Simulated Annealing Algorithm for Minimising the Surplus Parts in Selective Assembly-A Software Approach

Selective assembly is a precision assembly making process from wider tolerance components in which products are produced from randomly assembling the measured components from portioned bins. Very few papers were discussed in literature regarding cost aspects of manufacturing products in selective assembly techniques. Only procedural algorithms were used to calculate the cost of a product in these techniques. Moreover, tolerance cost curves and selection of tolerance for a component with alternative process were also not considered in most of the papers found in literature. The best combination of bins to make less scrap was computed only based on component's tolerance and not by considering the original dimension of the components. None of the paper used simulated annealing algorithm to compute the best bin combinations to make more number of assemblies in turn to reduce surplus parts. In this paper, these problems were removed using Power Builder, an optional programming language with Sybase SQL Anywhere as a database and Matlab software. Shaft housing assembly is considered as an example product to illustrate the effectiveness of the proposed method. Tolerance cost curves can be viewed within no time for individual process or together with a good number of processes. Moreover, for the same process and tolerance value, the best nominal dimension of the product can also be selected in this work and for the given tolerance of a component with alternative processes; the minimum cost process can be visualized and calculated.