Reliability Estimation of Retraction Mechanism Kinematic Accuracy under Small Sample

Abstract

Due to intricate operating conditions, including structural clearances and assembly deviations, the acquisition of test data of landing gear retraction mechanism is limited, posing challenges for reliability analysis. To solve the problem, a Bayesian-based reliability analysis method by fusing prior and test data is proposed, focusing on the mechanism kinematic accuracy under small-sample conditions. Firstly, a dynamic simulation model is established to collect prior data, and retraction tests are conducted to obtain test data. Then, based on Bayesian theory, the motion accuracy parameter estimation model integrating prior and test samples is established. To obtain accurate hyper parameters, the prior samples are expanded using neural network. Finally, taking the retraction mechanism as the research object, the kinematic accuracy reliability is quantified, and the impact of uncertainty factors is analyzed in depth. The results show that the proposed method is superior to the classical interval estimation method in stability and effectively mitigates the impact of uncertainty factors.