Entity

Time filter

Source Type


Yang T.,Hefei University of Technology | Yang T.,Key Laboratory of Process Optimization and Intelligent Decision Making of Ministry of Education | Fu C.,Hefei University of Technology | Fu C.,Key Laboratory of Process Optimization and Intelligent Decision Making of Ministry of Education | And 7 more authors.
Journal of Combinatorial Optimization | Year: 2016

This paper considers a closed-loop supply chain consisting of one-manufacturer and one-retailer. This supply chain provides single-kind products with reusable containers. The main purpose of this study is to explore and evaluate the value of recovery information captured by embedded sensors in the environment of internet of things. The recovery information of containers dynamically monitors recovery status and provides a reliable estimation of return quantity. The value of information is measured by the cost saving performances with full, partial or no recovery information. When the full or partial recovery information is available, the decisions are made based on the known quantities of the usable or total return flow. When no recovery information is available, the decisions are made based on the stationary distribution of the return flow. A periodic inventory model is built with uncertainties of forward and reverse flows. Then, a myopic order policy is proposed for the different levels of information utilization. Through the optimality analysis, we introduce a farsighted inventory control policy. Using the general result of Markov decision processes, the performance of heuristic policies is displayed. The farsighted policy performs better than the myopic policy. In addition, the farsighted policy helps to lessen the convex impact of utilization rate on the expected cost. Afterwards, we extend the model with the selective disposal behavior. A simulation study is used to depict sensitivity and robustness of the farsighted policy. Finally, we extend the simulation experiment with uniformly distributed in-use time for a more general applicability. © 2015 Springer Science+Business Media New York Source


Zhong J.,Hefei University of Technology | Zhong J.,Key Laboratory of Process Optimization and Intelligent Decision Making of Ministry of Education | Chu F.,University of Evry Val dEssonne | Chu C.,Ecole Centrale Paris | And 2 more authors.
Journal of Systems Science and Systems Engineering | Year: 2015

This paper addresses a dynamic lot sizing problem with bounded inventory and stockout where both no backlogging and backlogging allowed cases are considered. The stockout option means that there is outsourcing in a period only when the inventory level at that period is non-positive. The production capacity is unlimited and production cost functions are linear but with fixed charges. The problem is that of satisfying all demands in the planning horizon at minimal total cost. We show that the no backlogging case can be solved in ) O(T2) time with general concave inventory holding and outsourcing cost functions where T is the length of the planning horizon. The complexity can be reduced to O(T) when the inventory holding cost functions are also linear and have some realistic properties, even if the outsourcing cost functions remain general concave functions. When the inventory holding and outsourcing cost functions are linear, the backlogging case can be solved in O(T3logT) time whether the outsourcing level at each period is bounded by the sum of the demand of that period and backlogging level from previous periods, or only by the demand of that period. © 2015 Systems Engineering Society of China and Springer-Verlag Berlin Heidelberg Source


Pei J.,Hefei University of Technology | Pei J.,University of Florida | Liu X.,Hefei University of Technology | Liu X.,Key Laboratory of Process Optimization and Intelligent Decision Making of Ministry of Education | And 7 more authors.
Annals of Mathematics and Artificial Intelligence | Year: 2015

This paper investigates a scheduling model with certain co-existing features of serial-batching, dynamic job arrival, multi-types of job, and setup time. In this proposed model, the jobs of all types are first partitioned into serial batches, which are then processed on a single serial-batching machine with an independent constant setup time for each new batch. In order to solve this scheduling problem, we divide it into two phases based on job arrival times, and we also derive and prove certain constructive properties for these two phases. Relying on these properties, we develop a two-phase hybrid algorithm (TPHA). In addition, a valid lower bound of the problem is also derived. This is used to validate the quality of the proposed algorithm. Computational experiments, both with small- and large-scale problems, are performed in order to evaluate the performance of TPHA. The computational results indicate that TPHA outperforms seven other heuristic algorithms. For all test problems of different job sizes, the average gap percentage between the makespan, obtained using TPHA, and the lower bound does not exceed 5.41 %. © 2015 Springer International Publishing Switzerland Source


Pei J.,Hefei University of Technology | Pei J.,Key Laboratory of Process Optimization and Intelligent Decision Making of Ministry of Education | Liu X.-B.,Hefei University of Technology | Liu X.-B.,Key Laboratory of Process Optimization and Intelligent Decision Making of Ministry of Education | And 3 more authors.
Nanjing Li Gong Daxue Xuebao/Journal of Nanjing University of Science and Technology | Year: 2012

To study two-tier supply chain scheduling models comprised by multi suppliers and multi manufacturers, the scheduling optimization theory is applied into the supply chain domain. A math model is built by taking the minimization of total flow time of work pieces and total cost of delivery as optimization objectives. The dynamic programming algorithm is used to solve the problem based on analyzing the features of its optimum solution. Multi-objectives fusion is realized through analytic hierarchy process in the algorithm iterative process. The feasibility of this model and algorithm is illustrated by a numerical example. Source


Pei J.,Hefei University of Technology | Pei J.,University of Florida | Liu X.,Hefei University of Technology | Liu X.,Key Laboratory of Process Optimization and Intelligent Decision Making of Ministry of Education | And 5 more authors.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2014

This study addresses a two-stage supply chain scheduling problem, where the jobs need to be processed on the manufacturer's serial batching machine and then transported by vehicles to the customer for further processing. The size and processing time of the jobs are varying due to the differences of types, and setup time is needed before processing one batch. For the problem with minimizing the makespan, we formalize it as a mixed integer programming model. In addition, the structural properties and lower bound of the problem are provided. Based on the analysis above, a novel hybrid dynamic programming algorithm, combining dynamic programming and heuristics, is proposed to solve the problem. Furthermore, its time complexity is also analyzed. By comparing the experimental results of our proposed algorithm with the heuristics B F F and L F F , we demonstrate that our proposed algorithm has better performance and can solve the problem in a reasonable time. © 2014 Springer International Publishing. Source

Discover hidden collaborations