Customized configuration for hierarchical products: component clustering and optimization with PSO

Most of the current studies on product configuration suppose modules in configuration are well-defined. In fact, if a component is regarded as the basic configuration item in a hierarchical product, it is worth studying how to cluster components in a better way to form modules satisfying some requirements for product diversity and cost. The authors identify the possible computation scale of a component clustering issue, represent it with network diagram, and apply discrete particle swarm optimization (PSO) to seek a desired clustering way of components. A mobile printer is introduced to exemplify the calculation procedure of the proposed method, and some components of the product are decomposed further so to evaluate the adaptability of PSO to the change of computation scale. The effectiveness of the method is validated in a wide range.