Objective Due to inconsistent formats, nonstandard statistical practices, and disconnection from manufacturing processes, the nondestructive testing (NDT) data for EMU (electric multiple unit) bogies are insufficient to support the quality improvement of bogie components. To improve the quality of bogie products and enhance the controllability of component manufacturing and repair processes, it is necessary to investigate an effective statistical method for NDT data.

Method First, NDT data are systematically classified and standardized according to component attributes and testing methods. Second, for 'defect count' and 'non-conformity rate' two types of discrete data, an analytical framework centered on C control charts and P control charts is established, providing the data and theoretical foundation for constructing a quality improvement model. After the establishment of the model, a closed-loop improvement method covering the entire process of data collection, model monitoring, abnormal response, and corrective action verification is proposed. This method first standardizes data collection and preprocessing procedures, then establishes a real-time monitoring and abnormal response mechanism based on control chart models, and finally, in accordance with the PDCA (plan-do-study-act) cycle principle, verifies the effectiveness of the proposed quality improvement model in actual production scenarios.

Result & Conclusion  Through C control chart analysis of the defect data from 34 bogie frames, out-of-control frame defects caused by insufficient skills of the specific operators are identified, and the operators restored data to normal after training. Through P control chart analysis of the non-conformity rate of traction beams, out-of-control non-conformity rates caused by material issues are identified; after material optimization, the data returned to normal. The proposed quality improvement model can provide a theoretical basis and practical case studies in data analysis for the construction of digitalized and intelligent production lines in bogie manufacturing and repair processes.