Foam flotation operation performance evaluation method based on image feature joint distribution
The invention discloses a froth flotation operation performance evaluation method based on image feature joint distribution. The froth flotation operation performance evaluation method comprises the following steps of 1, obtaining preprocessed froth flotation image feature data; step 2, screening to...
Saved in:
Main Authors | , , , , |
---|---|
Format | Patent |
Language | Chinese English |
Published |
28.03.2023
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | The invention discloses a froth flotation operation performance evaluation method based on image feature joint distribution. The froth flotation operation performance evaluation method comprises the following steps of 1, obtaining preprocessed froth flotation image feature data; step 2, screening to obtain final evaluation characteristics for evaluating the flotation operation performance; step 3, obtaining a trained clustering algorithm; fourthly, a Gaussian distribution model under the froth flotation operation condition is established according to the result obtained through clustering; and 5, determining the category of the to-be-identified working condition according to the Gaussian distribution model. According to the method, the classification condition of different classification cluster boundary areas in K-means clustering is reasonably optimized by using the KNN, and inter-cluster diversity is enhanced on the premise of ensuring intra-cluster similarity; according to the method, high-dimensional Gau |
---|---|
Bibliography: | Application Number: CN202211645720 |