• 지능정보연구
• /
• 제1권2호
• /
• pp.123-137
• /
• 1995

Most of solution methods in scheduling attempt to generate good solutions by either developing algorithms or heuristic rules. However, scheduling problems in the real world require considering more factors such as multiple objectives, different combinations of heuristic rules due to problem characteristics. In this respect, the traditional mathematical a, pp.oach showed limited performance so that new a, pp.oaches need to be developed. Expert system is one of them. When an expert system is developed for scheduling one of the most difficult processes faced could be knowledge acquisition on scheduling heuristics. In this paper we propose a method for the acquisition of knowledge on the selection of scheduling heuristics using Dempster-Shafer Theory(DST). We also show the examples in the multi-objectives environment.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권10호
• /
• pp.5013-5034
• /
• 2019

In the Peer-to-Peer (P2P) streaming systems, peers randomly form a network overlay to share video resources with a data scheduling scheme. A data scheduling scheme can have a great impact on system performance, which should achieve two optimal objectives at the same time ideally. The two optimization objectives are to improve the perceived video quality and maximize the network throughput, respectively. Maximizing network throughput means improving the utilization of peer's upload bandwidth. However, maximizing network throughput will result in a reduction in the perceived video quality, and vice versa. Therefore, to achieve the above two objects simultaneously, we proposed a new data scheduling scheme based on multi-objective particle swarm optimization data scheduling scheme, called MOPSO-DS scheme. To design the MOPSO-DS scheme, we first formulated the data scheduling optimization problem as a multi-objective optimization problem. Then, a multi-objective particle swarm optimization algorithm is proposed by encoding the neighbors of peers as the position vector of the particles. Through extensive simulations, we demonstrated the MOPSO-DS scheme could improve the system performance effectively.

• 지능정보연구
• /
• 제2권2호
• /
• pp.69-83
• /
• 1996

The objective of this research is to develop a knowledge acquisition and refinement method for a multi-objective and multi-decision FMS scheduling problem. A competitive neural network and an inductive learning algorithm are integrated to extract and refine necessary scheduling knowledge from simulation outputs. The obtained scheduling knowledge can assist the FMS operator in real-time to decide multiple decisions simultaneously, while maximally meeting multiple objective desired by the FMS operator. The acquired scheduling knowledge for an FMS scheduling problem is tested by comparing the desired and the simulated values of the multiple objectives. The result show that the knowledge acquisition and refinement method is effective for the multi-objective and multi-decision FMS scheduling problems.

• 경영과학
• /
• 제13권3호
• /
• pp.173-186
• /
• 1996

This paper presents an approach for classifying scheduling problems and selecting a heuristic rule to yield best solution in terms of certain performance measure(s). Classification parameters are employed from previous studies on job shop scheduling and project scheduling. Neural network is used for learning and estimating the performance of heuristic rules. In addition, multi-criteria decision making techniques are employed to combine the preferences for each performance measure and heuristic rule for the problems with multi-objectives.

• Management Science and Financial Engineering
• /
• 제20권1호
• /
• pp.1-9
• /
• 2014

In multi-objective scheduling problems, the objectives are usually in conflict. To obtain a satisfactory compromise and resolve the issue of NP-hardness, most existing works have suggested employing meta-heuristic methods, such as genetic algorithms. In this research, we propose a novel data-driven approach for generating a single solution that compromises multiple rules pursuing different objectives. The proposed method uses a data mining technique, namely, random forests, in order to extract the logics of several historic schedules and aggregate those. Since it involves learning predictive models, future schedules with the same previous objectives can be easily and quickly obtained by applying new production data into the models. The proposed approach is illustrated with a simulation study, where it appears to successfully produce a new solution showing balanced scheduling performances.

• 경영과학
• /
• 제32권4호
• /
• pp.69-80
• /
• 2015

The semiconductor industry is one of those in which the most intricate processes are involved and there are many critical factors that are controlled with precision in those processes. Naturally production scheduling in the semiconductor industry is also very complex and studied by the industry and academia for many years; however, still there are many issues left unclear in the problem. This paper proposes an multi-objective optimization-based scheduling method for semiconductor fabrication(fab). Two main objectives are throughput maximization and meeting target production quantities. The first objective aims to reduce production cost, especially the fixed cost incurred by a large investment constructing a new fab facility. The other is meeting customer orders on time and also helps a fab maintain stable throughput through controlled WIP balancing in the long run. The paper shows a trade-off structure between the two objectives through experimental studies, which provides industrial practitioners with useful references.

• Industrial Engineering and Management Systems
• /
• 제10권2호
• /
• pp.134-139
• /
• 2011

The problem of scheduling in permutation flowshops has been extensively investigated by many researchers. Recently, attempts are being made to consider more than one objective simultaneously and develop algorithms to obtain a set of Pareto-optimal solutions. Varadharajan et al. (2005) presented a multi-objective simulated-annealing algorithm (MOSA) for the problem of permutation-flowshop scheduling with the objectives of minimizing the makespan and the total flowtime of jobs. The MOSA uses two initial sequences obtained using heuristics, and seeks to obtain non-dominated solutions through the implementation of a probability function, which probabilistically selects the objective of minimizing either the makespan or the total flowtime of jobs. In this paper, the same problem of heuristically developing non-dominated sequences is considered. We propose an effective heuristics based on simulated annealing (SA), in which the weighted sum of the makespan and the total flowtime is used. The essences of the heuristics are in selecting the initial sequence, setting the weight and generating a solution in the search process. Using a benchmark problem provided by Taillard (1993), which was used in the MOSA, these conditions are extracted in a large-scale experiment. The non-dominated sets obtained from the existing algorithms and the proposed heuristics are compared. It was found that the proposed heuristics drastically improved the performance of finding the non-dominated frontier.

• Industrial Engineering and Management Systems
• /
• 제12권2호
• /
• pp.151-160
• /
• 2013

In the past decades, several algorithms based on evolutionary approaches have been proposed for solving job shop scheduling problems (JSP), which is well-known as one of the most difficult combinatorial optimization problems. Most of them have concentrated on finding optimal solutions of a single objective, i.e., makespan, or total weighted tardiness. However, real-world scheduling problems generally involve multiple objectives which must be considered simultaneously. This paper proposes an efficient particle swarm optimization based approach to find a Pareto front for multi-objective JSP. The objective is to simultaneously minimize makespan and total tardiness of jobs. The proposed algorithm employs an Elite group to store the updated non-dominated solutions found by the whole swarm and utilizes those solutions as the guidance for particle movement. A single swarm with a mixture of four groups of particles with different movement strategies is adopted to search for Pareto solutions. The performance of the proposed method is evaluated on a set of benchmark problems and compared with the results from the existing algorithms. The experimental results demonstrate that the proposed algorithm is capable of providing a set of diverse and high-quality non-dominated solutions.

• 한국경영과학회:학술대회논문집
• /
• 한국경영과학회/대한산업공학회 2003년도 춘계공동학술대회
• /
• pp.1071-1076
• /
• 2003

Evolutionary algorithm is recognized as a promising approach to solving multi-objective combinatorial optimization problems. When no preference information of decision makers is given, multi-objective optimization problems have been commonly used to search for diverse and good Pareto optimal solution. In this paper we propose a new multi-objective evolutionary algorithm based on competitive coevolutionary algorithm, and demonstrate the applicability of the algorithm. The proposed algorithm is designed to promote both population diversity and rapidity of convergence. To achieve this, the strategies of fitness evaluation and the operation of the Pareto set are developed. The algorithm is applied to job shop scheduling problems (JSPs). The JSPs have two objectives: minimizing makespan and minimizing earliness or tardiness. The proposed algorithm is compared with existing evolutionary algorithms in terms of solution quality and diversity. The experimental results reveal the effectiveness of our approach.

• 한국경영과학회지
• /
• 제37권1호
• /
• pp.19-28
• /
• 2012

We consider a multi-agent scheduling problem such that each agent tries to maximize the weighted number of just-in-time jobs. Two objectives are considered : the first is to find the optimal solution for one agent with constraints on the other agents' weight functions, and the second is to find the largest set of efficient schedules of which corresponding objective vectors are different for the case with identical weights. We show that when the number of agents is fixed, the single machine case with the first objective is NP-hard in the ordinary sense, and present the polynomial- time algorithm for the two-machine flow shop case with the second objective and identical weights.