Real-time Determination and Correction of Multi-source Multi-modal Heterogeneous Data Quality

With the rapid development of industrial informatization, digital twins have achieved comprehensive dynamic simulation in the production and manufacturing processes in digital form. They cover the entire product lifecycle and the entire industrial chain, which is of significant importance in promoting the development of future intelligent manufacturing. Digital twin workshops measure the multidimensional attributes of the production process by deploying multiple independent heterogeneous data sources, resulting in multi-source, multi-modal heterogeneous perception data. However, due to inherent limitations of equipment and environmental noise, perception equipment inevitably encounters issues such as missed reads, multiple reads, and incorrect reads, leading to common data quality problems. Therefore, this paper focuses on the twin workshop as the main subject and multi-source, multi-modal heterogeneous perception data as the object.It explores critical technologies such as real-time data collection and data quality control during the production process. These technologies enable the assessment and optimization of real-time data quality, ensuring the generation of high-quality digital twin models.This allows the workshop to operate accurately and efficiently, providing a reference for advancing the research and application of digital twin technology in future manufacturing industries.