introduced. Then from the perspective of supplier, port logistics providers and client, this paper respectively give their optimal target and global optimal target, and presents a solving method for optimal target based on the variational inequality theory.Finally the efficiencies and rationalities of the model are verified by examples.2. In order to reduce the computational complexity and calculation amount of solving quadratic programming problems and improve the speed and precision of the algorithm, this paper proposes the training methods of the linear and approximate linear support vector machine (SVM) based on variational inequality. These methods can transform the problem of convex quadratic programming to the problem of variational inequalities in the process of training SVM, and obtain the optimal separating hyperplane. The solving process occupies less memory, which can improve the training speed of SVM classification. Simulation results show that SVM based on variational inequality has not only strong generalization ability but also faster convergence speed.3.In order to reduce the loss caused by various risks on port supply chain and improve the stability of the port supply chain management and operational efficiency, the port supply chain risk assessment model based on SVM is given, comparing with the models based on BP neural network and RBF neural network, the test examples show that the model based on SVM is more feasible. At the same time, considering the error of the acquisition data, a fault-tolerant value and a fault-tolerant interval of each of the data are introduced, and rough neural network method is applied to port supply chain risk assessment model for the uncertain analysis.The influence of different fault-tolerant values to prediction error is discussed, which can provide a reference for the establishment and development of the port supply chain risk assessment models. |