Data Governance of Electrical Material Properties Based on EBGA-HSA

Many electrical materials are used in the design of power equipment in the power system. However, the existing database systems for electrical materials are not only vast and complex but also contain erroneous information and missing data, which significantly hinders practical application. This paper proposes a method for data governance in the electrical material system by combining the Elite Breakthrough Genetic Algorithm and the Hybrid Similarity Algorithm (EBGA-HSA). The proposed algorithm introduces a hybrid strategy, where multiple independent single-similarity functions are integrated into a hybrid similarity function to enhance the algorithm's adaptability to the characteristics of the electrical material database. Subsequently, an elite breakthrough strategy is proposed and the genetic algorithm is incorporated to determine the optimal solution for similarity weights. Ultimately, the EBGA-HSA is utilized for filling in missing data and correcting errors in the electrical material database of power systems. The algorithm's effectiveness is validated by establishing a predictive model and comparing it with the cosine similarity algorithm, Jaccard similarity algorithm, hybrid similarity algorithm, genetic-hybrid similarity algorithm, and Elite Breakthrough Genetic Algorithm-Hybrid Similarity Algorithm. Experimental results demonstrate that EBGA-HSA improves the ability to discriminate attributes within the electrical material data system and outperforms existing methods in data governance, thereby enhancing prediction accuracy. Therefore, the proposed algorithm demonstrates practicality and feasibility in the data governance of electrical material systems.