Correlation modeling method for coupling failure of critical components of deep well hoist under incomplete information condition

摘要

#$%^&*AU2018374073A120190815.pdf#####ABSTRACT The present invention discloses a correlation modeling method for coupling failure of critical components of a deep well hoist in an incomplete information condition. The method includes the following steps: 1) acquiring fault data samples of the critical components of the hoist in different failure modes, and collecting statistics about statistical moment information of random response in each failure mode in the incomplete information condition; 2) obtaining a marginal probability distribution function of each failure mode by using a function approximation method; 3) analyzing probability correlation attributes between each pair of failure modes by using a goodness-offit test criterion, and determining best-fit copula functions for describing different correlation attributes of part failure; and 4) establishing a hybrid copula function model by combining the marginal probability distribution function of each failure mode with each best-fit copula function. By means of the present invention, an accurate marginal probability density function can be established in a small sample condition, and symmetric correlation, and upper tail correlation and lower tail correlation attributes possibly existing between failure modes are considered, thereby enhancing the flexibility and applicability of correlation modeling for coupling failure of critical components of an ultra-deep well hoist.