S. Saravana Sankar

Parallel Genetic Algorithm for VLSI Standard Cell Placement

This work addresses the methods to solve VLSI standard cell placement problem with the objectives of minimizing the wire length and computational time. In this work a parallel GA architecture is designed and the computational time of Genetic algorithm is reduced by means of parallel technique. The proposed algorithms are tested in IBM bench mark circuits. Proposed method is compared with Qplace5.1.67,, Dragon 2.2.3 and Fastplace are found superior in terms of computational time and solution quality.

Maximising manufacturing system efficiency for multi-characteristic linear assembly by using particle swarm optimisation in batch selective assembly

Quality of an assembly is mainly based on the quality of mating parts. Due to random variation in sources such as materials, machines, operators and measurements, even those mating parts manufactured by the same process vary in their dimensions. When mating parts are assembled linearly, the resulting variation will be the sum of the mating part tolerances. Many assemblies are not able to meet the assembly specification in the available assembly methods. This will decrease the manufacturing system efficiency. Batch selective assembly is helpful to keep the assembly requirement and also to increase the manufacturing system efficiency. In traditional selective assembly, the mating part population is partitioned to form selective groups, and the parts of corresponding selective groups are assembled interchangeably. After the invention of advanced dimension measuring devices and the computer, today batch selective assembly plays a vital role in the manufacturing system. In batch selective assembly, all dimensions of a batch of mating parts are measured and stored in a computer. Instead of forming selective groups, each and every part is assigned to its best matching part. In this work, a particle swarm optimisation based algorithm is proposed by applying the batch selective assembly methodology to a multi-characteristic assembly environment, to maximise the assembly efficiency and thereby maximising the manufacturing system efficiency. The proposed algorithm is tested with a set of experimental problem data sets and is found to outperform the traditional selective assembly and sequential assembly methods, in producing solutions with higher manufacturing system efficiency.

MIP model and elitist strategy hybrid GA-SA algorithm for layout design

It is most important for any manufacturing industry to have an efficient layout for their production environment to participate in global competition. One of the prime objectives of such an organisation is to decide an optimal arrangement of their facilities (machines or departments) in a two-dimensional planar region satisfying desired objectives, which is termed facility layout problem. To overcome the drawbacks of traditional layout design methodology, it is attempted to solve three important layout design problems such as inter-cell layout design, determination of optimum location for input/output stations and flow path layout design of material handling system simultaneously in an integrated manner. The quality of the final layout is evaluated by minimizing the total material handling cost, where the perimeter distance metric is used for the distance measurement. Sequence-pair, an elegant representation technique is used for layout encoding. The translation from sequence-pair to layout is efficiently done by longest common subsequence computation methodology. Due to the non-polynomial hard nature of the problem considered, an elitist strategy based hybrid genetic algorithm that uses simulated annealing as local search mechanism (ESHGA) is developed and tested with test problem instances available in the literature. The results indicate that proposed integrated methodology with developed mixed integer programming based mathematical model along with ESHGA could generate realistic layouts compared to reported result.

Bi-objective Optimization in Identical Parallel Machine Scheduling Problem

This paper presents bi-objective identical parallel machine scheduling problem with minimization of weighted sum of makespan and number of tardy jobs simultaneously. It is a known fact that identical parallel machine scheduling problem with makespan and number of tardy jobs based criteria is NP hard. Metaheuristics has become most important choice for solving NP hard problems because of their multi-solution and strong neighborhood search capabilities in a reasonable time. In this work Simulated Annealing Algorithm SA and Genetic Algorithm GA has been proposed to optimize two different objectives namely i minimization of make span ii minimization of number of tardy jobs using combined objective function COF. The effectiveness of the proposed algorithm have been analyzed by means of benchmark problem taken from the literatures and relative performance measures for the algorithm have also been computed and analyzed. Computational results show that GA outperforms SA by a considerable margin.

Bi-criteria optimization in integrated layout design of cellular manufacturing systems using a genetic algorithm

Traditionally the design of the physical layout of the manufacturing system and that of the material flow path and material handling system are carried out in isolation. In this work, an attempt was made on the integrated layout design, that is, to concurrently design the physical layout and the material handling system using a Genetic Algorithm-based methodology. The proposed algorithm was employed to simultaneously optimize two contradicting objectives viz., 1. Total material handling cost 2. Distance-weighted cost of closeness rating score. The algorithm was tested on four different benchmark layouts and with different initial problem data sets. It was found that the proposed algorithm is able to produce satisfactory solutions consistently within a reasonable computational limit.

VNS-Based Heuristic for Identical Parallel Machine Scheduling Problem

Minimization of make span and minimization of number of tardy jobs in identical parallel machine scheduling problems are proved to be NP-hard problems. Many researchers have attempted to solve these combinatorial optimization problems by employing different heuristic algorithms. While providing a satisfactory solution to the production environment for each of the above-said objectives, still remains as a challenge, most of the time, the need has been to have satisfactory solutions optimizing simultaneously the above-said two objectives. In this research work, an attempt is made to address this issue and heuristic algorithms using simulated annealing algorithm (SA) and variable neighborhood search algorithm (VNS) have been developed to provide near-optimal solutions. The developed heuristics are tested for their efficiency on a very large data sets generated as per the prescribed procedure found in the literature. Based on the results of experiments, it is inferred that the VNS-based heuristics outperforms the SA-based heuristics consistently both in terms of solution quality and consistency.

Testing and performance analysis of micro encapsulated rice bran distilled fatty acid

This work deals with the performance analysis of incorporation of a phase change material (PCM) in building construction as micro capsules instead of macro encapsulation. Micro capsules are used in various applications, particularly in construction for passive exchange of heat across walls and roof. Completion of basic study leads rice bran distilled fatty acid-grade-A (RBDFA)/melamine formaldehyde (MF) as the suitable core PCM and covering material, respectively for the test site. Micro RBDFA capsules thermo-physical properties were studied using thermo gravimetric analysis (TGA), Fourier transform infrared spectrophotometer (FT-IR) and differential scanning calorimeter (DSC) and all test results satisfied the basic requirements. Two model blocks were constructed, one with macro encapsulated RBDFA roof and another one with micro encapsulated RBDFA roof. Temperature variation inside the model blocks was observed. The results indicated that, inclusion of micro encapsulated RBDFA in construction gives positive effect to maintain less temperature variation at indoor that is 5-7°C less than ambient condition.

Multi objective integrated layout design problem

Traditionally the design of Inter-cell layout and Material Handling System (MHS) of the manufacturing system is being carried out in step by step. This leads to sub-optimal solutions for facility layout problems (FLP). In this work an attempt is made to concurrently design Inter-cell layout and the MHS using a Genetic Algorithm (GA) based methodology using simulated annealing algorithm (SAA) as local search tool for a Cellular Manufacturing System (CMS) environment under open field configuration. The proposed algorithm is employed to simultaneously optimize two contradicting objectives viz. 1. Total material handling cost 2. Distance weighted cost of closeness rating score. The algorithm is tested on two different bench mark layouts and with different initial problem data sets. It is found that the proposed algorithm is able to produce approximate pareto-optimal solutions.

Ant Colony Optimization to Improve Precision of Complex Assembly

An assembly consists of two or more mating parts. The quality of the assembly is mainly based on the quality of mating parts. The mating parts may be manufactured using different machines and processes with different standard deviations. Therefore, the dimensional distributions of the mating parts are not similar. This results in clearance between the mating parts. To obtain high precision assemblies, clearance variation has to be reduced. Selective assembly helps to reduce this clearance variation. In this paper, appropriate selective group combination for assembling the mating parts is obtained using an ant colony optimization (ACO). The combination obtained has resulted in an appreciable reduction in clearance variations.

A new approach to nullify surplus parts in selective assembly

Selective assembly is a generally accepted method for producing high precision assemblies from relatively low precision components. The mating parts are manufactured with wide tolerances. The mating part population is partitioned to form selective groups. The corresponding selective groups are then assembled interchangeably. The accuracy of selective assembly is mainly based on the number of selective groups (fixed before the assembly) and the range of selective groups. However, there are often surplus parts in some groups due to the imbalance of mating parts, especially in the case of undesired dimensional distributions. This paper presents a new approach to nullify the surplus parts in selective assembly.