National Science and Technology Innovation 2030 Next-Generation Artificial Intelligence Major Project,
当前, 智能制造面临的许多问题都具有不确定性和复杂性, 单纯地利用专家经验和机理模型难以有效解决. 本文面向跨层跨域的复杂制造系统网络化协同控制机制, 提出一种基于本体的人机物三元数据融合方法, 研究复杂制造环境下的人机物三元数据融合建模. 在抽取三元组时, 区别于传统的流水线式抽取方式, 提出一种基于实体-关系联合抽取的模型ErBERT. 该模型首先经过预训练模型BERT进行词序列化, 经过最大池化、全连接和Softmax等操作后, 完成实体识别和关系分类任务, 得到抽取完毕的人机物三元组. 将抽取好的三元组按照规则映射至OWL文件, 最终存储在图数据库中, 实现本体模型的构建. 经实验证明, 经过ErBERT抽取出的三元组有较好的准确率, 达到了通过本体来融合人机物三元数据的目标, 并为实现制造企业人机物三元协同决策与优化提供技术支撑.
Abstract: At present, many problems of intelligent manufacturing are uncertain and complex, which cannot be solved effectively by expert experience and mechanism model. In this paper, the human-cyber-physical data fusion modeling of complex manufacturing environments is studied, and the ontology is proposed as the approach of the human-cyberphysical data fusion. In the extraction of triplets, a model named ErBERT based on entity-relation joint extraction is proposed, which is different from the traditional pipeline extraction. After word serialization by pre-training model BERT, the model completes entity recognition and relationship classification by max pooling, fully connection and Softmax, and obtains the extracted human-cyber-physical triplets. The extracted triplets are mapped to OWL files according to rules, and finally stored in the graph database to realize ontology construction. The experimental result shows that the triplets extracted by ErBERT has a good accuracy and achieves the goal of fusion of human-cyber-physical data through ontology, which provides theoretical method support for realizing the ternary collaborative decision-making and optimization of human-cyber-physical data.