| With the continuous increase of product variety and the application of mass customization production, the cost of the whole supply chain network is also growing. In order to obtain sustainable competitiveness, a new logistics concept is put forward, namely just-in-sequence(JIS). JIS refers that the suppliers deliver a variety of components in accordance with the assembly sequence of manufacturers, and it is a kind of special delivery mode of just-in-time(JIT). JIS not only need to deliver parts in the right time, quantity and quality, but also need to deliver parts in the right sequence. There will be some unpredictable problems during the implementation of JIS, which are the unreliability of suppliers’ parts delivery and the temporary change of manufacturers’production sequence.This paper studies several relevant problems of JIS, and the automobile mixed-model assembly line is taken as the research object. Aiming at the possible problems during the implementation of JIS, both active and passive strategies are put forward from the perspective of manufacturers, which provide the guarantee for the smooth implementation of JIS, so as to give full play to the advantages of JIS. The main research work of this thesis is summarized as follows:(l)From the perspective of increasing suppliers’ delivery lead time, the sequencing requirements of the painting shop and assembly shop are taken into consideration at the same time to determine a car sequence which can simultaneously satisfy the production requirements of two shops, to improve the reliability of parts delivery and reduce the delivery risk of JIS. Therefore, three ideas have been proposed to solve the problem. Then according to these three ideas, three algorithms are designed based on the ant colony algorithm. Computational experiments show that the three algorithms have their own advantages, when the weight of the painting shop target is larger, the algorithm 2 is the best, and when the weight of the assembly shop target is larger, the algorithm 3 is the best. However, algorithm 2 and algorithm 3 have some limitations, which can not take the two targets into account at the same time. The algorithm 1 has high stability and wide applicability, which can meet the different demands when there is no clear relationship between the two shop targets.(2)When the parts delivery meet accidents in the JIS delivery mode that uses entrepot storages to supply parts, a resequencing method is used to avoid the negative impact on the mixed-model assembly line. The resequencing method is that the corresponding workpieces are removed from the predetermined sequence and are arranged to be produced in the next parts delivery cycle. The goal is how to insert the accident workpieces to the subsequent sequence to minimize the total cost of overload time and free time. Therefore, three greedy rules and a simulated annealing algorithm are proposed. Computational experiments show that the greedy rule 3 and the simulated annealing algorithm are both superior to the local search algorithm in the literature, and the simulated annealing algorithm is the best.(3) In the JIS delivery mode, the methods that combine the mix-bank method and the virtual resequencing method are used to restore the changed sequence according to a predetermined sequence. Two mix-bank resequencing methods and two virtual resequencing schemes are proposed, and the two kinds of methods are combined to four combined restoring strategies. Then the simulation is used to verify these methods and stratefies. Results show that in a certain repair rate and delay time, when the capacity of a mix-bank is relatively small, the combined restoring strategy 3 is the best; and when the capacity of a mix-bank is relatively big, the mix-bank resequencing method 1 is the best; in general, the lager the capacity of a mix-bank and the number of lanes, the better the recovery effect of a sequence. |
PDF Full Text Request