Search

In a large-scale construction project, the project manager needs to allocate many resources, including numerous workers, various kinds of materials, and a lot of cash. Therefore how to allocate resources and to control cash flow ahead of time is particularly important. The general contract and material price fluctuation make the project manager to take great risk. In this paper, we establish a two-stage stochastic programming model which takes price fluctuation, working capital and other necessary constraints into account. We employ a sample-based heuristic to study how price fluctuation affects the progress schedule of the project. We also conduct sensitivity analysis on the important parameters, including working capital and labor cost.

It is increasingly difficult for the traditional fault diagnosis technologies to meet the complex and automation requirements of electronic equipments, so the combination of artificial intelligence technology has become a development direction of fault diagnosis. In the fault diagnosis, BP neural network has also been widely used. As for the deficiency of BP network, the paper presented an improved BP network dynamic parameter adjust algorithm and applied it in the research of electronic equipment fault diagnosis. Proved theoretically and practically, the method can effectively overcome the deficiency of standard BP algorithm, and provides efficient way for the fault diagnosis of electronic equipments

Traveling salesman problem (TSP) is a very hard and classical optimization problem in the field of operations research, and often-used benchmark for new optimization techniques. This paper will to bring up multi-agent approach for solving the TSP based on data mining algorithm, for the extraction of knowledge from a large set of TSP. The proposed approach supports the distributed solving to the TSP. It divides into three-tier, the first tier is ant colony optimization agent; the second-tier is genetic algorithm agent; and the third tier is fast local searching agent. In using an ant colony algorithm (ACA) for the TSP, an attribute-oriented induction methodology was used to explore the relationship between an operations' sequence and its attributes and a set of rules has been developed. These rules can duplicate the ACA's performance on identical problems. Ultimately, the experimental results have shown that the proposed hybrid approach has good performance with respect to the quality of solution and the speed of computation.

An Improved Ant Colony Optimization (IACO) algorithm is proposed to the Flexible Job Shop Scheduling Problem (FJSSP) in this paper. IACO algorithm provides an effective integration between Ant Colony Optimization (ACO) model and knowledge model. In the IACO algorithm, knowledge model learns some available knowledge from the optimization of ACO, and then employs the existing knowledge to guide the current heuristic searching. The performance of IACO was evaluated by many benchmark instances taken from literature. Final experimental results indicate that the proposed IACO algorithm outperforms some current approaches in the quality of schedules.

In this paper, we study a periodic review, capacitated, make-to-stock system with two demand classes. The first class demand comes from customers who have long term and profitable relationships with the manufacturer. The second class demand comes from the retail market. The first class demand not satisfied immediately is backordered, whereas the unsatisfied second class demand is lost. There are two fundamental decisions for managing such a system, i.e., production decision and rationing decision. We prove that a base stock rationing policy is optimal. Furthermore, through an extensive numerical study, we study the cost savings of the optimal policy compared with two common used heuristics and show that the benefit of the optimal control policy can be quite significant.