Research On Multi-product Dynamic Lot Sizing Decision Based On Taboo Search Algorithm

Dynamic Lot Sizing Problem is a difficult research of a theory and algorithm, and also one of the great economic benefit in the practice of production. Appropriate Lot Sizing decision-making can avoid excessive inventory or frequent shortage, improve service quality, and reduce the cost of production operation.This paper studies the dynamic lot sizing problem with multiple products, Multi-product Dynamic Lot Sizing Problem with Remanufactured Constraints, Multi-product Dynamic Lot Sizing Problem with End-of-use Constraints Based on the service product as the object. Through the establishment and analysis of the three kinds of multi-product dynamic lot-sizing model under different conditions, and combine with the current research situation of the dynamic Lot Sizing model and the dynamic Lot Sizing model algorithm at home and abroad, presents an tabu search algorithm under the rules of improved subgradient. The algorithm first uses Lagrange relaxation to decompose dynamic lot-sizing model of product is established, and through the analysis of the decomposed sub-model, set up solving rules of the sub-model,then this paper adopts the random method to generate initial solution, and through improved subgradient rule neighborhood operation continuously on the initial solution optimization until satisfied termination condition.In order to verify the reliability of the algorithm, This paper applied the MATLAB programming language to the tabu search algorithm for programming, and introduces the concrete examples of simulation, compare with the dynamic programming method, the Lagrangian relaxation algorithm, the results showed:tabu search algorithm can obtain satisfactory solution in a short time. And, this paper compare Multi-product Dynamic Lot Sizing Problem with Remanufactured Constraints with traditional Multi-product Dynamic Lot Sizing Problem, verified the Multi-product Dynamic Lot Sizing Problem with Remanufactured Constraints is lower than Multi-product Dynamic Lot Sizing Problem by comparing the total cost.The main innovation of this thesis is:Research on Multi-product Dynamic Lot Sizing Problem of the service product, and constructs a tabu search algorithm which can provides a scientific and reasonable solution for the enterprise. And the paper compare the same remanufacturing enterprises with traditional enterprises under the same market demand show that the s recycling of service product can reduce the production cost, improve corporate profits, and provides a good guidance for traditional enterprises.