Robot Autonomous Assembly Task Understanding Based on Information Mining

Aiming at the problems in the production process of multi-variety, small-batch, and high flexibility products, the complex process files need to be translated into a data formatthat can be recognized by robot directly, this traditional way will lead to an inaccurate understanding of the process information, useful information miss, low efficiency and other issues, a semantic understanding method of complex product assembly based on information mining is proposed. Taking the assembly process document of a certain type of regulating valve as anexample, the key information of the assembly process documentis extracted and the domain professional lexicon is established, the apriori algorithm is used to mine the matching relation between production parts and tools by extracting historical process records. According to the information organization form of entity-attribute-entity, the knowledge graph is established to associate the multi-source heterogeneous information needed by the robot assembly.