An active learning Kriging‐based multipoint sampling strategy for structural reliability analysis

Abstract

In order to effectively and accurately assess the failure probability of mechanical structures, this paper proposes a multi‐point sampling active learning reliability analysis method called AKMP. First, a GA‐Halton sequence is introduced to make the initial samples well dispersed and homogeneous in the design space. Second, a new learning function FELF is constructed to efficiently update the Kriging model, which takes into account the relationship between the location of the sampling points and the performance fun. Then, a combination of NCC criterion and multipoint sampling strategy is proposed to further improve the convergence efficiency, which can effectively terminate the active learning process. Finally, numerical and engineering cases are tested to verify the application performance of the proposed AKMP. The results show that the method has superior performance in terms of both accuracy and failure probability efficiency, and can reduce the computational resources of the active learning process.

Reliability Analysis of a Three-Engine Simultaneous Pouring Control System Based on Bayesian Networks Combined with FMEA and FTA

“…The operating failure rate of the low-voltage DC power supply is determined as λ py = 0.857 × 10 −6 (8) The working failure rate of the fuse The working failure rate of the fuse is given by λ pr = 0.150 × 10 −6…”

“…Reliability analysis methods such as Fault Tree Analysis (FTA), Failure Mode and Effect Analysis (FMEA), Reliability Block Diagram (RBD), and Markov Models have been widely used in rolling stock [7]. Reference [8] describes a reliability analysis method called AKMP, which is used to evaluate the failure probability of mechanical structures. This method utilizes multi-point sampling and active learning, which combines the GA-Halton sequence and a new learning function, FELF, to update the model.…”

Research on Optimal Preload Method of Controllable Rolling Bearing Based on Multisensor Fusion

“…The bearing preload is a crucial mechanical connection technique that establishes a robust link between the shaft and bearing by applying an appropriate pressure or tension. It effectively minimises clearance, enhances bearing rigidity and precision, reduces operational vibrations and noise, improves the load capacity of the machine, and ensures stable machinery operation [4,5].…”