Quality-integrated diagnostic platform for aerospace complex product assembly processes

Abstract

Addressing the challenge of tracing and quantifying quality issues that emerge during the assembly process of intricate aviation products, this paper introduces a Quality-Integrated Diagnostic Platform (QIDP). The platform effectively merges machine learning algorithms with the Fuzzy Analytical Hierarchy Process (FAHP) to provide comprehensive solutions for complex aviation product quality assessment. Firstly, the viability of amalgamating assembly quality management with the random forest diagnostic algorithm is presented. Secondly, leveraging the attributes inherent to the intricate assembly process, alongside algorithmic model development and the underpinning of an information technology platform, a novel architecture is introduced for diagnosing the sources of quality issues within the aviation complex component assembly process. Then, building upon the architectural framework, this paper elucidates the design and implementation procedure of the diagnostic algorithm that amalgamates the FAHP with the random forest technique. To illustrate, the bushing assembly process within the aerospace sector is taken as an example. Finally, grounded in the model framework and fusion algorithm, the Assembly Process QIDP is conceived and actualized. The platform seamlessly integrates data provisioning and retrieval components for initial deployment within the aerospace cluster. These components encompass data acquisition, data processing, algorithmic construction, and the visualization of diagnostic outcomes. The achieved diagnostic results effectively fulfill expert evaluation criteria and offer a swift, pragmatic, and intelligent resolution for addressing quality fluctuations within the assembly process of aerospace enterprises.