Research on acquisition method of robotic shot blasting process knowledge based on PCA-rough sets

At present, the derusting of ship deck has been replaced by the low-efficiency, serious pollution, and backward manual hand-held grinder grinding to the high-intelligence,high-efficiency,and environmental-friendly deck shot blasting robot shot blasting. However, due to the different requirements for derusting grades, not all parts need deep cleaning. Excessive blasting will not only slow down the working speed, but also increase the consumption of shot materials. Choosing a reasonable shot blasting process for shot blasting robots can improve the quality of shot blasting. Improve and reduce production costs. Therefore, this paper proposes a shot blasting process data acquisition method based on principal component analysis-rough set, which is used to guide the robot shot blasting. First, select the target data set from the robot shot blasting process database, check and preprocess the selected target data set; use principal component analysis to reduce the dimensionality of the deck shot blasting multi-source heterogeneous data, and construct a decision table Attribute, sample attribute parameter selection, establishment of decision table and value reduction algorithm based on decision matrix to remove redundant condition attributes, explore the influence of equipment parameters, operating specifications, environmental conditions, etc. on the rust removal level, and obtain shot blasting process knowledge. Knowledge has important guiding significance for improving the efficiency of shot blasting robots and reducing costs.