Supply Chain Problem Solving for Manufacturing Companies based on Recursive Models, Objective Programming, and Backward Decision Algorithms

The ordering and transport of raw materials for a production company is tightly linked to suppliers and forwarders, creating a complex supply chain problem. There is a matching problem between the supplier’s delivery and the manufacturer’s order quantity, and the forwarder has a certain transit loss problem in the process of transit. In this paper, the evaluation of the production importance of suppliers is performed via the TOPSIS algorithm, implemented by the entropy-based optimization method in feature engineering. Furthermore, grey correlation analysis is employed to analyse the pairing propensity between suppliers and forwarders. Secondly, the ARIMA method is utilised to predict the future loss rate and supply quantity, which are used as subsequent prophetic factors and search convergence points respectively to overcome the NP-hard problem. For the supplier selection problem, the optimisation objective is constructed based on 0-1 planning, a recursive model is built, constraints are constructed with sustainable production requirements, and search is used to solve the problem. For the raw material ordering solution problem, a single objective optimisation model based on least cost is constructed and solved using Monte Carlo algorithm. For the transit solution formulation problem, a loss minimisation-based decision scheme is constructed, based on the idea of priority levels, a single step decision is made and a backtracking method is introduced to solve for the constraint limitations. Based on the above model and algorithm, the supply chain problem between suppliers, forwarders and enterprises can be well solved.