摘要:

A procedure is proposed for screening applications in quality control that eliminates overselection (namely, the condition where the proportion of accepted screened items is greater than the proportion of conforming items in the original population) and controls for individual unit misclassification error (IME), namely, the probability of an accepted/selected item being below specification. Screening involves selecting items whose performance is within specifications based on observing one or more correlated screening variables in lieu of observing the performance variable directly. Overselection and an uncontrollably high IME can occur when (1) there is a high correlation between the performance and screening variables and (2) the proportion of conforming items among accepted items (or its complement, average outgoing quality (AOQ)), is prespecified as in, for example, Owenet al. (1975). We instead propose a procedure thatprespecifiesa maximum tolerable IME (emaxa) for accepted items, which eliminates overselection and assures that the IME of each outgoing item as well as the AOQ are both equal to or less than emaxa. The quality of eachindividual itemas well as AOQ is thereby explicitly controlled. Only the use of standard normal tables is required to implement our proposed procedure.

ISSN:0740-817X    

DOI:10.1080/07408179508936786

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

A hybrid method is presented for the design of inspection stations in serial production systems. The hybrid method combines a constrained bottleneck shortest path algorithm and discrete event simulation. The network method determines the allocation of the inspection stations and the simulation helps the decision maker estimate accurately the possible outcome of the allocation. A computational study with the method is performed based on a practical problem. Other aspects related to the application of the method to real problems are also discussed.

ISSN:0740-817X    

DOI:10.1080/07408179508936787

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

Conventional learning curve models are able to deal only with past data that includes an integer number of cycles and time per cycle. However, in real-life situations, the data collected are of a different nature: periodical information, which includes total work time, in-process inventory and completed units, for example. In such circumstances it would be wrong to consider just the completed units, disregarding the in-process inventory. The concept of Equivalent Number of Units (ENU) has been introduced to permit one to sum up all the work performed, and express it in a manner that allows one to use the learning curve model for such cases. Also, time per unit may not be given for single units and the ENU produced per reported period may not be an integer. In order to solve the above problems, a simple procedure was developed for periodical non-integer data. The end result of the procedure is the product learning curve parameters.

ISSN:0740-817X    

DOI:10.1080/07408179508936788

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

Most research in group decision support systems (GDSSs) has focused on small-group, single-site activity using a Decision Room concept. ‘Computer-mediated meeting management’ (CM3) is a new group decision support system designed to address the challenges of meetings distributed in time and space. Design principles, operational characteristics and experiences gained from sessions using the CM3system are discussed. In addition, the fundamental issues related to decentralized meetings, including mechanisms for identifying conversational threads in brainstorming sessions, are presented. Finally, the unique features of the CM brainstorming and voting modules are discussed in detail.

ISSN:0740-817X    

DOI:10.1080/07408179508936789

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

Coordination among computer integrated manufacturing databases has become one of the key difficulties in the implementation of an integrated system. The issue of how to implement interdependency among CIM databases has some alternative technological solutions, trading off various levels of mutual consistency against various levels of the system throughput. Although these alternatives and their formalization are important issues in the database theory perspective, this research deals with an engineering issue of devising a methodology to select among these alternatives in any given case. A heuristic optimization model, with worst time complexity ofO(N2) is presented. Its effectiveness is analyzed with a case study and a series of sensitivity analysis activations. This is a first attempt to solve this problem. Our empirical results show that the overall system performance, as reflected by a system's goal-function, is substantially improved by the use of our model, relative to intuitive or trivial solutions.

ISSN:0740-817X    

DOI:10.1080/07408179508936790

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

A new approach to training backpropagation neural networks for identifying cyclic patterns on control charts is presented. The objectives of this research are to show that building an effective cyclic-pattern-recognition neural network requires proper training strategies and to demonstrate how these strategies, namely, incremental and decremental training, should be applied and how the performance can be improved with additional statistics. A series of experiments were conducted to study the effect of the number of output pattern classes and the effect of noise on network training and performance. Experiments show that reducing the number of output pattern classes to a small number, e.g., four or fewer, does not guarantee effective learning, and that the noise added to the training data should be maintained at a reasonable level to achieve a balanced performance. Further incorporation of harmonic amplitude statistics (HAS) also proved that the proper use of statistics adopted from Fourier analysis can improve the performance of a cyclic-pattern-recognition neural network. This study offers valuable insights as to how to construct and train a back-propagation neural network properly and effectively for detecting cyclic patterns.

ISSN:0740-817X    

DOI:10.1080/07408179508936791

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

The most commonly used model of ‘statistical’ control' for industrial production processes is that observations taken over time are independent and identically distributed. Thus, for example, the estimate of variability used to establish Shewhart X¯-chart control limits is almost always based on within-subgroup variability. We review and present generalizations of this in-control model to include multiple components of common cause variability, such as batch-to-batch variability. An example involving integrated circuit processing, with three components of variability, is given to illustrate the proposed techniques. The usefulness of the multi-vari chart is emphasized.

ISSN:0740-817X    

DOI:10.1080/07408179508936792

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

Trade-off analysis after optimization of an economic CUSUM control chart is proposed. This procedure exploits the relatively flat response surface in the neighborhood of the optimal design. Comments are made with respect to criticisms of pure economic models and the difficulties with constrained optimization. Two previously published numerical examples are used to demonstrate that, by using trade-off analysis, large gains in statistical properties are attainable with little increase in cost. By focusing on designs with improved in-control and out-of-control run length performance, false-alarm rates can be reduced and the ability to detect process shifts quickly can be increased.

ISSN:0740-817X    

DOI:10.1080/07408179508936793

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

The automated guided vehicle (AGV)system is emerging as the dominant technology to maximize the flexibility of material handling, while increasing the overall productivity of manufacturing operations. This paper presents a new way of finding the shortest flow path for an AGV system on a specific routing structure. An optimal solution of the system is determined by using an approach based on the Hopfield neural network with the simulated annealing (SA) procedure. In other words, the proposed approach reduces the total cost of an AGV delivery path from one workstation to another on the shop floor. By changing the temperature of the two-stage SA, a solution can be found that avoids potential collisions between AGVs. Both the flow path and the potential collision, which are major problems in AGV systems, may be solved simultaneously by the proposed neural network approach. Other advantages offered by the proposed method are its simplicity compared with operations research (OR)methods and a decreased number of needed AGVs. The performance of the approach is also investigated.

ISSN:0740-817X    

DOI:10.1080/07408179508936794

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor

摘要:

When dealing with maintenance and replacement policies for finite or infinite periods, the cost parameters tend to vary owing to inflationary trends. If the objective is to minimize the total cost incurred during a given planning period, the computational time required to obtain the solution increases as the planning horizon increases. A similar effect is also seen for other parameters. In such cases it is preferred to obtain solutions of good quality that require less computational time. This paper discusses the benefits drawn from the already available Branching Algorithm. With minor modifications to the algorithm, the reduction in the number of nodes created and the computation time required are very significant. The solutions are compared with the solutions of the Branching Algorithm for a wide range of parameters.

ISSN:0740-817X    

DOI:10.1080/07408179508936795

出版商:Taylor & Francis Group    

年代:1995

数据来源: Taylor