SCIENTIA IRANICA
Year:2002 | Volume:9 | Issue:3|Papers Count:13

With the changes in the workplace and the added element of competition, it is apparent that the efficient scheduling of resources such as machines, labor and material is extremely important. Scheduling covers a broad array of topics, such as machine scheduling and sequencing, vehicle and fleet scheduling, crew scheduling, employee scheduling, project scheduling, lot-size scheduling, interval (or reservation) scheduling and timetabling. This article reviews the literature to identify the state-of-the-art research on scheduling and routing problems.

This article provides an overview of information technologies in support of the effective management of organizational knowledge resources. A variety of information technologies are used for knowledge management such as Enterprise Resource Planning (ERP) systems, Decision Support Systems (DSS) and Knowledge-Based Systems (KBS). Application of information technology towards managing explicit knowledge is well-known and has been implemented in many enterprises. However, the management of tacit knowledge through information technology is a challenge that has yet to be met.

In this paper, a multi-objective Hierarchical Production Planning (HPP) model is presented for a single stage problem with stochastic demands. The HPP model is assumed to have two levels: Product type and product family, where a type is a set of families with similar costs per unit of production time. A chance-constrained goal programming approach was suggested for the first level of the HPP model, namely, Aggregate Production Plan (APP). For the second level, Disaggregate Production Plan (DPP), a cost modified function was proposed based on set-up cost and expected shortage cost within the families. The problem is also formulated as a Mixed- Integer Programming (MIP) model. Considering three alternative formulations for APP, the results of the models are compared in three directions: The versions of APP formulation, HPP versus MIP and adjusted DPP versus unadjusted DPP. The computational results demonstrate the effectiveness of the proposed approach.

In this paper, an introduction is given to the decision-making theory, the Analytic Hierarchy Process (AHP). In the AHP, a decision hierarchy is constructed with a goal, criteria and alternatives. The criteria are pair wise compared for their importance with respect to the goal to derive a scale of relative importance and the alternatives are pair wise compared with respect to each criterion to derive relative scales. The relative scales are synthesized using a weighting and adding process to show which is the best alternative. Judgment is used to make the pair wise comparisons. Data can also be used directly in the pair wise comparisons if there is no need to assessor interpret the value represented by the numbers. This process works well for intangibles with no existing scales of measurement. Even when measurement scales and data using these scales exist, it is often advantageous to use judgment to interpret it. Two concepts integral to the AHP, measurement and consistency, are explained. Finally, the ideas and the method are illustrated through an example.

Dynamic Cournot oligopolies are examined when the firms do not have accurate information about price function. After the mathematical model is formulated, the local asymptotical stability of the steady-state is proved under realistic conditions. Instability occurs when information lag is assumed. The possibility of limit cycles is examined, based on the Hopf bifurcation theorem.

In this paper, the optimization of long term operation and determination of the reliability level of a hydro-thermal power system are integrated into a unified model. Inflow to reservoirs is modeled as a random variable. Furthermore, the demand for energy is also assumed to be a random variable with normal distribution. In order to minimize the total cost, the reliability level of the system is determined, rather than considering it as a priori input data. Since the resulting model is a large-scale stochastic nonlinear programming, it is necessary to develop a special method to solve it. This method, which provides an optimal solution within three stages, consists of a decomposition technique, Lagrangian relaxation and nonlinear and dynamic programming methods. To test the method, it has been implemented in Khuzestan power system and the results are compared with the existing operation procedures.

The purpose of this study is to determine productivity indexes and productive efficiency in the container terminal at Port Rajaei of the Ports and Shipping Organization (PSO) in Iran.Parametric and nonparametric approaches are used to construct a frontier to be used as a yardstick of productive efficiency for the period of 1994-1999. Moreover, indexes with respect to the goals are identified and, then, are measured for the period of 1990-1999.

Selecting an optimum combination of items from a set of items, known as knapsack problems, is an important issue in the decision making domain. In this paper, a new approach is developed to solve a Multiple Attribute Knapsack Problem (MAKP) in which each combination of items is evaluated using some quantitative and qualitative attributes. The assumed qualitative attributes cannot be measured by a mathematical formulation but by a DM/expert. In this paper, a Genetic Algorithm (GA) model has been developed to generate different combinations as the sequential population of the GA model. To rate the qualitative attributes for each chromosome (or combination) in the population, a Neural Network (NN) model has been developed. The ratings (or scores) resulted from quantitative attributes (by NN) and qualitative attributes (by mathematical formulation) for each chromosome form a row of a decision matrix. Having the decision matrix and known weights of attributes, the combinations in each population are ranked by applying a MADM model. The ranks obtained for each chromosome shows the fitness of that chromosome. Using the GA model, the best combination is achieved. The results of conducted experiments show the capability of the proposed approach to deal with MAKP problems.

The fundamental step in leather production is "tanning" or "processing" the animal hide. This process is performed in a wooden cylindrical device called a "drum" or a "Ianner". The raw hide is processed inside the drum with a variety of mixtures of chemicals and water under certain environmental conditions and controlled temperatures. During the process, controlling the temperature and pH of the mixture is of utmost importance and directly affects the quality of the leather produced. In many developing countries, including Iran, control is exerted manually by an experienced worker. Since the quality of leather is directly linked to the level of control maintained during the process, a numerical approach and a precise computer controller could not only increase the quality of the final product but would also make the process more efficient. In this paper, the benefits of a computer control system will be studied, the required software will be designed and, finally, the required computer hardware will be addressed.

A new procedure is presented for finding optimal solutions to resource constrained project scheduling problems. It traverses the decision tree by investigating the most promising partial schedule at each iteration, avoiding two main drawbacks associated with other similar procedures. These drawbacks are huge memory requirements and a time consuming task of selecting the most promising partial schedule among many candidates in each iteration.Computational experience indicates that the procedure finds optimal schedules for 100-activity projects, regardless of their strength of resource constraints.

In this paper, a Hybrid Genetic Algorithm (HGA) model is presented to solve two kinds of flow-shop problem and minimize makespan. These problems are: Regular and block flow shop problems. The flow-shop problem studied here is shown as n/m/ F/Cmax. Genetic Algorithm (GA) technique is one of the efficient and robust techniques for solving combinatorial programming problems, which can be used either alone (Simple GA) or along with other techniques (Hybrid GA). The latter can benefit from other heuristics to reach the best solution more efficiently. The HGA model presented in this paper uses two heuristics developed already to solve the n/m/ F/Cmax problems. The Palmer and CDS heuristics are embedded in the model to generate the initial population. The model works based on elitism. The elitist strategy takes special care to preserve the best solutions and inserts them in the next generation. By analyzing the conducted experiments, it is found that the HGA is a powerful technique for solving flow shop problems. The proposed GA model outperforms the well-known heuristics, both in terms of minimizing makespan and satisfying the closeness limitation in the block problems.

In this paper, the formulation and creation of an enterprise-wide global production planning system for the semiconductor industry are described, using linear optimization as the core planning technology. The objective is to generate production plans on a weekly or daily basis which are optimized to maximize customer service and cash flow in response to a prioritized statement of demand. When evaluating production alternatives, the system observes defined product and process structure relationships, as well as capacity and materials limitations within the various stages of manufacturing.

Each dynamic job shop can be represented as a network of queues, in which each service station indicates a machine or a production department. Now, assume that the service rates of these service stations can be controlled. In this paper, a new model is developed for bicriteria optimal control of service rates of all service stations in a class of Jackson networks, in which the expected value of the shortest path of the network and, also, the total operating costs of all service stations of the network, per period, are minimized. The expected value of the shortest path of such a network of queues is equal to the expected value of the time that the first product is completed.This is an important factor in production systems, because, related to design and manufacture of a new product, the first manufactured product often has the maximum flow time, in effect, which can be minimized through the model. The networks of queues analyzed in this paper, have all the specifications of Jackson networks, except for not containing M/M /C queueing systems.