Assessing Embedded Software Contribution Under Data Scarcity: An Improved Grey PCA and GRA Framework

With the intelligent and digital upgrading of modern industrial system operation modes, software, as the core carrier of intelligent system equipment, has shown increasingly prominent application value. The traditional cost-based pricing and evaluation system for system software fails to effectively quantify its innovative value and intellectual attribute contributions, which makes it necessary to establish a performance-oriented comprehensive evaluation model. Aiming at the defects of existing contribution rate evaluation methods in small sample size and heterogeneous system assessment scenarios, this paper proposes a software contribution rate evaluation method based on grey correlation analysis and improved grey principal component analysis. The proposed method first screens core characteristic indicators via grey correlation analysis, adopts the improved grey principal component analysis algorithm to realize dimensionality reduction of high-dimensional characteristic data, and takes the grey correlation degree between principal components and system performance indicators as the weight to complete the quantitative calculation of the software contribution rate.