Reliability Modeling and Closed-Loop Optimization of a Non-Contact Profile Detection System Using Fuzzy AHP and Grey Relational Analysis
Limited failure data and uncertain expert judgments complicate reliability allocation and prediction during preliminary design. This study develops a closed-loop framework for a non-contact profile detection system for axisymmetric thin-shell workpieces. Five functional subsystems are modeled using a series reliability block diagram and fault tree analysis. A two-level comprehensive factor model with fuzzy analytic hierarchy process (AHP) allocates failure-rate targets, while similar-product prediction integrating ordered weighted averaging (OWA), intuitionistic trapezoidal fuzzy numbers, and grey relational analysis (GRA) corrects historical rates. Comparison identifies the optical measurement and numerical-control rotary table subsystems as having the smallest reliability margins. Under the benchmark scenario, optical power control, dual-camera redundancy, sealing, thermal management, and dual-grating redundancy reduce the predicted baseline system failure rate from 775.007 × 10-6 h-1 to 650.031 × 10-6 h-1 and increase mean time between failures (MTBF) from 1290.3 to 1538.4 h. Sensitivity scenarios yield MTBFs of 1421.5–1676.3 h; For the same increment in subsystem failure-rate reduction fractions, the sensitivity of the system failure rate to S4 improvement is approximately 2.08 times that to S3 improvement. The framework supports allocation, prediction, redistribution, and prospective optimization without complete life-test data.
Authors
- 胡启国
- Huoran Hu
- Yasen Feng
Institutions
- State Grid Corporation of China (China) (CN)
- Chongqing Jiaotong University (CN)
Publication Details
- Journal
- Transactions of the Canadian Society for Mechanical Engineering
- Published
- 2026-11-21
- DOI
- https://doi.org/10.1139/tcsme-2026-0119
- Primary Topic
- Reliability and Maintenance Optimization
- Type
- article
- Field-Weighted Citation Impact
- 0.00