Sanjay Kumar Shukla

Optimization of the supply chain network: Simulation, Taguchi, and Psychoclonal algorithm embedded approach

In todays market increased level of competitiveness and uneven fall of the final demands are pushing enterprises to make an effort for optimization of their process management. It involves collaboration in multiple dimensions viz. information sharing, capacity planning, and reliability among players. One of the most important dimensions of the supply chain network is to determine its optimal operating conditions incurring minimum total costs. However, this is even a tough job due to the complexities involved in the dynamic interaction among multiple facilities and locations. In order to resolve these complexities and to identify the optimal operating condition we have proposed a hybrid approach incorporating simulation, Taguchi method, robust multiple non-linear regression analysis and the Psychoclonal algorithm. The Psychoclonal algorithm is an evolutionary algorithm that inherits its traits from Maslow need hierarchy theory and the Artificial Immune System (AIS). The results obtained using the proposed hybrid approach is compared with those found out by replacing Psychoclonal algorithm with the Artificial Immune System (AIS) and Response Surface Methodology (RSM), respectively. This research makes it possible for the firms to understand the intricacies of the dynamics and interdependency among the various factors involved in the supply chain. It provides guidelines to the manufacturers for the selection of appropriate plant capacity and also proposes a justified strategy for delayed differentiation.

Six sigma-based approach to optimize deep drawing operation variables

The six sigma approach has been increasingly adopted worldwide in the manufacturing sector in order to enhance the productivity and quality performance and to make the process robust to quality variations. This paper deals with one such application of six sigma methodology to improve the yield of deep drawing operations. The deep drawing operation has found extensive application in producing automotive components and many household items. The main issue of concern of the deep drawn products involves different critical process parameters and governing responses, which influences the yield of the operation. The effects of these parameters are analysed by the DMAIC (Define, Measurement, Analyse, Improve, Control)-based six sigma approach. A multiple response optimization model is formulated using the fuzzy-rule-based system. The functional relationship between the process variables and the responses is established, and thereafter their optimum setting is explored with the aid of response surface methodology (RSM). Rigorous experimentations have been carried out, and it is observed that the process capability of processes is enhanced significantly, after the successful deployment of the six sigma methodology.

Soft decision trees: A genetically optimized cluster oriented approach

When descriptions of data values are too detailed, the computational complexities involved in mining useful knowledge from the database generally increases. This gives rise to the need of tools techniques which can reduce these complexities and mine the valuable information hidden behind the database. There exists number of such techniques viz. decision trees, neural networks, rough-set theory, rule induction, and case-based reasoning which are able to meet the aforesaid objective up to some extent. Each of these techniques has its advantages and limitations that motivate researchers to develop new tools for the mining tasks. In this paper, we have developed a novel methodology, genetically optimized cluster oriented soft decision trees (GCSDT), to glean vital information imbedded in the large databases. In contrast to the standard C-fuzzy decision trees, where granules are developed through fuzzy (soft) clustering, in the proposed architecture granules are developed by means of genetically optimized soft clustering. In the GCSDT architecture, GA ameliorates the difficulty of choosing an initialization for the fuzzy clustering algorithm and always avoids degenerate partitions. This provides an effective means for the optimization of clustering criterion, where an objective function can be illustrated in terms of clusters center. Growth of the GCSDT is realized by expanding nodes of the tree, characterized by the highest inconsistency index of the information granules. In order to validate the proposed tree structure it has been deployed on synthetic and machine learning data sets. Moreover, Results are compared with those produced by standard C4.5 decision trees and C-fuzzy decision trees; further student t-test is applied to show that these differences in results are statistically significant.

Chaos-based improved immune algorithm (CBIIA) for resource-constrained project scheduling problems

This paper introduces a novel meta-heuristic, the chaos-based improved immune algorithm (CBIIA), for solving resource-constrained project scheduling problems (RCPSP). In RCPSP the activities of a project have to be scheduled with the objective of minimizing total makespan subject to both temporal and resource constraints. The proposed CBIIA is based on the traits of an artificial immune system, chaotic generator and parallel mutation. CBIIA is different from the traditional immune algorithm in its initialization and hypermutation mechanism. Initialization in CBIIA is done by using chaotic generator (Logistic, Tent, and Sinusoidal) instead of conventional random number generator (RNG). The hypermutation is performed by parallel mutation (PM) operator rather than point mutation. Parallel mutation comprises two mutation strategies viz. Gaussian and Cauchy. Gaussian strategy is utilized for small step mutation and Cauchy strategy is for large step mutation. In order to demonstrate the efficacy of the proposed algorithm, Pattersons test suites are worked out. This study aims at developing an alternative and more efficient optimization methodology and opening the application of variants of artificial immune system for solving the RCPSP.

GA Guided Cluster Based Fuzzy Decision Tree for Reactive Ion Etching Modeling: A Data Mining Approach

There are various data mining techniques that are frequently used for the mining of vital patterns embedded within bulk data. These techniques include neural network, regression analysis, rough set theory, Bayesian network, decision trees, and so on. In this research, a novel data mining technique, genetically guided cluster based fuzzy decision tree (GCFDT), is introduced for the mining task. In order to test the efficacy of GCFDT, it is employed for building the predictive process models of reactive ion etching (RIE) with the aid of optical emission spectroscopy (OES) signals. This model endeavors to predict the wafer surface conditions for the new incoming set of process parameters. OES is an efficient tool for monitoring plasma emission intensity. In contrast with the C-fuzzy decision tree where granules are devolved through fuzzy clustering here, granulation is practised through genetically guided fuzzy clustering. The growth of the tree is governed by expanding the node having highest diversity. The results obtained by employing CGFDT in RIE process modeling reveal that it dominates both the traditional C-fuzzy decision trees and C4.5 decision trees in terms of both the accuracy and compactness.

Predictive modeling of high-performance concrete with regression analysis

High-performance concrete (HPC) is a very complex material and hence very hard to predict its compressive strength. This paper deals with building a regression model for predicting concretes compressive strength. First of all, eight process variables are identified as determinants of Concrete Compressive Strength (CCS). These variables are Cement, Blast Furnace Slag, Fly Ash, Water, Superplasticizer, Coarse Aggregate, Fine Aggregate, and Age. Further, correlation among these variables is computed and it is found that a few of them are highly correlated. Therefore, interactions among these variables are taken into account. After that, a regression model is developed by regressing CCS against all process variables and significant interactions. Finally, diagnostics are conducted to fine tune the model and a parsimonious model is obtained with 84.37% coefficient of determination. Appropriateness of the model is investigated by testing it against unseen data points.

Pulling the Value Streams of a Virtual Enterprise with a Web-based Kanban System

A Kanban system is one of the major enablers of lean manufacturing implementation. With the aid of information technologies, web-based Kanban systems inherit the benefits of using physical Kanban cards and meanwhile eliminate several limitations such as the scope and distance of applied areas, amount and types of information contents, real-time tracking and monitoring, and flexibility for adjustments. With the enhanced functionality, the Kanban system is no longer merely for shop floor control. The web-based Kanban system is ready to bring the “pull” concept into a distributed manufacturing environment to make a virtual enterprise. This chapter presents the functionality requirements and available solutions of web-based Kanban systems. The applications of a web-based Kanban system in various environments, from manufacturing cells to virtual enterprises, are explored. A web-based Kanban system provides great visibility of the production flows in an enterprise system. It can deliver a clearer picture of the up-to-date status of the system as well as the dynamics over time. Using the enhanced information, decision makers will be able to plan and manage production flows of a virtual enterprise more effectively.