Pareto-optimization of complex system architecture for structural complexity and modularity

Abstract

Due to ever-increasing complexity of cuttingedge engineering systems, the need for managing structural complexity and modularity of such systems is becoming important. The complexity of the overall system architecture is mostly decided during the initial concept generation stage, when configurations of major modules within the system are determined. In this paper, we present a multiobjective optimization framework for (1) minimizing the variation in complexity allocation to individual modules, while (2) maximizing for the degree of modularity. The optimization framework was applied to a case study, where a trailing bogie system for railroad train was optimized for structural complexity allocation among individual modules and overall system modularity. The modularity maximizing decomposition is shown to induce a large variation in module-level complexity distribution with a small fraction of modules sharing a disproportionately large chunk of overall system complexity, while equitable distribution of module-level complexity leads to erosion in the degree of modularity achieved for the resulting system decomposition.

Design structure matrix‐based modularization approach for complex systems with multiple design constraints

“…Works by Asadi et al and Suh et al focused on modular system design to improve various aspects of the system flexibility. Sinha and Suh proposed Pareto‐optimization methods for optimizing simultaneously the modularity of a system and the structural complexity allocation. Baylis et al introduced a Pareto‐optimization approach for product platform modularity and commonality.…”

“…Heat (H) 6,8,9,10,16,17,18,23,30,31,32 Specified components are heat-generating components and need to be assigned to separate modules if possible Maintenance (M) 1 day 1,19,33 Specified components require thorough inspection every 3, 5, or 6 years. Components that require daily maintenance are inspected every day.…”

Entropy‐based system assessment metric for determining architecture's robustness to different stakeholder perspectives

“…System architecture is the synthesis, screening, and selection of design concepts wherein designers incorporate the functional characteristics of a developing system into a physical form such that the interfaces between form and function are clearly defined to specifically govern the behavior of the system . For the development and operation of large and complex systems, architectures are decomposed, analyzed, and selected with the aims of appropriately distributing the complexity of the system into a set of manageable subsystems . System architectures are defined by decomposing the system, wherein architects model the interdependencies between system components, organizations, and processes .…”

“…While modularity can be a way of managing complexity, complexity is distributed across a set of modules and therefore, the overall complexity of the engineered system remains unchanged . An example of a modular architecture is a desktop computer, wherein specific functions are carried out by different modules . Continuous efforts have been made to aid complex system decomposition and modularization by implementing various clustering algorithms to generate an optimized set of modules .…”

“…2 For the development and operation of large and complex systems, architectures are decomposed, analyzed, and selected with the aims of appropriately distributing the complexity of the system into a set of manageable subsystems. 3 System architectures are defined by decomposing the system, wherein architects model the interdependencies between system components, organizations, and processes. 4 On a physical level, system architectures can be modeled and explored using the product-architecture design structure matrix (DSM), which is a matrix-based visualization of a network model that helps in mapping the physical relationships between the components of a complex system.…”