衬衫缝制生产关键工序的识别与诊断OA

To address the problems of inaccurate identification of key processes and low efficiency of quality problem traceability in garment sewing workshops,this paper constructs an integrated identification-diagnosis method for the sewing process chain of men's shirts.As the core link affecting product quality and production efficiency in the garment manufacturing industry,the accuracy of key process identification and the efficiency of quality traceability directly determine an enterprise's process management level and product competitiveness.Traditional key process identification relies on the subjective experience of experts and lacks systematic quantitative evaluation.Meanwhile,quality traceability mostly adopts a post-event analysis mode,which makes it difficult to quickly locate the root cause of problems.The integrated method proposed in this paper fills the gap between theoretical research and practical application in this field. Firstly,this paper establishes a process criticality evaluation system covering four first-level indicators and 12 second-level indicators based on the influencing factors of sewing processes,which fully includes qualitative and quantitative factors and avoids the one-sidedness of single-factor evaluation.The fuzzy analytic hierarchy process(FAHP)is used to determine the indicator weights,solving the fuzziness and subjectivity in the weight determination process.Combined with the grey comprehensive evaluation method,the criticality of 28 sewing processes of men's shirts is quantified,and key processes are selected according to the criticality ranking.Secondly,a Bayesian network(BN)diagnosis model is constructed by integrating man,machine,material,method,measurement,environment(5M1E)and fault tree analysis(FTA).Potential fault factors are identified via the 5M1E theory,and the logical relationship between quality problems and root-cause factors is clarified through FTA,providing reasonable structural support for the Bayesian network model.Finally,parameter learning and backward reasoning are realized through GeNIe software to conduct probabilistic traceability of quality problems in key processes. Verification is carried out based on three months of production data from the HL men's shirt production line of an enterprise.The results show that six key processes are identified,with an 89%consistency rate compared with the judgment results of enterprise quality experts,verifying the practical applicability of the proposed evaluation system and identification method.The root cause diagnosis accuracy of the Bayesian network model for quality problems in the key process of cuff reaches 93.20%,proving that the model can realize rapid and accurate traceability of quality problems. The proposed integrated method is applicable to the screening of key processes and the rapid location of quality root causes in garment sewing workshops.It not only provides a scientific and reliable technical tool for enterprise process management,but also offers decision support for process control under different product types and customer standards.This method is of great significance for improving the overall quality management level of the garment manufacturing industry.