Simultaneous coordination of capacity building and price decisions in a decentralized supply chain

In this paper, we investigate the simultaneous coordination of price and capacity building decisions in a dyadic supply chain. This problem is a combination of capacity reservation problem and pricing problem. While the coordination of supply chain with stochastic demand and fixed prices has received much attention in the literature, price-dependent and stochastic demand has been less considered. We study the latter case where a price-setting retailer faces a linear decreasing demand with respect to price. To capture the uncertainty of the demand, we add a stochastic variable to the demand function. In addition, we incorporate production rate and inventory cost on the supplier side. We propose Revenue Sharing Reservation Contract with Penalty (RSRP) as a coordination mechanism to align the price and capacity decisions. We then extend the model to include multiple retailers which are geographically dispersed. We next conduct a comprehensive numerical example with an extensive sensitivity analysis to understand the behavior and robustness of the supply chain under a RSRP, and finally, we draw some managerial implications.     

A multi-criteria decision-making approach to solve the product mix problem with interval parameters based on the theory of constraints

The product mix problem is one of the most important problems in production systems. Several algorithms to determine the product mix under the theory of constraints have been developed. Most of the previous works focused on one bottleneck (dominant bottleneck) and considered the product mix problem with exact data. In this paper, all bottlenecks are used in order to determine the aggregated priority of each product, and a multi-criteria decision-making approach is proposed for product mix problem with interval parameters. The proposed approach involves bottlenecks identification, determination of the production priority and the weights vector of bottlenecks, application of interval TOPSIS to calculate the aggregated priority of each product, and use of reducing and increasing process to improve the production plan. At the end, a numerical example is presented to illustrate the procedure of the proposed approach. The results obtained from the computational study have shown that the proposed algorithm is an effective approach to solve the product mix problem.     

Metaheuristics for fuzzy dynamic facility layout problem with unequal area constraints and closeness ratings

Dynamic facility layout problem deals with the problem of arranging and rearranging the layout plan of a system throughout several periods. In each period, the material handling costs are different from the past period due to the change in the market demand and product mix. In this paper, the uncertainty that exists in transportation values’ forecast is modeled by fuzzy theory. In this paper, departments have unequal areas. This means that in each period, departments can be placed only in certain places due to different spacing requirements. In addition, closeness rating matrix is considered in order to model the goodness of different layout plans with regard to qualitative factors according to decision maker. Accordingly, fuzzy dynamic facility layout problem with unequal areas and closeness rating matrix is considered that aims to minimize the uncertain material handling costs as well as the shifting costs, and maximize closeness rating with regard to space requirements of unequal area departments. A number of fuzzy algorithms are developed in order to deal with the problem. A number of ranking criteria from the literature are implemented in order to compare the performance of the developed algorithms.     

Manufacturing-to-Sale Planning Model for Fuel Oil Production

The fuel oil refinery production industry in Taiwan has entered a completely competitive free open market. In such an environ-ment there exist a variety of uncertain factors, and a traditional production planning model will not be able to deal with the new situation. This study develops a responsive and flexible manufacturing-to-sale planning system to deal with uncertain manufacturing factors. The major objective of this study is to model the problem of the uncertain nature faced by the Chinese Petroleum Corporation (CPC) and to establish a manufacturing-to-sale planning model for solving the problem. A linear programming technique is suggested for developing the optimal strategy for use in production plans. Fuzzy theory is adopted for dealing with demand/cost uncertainties. A possibilistic linear programming model is thus formulated in this study. The possible uncertain fluctuations on demand/cost are included in the model. Therefore, the strategy for creating maximum profit for the company can be obtained via the proposed modelling procedures.     

A new model of dynamic cell formation by a neural approach

This paper proposes a nonlinear integer model of cell formation under dynamic conditions. The cell formation (CF) problem is a portion of a cellular manufacturing strategy (CMS), in which the parts and machines are clustered with the aim of minimizing the material handling cost. In most previous research the cell formation problem has always been under static conditions in which cells are formed for a single-period planning horizon where product mix and demand are constant. In contrast, in dynamic conditions, a multi-period planning horizon is considered, where the product mix and demand in each period is different. This occurs in seasonally or monthly production. As a result, the best cell design for one period may not be efficient for subsequent periods. To verify the presented model, different problems have been solved and results are reported. Where the cell formation problem belongs to NP class, the use of a novel approach is necessary. In this research, we apply a neural approach based on mean filed theory for solving the proposed model. In this approach, the network weights are updated by an interaction procedure. The proposed model is solved by LINGO software and an optimum solution is obtained. Comparison of optimum and neural approach solutions shows the efficiency of the presented neural network approach.     

Possibilistic inventory and supplier selection model for an assembly system

In this paper, the problem of inventory lot-sizing and supplier selection for an assembly system is considered, where the supplier available capacities are assumed as ambiguous dynamic parameters. In this scenario, which is a frequent case in large assembly-based factories such as automobile manufacturers, the final product is assembled from multiple components with different conversion factors, which can be sourced from multi-capacitated suppliers through the multi-period horizon of imprecise demand. Due to high shut-down costs of assembly lines, it is assumed that production never stops even though some components may not be available. Therefore, the unfinished products are transferred to a buffer zone and preserved there until the lacking components become available. In this study, a possibilistic mixed integer mathematical model, with fuzzy objective function and soft constraints, is developed to determine which component in what quantities, from which suppliers, and in which periods should be ordered. The model, inspired by the real case of the Iran Khodro Car Company, aims to maximize the profit while keeping a high customer service level by avoiding shortages. This model also considers the ambiguity of dynamic parameters such as demand, suppliers’ available capacities, prices, and holding and shortage costs. To solve the problem, the possibilistic model is first converted into an auxiliary crisp multi-objective model. Through an interactive fuzzy approach, the suggested multi-objective problem is then transformed into an equivalent single-objective model. Finally, a particle swarm optimization is proposed to achieve the overall satisfactory compromise solution. A numerical sample is used to validate the proposed model.     

Efficient management of production-inventory system in a multi-item manufacturing facility: MTS vs. MTO

Suppliers produce a variety of products to serve both large and small customer orders with unreliable demand information. Furthermore, suppliers also face customer pressure to improve quality, lower cost, and reduce delivery delay. These conflicting objectives lead firms to use both make-to-stock and make-to-order production strategies together. These manufacturing strategies were known to be competing policies and in some cases the choice depends on characteristics of the product. In this study, the firms that manufactured multiple-item types are considered to be free to choose either production policy for each product type (no product-specific requirements are present). Then, using the order-arrival characteristics and cost parameters for each product type, the firm wants to decide which production/scheduling policy to use for each product type. Two production policy (MTS vs MTO) and two scheduling strategy (FIFO vs cyclic) are considered in this study. The analysis and numerical study show that there is no dominant strategy neither for production policy, nor product scheduling policy.     

A tool planning approach considering cycle time constraints and demand uncertainty

The tool planning problem is to determine how many tools should be allocated to each tool group to meet some objectives. Recent studies aim to solve the problem for the cases of uncertain demand. Yet, most of them do not involve cycle time constraints. Cycle time, a key performance index in particular in semiconductor foundry, should not be ignored. The uncertain demand is modeled as a collection of scenarios. Each scenario, with an occurrence probability, represents the aggregate demand volume under a given product mix ratio. A genetic algorithm embedded with a queuing analysis is developed to solve the problem. Experiments indicate that the proposed solution outperforms that obtained by considering only a particular scenario.     

A tool planning approach considering cycle time constraints and demand uncertainty

The tool planning problem is to determine how many tools should be allocated to each tool group to meet some objectives. Recent studies aim to solve the problem for the cases of uncertain demand. Yet, most of them do not involve cycle time constraints. Cycle time, a key performance index in particular in semiconductor foundry, should not be ignored. The uncertain demand is modeled as a collection of scenarios. Each scenario, with an occurrence probability, represents the aggregate demand volume under a given product mix ratio. A genetic algorithm embedded with a queuing analysis is developed to solve the problem. Experiments indicate that the proposed solution outperforms that obtained by considering only a particular scenario.     

Optimal production control and marketing plan in two-machine unreliable flexible manufacturing systems

In this paper, we have developed a production planning and marketing model in unreliable flexible manufacturing systems with inconstant demand rate that its rate depends on the level of advertisement on that product. The proposed model is more realistic and useful from a practical point of view. The flexible manufacturing system is composed of two machines that produce a single product. Markovian models frequently have been used in modeling a wide variety of real-world systems under uncertainties. Therefore, in this paper, the inventory balance equation is represented by a continuous-time model with Markov jump process to take into account machines breakdown. The objective is to minimize the expected total cost of the firm over an infinite time horizon. While the total cost consists of the cost of the product surplus, the cost of the production, and the cost of the advertisement. In the process of finding a solution to the problem, we first characterize an optimal control by a class of linear stochastic system where some parameter values are subject to random jump. By defining quadratic cost functions and characterizing the associated limiting optimal control problem, a discrete-time approximation model and an asymptotic optimal control model are developed. It is clear that such a solution exists and can be obtained as a limit of a monotonic sequence with solving the steady-state Riccati equation.     

Production control in a failure-prone manufacturing network using discrete event simulation and automated response surface methodology

In this paper, a system consisting of a network of machines with random breakdown and repair times is considered. The machines in this system can be in one of four states: operational, in repair, starved, and blocked. Failure and repair times of the machines are exponentially distributed. Previous research on multi-machine failure-prone manufacturing systems (FPMS) has focused on systems consisting of machines in series or in parallel. This paper considers a network of machines with relationship constraints. Additionally, the system under study models work in process for multiple products, intermediate and final buffers and one type of final product. The demand rate for the final commodity is constant and unmet demand is either backlogged or lost. The objective of this control problem is to find the production rates and policies of the different machines so as to minimize the long run average inventory and backlog cost. The applied control policy is the hedging point policy that is determined by factors representing the level of buffer inventory for each machine. Obtaining analytical solutions is generally impossible for such complex systems. To simultaneously control the production rates of the machines we have therefore developed a method based on a combination of stochastic optimal control theory, discrete event simulation, experimental design and automated response surface methodology (RSM). The application of an automated RSM for Network FPMS is another contribution of this paper. The model can be extended easily to systems with age-dependent failure rates, a preventive repair maintenance policy and non-exponentially distributed up and down times.     

Application of chance-constrained programming for stochastic group shop scheduling problem

In this paper, we study a group shop scheduling (GSS) problem subject to uncertain release dates and processing times. The GSS problem is a general formulation including the other shop scheduling problems such as the flow shop, the job shop, and the open shop scheduling problems. The objective is to find a job schedule which minimizes the total weighted completion time. We solve this problem based on the chance-constrained programming. First, the problem is formulated in a form of stochastic programming and then prepared in a form of deterministic mixed binary integer linear programming such that it can be solved by a linear programming solver. To solve the problem efficiently, we develop an efficient hybrid method. Exploiting a heuristic algorithm in order to satisfy the constraints, an ant colony optimization algorithm is applied to construct high-quality solutions to the problem. The proposed approach is tested on instances where the random variables are normally, uniformly, or exponentially distributed.     

A particle swarm optimization approach for constraint joint single buyer-single vendor inventory problem with changeable lead time and (r,Q) policy in supply chain

In this paper, the chance-constraint joint single vendor-single buyer inventory problem is considered in which the demand is stochastic and the lead time is assumed to vary linearly with respect to the lot size. The shortage in combination of back order and lost sale is considered and the demand follows a uniform distribution. The order should be placed in multiple of packets, the service rate limitation on each product is considered a chance constraint, and there is a limited budget for the buyer to purchase the products. The goal is to determine the re-order point and the order quantity of each product such that the chain total cost is minimized. The model of this problem is shown to be an integer nonlinear programming type and in order to solve it, a particle swarm optimization (PSO) approach is used. To assess the efficiency of the proposed algorithm, the model is solved using both genetic algorithm and simulated annealing approaches as well. The results of the comparisons by a numerical example, in which a sensitivity analysis on the model parameters is also performed, show that the proposed PSO algorithm performs better than the other two methods in terms of the total supply chain costs.     

满足产品需求条件下的车间最优随机生产计划与控制

根据实际需要建立关联方程有延迟且以正好满足产品需求为约束条件的车间生产计划与控制的随机非线性规划模型，即一种求解动态优化问题的静态优化模型，为求解方便将其转化成线性规划模型。提出分别用卡马卡算法和基于卡马卡算法的关联预测法来求解柔性自动化车间（FAM）最优随机生产计划与控制问题，并编制了相应软件。通过算例研究，比较了上述2种方法和Matlab中的线性规划法，结果表明所提方法非常适合将不确定性环境中的FAW产品需求计划最优分解成由FAW中各柔性制造系统（FMS）执行的短期随机计划，尤其适合FMS之间工件传输需经出入库并有1个生产周期延迟的情况。     

不确定需求下的设备动态布局方法

针对现有车间设备动态布局方法存在的不足,在考虑产品需求不确定性对布局性能稳定性影响的基础上,提出了一种结合模糊理论与改进遗传算法的不等面积设备动态布局方法。分析了产品需求不确定性及其随时间变化特性,引入了三角模糊数描述不确定产品需求;通过分析各生产阶段间的设备重组过程,将动态布局转化为数个静态布局,构建了基于柔性区域结构的不确定需求动态布局模型。结合三角模糊数运算及排序方法与自适应局部搜索机制提出了改进遗传算法,以物料搬运及设备重组费用总和为优化目标,解决不确定需求下的不等面积设备动态布局问题。通过算例测试和实例分析,验证了所提方法的有效性和实用性。     

Job releasing and throughput planning for wafer fabrication under demand fluctuating make-to-stock environment

In the semiconductor industry, dynamic changes in demand force companies changing the product mix makes the production planning challenging. This paper aims at an environment where product mix changes periodically and presents a production scheduling system to plan the wafer lot release and throughput. The proposed system is designed on the make-to-stock basis with the objective of meeting demand forecast while maintaining production smoothness. Two modules are included in the system. Preliminary analysis module analyzes throughput and cycle time distributions for different product mixes so as to determine relative parameters to be used as the inputs to the job releasing plan. In the production scheduling module, with the considerations of the attainment of demand forecast, production smoothness, and commitment of the due dates of the released job orders, the job release schedule and completion time table are prepared. A simulation model of a semiconductor fab is used as the base case to demonstrate the effectiveness and efficiency of the proposed system.     

Multi-agent-based proactive–reactive scheduling for a job shop

Multi-agent-based proactive–reactive scheduling for job shops is presented, aiming to hedge against the uncertainties of dynamic manufacturing environments. This scheduling mechanism consists of two stages, including proactive scheduling stage and reactive scheduling stage. In the proactive scheduling stage, the objective is to generate a robust predictive schedule against known uncertainties; in the reactive scheduling stage, the objective is to dynamically rectify the predictive schedule to adapt to unknown uncertainties, viz. the reactive scheduling stage is actually complementary to the proactive scheduling stage. A stochastic model is presented, which concerns uncertain processing times in proactive scheduling stage on the basis of analyzing the deficiencies of a classical scheduling model for a production schedule in practice. For the stochastic scheduling problem, a multi-agent-based architecture is proposed and a distributed scheduling algorithm is used to solve this stochastic problem. Finally, the repair strategies are introduced to maintain the original proactive schedule when unexpected events occur. Case study examples show that this scheduling mechanism generates more robust schedules than the classical scheduling mechanism.     

Inspection Allocation Planning for a Multiple Quality Characteristic Advanced Manufacturing System

Producing products with multiple quality characteristics is always one of the concerns for an advanced manufacturing system. To assure product quality, finite automatic inspection systems should be used. Inspection planning to allocate inspection stations should then be performed to manage the limited inspection resource. Except for finite inspection station classes, in this work, the limited number of inspection stations, of each inspection station class, is considered for solving the inspection allocation problem in a multiple quality characteristic advanced manufacturing system. Since the product variety in batch production or job shop production increases to satisfy the changing requirements of the various customers, the tolerances specified will vary from time to time. This inspection allocation problem is solved using a unit cost model in which the manufacturing capability, inspection capability, and tolerance specified are concurrently considered for a multiple quality characteristic product. The situation of unbalanced tolerance design is also considered. The inspection allocation problem can then be solved according to customer requirements. Since determining the optimal inspection allocation plan seems to be impractical, as the problem size becomes large, two decision criteria (i.e. sequence order of workstation and tolerance interval) are employed separately to develop two different heuristic solution methods in this work. The performance of each method is measured in comparison with the enumeration method that generates the optimal solution. The result shows that a feasible inspection allocation plan can be determined efficiently. ID="A1"Correspondance and offprint requests to: Dr Yau-Ren Shiau, Department of Industrial Engineering, Feng-Chia University, 100 Wenhwa Road, Seatwen, PO Box 25–097, Taichung 407, Taiwan     

实时动态排产系统研究

针对众多企业在多品种小批量生产且缺乏稳定的主生产计划时难以实施MRPⅡ与JIT的困境，研究了一种新的生产计划与控制体系——实时动态排产系统。详细介绍了该系统的整体架构、订单处理流程和排产与插单算法。对所提出的两种排产算法及相应的插单算法进行了数值模拟，给出了不同算法的优缺点及应用条件。受运算速度的制约，该系统目前较适合产品结构简单、生产工序较少的企业。但随着计算机技术的发展与排产算法的进一步优化，该系统将会有更加广阔的应用前景。     

Stochastic models for transfer point location problem

The transfer point location problem has been introduced recently and for the case of minimax objective and planar topology, has only been studied for situations in which demand points are not weighted and have known coordinates. In this paper, we consider the case in which demand points are weighted and their coordinates have bivariate uniform distribution. Also, the problem is developed from a conceptual view and different distance measures are used to make models more applicable in real world situations. The problem is to find the best location for the transfer point such that the maximum expected weighted distance to all demand points through the transfer point is minimized. Depending on assumptions for uniform distributions, two models are considered, convexity conditions are discussed, properties of the optimal solution are obtained and methods to solve the problems are proposed. Finally, numerical examples are given.