Saeed Maghsoodloo

Strengths and limitations of taguchi's contributions to quality, manufacturing, and process engineering

This paper reviews Genichi Taguchis contributions to the field of quality and manufacturing engineering from both a statistical and an engineering viewpoint. His major contributions are first listed and then described in a systematic and analytical manner. The concepts underlying Taguchis univariate quality loss functions (QLFs), his orthogonal arrays (OAs), robust designs, signal-to-noise (S/N) ratios, and their corresponding applications to quality and process engineering are examined and described in great detail. Some of Taguchis OAs are related to the classical (fractional) factorial designs (a field that was started by Sir Ronald A. Fisher in the early 1920s). The applications of Taguchis robust (parameter and tolerance) designs to manufacturing engineering are illustrated through designed experiments.

Optimization of mechanical assembly tolerances by incorporating Taguchi's quality loss function

Abstract In an end-to-end mechanical assembly, optimization of tolerance allocation is a critical issue. Many papers have concentrated on minimum cost-tolerance allocation without considering the quality of the final assembly. The objective is to examine the optimal tolerance allocation by considering both tolerance cost and quality loss so that the total assembly cost is minimized. In this paper, a sensitivity analysis is also conducted to determine the effects of shifts in a components mean and variance. We find that if a components mean varies, only the quality loss associated with that component will be changed. If a components variance shifts, the optimal allowance, tolerance costs, and quality losses associated with each component will be affected.

Optimal asymmetric tolerance design

An asymmetric tolerance design occurs when deviation (from the ideal target) of a quality characteristic in one direction is more harmful than in the opposite direction. Asymmetric tolerances are common in many manufacturing processes. Traditionally, the designer of a manufactured component either would choose the smaller tolerance as the tolerance for both sides of the ideal target, or would set a process mean at the middle of the tolerances. Both methods fail to minimize the expected value of Taguchis societal quality losses when the quality loss function is asymmetric. Linear and quadratic quality loss functions are considered to determine the optimal value of a process mean that minimizes the expected value of the quality loss function. Also, a quality loss model involving a pokayoke defect prevention procedure is investigated.

The safety impact of the 65 mph speed limit: a case study using Alabama accident records

This study investigated the effects of the change of the speed limit on rural Interstate roadways from 55 to 65 mph in Alabama, which took effect on August 1, 1987. The primary source of data was the accident records, which were compared for several test and control areas as well as one year time periods before and after. Traffic volume and speed data were also analyzed to properly interpret the results of the accident data analyses. Both average speed and traffic volume increased more on the rural Interstates than on other roadway types. Significant incrases in PDO and injury type accident frequencies were correlalted with the speed limit change. Spillover effects were found on 55 mph Interstates but not on other roadway types. Several other factors were analyzed, inlcuding weather, vehicle type, roadway character, number of vehicles, collision events, and the use of seat belts.

Quadratic loss functions and signal-to-noise ratios for a bivariate response

Abstract In this paper, the quadratic quality loss functions and signal-to-noise (S/N) ratios for a bivariate response are developed. Quality characteristics by variables are divided into three types: (1) Smaller the better (STB), (2) Larger the better (LTB), and (3) Nominal the best (NTB). The focus is on the bivariate quality characteristic response combinations (NTB, NTB), (STB, STB), and (LTB, LTB) cases. The treatment of mixed bivariate (STB, LTB), (STB, NTB), and (LTB, NTB) response cases will be forthcoming as a sequel to this paper in the next issue of this journal (Vol. 20, No. 2, 2001). The relationships among quality loss constants are also discussed in each case. An example of a robust parameter design experiment using simulated data is provided to illustrate the use of signal-to-noise ratios, which were developed to identify the optimal factor settings. The design matrix used in the example was an L 8 (2 7 ) Taguchi orthogonal array (OA) as a one-half fraction of a 2 4 full factorial experiment.

Simulation optimization based on Taylor Kriging and evolutionary algorithm

This paper develops a simulation optimization algorithm based on Taylor Kriging and evolutionary algorithm (SOAKEA) for simulation models with high computational expenses. In SOAKEA, an evolutionary algorithm is used to search for optimal solutions of a simulation model, and Taylor Kriging temporarily serves as a surrogate fitness function of this evolutionary algorithm to evaluate solutions. Taylor Kriging is an enhanced version of Kriging where Taylor expansion is used to approximate the drift function of Kriging, and it improves the interpolation accuracy of Kriging. The structures and properties of SOAKEA are analyzed. A combination correction strategy is created, and it effectively reduces the computational expense of SOAKEA. The empirical comparison of SOAKEA with some other well-known metaheuristics is conducted, and the proposed SOAKEA uses particle swam optimization, a population-based evolutionary algorithm, to solve four simulation problems based on multimodal benchmark functions. The results indicate that SOAKEA has significant advantages in optimizing simulation models with high computational expenses.

The Control Structure Diagram: An Overview and InitialEvaluation

A new graphical representation, the Control Structure Diagram (CSD), has been created to visualize software at both the source code and program design language (PDL) level. The primary impetus for creation of the CSD was to improve the comprehension efficiency of software and, as a result, improve reliability and reduce costs. The CSD has the potential to replace traditional prettyprinted source code. As part of the GRASP (Graphical Representations of Algorithms, Structures, and Processes) research project at Auburn University, the GRASP software engineering tool has been successfully developed. GRASP automatically generates CSDs from source code written in Ada, C, C++, Java, and VHDL. The emphasis to this point has been on the automatic generation of the CSD to support development, maintenance, reverse engineering and reengineering through the use of GRASP. GRASP has been applied successfully to numerous programs ranging in size from several hundred to several thousand lines of source code and is efficient and sufficiently flexible for use in a production setting. To demonstrate the potential benefits of the CSD and its automatic generation using GRASP, a series of empirical studies has been planned and initiated. First, as reported in this article, the perceived usefulness of the CSD was evaluated using a preference instrument based on eleven performance characteristics in which a comparison was made with other well-known graphical representations for algorithms. Statistical analysis indicated numerous significant differences with a clear preference for the CSD in seven of the eleven performance characteristics. Further empirical studies, currently being implemented, will examine the effect of the CSD and GRASP on objective measures such as comprehension efficiency and effectiveness.

A cost model for determining the optimal number of software test cases

A probabilistic model is presented that demonstrates the optimal number of software test cases required in situations where the following can be estimated as independent parameters: (1) the cost per test; (2) the cost per error if undetected until field implementation; (3) the number of software executions over its lifetime; (4) the number of possible different executions; and (5) the number of faults embedded in the software. A formula is derived by the use of calculus and is solved by approximation techniques. Tables of the optimal number of tests over a range of parameter values are presented to illustrate the results. The model serves as a basis for further research efforts to improve the accuracy of input variable estimation. >

Estimates of the quantiles of kendall's partial rank correlation coefficient

The sampling distribution of kendalls partial rank correlation coefficient, Jxy∗z, is not known for N>4, where N is the number of subjectts. Moran (1951) used a direcr conbinatorial method to obtain the distribution of Jxy∗z forN=4; however, ten minor computationa; errors in his Table 2apparently resulted in how erroneous entries for his frequency table. Since the parctial limits of the direct combinatorial approach have been reached once N>4, the first main objective of this paper was to obtain the exact distribution of Jxy∗z for N=f, 6, and 7 using an electronic computer. The second was to use the Monte Carlo method to obtain reliable estimates of the quantiles of Jxy∗z for N=8,9,...,30

A Simple Method for Obtaining Resolution IV Designs for Use with Taguchi's Orthogonal Arrays

A concern expressed in the literature about Genichi Taguchis methods for experimental design is that many such designs are not optimal in the sense that they are not the maximum resolution possible for a given number of main effects and array size. Thi..