Failure sensitivity analysis of safety belt guide ring parameter design based on BP neural network

When the car suddenly collides, the webbing bunching phenomenon will occur at the upper guide ring (D-ring) of the height adjuster of the seat belt, causing serious damage to the human chest by local positive pressure. In this paper, finite element method is used to calculate lateral displacement of seat belt. The BP neural network training method was used to fit the relationship between the displacement response and the random parameters, and the reliability function was established. The reliability and reliability sensitivity of random parameters were calculated by first-order second-moment method, and the influence rule of belt guide ring and belt parameters on reliability was obtained, which provided a theoretical basis for the parameter reliability design of safety belt guide ring.