Identification of Influential Modules Considering Design Change Impacts Based on Parallel Breadth-First Search and Bat Algorithm

Xianfu Cheng, Zhihu Guo, Xiaotian Ma, Yuan TianFrontiers in Bioengineering and Biotechnology 2022 · DOI

Abstract

Modular design is a cornerstone of modern engineering, allowing for flexibility and customization. However, the interconnectedness of modules introduces the risk of change propagation, where a modification in one component triggers a cascade of necessary updates throughout the system. This paper presents a framework for identifying "influential modules"—those with the highest potential for systemic impact. By integrating Design Structure Matrices (DSM), parallel breadth-first search (BFS), and the bio-inspired Bat Algorithm, the proposed method quantifies change impact degrees. A case study on a crane grab demonstrates the framework's ability to distinguish between high-impact influential modules and "absorption modules" that contain changes, providing a valuable tool for risk management in complex product development.

Introduction

In complex engineering systems, modularity facilitates the management of diversity and complexity. While modules are designed to be relatively independent, functional and physical dependencies often persist. When a design change occurs—due to updated customer requirements or performance improvements—it can propagate through these dependencies. This "ripple effect" can lead to significant cost overruns and schedule delays if not properly anticipated. Identifying which modules are most likely to propagate changes is critical for robust product architecture.

Methodology

The proposed framework utilizes a multi-step approach to map and quantify these risks.

1. Design Structure Matrix (DSM)

The process begins with a DSM to represent the dependencies between individual parts. Each cell in the matrix contains a numerical value representing the probability of change propagation between two components. This provides a granular view of the product's internal relationships and helps group parts into modules based on their connectivity.

Design Structure Matrix (DSM) showing dependencies between parts and modules.

2. Change Propagation Network and Parallel BFS

To understand how changes move beyond direct neighbors, the framework constructs a change propagation network. Because the number of possible paths grows exponentially with complexity, a parallel breadth-first search (BFS) algorithm is employed. This allows for the simultaneous exploration of multiple propagation branches, identifying all reachable parts and the cumulative probability of impact along various paths across module boundaries.

Change propagation network illustrating paths and probabilities between modules M1 and M2.

3. Optimization with the Bat Algorithm

Calculating the "comprehensive change impact degree" for every module is a complex optimization problem. The researchers utilize the Bat Algorithm, a metaheuristic inspired by the echolocation behavior of bats, to efficiently search the network for global maximum impact values. This ensures that the most critical "bottleneck" modules are identified even in highly complex architectures where manual calculation would be mathematically exhausting.

Case Study: The Crane Grab

The framework was validated using a heavy-duty crane grab consisting of 42 key components grouped into 9 modules. These modules include the Lower Pulley Group (M1), Upper Bearing Beam (M2), and various struts and pulley systems. By applying the DSM and Bat Algorithm, the researchers were able to rank the modules based on their potential to disrupt the entire system.

Results and Discussion

The analysis identified Modules 1, 6, and 7 as the most influential. Specifically, M1 (Lower Pulley Group) exhibited the highest comprehensive impact degree (9.5863), suggesting that any modification to this module necessitates extensive redesigns elsewhere.

Conversely, the study identified "absorption modules" (M4 and M5). These modules had impact degrees of zero, meaning they are "sinks" that can absorb changes from other parts of the system without passing them further along. Identifying these modules allows engineers to focus their risk-mitigation efforts on the high-impact areas while understanding where the system is naturally resilient.

Conclusion

This research provides a systematic, computationally efficient method for predicting the consequences of design changes. By moving from part-level analysis to module-level impact degrees, the framework offers project managers a high-level map of architectural risk. Future work will focus on improving the accuracy of initial probability estimates and extending the model to dynamic manufacturing environments and entire product families.

Summary updated

Abstract

Modular design is a widely used strategy that meets diverse customer requirements. Close relationships exist between parts inside a module and loose linkages between modules in the modular products. A change of one part or module may cause changes of other parts or modules, which in turn propagate through a product. This paper aims to present an approach to analyze the associations and change impacts between modules and identify influential modules in modular product design. The proposed framework explores all possible change propagation paths (CPPs), and measures change impact degrees between modules. In this article, a design structure matrix (DSM) is used to express dependence relationships between parts, and change propagation trees of affected parts within module are constructed. The influence of the affected part in the corresponding module is also analyzed, and a reachable matrix is employed to determine reachable parts of change propagation. The parallel breadth-first algorithm is used to search propagation paths. The influential modules are identified according to their comprehensive change impact degrees that are computed by the bat algorithm. Finally, a case study on the grab illustrates the impacts of design change in modular products.
BibTeX
@inproceedings{cheng2022identification,
  author = {Xianfu Cheng and Zhihu Guo and Xiaotian Ma and Yuan Tian},
  title = {{Identification of Influential Modules Considering Design Change Impacts Based on Parallel Breadth-First Search and Bat Algorithm}},
  booktitle = {Frontiers in Bioengineering and Biotechnology},
  year = {2022},
  doi = {10.3389/fbioe.2021.791566},
}
Figure